Loading

MakerBot Print is our newest print-prepration software, which supports native CAD files and STL assemblies,
allows you to interact with all your printers via the Cloud, and many other exciting new features.

Download Now

thorgal

ToyREP 3D Printer

by thorgal Jul 28, 2015
Download All Files

Thing Apps Enabled

Please Login to Comment

Hi it's me again
Could you tell me where I could buy all the bolts, nuts, rods(metal parts), on the internet.
Thank you very much

Hi,
I honestly don't know. Bought mine in local hardware stores. And the few e-shops I know are in Czech Republic and probably don't ship abroad. Sorry.

ok thank you very much
And thank you for the design, I really like it.

I would like to make it but im not sure if the printing quality is good?
Could you tell me more about it?
Thanks

Check the videos on youtube. Don't expect it to print like i3. Round shapes have slight problem beacause of the backslash in internal gearbox. And it prints slowly.

Hello dears! help please, from whom there is ready a sketch on dannuya staff the: Mega 2560 R3 CH340; A4988; 12864 controller; engines 28BYJ-48. which is on the site sketch, stitched but the screen just tan and all. thanks for earlier!

by "sketch" you mean firmware changes required or drawing of electric connection?

changes required

Necessary changes for Marlin Firmware are on webpage https://reprap.org/wiki/ToyREP_User_Manual

Edit: But there is not chapter dedicated to 12864 display. That you have to find elsewhere.

Hello. Is there a possibility for you to mod Z axis to use NEMA17 motors? There's a mod for X and Y axis but not for Z :(

Z axis is not possible to modify without changing most of the printed parts and many rods. Actually it is not so crucial as Z axis is used only when layer is changed and the backslash of gearbox don't make artefacts on prints. If you want to speed up the movement of Z axis, it would be a better idea to change Z axis to use trapesoidal rods 6mm dia 12mm pitch. It will be less complicated as only two printed parts have to be changed (along with the threaded rod).That will result in 3mm/s with resolution of 0.003mm/fullstep. I have the rods and nuts at home ready for this upgrade, but not enough time to make models and publish this.

Hi , Iam very beginner in electronics & arduino .
can send me the Setting file upload it directly to arduino .

Thanks for your help , but now motor is 5 V and my ramps driver 12V , WHAT CAN I DO?

Please read the instruction and comments first. That is because of the bipolar hack. http://i.imgur.com/zcWIgva.png You join two 5V coils togehter.

is this motors can support long time and precise printing or better use large one for 200 X 200 x 200 Printing size

Well, I would rather say no. Technicaly it is possible, but longest print I did on ToyREP was around 5 hours. And precision is influenced by the internal gearbox of 28BYJ-48. You can use the Nema 17 tweak for X and Y axis along with Nema 17 bowden extruder. That can do the job but you will spent $30 more.

There is configuration.h file in Thing Files if that's what you're looking for. However, it may not work if you just upload it and you may have to edit some things. At reprap wiki there is a manual on how to configure the firmware.

Hello Thorgal!

I've been building the toyREP for a while and today is the day I wanted to make it print. BUT, I've come to a problem. The stepper motors (I, of course, used the bipolar hack - cut the connection and desoldered the red wire) simply don't work. When I connect NEMA 17 to the Ramps, it works flawlessly, but other motors simply don't.
Z and Y - Turns okay most of the times, but in random direction. I tried various acceleration speeds (from 20 to 200) but nothing changed.
X used to "work" the same way, but now I unscrewed it to look inside, it seems like it's trying to do anything and I can feel it resonating, but the shaft doesn't turn.

I'm just wondering if you came across this problem too. If not, don't read more, but I'm gonna write something I realised, in case you know about this.

First things first: While building the ToyREP, I wanted to try the motors. They didn't work, so I was searching and searching till I found that I have to short all 3 wires under the motor driver to get it going and that's what helped me. The question is, should the pins be shorted?

When I connect a multimeter between two cables in the middle of the stepper motor (pink and orange), it starts beeping - I assume they are connected, but should they be? I tried different motor, new one. I carefully "hacked" it and it does the same, and I doubt all the motors I ordered are bad :D

Thank you very much,
Appreciate what you did for this community :)

■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■
■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■SOLVED!!!■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■
■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■■
After desperately trying to solve this with help of Thorgal, we solved it and printer now works.
The only problem was that two cables of the stepper motors had to be swapped. I thought they are connected as they should, but when I tried and swapped the blue and orange cables, the motor started working perfectly.
Sample of the very first print: https://gyazo.com/52f1066101cb32c59f04a09e88306fcb
Once I have more tests, I'll post my ToyREP in the makes section as well.

Good luck everyone building this printer and thank you Thorgal! :)

Well, that's confusing. I now tried to remove the pins for shorting under the drivers, and the motor works. Yet, it still spins in random direction..

30min later, I have the same problem after editing the marlin numbers to the ones on your site (your ones). Z didn't work from the start, but now it does (but it moves pretty weird, compared to back then when it moved smoothly).
And now that I tried connecting nema 17 as before, it works really weird (but I guess it should, it's a different motor after all).

EDIT: I changed some firmware settings (

define DEFAULT_AXIS_STEPS_PER_UNIT { 650, 650, 2700, 500 }

define DEFAULT_MAX_FEEDRATE {On the way, 1, 0.35, 4.17}

)
And Z with Y turn pretty smoothly. Just gonna find out the right number to put instead of "On the way" and I'm good I guess. But one problem still remains - it turns in random direction. Acceleration is 20mm/mm/s so I don't know how much lower should I go.

There are three things that will influence the motor movement. And once you will set everything for 28BYJ-48 it is not so easy to put there NEMA 17 a run.
1, reference voltage on driver - for 28BYJ-48 I think you can safely use 0.09A to 0.15A per one motor (0.18-0.3A in Z axis). That is Vref 0.07-0.12V with common A4988 or 0.045V-0.075V with DRV8522. Nema17 needs roughly ten times more.
2, movement speed - for 28BYJ-48 inToyREP is safe to go 12.5, 12.5, 0.2, 1.37 mm/s in X, Y, Z and E. Everything above is matter of tweaking and testing
3, acceleration and jerk - start with someting acound {300, 300, 1, 17}, and jerk 5, 5, 0.01 and 0.1 for X, Y, Z and E. You can change theese once your motors are working. This is again totaly different from numbers ysed for NEMA 17.

Now, if your motors are chopping and changing directions randomly, there is either problem with wireing coils of the motor, pin connector or reference voltage. If they will try to move, but stall, it is probably problem of speed or acceleration and jerk.

Mine 28BYJ-48 after the bipolar hack have this order in header pin: Blue, Yellow, Pink, Orange. Blue-yellow is one coil, pink-orange is the second one and you will either detect short or measure same resistance with multimeter on each coil. There must not be any short between the coils otherwise the hack is not done properly.

Well, something doesn't feel right. I measured the voltage between - and the potentiometer on the DRV8825 and it is 1.6V. Wut?

  • When I disconnect the usb cable from arduino so my motors turn off (I temporarily left the motors on even while not moving), the voltage is 0.36V and it doesn't change even when I rotate the potentiometer. And when I plug the usb in, I can adjust it from 0.04 to like 10V? While 1.6V is in the middle (It was set to 1.6 when manifactured).

There must be something wrong either with your ramps (I suppose you have it along with arduino mega) or more probably the drivers.
Or you measure the Vref voltage on wrong places. Place one probe in the middle of trimpot and the other one on Gnd pin of the driver.
https://i.stack.imgur.com/xlrt3.png

Popolu driver with DRV8825 is typicaly made in such way, it prepares its own voltage for Vref from motor Vmx. The up most V3P3out it can give is 3.4V/0.1mA (typicaly 3.3V) and maximal Vref you can return back to the driver and the driver can survive (when you don't use V3P3out) is 4V.

So 1.6V in the middle of the scale of trimpot is perfectly OK, but you shall never measure more than 3.3V in the maximal position. And this can be only measured at the moment when 12V (which is your Vmx) is connected to the ramps because DRV8825 don't use 5V comming from Arduino as it have Fault pin instead of Vdd (compared to A4899, which relies on 5V Vdd from Arduino).

Now, current for motor Ifs is calulated as: Ifs= Vref/(5xRsense) where Rsense are two small SMD resistors next to the tripot. Typical Chinese drivers have Rsense 0.1OHM, whitch is nonsense, because than the maximal current for one coil can be up to 6.6A!!! and that is crazy. Even preset 1.6V means 3.3A per coil, way above the specified 2.5A this driver can temporaly give. So the driver is probably in some error state and resets it self frequently.

Try to turn the trimpot as much to the minimum as possible (beware, these trimpots can go 360°, so you can get to place where it is off and than straight to maximum). Plug 12V (without the USB cable) and measure the Vref (BTW the ramps in default setup will power up the arduino instead of USB so it will prepare 5V anyway). Try to get close to the 0.08V on X, Y and E axis and 0.16V on Z. It is good idea to use screw driver with ceramic tip or atleast with insulated handle. Turn it, than remove screw driver and measure, Repeat it until you will get there. Don't measure with the screw driver on trimpot as it can get some parasitic capacity.

Hopefully the drivers and motors are not gone yet.

Yes, Ramps is along with mega.

When I have the USB cable disconnected (when the motors are not active):
Well, I cannot measure any voltage between the trimpot and gnd of the driver - it shows no sign of voltage.
However, I can measure the voltage between the trimpot and "-" of the PSU (and so I did, all the numbers I wrote before are measured by that) - it says 0.04V and it doesn't change even if I move the trimpot to it's ends (mine stops when you reach the max/min). Also the motors are equally strong and it doesn't how I turn the trimpot - only at a tiny point - when I rotate it to it's minimum, the motor stops working until I set higher voltage again.

Now I tried brand new driver (for times like these I like to order one more piece of everything). I screwed the trimpot to the minimum voltage and turned on the PSU while USB is not connected. Between the GND and trimpot there was 0V, but between PSU and trimpot there was 0.018V. This finally seemed right to me.. Until I realised that when I turn the trimpot, the voltage stays the same. Still 0.018V. (Turning the PSU off, moving the trimpot, turning PSU on and measuring it - doesn't change a thing. Still same measurements).

It's only when the motor is running that I can adjust it I guess?

BTW: My minimum of the trimpot seems kinda opposite of where it should be. When I screw - I get to the minimum; When I unscrew - I get to the maximum.

But it doesn't seem probable to me that the drivers are gone, since the motors move nice and smoothly (if I have the right settings for them).

You don't need to see any pictures or the video if you can imagine the problem I have, I'm including them just in case you wanted to see it :D.
Here I got some pictures of measuring for you (they are just showing what I wrote here):
https://ibb.co/gZV6Fq
https://ibb.co/eYaHMV
https://ibb.co/jZH8TA
https://ibb.co/fDU0aq
And one video of the Y-motor moving - though with every axis, the same thing happens (with ultra low acceleration...):
https://youtu.be/0smqlpRj9i4

PS: Should there be any of the pins under the driver shorted or not?

Sincerely thank you even though we haven't solved much yet. I didn't expect responses kind like that :)

I don't have any spare arduino mega and ramps to check this out (burned 2 arduinos in past two years :D), but I am almost 100% sure Vref on drivers can be changed without motors running as it is recommended to turn the motors off (for eg. in Pronterface), when you change the Vref. The motor even don't have to be connected to the driver when you set the Vref.

Yes you can measure Vref between the center of trimpot and any GND you have in machine. So "-" of PSU or the metal cover of USB connector is just fine. BUT ... Strange is the fact you cannot measure any voltage between trimpot center and GND pin on driver. Please check with electronics disconnected from any power supply if GND of driver and other GND ( or "-") on ramps is connected (with multimeter it shall look like short circuit). If GND is not connected to driver pin, there is probably some problem with ramps or driver, like something is not soldered well.

Yes DRV8825 have opposite direction of trimpot compared to A4899. It goes to minimum when you go clock wise.

On this picture https://a.pololu-files.com/picture/0J4233.600.png?665d623ba84232de64511d8aa6644836 is schematic of the driver. GND pins shall be connected together and to "-" of PSU. MS0, MS1, MS2, will be either open or connected to Vdd = fault pin on DRV8825 (depends on microstepping you have set). Step, Dir, Enable are connected to microcontroler (hard to check). Sleep and Reset are either joined together or connected to Vdd (I am not sure here). A1 can be connected with A2 and B1 with B2 only when motor is connected (you can measure the resistance of the each coil here). Vmot is conected to "+" of PSU. Check other driver positions too.

No problem. I've learned a lot through helping other people out with ToyREP and this is just the kind of experience I am looking for ;).

Wow, what?
Between the trimpot and the PSU's GND there is a voltage of 0.018V
Between the trimpot and the USB's shell there is no voltage
Between the trimpot and the DRV's GND there is no voltage
But, USB's shell and the GND of PSU are connected, as well as the USB's shell and DRV's GND is connected. Also all GNDs of DRV's are connected.

I wonder how is that even possible? And I have an answer, it is not (as I was taught...). Well, I tried to measure it again, and again, but I get the same result every time.
So I assume the only faulty thing here must be my multimeter? Indeed, I don't think so - the multimeter is fairly new and disappoints me only when measuring resistance. It has always given me good values of voltage.

Other measurements:
Between fault and GND - 0.4V (logic power supply) - it should be 2.5-5.2V..
Both GNDs on the DRV are connected.
Reset (I couldn't identify sleep pin) is not connected to Vdd.

B1, B2; A1, A2 - B with B and A with A show the same resistance (although, it is about 1710ohm? well, maybe my multimeter is wrong). What's interesting is, that when I connect the motor, this resistance drops to +-1660. Like, what? BTW, when I turn on the PSU, I can't measure any resistance here (even with motors connected).
Motor coil resistance is +-46ohm for both.

I guess RAMPS and DRVs are ignoring all the laws of physics :D

Told myself to do some more measuring:
Voltage between all trimpots and PSU's GND is 0.018V as I wrote before. With the exception of Extruders DRV - 0.01V here while trimpots are turned the same on every driver. You might think that it is because I only tested XYZ motors and drivers and E is the only working now, but nah, the E motor was running as well.
So I measured the resistance on D10 - Extruder's heater - to test my multimeter. It's 4Ohm (+-); 1212/4 (UU/R=P) is 36W, so the multimeter seems to be all right...

I'm feeling kinda hopeless now :D I have to go to sleep now, but tomorrow I will probably measure same things on my geeetech printer (although, I believe I have A4988 there :/ ), hopefully something will ring a bell.

I (we) have to figure something out, I don't want this to be wasting your time.

PS: Should I replace the DRVs with A4988 from my geeetech or vice versa? Seems like good idea to me.

I think a have an explanation, and You will not like it. Don't measure voltage, just check "short" connection by multimeter. Remove RAMPS and Arduino out of the printer. Disconnect it from everythig and remove drivers. Use it to check the GND pins under the drivers. If you will find them "disconected" from "-" on PSU clamp or USB cover (no short detected) than the pins or traces on RAMPS burned when the drivers were set to Vref 1.6V = 3.3A.
If that is the case, remove RAMPS out of Arduino and check it from the bottom side. If you can measure short between soldering of GND pin and "-" on the clamp, traces are good and the pins are gone otherwise you will have to find where the trace is interrupted.
http://domoticx.com/wp-content/uploads/2017/05/RAMPS-Shield-1.4-print-layout-beide-kanten.png

If the GND line on RAMPS is OK, GND pins (just the pins) on your drivers are gone. You can check them separately by checking short between tip and soldering.

Anyway you will have to solder a bit or buy some new parts.

EDIT: Last thing I can come up with is that the RAMPS is not inserted into the Arduino properly or some pins are gone there, as it is strange you cannot measure Vref across trimpot and USB cover and there is not 5V on Fault pin of the driver.

Would that be even possible (traces burning out) if my motors are working (kinda)?
Well, I'm going to try it (I will also have my chance to do the cable management, since from the bottom it looks really bad :D ), will edit soon. And yeah, I'll also check if all the RAMPS's pins are connected to MEGA's pins.

Edit1: All pins of MEGA are shorted with pins of RAMPS as they should be. No problem here.
Also I found out why there was only 0.01V on the E's DRV while on every other there was 0.018V - I had all 6 jumpers under the DRV connected so the microstepping was 1/32 I guess.
Btw, choosing full step (no shorting under the DRV) is what I should have done, ain't it?

Edit2: I put one probe of the multimeter (multimeter is set to short detection) on the PSU's GND and when I put the other one on GNDs of DRV's or USB's shell, the short was detected and the multimeter started beeping, so I assume everything's all right here.
I'm now thinking of everything that could make this stuff happen, so, it's okay to use the same pins of PSU for both terminals on the RAMPS? Like, I just coupled the same 2 cables (+ and -) of the PSU so I have all the power pins on the RAMPS connected. I'm just wondering if they don't have to be "galvanically separated" (not sure what's it called, but I think you'll understand).
I'm now afraid that motors are not connected well. And 1A might be connected to 1A but 2A might be connected to 1B for example. Something tells me that swapping two (neighbour) cables will only change the motor's direction, but I think it's worth taking a look at, or trying it..

Edit3: Swapping the motor's pins didn't change a thing. The shaft still turns randomly.
I'm wondering, isn't there one pin of the DRV that tells it the direction for turning the motors? The DIR Pin? Maybe I can measure it.
Using two PSU's didn't help either. The first 7 turns were in the same direction but it again went to rolling a dice...

Well motors will try to "ground" through some other pins. That may be the reason why you don't have 5V on Fault pin for eg. It's going to burn your electronics sooner or later ...
Mictostepping pins shall not be an issue. They are either not connected or, giving 5V signal to driver. If jumpers installed on MS0, MS1 and MS2 pin you will get short with 5V pin on RAMPS.

What do you mean that they try to ground trough some other pins? The way I understand it is that if they don't have ground, more current will flow trough other pins because that's where the ground is (although with some more resistance)?
And, so wouldn't it be a good idea to put one DRV instead of the (assuming) A4988 drivers on my geeetech printer to test if the voltage (on FAULT) is different?
Just wanna know if by doing this I won't blow something up.

Also... I'm sceptical now. On the photo, there is 0.378V, right? I fear that it is 3.78V so that's why I'm asking.
https://bit.ly/2z23A4d

Sorry it took so long to reply. My printer at work is under maintenance and I had to throw my spare time there. Funny what one fan in PSU, that is not runnig, can cause.

Your understanding is correct. It is groungind elsewhere. But that results in unpredictable behavior and overload of some other parts, that may not survive it in long therm. The easiest way to correct it is to ground it as it should be done. So rather don't put any good parts in it unless you will find the cause or you may just loose more parts on the way.

I found one spare RAMPS at home and Arduino Mega is on the way from China. I will need it anyway as right now my ToyREP is not working (it was disassembled to give parts to other printer). Than I will be able to show you, how the electronics shall work.

Despite your worries I tried putting one A4988 instead of the DRV (tried Z and X axis) and it didn't help. Exactly the same thing was happening as before.
So I'm starting to fear that arduino or RAMPS is faulty...
I have no problem ordering another one, but I'd like to test which one is faulty. Well, I don't think that the software is wrong, or all 6 motors are the problem. But I might take a look at the software once more, deeply.

Software will not change the way your drivers ar grounded. I can show you on Arduino Uno and CNC sheeld what and where can be measured, to check your hardware. I will try to make a video.

No problem with that, but as I wrote, grounding the driver doesn't seem to help either.
Guess I'll wait for it then.

Hi, I am still waiting for Arduino Mega, so for now there is video with driver installed on Sanguinololu.
https://youtu.be/em-JaAsMmg8

Video: https://youtu.be/zbYMTAb4XsA

Tl;dr
So I tried measuring everything you covered in the video. Everything but this looks good:
Where I should have 5V, I have 4.3-4.5V
I have set the Vref to 0.1-0.15V, it's almost at the end (of the rotation)
You were showing that you shorted all the pins under the DRVs, so how should it be on the toyrep? I know you said it shouldn't be an issue, but I still don't know which should I use :D
Sorry for responding 5 days later, I've got a deadline on two another projects so I had to put some time there.

I think we are getting close.
1, 5V are made by Arduino Mega. Try to measure 5V on Arduino alone (without Ramps on it). If it will be around 5V Arduino is OK and something on your Ramps is shorted. Than try to measure 5V on Arduino with Ramps but without drivers. If it will drop to your 4.3V, check Ramps for some leftover solder, small balls of solder, unsoldered connections, and wash away residual flux with toluen. If Ramps don't cause the drop, than check drivers one by one.
2, If there is less than 4.5V, the ATMEGA2560 is undervolted and may do some unpredictable errors The processor may work with lower voltage, but it also have to have lower clock speed. See page 22: http://ww1.microchip.com/downloads/en/devicedoc/atmel-2549-8-bit-avr-microcontroller-atmega640-1280-1281-2560-2561_summary.pdf Arduino have to be manufactured as such (you cannot change clock speed easily).
3, For microstepping settings of DRV8825 see this picture: http://doku.radds.org/wp-content/uploads/2015/02/MicroStepping_Modes_tbl.png
(The pins are not "shorted". Either the MS0-2 are open=OFF or connected with 5V = ON). For Z asis you don't need microstepping at all so you can do fullstep = 1 or halfstep = 1/2. For other axis (X, Y, E) = 1/8 or 1/16 is sufficient. You have to change the DEFAULT_AXIS_STEPS_PER_UNIT settings in firmware accordingly (also use correct movement speeds in Cura or Pronteface - beware, these are typically in mm/min) otherwise the motor may try to move too quickly and stall.
4, This was not on my video clearly explained - The Dir, Step and Enable pins of the drives are connected do ATMEGA data pins. To set them HIGH voltage between 2.2-5.2V have to be used (in my case 4.2V). To set them LOW less than 0.7V have to be measured on these pins. To make the motors move Enable have to be LOW and Reset have to be HIGH (reason why it is connected straight to 5V). Dir depends on the movement direction and Step gives short pulses that can be checked only by oscilloscope.

No problem with late response, I am quite busy too and the video took me a while either ;)

Yeah, it's often the little and unexpected thing that is responsible for all the problems :D

Well, I'll try to connect PSU's ground to GND of the DRV to see if anything changes. Will edit soon.

Edit: No change. After connecting PSU's GND to DRV's GND by cable, same things were happening.
Well, changing DRVs and trying them on another printer, or simpler - trying A4988 on the toyrep - seems to me like the fastest and the only reliable way to find out if the RAMPS or DRV is bad (or neither of them).

All right, I just wanted to try if testing it with NEMA won't ring a bell in my mind.

  1. Oh, forgot about the voltage and current in the motors. I'll take a look at that and adjust it, since I didn't change anything on the drivers, and used them as they came.
  2. Well, as I try different speeds (Hopefully feed rate is the same as movement speed), the motors don't work. For X and Y I had the feed rate 10 for example, and it moved a little little bit and made high-frequency noise. So I found a sweet spot of 1. But now when I think about it, if I have too low current flowing to the motors, this may change.
  3. I tried different accelerations before to make them work in the direction I want, and I tried like 10 numbers between 20 and 300 and none of them helped, so yeah, I think I'll get back to it when all the motors are running smoothly.

To the bipolar hack. I did the same thing with 6 motors (5 by using drill, one by using knife) - cut the connection in the middle. I then measured every of them and confirmed that the connection was broken. And then desoldered the red (middle) cable. I think that is all that is necessary.
Thank you, I realised that I wasn't measureing two different coils, but the same one... So it's the resistance of the coil that makes my multimeter beep.

Will update soon, thank you for now :)

Hey Thor how have you been?

What are you (or anyone else) using for retraction settings? speed and distance? It seems mine is very slow and I cannot tweak it too much or it won't reverse at all.

Best!

Hi,
With ToyREP I don't retract at all. It have no sense as the speed of extruder is around 4mm/s. If you really have to try very short reverse movement of around 1-1.5mm and no z-hop.

Also may I suggest some "stopper-bolts" for X-axis? I dunno if anybody else got this, but when i printed thos parts - the smooth 8mm rod just slides in&out and in there is a little wobble. Checked dimensions - I'll try to scale those parts and think of corks.

Please remove supports on big pieces, so I can use supports from Cura. This is what happens when those thin wallssupports get ignored

Published. Unfortunatelly it is a flaw of Cura, as it can't print sub nozzle dia. thick walls. It is not exactly Rev1, as I lost some source files, but it should be close to it.
Concerning your second question, all 8mm holes have additional 0.5mm clearance for overextrusion. 8mm precision rod was siting there just fine. Not too tight but no wobble either. If you measure internal dia on plastic part close to 8.5mm than there is something wrong. Maybe just a bit higher extrusion multiplier will help.

0.31 layers at 60 mm/s you need to make this many mm/s from your extruder

3.32349725638639 for 1.75 mm filament
1.12377245359828 for 3mm filament

im glad the 4mm/s is all one would need to make extruders for filament, these could be used on many other printers.

Hi,
I guess you are talking about 0.4mm nozzle (trace width). Although your volumetric calculation is right, there is one slight problem - maximal turning speed of the motor. On the output shaft 28BYJ-48 can do only 15RPM.

The only way you can achieve 60mm/s print speed with 0.4mm nozzle a 0.31mm thick layer on ToyREP extruder is with 3mm filament as you will achieve desired volume at 14RPM (1.75mm filament will need almost 40RPM!). For the extruder is 1.18mm of filament per second the speed limit.

https://www.thingiverse.com/thing:1597741 has one pushing 3mm filament at 10 mm/s probably not torque enough but maybe the big gear on this one could be reduced so it can do 4mm/s to make it spin a bit faster and still have enough oomph to make awesome remote extruders.

28BYJ-48 Extruder

Hi!

nice project

thank you

Hello Thorgal,

I could recommend that you check out the Klipper firmware. It has done wonders for my little 28byj-48 steppers both in performance and noise. It would be exciting to see what you can do with it on the toyrep

/Jonas

Thank you for this advice. Unfortunatelly I don't have spare RasPi and as keeping this project as cheap as possible, it will have to manage from computer/notebook.

But I will try it on some other machine I have. The command rate sounds very interesting.

I have a three 28-byj48 motors and all has turnable backlash on shaft. How printer it compensate?

Hi,
Sorry to say, printer cannot compensate backslash of any type of the motor alone. You can use encoders to detect if you moved to the correct position and if not, firmware can compensate. By doing this you will make stepper motor behave like servo. I've seen this on experimental DC driven ToyREP with encoder strips from inkjet printer. But this is not very common solution and support in firmware is still minimal (or none at all). Least to say, you will need additional hardware. In the past I was thinking about hacking digital calliper to do the same, but you can read it only few times per second while printer have to change position much more often.

If you want to avoid this, I would recommend using NEMA17 upgrade in X and Y axis. In Z axis and extruder you can keep 28BYJ-48.

Hello. I have a problem with Z direction. As you know it has two motors. And my problem is that one sometimes goes to the different direction (so basically one is screwing and another one unscrewing). It is like random, sometimes it is ok, sometimes this one goes to the different direction. Could you help me, please? Wiring is like in building tutorial.

hi,
it's most probably low current for z axis or one motor have a bad connection in header pin. regards
martin

Unfortunately, it isn't so easy, the problem is somewhere in variables in the code, something with speed or acceleration. Could you help me, please?

In that case try to check, if you don't have commas instead of dots in numbers, or if the acceleration is not too high.
But I'm still convinced it is one of the reasons above. Too low voltage can result in unpredictable movement when motor is under load. Bad contact can also make the motor go crazy. It do not have to be in motor pins abut somewhere in wiring on in the motor it self.

hi thorgal,
I was trying to add Belt tensioners in the x and y axis but is getting quite difficult to edit the .stl files in thinkercad. So i wanted to ask you if you could shared the x2 and y2 cad files in order yo make the improvements . I promise that i Will share the files if ir works
Best, Franco
PD i have tried to make a bowden direct extruder but i had no success. The motors have just not
enough torque ( it skipped steps)

I've seen dual 28BYJ-48 direct drive just recently https://www.thingiverse.com/thing:1597741 , but I'm convinced that even this will not work in long therm. The problem is in the side pressure to the motor shaft. It will require some bearings to support the hobbed drive gear as was in the original delta extruder https://www.thingiverse.com/thing:200078 .

Send me a PM with your mail and I will send you DWG file with required parts. But again be aware that any belt tensioning will put side pressure to the motor shaft which will result in motor jamming, so it is probably not good way to go unless you will add also a bearing on top of the motor pulley.

28BYJ-48 Extruder
Delta extruder with 28BYJ-48 stepper

Hello, I've got a huge problem after changing belt tension on y :-( I removed the elecronics, changed the tension on the carriage and I tightened all bolts.

now, if I move manually y carriage the belt goes out from the motor pulley! Do you have any idea on how to fix it?

Maybe the belt is too tight and I'm getting a tilt on the bearing shaft as result?

Than you in advance!

Yes, that is probably too tight for the motor shaft. Also try to get the belt into one layer.
And most of all, put the belt on after the nuts are tightened otherwise the front and rear legs won't be perpendicular to the table. Make the frame square, tight and then put the belt on.

Thank you for the reply, but I still haven't fixed it. Maybe I'm moron a little, but I've no clue on how to fix this. I've started again from y and smooth rod installation, cheched everything, tightned all bolts and so on... If i move y carriage manually, the belt goes out, I printed a 2 layers square and the printer worked. I'm gonna check it with a longer print tomorrow.
I also noticed that the belt ins't perfeclty a stright line, I think it needs to be higher on the motor and lower on the bearing... I'm so confused!
Thank you for any idea that you can give me

EDIT: I've just discovered this, maybe I can try it https://www.thingiverse.com/thing:1032224/#files

2.5 ToyRep pulley for Y axis
by sobo84

Try not to push the Pulley on motor so deep. Also Sobo84's pulley may help as it have bigger hem on the top, that will keep the belt on it.

If the bearing on the other side is too high, you can grind off the plastic part a bit, so it will sit a bit lower.

Also check if the belt is not pushed too deep in the Y Frog. Try to make its side even with the bottom side of the carriage.

Hello. I am making the ToyRep according to reprap site. It says I need a ToyREP-PCB1.stl but there are two files ToyREP-PCB1-1.stl and ToyREP-PCB1-2.stl. Same with ToyREP-PCB2.stl . Could you help me please, which one to choose?

it is in the table with printed parts on page: http://reprap.org/wiki/ToyREP_Buyers_Guide

You need both of them. Either ToyREP-PCB1-1 and ToyREP-PCB1-2 for ToyREP-PSU1 PSU cover or ToyREP-PCB2-1 and ToyREP-PCB2-2 for ToyREP-PSU2.

Reason is that PSUs are typically connected by three screws on side and the screw holders are also PCB holders for Arduino and RAMPS. So one of these L shaped parts have one hole and the other one two holes for PSU screws. Smaller and bigger PSUs have different positions of these screws.

If you won't install PSU into the printer, you can pick any of the PCB holders and print it twice.

Is there a version to build it not on the power supply? Because on RepRap site they are using power supply in the bottom.

PSU don't have to be part of printers construction. Sturdiness relies entirely on the threaded rods. You will still need some printed parts for PSU, as they are shared with Arduino and RAMPS, but not the PSU cover.

So it doesn't hold all of the construction?

Yes, on the fourth picture from left, you can see the first prototype even without PSU embeded.

Is there a complete list of all of the parts I would need to buy anywhere? I kind of want to build one but don't know where to start.

Yes, BoM is uploaded in thingfiles and also there is page on reprap.org with all possible information: http://reprap.org/wiki/ToyREP_Buyers_Guide

https://photos.app.goo.gl/BSTpuexGkjoEixYg1

hi i made toyrep.

i use sd slot and made a circle, it was made oddly.

it's not a perfect circle.

Is this because of backlash?

can u help me?

Comments deleted.

Hi,
Thank you for feedback. As I stated after printing Marvin, ToyREP is not good at printing round things, and I am still not sure why as you proved one of my theories with boudrate wrong (although I've experinced similar effect on other printers with NEMAs too, and SD card made circles more regular on them for sure).
Now It can be:
1, backslash in internal gearbox (that can be changed only with more precise motors)
2, backslash on belt (make them a bit tighter)
3, irregular shape of pulley (would be hard to get a better one without different 3D printer)
4, too high microstepping (try full step in X and Y axis to see a difference)
5, wrong decay on stepper driver (common problem on DRV8825 with 1/32 microstepping but hard to trace without oscilscope)
6, something wrong in Marlin firmware (hard to say, as almost all other firmwares have the same base except for Teacup which is quite hard to use for a newcomer)
Can you please let it print a bit higher to find out if this is regular or random (and also if you would not get layershift due to low current) and post me the result?
Thank you once again and I am sorry I can't help you right away.
Best regards
Martin

Hi all, I've noticed almost the same problem, I can't get good circles with small diameters, but I get good circles with 3cm diameter or more. For example, a 5cm circle is perfect! I set up 1/16 step for x and y, If I shift to fullstep, should I calculate stepp/mm again?

Understand. Small circles usually have very sort traces. Either the gear box backslash or firmware is not able to react to it.
For full step set MS1, 2 and 3 pin on electronics to open and recalculate firmware. You will get around 41 microsteps per mm in X and Y axis. Than upload it again.

Wow very nice I will try it

Hello, I have been very interested in the design and how it works.

What do you think an estimated cost of all of the parts (excluding the printable ones) would be?

Not sure about the prices today, but there is OpenOffice sheet with prices among thing files. Without the plastics parts it was around 76EUR two years ago.

I really LOVE this design!

I ordered everything today and hope to have this printer soon too.
I was luckly and got up to 11% on every part i ordered at Banggood:

https://goo.gl/9W9CmA

Thank you so much!

Comments deleted.

I need some advice: My motors for the X and Y Axis are only turning one direction and in the other direction there's just a lot of noise but no movement. But its not a problem of the motors, because if I connect them to the Z axis they work fine - so they work with the settings for Z axis for some weird reason. Anyone had this problem before?

I've had this before with other 3D printers. I ended up just having to power everything off completely for about 30 seconds and restart it. You probably tried that already but, just in case...

What suggested JCriotz may solve the problem. If it will persist, there may be some problem on the X and Y motor connector or stepper driver on your electronics. Try to pull the stepper drivers out and and push them back (maybe also swap the drivers).

Also check by hand if there is the same friction when moving the carriages there and back.

Hi Thorgal,

Decided to design a z-probe for my toyrep. Seems to be working OK (just sorting out the offsets, but the Z-probe it's self appears to be working well). It's up on thingiverse here: https://www.thingiverse.com/thing:2405760

Thanks for the ToyREP design, It's been fun to build.

Martyn

ToyREP Z-probe

Good job Martyn, Thank You.
If possible I would rather turn it to the other side of the extruder, as on this side, it may hit the motor or cut off a bit of the print area. On the other side where are endstops, there should be more space.
I will add it among improvements soon.
Regards
Martin :D

It's designed to be thin enough to clear the X axis stepper motor. Unfortunately the extruder stepper is in the way on the other side (the arm needs to be at least this length to reach below the hotend).

One thing I have noticed whilst trying to calibrate this is that there is quite a bit of slop in the shaft of the steppers I have. I guess the addition of the bearings at the top of the Z axis threaded rods helps with this at the risk of adding artefacts in the print if the rods aren't perfectly straight.

Fair enough.

Well originally the bearings on top were supposed to hold threaded rods in the same height, while motors will just turn them, but it shows a bit of wobbling.

Gluing nuts into plastic parts helps with horizontal alignment too.

Hello Thorgal,

I am making a printer (HangAround) using the 28BYJ Steppers in combination with a RAMPS board but it cannot be run above 100mm/min before losing steps. This seems far below what is possible.

The odd thing is that it seems possible to run at some higher speeds. 1000mm/min works fine while others (200 mm/min) does not. Any hints what could cause this?

The stepper drivers are tuned down to a level where the steppers run hot to the touch but I can still keep my finger on them so there should be enough current available for speed. Maybe I should try microstepping? Right now they are running at full steps

I use the forum for HangPrinter for my design. It is a cheap, low budget version of that design. https://www.facebook.com/groups/hangprinter/

Any tips you can give would be appreciated. Thanks
/Jonas

Hi Jonas,
There are four things that come to my mind.
1, Hangprinter designs with Nema17 motors are useing quite hugh reduction gears to gain power and resolution (steps/mm). How big are your pulleys and what reduction gear you use?
2, If you rely just on the internal gearbox froget 1:16 version of 28BYJ-48 and 12V rated motors. The only version that can be used with bipolar hack in printers is 1:64 5V version. And even than, you realy need to get them runing because output power = speed * torque. That explains why you can run 1000mm/min, but not 100mm/min. You simply do not have enough power at low speeds.
3, The torque of 1:64 version is good enough to move slowly light weight carriage with linear bearings on precision rods by slightly tensioned timing belts. You are on the other hand moving the whole printer on outputshafts of three motors, that do not have even bearings inside. You are bending the shafts to the side, possibly damaging the internal gearbox as the teeth are jamming. Is there any way how to catch the radial load comming from pulleys with some bearing and use the 28BYJ-48 only to turn the pulleys?
4, Possibly there is some problem in the firmware that is causing this strange behaviour or you are simply lacking steps/mm so what you expereince is printer jumping from one step to another. But microstepping will help you only to some extent.

I would definetly recomend to add 6 bearings (2 form each side of every pulley+use some screws as pulley shaft and join them with motor shafts by some flexible coupling. Also use geared extruder instead of direct drive. Feel free to remix mine. Again for the same reasons. Shaft of the motor is not under side stress and also the motor is turnig four times more compared to direct drive so it is four times more powerfull when extruding.

Hope that will help
Rergards
Martin

I've managed to reach 1000 mm/min now. The voltage over my steppers had to be 0.15V for it to work which is a bit higher than what you recommend. One of them had only 0.12V and lost steps when the speed increased above 400mm/min.
I also had some issues with the software as seen in the other posting. I can run the steppers much faster with the g6 command rather than through the G1 command.

There is an interesting effect where it seems that the Arduino may not be able to keep up with quick feedrates.
I lowered the SEGMENTS_PER_SECOND and it gave me higher speeds without losing steps so there is something there as well. The internal calculations which are a bit different for a Hangprinter seems to throw the steppers off somehow or at least make them more sensitive.

I will redesign for minimal radial load on the spools. The loads on the steppers are small in general because the printer hub is very light. Still, I will do what I can to ease the load.

Many thanks Martin

Well 400 or 1000mm/min do not say much without more information (at least about the pulley diameter and gear ratio if any or steps per milimeter). How many turns per minute of output shaft is it? If the motor is doing 15RPM at 400mm/min, than you are already hitting the limit there.

Vref 0.15V is again not enough to judge. What drivers are you using (A4988, DRV8825, or other), what are the sensing resistors on it? Please follow webpage: http://reprap.org/wiki/Pololu_stepper_driver_board rather than a recommendation for some specific situation.

Firmwares on Arduino boards are usually highly optimised as 8bit hardware lacks computing power and for that reason prone to error under specific situation. From what I have tested, Mega 2560 with DRV8825 can do hardly more than 9000 microsteps per second and axis reliably. You can get a bit more from A4988, not by much though. But G6 and SEGMENTS_PER_SECOND have no firmware I am familiar with. That is probably some speciality of Hangprinter so it is hard to say if these limits are working there as well.

If (judging by the name) SEGMENTS_PER_SECOND will give you lower resolution hence less calculations per second that your problem is definitely in computing power.

You are welcome, but actually I am not very helpful as this type of printer is strange to me.

Sorry about that.
1000 mm/min with a spool diameter of 24mm equals to approx 13 rpm on the stepper.
The drivers are A4988 on a Ramps board (Arduino Mega2560) and the firmware is Marlin with a Delta configuration which has been custom modified. I run with full steps. No microstepping.
The spec mensions that most A4988 knockoffs have 0.1mOhm for the Vref calibration making my set current to approx 190mA. Unfortunately I cannot measure the resistor but it seems sensible. 0.19A is above your recommended max, but I will try to bring it down as I go along.

You are correct that the parameter will change resolution and is directly coupled to the computing power. It would be interesting to find an alternative to the RAMPS which runs faster. Due to the nature of the hanging printer I need two extruders. One is used for the z-axis. This limits the amount of available alternative boards unfortunately,

No need to be sorry. But I have probably bad news for you. In attempt to speed TouREP up there were some unsuccessful designs with 30 teeth T2.5 pulleys. That is almost equal to your 24mm diameter string spool. Seems to me that while 50 mm circumference pulleys are working just fine, 75mm circumference pulleys already produces so much momentum that the torque of these tiny motors is not enough. It may get better with the bearings we already discussed, but be prepared to change the design to use slightly smaller pulleys with circumference of about 50-60mm instead of 75.4mm you have there now.

You can try to use Arduino DUE that have 32bit ARM processor instead of 8bit ATMega architecture. Some versions of Marlin are tweaked so they can be compiled to ARM in Arduino IDE. From what I know Arduino DUE can be used with RAMPS boards as they have similar pins as long as you are not using LCD, thermocouples, optical enstops and motor drivers that needs 5V (DUE have only 3.3V instead of 5V on MEGA). In such case you would have to use DRV8825 drivers as they prepare their own 5V from motor power supply. Or you can go with RADDS or RAMPS-FD, as they are prepared to work with DUE with no such issues. But please take that information from me with a grain of salt and do some research on your own (as I never used DUE and I may got it all wrong).

Edit: Check this forum for what have to be changed on RAMPS to make it run with DUE: http://forums.reprap.org/read.php?219,479626,page=1

Comments deleted.

Hi
First to thank you for the nice presentation of your project

One Question ( i havent build it yet) how do you handled the backlash of these motors ? (28BYJ-48)
I have lot of them and all of them they seem to have a backlash that is no problem for other applications but for a 3d printing i think will be when the have to change direction every time..

did you used any special tool to add this in count when you slices your object and make the Gcode ?

thanks in advance

I do not handle the backslash. It is there. The only way how to deal with it is to use motors that do not backslash too much, keep reasonable tension of the belt and rely on the resistance of linear bearings on precision rods that will not let the carriages move freely but only when you turn the motor, but still are slippery enough so these motors will suffice to get them moving. As they can move in X and Y axis only at low speeds, the backslash is not exaggerated by inertia, too.

So it will probably not show up. Rectangular shapes are just fine, there may be a problem with iregular circles but it is still not clear if it is due to backslash or some other issue. No special gcode generator is needed. I use Slic3r. You may try to use Repetier firmware with backslash compensation, but i have no personal experience with it.

Still, this printer is not made for high precision prints for sure, but it will do the job. You can see some videos on youtube. There are not only my examples so you can be asured I did not made it up or manipulate the results. ;)
Regards
Martin

Hi Thorgal, I've just deigned a little gift for my ToyRep! :-P Thank you again for this amazing project!
http://www.thingiverse.com/thing:2294654

RAMPS "LCD & SD-Card reader" M8 rod mount
by 3D7

Hey
im buying all the components but i cant find any t2.5 Belt in my country. Do you think it would work with a gt2 Belt?
Thanks again

I think you can... You can print "ToyREP-GT2_Pulley" instead of "ToyREP-T2.5_Pulley"

Hi, great job!
i wanted to know if i can build this printer with an olp PC power supply instead of the PSU used here.
does it affects the printer stucture not having a PSU?
thanks!

Hi,
Sure. you can build ToyREP without PSU. It have no structural purpose, but to make the printer easily transportable and compact.
Regards
Martin

Having designed 3 of my 3D printers, 2 of which use 5 volt 28BYJ-48 steppers , there is one thing that does not add up in the design of ToyRep that has bothered me since I first read about ToyRep, and that is the Bipolar hack.

If you use the 28BYJ-48 stepper motors as unipolar stepper motors they produce more torque than if modified for bipolar operation. Thus I ask why does ToyRep use drivers designed for Bipolar steppers, and why do the bipolar hack?

Both of the printers of mine that use the 28BYJ-48 steppers use them as unipolar steppers. When you write the firmware for the MCU of choice control the the printer (interpret the G-Code, read an SD-card, com with computer, and drive the printer) it is easy to assign the needed pins to drive unipolar steppers with ULN2803's or ULN2003's, as the current drivers. As such it makes no sense to use bipolar drivers when using the 28BYJ-48 stepper, it is much better to use ULN2803/ULN2003 drivers with an MCU that has at least the 16-GPIO's left over to run the 4 sets of steppers (X, Y, Z, E), so at least 20 GPIO's without a heated bed, and at least 23 GPIO's with a heated bed.

Also why use a Arduino based controller? Is it just to increase the price? Just an ATMEGA + Crystal will do the job for under $2 for the MCU, the ULN2803 drivers cost about $1 each and each will drive 2 sets of steppers, thus needing only 3 of them total (one for X,Y, one for Z,E, and one for PWM to control the hotend temp and to power the fans), then some simple passives to do the rest of the interfacing, add a simple board to wire everything up and it costs a total of about $7 including SD-Card socket to make a good working controller board. Even if you chose a Parallax P8X32A Propeller MCU it is still lower cost than an Arduino.

Hi Thorgal, almost done everything... But... maybe there's something wrong with motors! :-(

It's the first time i use Pronterface, have you some some links to share so I can learn something more about printer calibration?

The only motor I can move is the extruder motor, x motor can moves just 10mm L to R, and 10mm R to L, the same for Y motor except for sometime when it can go from the eletronic side to the endstop side without clicking it. Z movements are completery weird, they move only from top to bottom. and endstops seem to don't work. Motors are in general very noisy but they don't get hot. I connected all endstops (NC) to gnd and signal.

vfref are 0.05v for X,Y and E
vref is 0,1v for Z

here's my configuration.h --> https://pastebin.com/FNHRDhS1

here's my esteps:
JUMPER per E STEP
1/16 x = no no yes
1/16 y = no no yes
1/4 z = no yes no
1/16 extruder = no no yes

any clue?! :-(

Another question: please look at this picture, http://imgur.com/a/sEG8F , are the direction right for x, y and z? If moving +X on pronterface and I get an x movement from R to L, what I should do? Invert motor connector on ramps or change the firmware, is it right?

I didn't understand also how to set the M3 threaded rod for the Z endstop... But at the moment it isn't so important.

Hi,

I've walked through your configuration.h file.
On line define DEFAULT_TRAVEL_ACCELERATION 3000 should be rather 300 instead of 3000.

Also if you really have 1/4 microstepping in Z axis the values in line define DEFAULT_AXIS_STEPS_PER_UNIT { 655.36, 655.36, 40960 , 6953.59 } should be { 655.36, 655.36, 10240, 6953.59 }

You can also use define MOTHERBOARD BOARD_RAMPS_14_EFB - there is no need to go with BOARD_RAMPS_13_EFB unless you have this older version of the RAMPS board.

More next time.

Thank you so much! I have been pretty fool on ramps 13 and the acceleration value! Now I understood how to calculate default_axis_steps_per_unit. At the moment I'm able to move motors, but I'm quite confused about homing because I had to invert motor's connections to get it ... I'll try to explain myself better: If in Pronterface I click Z home I get the right movement to z-min endstop, If I click X home I get a movement to the x-max endstop and if I click Y home I get a movement to the y-max endstop... But, concerning Y axis, by clicking Y+ I obviously get a -y movement! Should I invert the y motor connection? Or is this right? I don't understand If I should look to xyz cartesian reference or I should look as home directions the endstop positions, I also don't get how to measure real coordinates of hot end and print area :-( . I'm really sorry for these stupid questions, but this is my very first experience. I hope you can forgive me Lord Thorgal! ;-)

First, forget about lord thorgal :D

Second, as I already told you your questions are perfectly ok. They are going to the right spot.

Third, Cartesian XYZ on 3D printers is done the typical way, X+ is going to the right, Y+ is going from you and Z+ is going up. When looking at ToyREP from the side of Y-motor (having ramps in the back) than both X and Y axis are homed in max in distant right corner. Than when homed and size limits are set, the only way printer will allow you to go is X- and Y- until you hit 0,0 coordinate in the closest left corner. Than the motors will stop and will let you go only X + and Y+ until you almost hit endstops again.

If the motors are working but in wrong direction, there are two ways to change to the direction. One mechanical as you suggested is to turn the connector and second software to change the line in configuration.h => define INVERT_Y_DIR to false or true (false is recommended, if the connector is installed the same way as on other motors).

Thank you again! I think I made it right, I've also noticed wrong speed values on Pronterface. But I still have some questions:

1- I really don't get these values about home_pos:

define X_MAX_POS 108

define X_MIN_POS 8

define Y_MAX_POS 101

define Y_MIN_POS 1

define Z_MAX_POS 100

define Z_MIN_POS 0

How do I measure my values?

2 - I understood how to calculate DEFAULT_AXIS_STEPS_PER_UNIT, but I'd like to know how to do a good measurement, on z axis above all (because I can move only by 10mm). I've just took a look on X axis and I think I'm gonna have around 90mm, it might be possible? It seems to be very far from 100mm.

When I'll get my first print, tell me your paypal account, I really would like to offer you a big beer at least! :-P

Hi Thorgal, concerning point no.1, maybe I got it! It took a lot because sometime Pronterface didn't respond to commands so I had to disconnect/connect it quite often... Despite this, I've found these values: x=108 and y=108. Now by clicking the red button on Pronterface controller I can see the nozzle moving on the center of the bed, and then moving where it's supposed my origin of the print area, almost like this :-P https://ibb.co/cOYMVa . Am I right?

Concerning point no.2
I've covered the bed with a paper tape, I drawn starting and ending points on it, so I measured the distance. I think it worked on x axis... I keep going! :-)

Last stuff: I noticed tha when I want to home X and Y I get a +z movement before X or Y gets homed. Is it the "safe homing" ? Do you reccomend to keep it turned on?

Hi,
Point 1, is correct. Print area is between the heads of the screws so you have nice flat surface. You can extend it slightly if you want more space but than at the corners be aware the countersunk screw heads have to be a bit under the surface so you won't damage the nozzle and printed filament is not going to stick to them. Than you may have almost 116x116mm big print area.

2, Yes you will have to recalculate (calibrate) the steps per mm to get correct values for your machine. The real size of printed pulleys, belt tension and real gear ratio of your motors is influencing the values a lot. Initial numbers are there only to get thing moving.

For Z axis it is best to print some tall calibration cube of known height, measure it and recalculate the steps per mm accordingly. You can also follow my reply to Eddiiie: http://www.thingiverse.com/thing:945156/#comment-1191875 where is the calibration of movement explained in detail (you will have to copy the link as it won't open correctly)

I am not aware of safe homing behaviour as I am mostly using Marlin 1.0.0 and 1.02 where it is not implemented yet. I think you can keep it if works correctly for you.

ToyREP 3D Printer
by thorgal

Hi Thorgal, tonight I started to print a 10mm cube to calibrate Z steps and I think I'm not so distant to the right value, but I've worse problems : -P

First of all, the printer is not printing on the bed center but on the far (from y endstop) right corner, and I don't know why! :-(
On Slicer & Pronterface the cube is sets in the center of the bed,
here's some pictures: https://goo.gl/photos/oTqvVmBPbw8NN8xH8
here's my actual config.h: https://pastebin.com/JchPG8m2

Second: I surely should set better slicing settings and I've to set the proper speeds, but as you can notice, I think there's something wrong with x axis because it causes a layer shifting. I don't think I've to tight belts even more, x belt is tighed almost like y's... Any suggestions?

comment
// #define THERMAL_PROTECTION_BED

Change lines as follows, because once you homed the printer there is no need to have offset in X and Y MIN or run back to the homing position at 108,108. Your print area is from 0,0 to 100,100 only homing position having offset out of the print area at 108,108.
define X_MIN_POS 0
define Y_MIN_POS 0
define Z_MIN_POS 0
define X_MAX_POS 100
define Y_MAX_POS 100
define Z_MAX_POS 100

define MANUAL_X_HOME_POS 108
define MANUAL_Y_HOME_POS 108
define MANUAL_Z_HOME_POS 0

Homing feed rate should be calculated a bit differently although it will give you the same numbers. HOMING_FEEDRATE_XY is rather multiplication of 12.5mm/s by 60s ... get use to this, as once you will find maximal feedrates they will be in mm/s. Same in Z 0.2mm/s *60s

define HOMING_FEEDRATE_XY (12.560) //=750mm/min //(5060)
define HOMING_FEEDRATE_Z (0.2*60)

Now layer shift is probably caused by the fact, your machine is trying to run quicker then is capable or current is too low. Check the Vref on driver. If it is set correctly than you will have to find maximal feed rate as per this page: http://reprap.org/wiki/ToyREP_User_Manual#Speeds_and_units
the DEFAULT_MAX_FEEDRATE { 50, 50, 1, 5 } is there as initial setup, only to let you find your maximal values. You will end up with something like:
define DEFAULT_MAX_FEEDRATE {30, 30, 0.35, 4.17}

Hope it will help to have better results.
Martin

Thank you Martin! Finally I got it! Your advices about HOME POSITION were right to get the the print on the bed center!

After I've made all improvements on my config.h I didn't get any results concerning "layers shifting"... I had the doubt that the problem might have been on gcode generted from Slic3r, so I decided to make a gcode from Cura and I got my cube!

Cube measurements are pretty weird, I need to do a fine tuning on the printer, but I didn't have the shifting problem!

Thank you so much my friend! I'll add my remake on my profile very soon!!!!

Hi Thorgal, I'm very happy of my first results, but I've found something strange while I'm trying to find proper DEFAULT_MAX_FEEDRATE. I can set x/y speeds from 2000 to 10.000 on printrun without see any difference on motor's speed. Why?

Another question is about retraction's settings on Cura (or other slicer). When I turn retraction on I get a layer shifting on Y axis and extrusion problems with small prints! It's weird! :-( Y drive it's set at 0.05v, should I give to it a little more?
I tryed to set retraction speed at 10mm/s and distance from 0.5mm to 4.5mm without any significant improvements. Any suggestion?

Speeds in Printrun are in mm/min. I presume you have already set DEFAULT_MAX_FEEDRATE {30, 30, 0.35, 4.17} in firmware, right? That will allow you to reach speeds around 30mm/s*60 s = 1800mm/min in X and Y axis. Anything above this is reduced by firmware for safety reasons no matter what you set in Printrun. If you want to test higher speeds you have to change DEFAULT_MAX_FEEDRATE to something like {50, 50, 1, 5}. That will let you do up to 3000mm/min in X and Y, 60mm/min in Z and 300mm/min with extruder, but ...

Concerning Cura. Well if maximal feedrate for extruder is set to 4.17mm/s in firmware, it will hardly do 10mm/s retraction in any slicing program. Even if 4.17mm/s is too much, and your machine can do only let say 3.5mm/s, you are asking for trouble.

Pumping Amps to motor will help only if the motor can run that quick. If not, it will just heat more and eventually burn while chopping. Vref depends on the driver you are using and sense resistor, that is near the trimmer. Check this web page for correct settings: http://reprap.org/wiki/Pololu_stepper_driver_board I think you can safely use 0.09A to 0.15A per one motor (0.18-0.3A in Z axis). That is 0.07-0.12V with common A4988 or 0.045V-0.075V with DRV8522 (you will have to dig up the datasheet for this one).

Actually retraction is not causing the layer shift. It is the movement speed of X and Y axis or movement acceleration. Remember that these motors in ToyREP can do safely only 12.5mm/s (doesn't matter if it is print speed or movement speed) and everything above this value is a bonus. Try to set speeds in slicer to this and than go up in small steps (like 2.5mm/s) while printing the cube or what ever test model. Than set your final numbers in firmware, so you will not accidentally try to push the machine above the limit with some wrong settings in slicer. If that will not help problem is probably in too high acceleration and/or jerk.

Lord Thorgal,
I would like to suggest a small change to the build BOM that will reduce the friction on the stepper motors.
From my little experience, not all 8mm smooth rods are created equal!! There was much resistance on the LM8UU bearings, eventually it wears in, but after many hours of printing, the bearings started to cut a groove into the smooth rods. This cutting into the rods started to cause one of the axis to jitter and eventually stripped the gear. Anyway, during this time I started building another printer, the microcube (thing #1938877) and it called for 8mm tube smooth rod, not solid. For whatever reason, the LM8UU bearings love this tube and slide up and down with no resistance at all. Here is a link to a seller I used for the 8mm smooth tube - http://www.ebay.com/itm/271607850532
I have replaced the 8mm smooth rod with these tubes (also come in good size, no cutting necessary..) and am very happy with how smooth the bed and carriage moves. This will no doubt make life easier ont he steppers. It took about 10 motors before everything was worked in... These tubes could have saved so much work, in my situation. All I had to do to change to them was loosen the nuts holding them in place and I tapped them in with a hammer and that also pushed the old one's out.

If you are ok with it, I'd like to add this to the Wiki in some way, but want your blessing, of course..
:)

Sure! Anything that works sounds good to me.

Small comment from my side. In BoM, I have an error anyway. instead of "smooth rods", there should be "precision rods", so I will have to correct it on a occasion anyway. Precision rods I use have much harder surface, I have them installed in other machines too and they are not wearing off as you described.

Thank you for pointing this out.

hey guy, i wanted to try this, but I just can't get my hacked steppers to work propperly PLS HELP. when i click to move 10mm it sometimes does, sometimes it just jumps a little, but almost everytime the first time after poweing up it moves as i wish. I have checked with multimeter that i did hacking right, but although i have removed a gearbox from one motor to see if it wasn t just stuck it works just as it was with gearbox. BTW I have tried controling it with mega+ramps, then UNO+cnc shield and teacup (does anybody know why the endstops aren¨t ever triggered although i have tried every combination of connecting and pressing them?)

If I got you right, you have assembled the printer, endstops are installed on the machine, you have uploaded firmware modified for ToyREP (with altered accelerations and movement speeds), measured the Vref voltage on drivers and you are still not able to move the printer?

Now, when you measure the endstops, are they normally open or normally closed? And do they change their state when pressed?

Marlin on default is set to use NC - normally closed switches. If yours are normally open, you will either have to resolder the wires to so the give normally closed signal or change in configuratuon.h the lines referring to ENDSTOP_INVERTING to true.
define X_MIN_ENDSTOP_INVERTING true
define Y_MIN_ENDSTOP_INVERTING true
define Z_MIN_ENDSTOP_INVERTING true
define X_MAX_ENDSTOP_INVERTING true
define Y_MAX_ENDSTOP_INVERTING true
define Z_MAX_ENDSTOP_INVERTING true

Than you should be able to home each axis and after homing move in the axis freely up to the point where you will hit maximal or minimal distance set for each axis.

Hello! I've solved all the problems I had, but there is one that appeared recently : when I want to put a filament in the extruder, it doesn't go trhough : the problem is it seems that my hobbed bolt is covering the bottom hole leading to the coldend/hotend. Maybe the hobbed part is'nt deep enough ?

I don't know if you see what I mean about the problem ? Or if you have any solution about it ?

Thank you very much !

Hi,
I understand. It is done on purpose so the hobbed bolt will stay on its position held just by the filament - no need to secure it by nut on the other end. Deeper groove is better for this concept (like hyena hobbed bolt have). With shallow black bolts from China the simplest solution is to open the idler, help to guide the filament down to the hotend, than close idler and fix it by compression springs.
Hope it will help.

Ah ok I see, I wasn't used to this with another printer.. Thank you !

Now the thing is the extruder, when I hit it the button to extrude 100mm will randomly stop and won't answer anymore until I restart the connection... I suppose there is a problem in the firmware like speed or I don't know ?

Thank tou again, and sorry for all those questions :/

Hi,
Don't worry. Your question are appreciated as with such good feedback I will learn a lot too.
Now this is a bit more complicated. What speed you set in software you use to control the printer and what are the values you set in firmware?
My working values are around 4mm/s (240mm/min) for maximal extrusion speed and 17mm/s^2 for acceleration and retraction. With 1/16 microstepping I have 7026.994 microsteps/mm on extruder. You will probably have slightly different number as my hobbed bolt diameter is around 6mm. With shallower groove it may closer to 6000 microsteps/mm.

Also check my conversation with Eddiie: http://www.thingiverse.com/thing:945156/#comment-1191875, where are some more things explained. (you will have to copy paste the link)

ToyREP 3D Printer
by thorgal

Yes I figured out it was related with those parameters, I set the speed a bit lower while I will do all the parameters correct... Thank you very much for your help !

That's not so common for a maker do answer that much to the comments, really appreciated though ! :)

I saw on the manual about the acceleration settings, and I have to say I don't know if I should touch it or leave it this way... ? Thanks :)

If you are asking about DEFAULT_MAX_ACCELERATION {300, 300, 1, 17} than this is the basic setup. You can rise these numbers a bit while the motors are still not chopping.

Thank you ! I finally got it running !!! I wanted to print this : https://www.youtube.com/watch?v=1FYp1-1CTlg
And in the end I got this : http://img4.hostingpics.net/pics/56141917506450102030905130995641780374116n1.jpg

I solved the mirror problems and the overextrusion problems, but I don't know what to do about the circle thing which isn't a circle ? Thank you !

Congratulations! Link to video is probably wrong. Never mind, the picture is pretty descriptive.

I suppose You have the printer connected directly to computer and sending gcode through COM port emulated on USB, right?

I had problem with circles when printing Marvin. Back than I thought it is gearbox backslash or some other mechanical problem. Than I experienced the same problem on my other machines too. Find out the problem is caused by communication through COM port. When circle is made of too many short lines the buffer in printer electronics will run empty before new commands can be uploaded.

There are two possible solutions:
1, Export the model to STL with longer lines (circles divided to bigger angles). You will loose resolution hence circles will seem like polygons, but should keep regular shape. Try to reduce the print speed when doing small holes too.
2, Print from SD card as the communication between SD card and electronics is either quicker than through COM port or the electronics can handle the buffer better this way.

Actually I tightened the belts and it got very much better! But the circle is still not perfect thoug :/ As I can't modify the .STL I think I will invest in a screen with an SDcard then?

Or maybe changing the baudrate will do something?

Can you try some other file or prepape your self a simple cylinder to test if it will help?

Sorry to say, boudrate won't help. Screen with SD slot is going to be better solution.

Ah damn :/ I guess I will invest in some screen then ^^ Maybe before, I should tighten even more the belts?

Thank you very much for your help!

Will you make version with nema 17 support?

Sorry, I've decided not to. It would confuse a lot of people.

Hi Thorgal, I've some questions about "Y Axis Completion" step:

1) The frog has a friction while is moving on smooth rods, I've to push the frog to feel less fricion... Is it ok? Or I've got bad lm88u bearings? Or I've to grease them?

2)I didn't understand how to get 113mm and 47mm from Z to legs. Should I unscrew Y axis, so move z axis and then thight everything keeping in mind to get 197mm between the legs as mentioned on "Y Axis" step?

Thank you in advance, and thank you for sharing this amazing project!

Hi,
1, LM8UU will always have some reasonable friction. You will have to grease them for sure. Yet if it is still too different from four bearings on one rod, than something is probably misaligned. Check if the precision rods are going straight and the frog is sitting well on the bearings. But don't worry too much about it. The motors are strong enough to overcome the friction.
2, I suppose you are referring to these measurements:
197mm ... http://reprap.org/wiki/File:ToyREP-27-Distance_Between_Front_And_Rear_Frame.jpg
113mm ... http://reprap.org/wiki/File:ToyREP-38-Adjusting_Y_Frame_Distance.jpg
47mm ... http://reprap.org/wiki/File:ToyREP-39-Space_For_Arduino.jpg
The distance between inner flat surfaces of legs is supposed to be 197mm. This distance can be tightened so you have a rigid and square (perpendicular) Y frame. Then you will insert RAMPS with Arduino and push the Y frame toward Z motor mounts so the electronics is fixed in grooves between Rear Y Frame and Bottom Z Frame (PSU mounts). For this you will have to loosen nuts around Z motor mounts, move the Z frame and tighten them it on correct position. When you will do this you will get ca. 47mm distance between inner flat surface of Rear Y frame legs and Z motor mounts and 113mm between Front Y Frame and Z motor mounts. These two measurements are there only for reference to make sure Y Frame is perpendicular to Z Frame - this can be also checked by steel square, but you will have hard time to squeeze it somewhere.

Thank you very much!

Comments deleted.

I have big problems with heat creep...

I have already conquered a lot of problems, but i'm afraid this will be my end..

I am trying to print the calibration cube in PLA 1.75, but despite different settings, i will stop after approximately 15 seconds.

I have tried following settings:
220 degrees celcius
200
185
Speed from 1mm/s to 12mm/s. (Eventhough he said not higher than 1mm/s)
Tried different heights so i wasn't because filament couldn't come out.
2 fans blowing on the part clossest to the hotend

When i do it in air with pronterface 150mm/min it is no problem. But when it starts to print suddenly it is..

In the beginning it will extrude a normal amount and very quickly it becomes less and less that gets extruded out. And then it is just small drops sometimes, getting out.

When i pull it out, while it is still hot, there will be this really thick end coming out. So my guess is that this is heat creep?

What can i do? i am totally lost here.
I'm afraid the low speed is causing this behaviour, can i change it somehow?

I think it is the same problem as had Eddiie. Please see my reply four days ago.

Hello, I've just finished setting everything up, but when I try to move any axis with repetier, I hear for example the X axis like vibrating but not moving at all...

It is set to 1/32 microstepping and with 56mV on the stepper...

Otherwise I followed everything, but I can't get from where it wouldn't work ?

Thank you in advance, awesome project :)

I figured out what was the problem, the belt is getting out of the pulley and makes friction with the support. Then why is the belt going away ? Is it to tight ?

Hi,
The problem when belt is going out of the pulley is caused by a slight tilt of shaft either on the side of the motor or ball bearing. That is the reason why is there the washer and why the printed pulley on the stepper motor have flat sides. If possible use washers on both sides of the ball bearing so the belt won't slide down to the rough surface of the plastic part, but will run on flat surface of the iron washer.

Also try to set lower microstepping. You should be fine with 1/8 concerning the precision and the motor will gain a bit of torque.

Ah thank you, but the belt slides on the motor side, on the pulley... And there isn't place enough within the X carriage part to put a washer on the pulley :/

Thanks for tha anwser anyways :)

That is bad. Try to put the pulley to the same position as the carriage and the ball bearing on the other side, so the belt is going in one straight line. And possibly replace the motor as it is probably badly assembled.

Ok thank you very much ! I'll try everyhting and keep you in touch :)

Also, in repetier, I found out that when I ask my Y axis to move 50 mm in one direction, it works, but another 50mm, it goes backward... One time out of two is the right direction... Strange ?

And now, when I ask the extruder to go backward, the Z motors will go up, and the heat cartridge won't be regulated anymore !
I don't know what's happening with this card, but it scares me for sure ! never did that to me on my prusa :o

Both seems to me as firmware related problems.
Thermal regulation went crazy when I messed up with thermal runaway protection.
Strange behaviour of steppers is mostly due to some wrong settings, or missing/malfunctioning endstops. You should home the printer before any test movement.

Ok I changed to Simplify 3D instead of repetier, and it works much better, just still has to wonder about this pulley problem :/

Thank you very much !!!

Woo!! Got my motors in (12vdc ones) those should work ok with the bipolar mod right?
Ok...a crazy question
These steppers are geared, they are built inside for up to 32 steps per rotation using full steps
With gearing and then 1/2 stepping, or lower stepping it's capable of getting what 32x64=2048
If we were to go full stepping instead of microstepping, would it keep the motors from overheating?
Is this printer (due to gearing and backlash) even capable of such accuracy with so many steps?
Would setting the firmware in ramps to use full stepping fix the overheating without really killing accuracy?

https://www.youtube.com/watch?v=15K9N1yVnhc

https://www.youtube.com/watch?v=B86nqDRskVU

OH! 12V 28BYJ-48 are not good choice as you will need at least 24V PSU to make them move with bipolar hack.
Otherwise you are right that 1/2 microstepping will be just fine. You will have 23264/50 = 81.92 microspteps per mm of movement in X and Y axis. But that will not have influence on overheating as that is dependent on the current you will set on drivers and speed of the motor you will use to move them.

Hi, unfortunately I've read about "5v steppers" after have buyed 12v version... I've just placed a new order for the right ones.
But... I understood I could use hacked 12v steppers with a 24v psu, but what's will happen using hacked 12v stepper with 12v PSU?
May them drive to slow? May them don't drive at all?
Thanks

If you will dig deep in the comments, you will find out that Sobo84 already had this issue. It is called inductance, and it is not that simple as going from 12V to 24V as actually the voltage of the motor have nothing in common with PSU voltage.

For eg. Nema17 motors have some rated voltage of around 2.5-5.0V but they are perfectly capable of 12V, 24V, 36V and even more, if your electronics, especially stepper drivers can survive it. Actually more volts, the better, as you can utilise more amps and thanks to it, get more speed, torque and that means better precision with higher and stable microstepping. And it is the property of coil inductance. It have a drawback though and that is the resistance that causes heating of motor. While Nemas can be cooled quite well with 28BYJ-48 it is not that simple.

Now the properties of 5V and 12V 28BYJ-48 are way different. The inductance and resistance of 12V version makes it hardly usable. You can try higher voltage but there is no guarantee it will work well.

Hi! Very nice project, I've already started to print parts... I've a question about BoM: Why there are 5 stepper and 4 drivers? Am I misunderstanding something?

I've also a question about hotend because I'm quite confused on what to choose. :-(
Can you explain me the difference beetwen the following items, and which would fit on toyrep?
Thank you in advance.

1) "E3D v5 direct" - http://www.ebay.it/itm/Estrusore-E3D-V5-all-metal-Hot-end-j-head-kit-Prusa-RepRap-DIRECT-Extruder-/262814443384?hash=item3d30f62778:g:gjsAAOSw5cNYgKeLhttp://www.ebay.it/itm/Estrusore-E3D-V5-all-metal-Hot-end-j-head-kit-Prusa-RepRap-DIRECT-Extruder-/262814443384?hash=item3d30f62778:g:gjsAAOSw5cNYgKeL

2) "short range" - http://www.ebay.it/itm/J-Head-Hotend-0-4mm-Nozzle-1-75mm-Short-Range-3D-Printer-Makerbot-Extruder-BI94-/192069129027?hash=item2cb8365343:g:lzMAAOSwUKxYa-dO

3) "short distance", I've seen a difference on the top hole in comparison with 2) - http://www.ebay.it/itm/3D-New-Metal-J-head-Extruder-Kit-Short-Distance-per-3D-printer-with-wire-TE429-/172521890903?hash=item282b1b2457:g:iJAAAOSwjDZYm1tW

Hi,
From all of the above I think number 2, is suitable.
Number 1 and 3 are in fact (per pictures) E3D V6 clones. They have different number of ribs on the cooled and are equiped with v6 heat block. No. 3 is probably threaded on top = prepared to be used with pneufit fiting for bowden extruder. No. 1 have also v6 throat and size of the cooler is big enough just for 30mm fan = won't fit in ToyREP as it needs cooler longer to fit under 40mm fan.
Unfortunately non of the hotends have any measurements in description, so I would suggest to ask the vendor for some measurements if possible.
Regards
Martin

Thank you Martin!

I have the hotend similar to the one listed in number 1. It sucks. The filament gets jammed with constant retraction gets bloated at the end and causes slipping/grinding/fail. Maybe it is the nozzle. Granted I ran the printer so many times without filament working out bugs with the steppers that maybe the inside got cooked. So, I ordered another hotend that looked good to me from folger tech ( https://folgertech.com/collections/hot-ends/products/all-metal-hotend-for-abs-pla-for-e3d-or-j-head-type-filament-extruder-1-75mm <-- THIS DID NOT FIT). Will use the throat, heat block, and nozzle. I had a hard time finding the correct top-part of the hot end to fit the printed parts, but finally got the right one after the 3 or 4th try.

Why this extruder doesn't fit? It seems to be the same used on this toyrap... And I've bought the same...

There are two issues.

1, PLA have very high friction on bare metal. The hole for filament is not exactly 1.75mm but rather around 2mm. When it get warmer and is under pressure (as it do above heatblock on the bottom of cooler) it will expand, stick to the inner sides of the metal throat, and stuck there. This will not show unless you try to print. When you extrude filament into free air, you usually extrude quickly, so the fillament don't have time to expand. Troubles starts when you print slowly.

2, The Chines clones of E3D v5 have a gap above throat in the aluminum cooler. You can check it by caliper. The threaded part of throat that is inside of the cooler is typicaly 20mm, but when you measure the hole for throat, it is around 23-24mm. Throat won't go deeper because the thread is not made all the way down (it is not fault of Chinese, it can't be technicaly done). For this reason E3D v6 have a teflon tube 4/2mm (OD/ID) going through cooler to the upper part of throat. And for the very same reason throat of v6 have M7 thread filled flat on the top.

To prevent filament jamming try to use throat with teflon tube inside and if it is possible drill bigger hole in cooler too, so this teflon tube can go through the whole aluminium cooler and throat in one piece and guide filament all the way down to the nozzle.

Comments deleted.

The reason for this is that you need 2 steppers for the Z Axis.

So:

X-Axis = 1 Driver, 1 Stepper
Y-Axis = 1 Driver, 1 Stepper
Z-Axis = 1 Driver, 2 Stepper
Extruder = 1 Driver, 1 Stepper

Thank you so much!

Hi, i am building the toyrep, and i have a e3d v6 hotend, can i use it in the toyrep?? what should i modify for fitting in??

Thankss

That will be a problem. E3D v6 is a bit shorter than v5 that is used in ToyREP. Fan, Extruder and Carriage models would have to be changed.

okkey Thanks, if i finally get it, i will tell you. Thanks for the toyREP design, it is a very good job!!! :)

Comments deleted.

Hi,

I've bought similar one year ago: https://www.aliexpress.com/snapshot/7229641759.html?orderId=72075492388935&productId=32518167514

Did not tested them yet though. Not sure how will perform with microstepping as they may have quite high induction. Still, I'm saving them for a different kind of printer as they would not fit into ToyREP.

After months of printing, getting parts, fitting together, starting with small power supply then deciding to upgrade to bigger power supply (START WITH THE BIGGER POWER SUPPLY - Don't even bother with the smaller one. As noted, changing the power supply later in the build is a freaking nightmare). Anyway, the printer is DONE! I have not gotten past Homing and am having trouble with the Z motors.

I've replaced about 3 Z motors on the left side, the one on the right seems to be fine. The motor on the left just stops turning after awhile and the motor on the right goes fine. Increasing the amps on the driver seems to make it worse. It seems the gears inside the motor give out. One complete movement from Z Home to Z Max and Home again will kill the motor.

I added much oil, too much oil and it has made the nuts come free on the Z parts. Gone through about 10 motors in various parts of the build, have more coming in the mail.

Questions if anyone can please help -
1) Why is one side of the Z axis always harder on the motor than the other side? Suggested solution? Moving the carriage up and down by hand seems to have no binding that would cause stress, something when the threaded rod is added in... Latest version of Marlin now supports dual Z motors, would this help?

2) I used super glue and JB Weld to hold the nuts in place on the Z axis, oil has made the glue from free. Is there a better way to secure the nuts to the Z parts?

3) Any chance of a remix on the Z motor parts to use NEMA 17 motors? :)

Thank You for feedback. Honestly I am a bit clueless, but I will try my best.

Add 1, I am aware of the fact, that for each and every stepper motor you should have one dedicated stepper driver. BUT ... one driver for two motors is working for NEMA17 and also with hacked 28BYJ-48 on other set-ups. So this should not be the case especially when both motors together need few mA only. Although you won't loose much trying Marlin with dual Z drivers.

Add 2, I use hot melt glue to attach M5 nuts to their position.

Add 3, Already sent source models to few people who wanted to do NEMA17 remix, but so far no one published the result.

I have few more questions for you
1, What drivers are you using? What is the value of sensing resistor on it? What is the reference voltage set for your motors?
2, Do both Z motors spin when not attached to the threaded rods? Is their movement fluent without chopping?
3, How are Z motors connected to RAMPS as there is not enough space to use the original connectors?
4, What values you have set in firmware concerning Z movement - max. speed, acceleration, etc.?

You can do NEMA11 or NEMA14's
Same steppers they put on the Tinyboy (hobbyking Fabrikator Mini)
https://hobbyking.com/en_us/turnigy-mini-fabrikator-3d-printer-v1-0-spare-parts-stepping-motor-for-x-y-or-z-axis.html

Getting NEMAs to ToyREP is not that easy. It was never designed for this and such upgrade would require many new plastic parts (almost all of them).

It is probably too soon to shout this loud, but there will be (probably, one day) a new version of ToyREP 2, that will be able to use 28BYJ-48, NEMA 17 and probably NEMA 14 too. I am still skeptical about NEMA 11 and smaller, as it is too weak to drive 3D printer (but with new version such upgrade would be possible to). I am waiting for some part from China right now so I can continue on this project.

THOR - thank you.

Update:
I found problems on z axis, fixed them, replaced motors and they work great.
This allowed me to work out all the other kinks in my build.

I go to print my first test print............................. (drum roll) Heat bed warms up. Hot end heats up. Nice new smell.
Printer homes, X and Y move to about the center of the print bed and .......................... STOP. The arduino locks up. Everything stops. It was late and I go to bed.

I suspect my values are wrong in firmware.

Marlin RC6

configuration.h items of interest -
#define HOMING_FEEDRATE {750, 750, 12, 0}
#define DEFAULT_AXIS_STEPS_PER_UNIT {655.36, 655.36, 40960, 6953,59} // ToyREP
#define DEFAULT_MAX_FEEDRATE {30, 30, 1, 5}
#define DEFAULT_MAX_ACCELERATION {300, 300, 1, 17}
#define DEFAULT_ACCELERATION 150
#define DEFAULT_RETRACT_ACCELERATION 17
#define DEFAULT_TRAVEL_ACCELERATION 30

#define DEFAULT_XYJERK 5.0
#define DEFAULT_ZJERK 0.01
#define DEFAULT_EJERK 0.1

I also added inductive Z probe for bed leveling, this is the first time I have added this, so maybe something there too?

The printer is close enough to post make pictures, however. It looks cool.

Hi Eddiie,
You have beautiful build!
I see only one mistake in line:
define DEFAULT_AXIS_STEPS_PER_UNIT {655.36, 655.36, 40960, 6953,59} // ToyREP
There should be 6953.59 at the end of the line instead of 6953,59 - the comma in the last number should be replaced with dot.
Rest of it I will check later.
Regards

I made the typo fix to DEFAULT_AXIS_STEPS_PER_UNIT and it has no impact.

Maybe it is the slicer settings.

It may be also slicer, but this seems to me as firmaware related thing or overcurrent protection in PSU.

Do you use Pronterface? If yes, please turn on Debug communication in Settings menu. Next time when you will try to print, it will write you why it stopped in the text console. If you don't see text console, then in Settings/Options/User interface set Interface mode to Standard or Compact and restart Pronterface.

Second question, did you do the PID calibration on both heatbed and extruder?

If you want to rule out the Slic3r, in Filament Settings set temperature on Extruder and Bed to 0°C. It will than use the temperatures you set in Pronterface and won't wait for heaters to reach any preset value.

And please send me PM with your email. I would like to see the whole configuration.h of your firmware.

Update: What is the real resistance of the 120x120mm heatbed? If it is around 1.0Ohm as I found somewhere on internet, then it would consume about 144W! Can your PSU deliver such power? If this is the case you have to set #define MAX_BED_POWER 128 or so (instead of 255). It will heat up much slower but should not crash when more things like extruder and motors will start running.

Hi!

Here is the output from ponterface, debug communications.
I resliced using Slic3r and no difference, still locks up.

Yes, I did the PID calibration, that all works fine, I have an amp meter on the PSU, there is plenty of power/amps available.

I also have inductive Z probe, this is my first probe, maybe I disconnect and disable in firmware.

Perhaps output below has clues......

RECV: echo:busy: processing
RECV: echo:busy: processing
RECV: echo:busy: processing
RECV: echo:busy: processing
RECV: echo:busy: processing
RECV: X:120.00 Y:120.00 Z:0.00 E:0.00 Count X: 78643 Y:78643 Z:0
RECV: ok
SENT: M105
RECV: ok T:207.3 /208.0 B:48.2 /48.0 B@:33.45W @:101
SENT: M105
RECV: ok T:207.3 /208.0 B:48.2 /48.0 B@:33.45W @:101
SENT: N4 G1 Z5 F50000
RECV: ok
SENT: N5 M109 S208
103
RECV: T:207.5 /208.0 B:48.2 /48.0 B@:
RECV: 32.44W @:98 W:?

The last SENT: command is not pasting correctly. there is an asterisks between S208 and 103
Can someone please PM me their configuration.h for their printer? Marlin 1.1.0 RC6

In the meantime, I will review settings.

This is not as complete as I was hoping for, but better than nothing.
The temperatures were set to 208°C on extruder and 48°C on Heatbed. For heating you are drawing 135W. The printer got past homing, lifting Z 5mm heigh and crashed when waiting for extruder to reach and stay on 208°C, right?

Multimeters are not measuring peak values, especially on switching PSU. Your PSU is rated for more than 11.25A = 135W?

You managed to home somehow or atleast your printer think so. Do you have only z-probe or also endstop?

No, PSU rated for 8.4 amps. Volt/amp meter hits 7.2 amps max when heatbed and hotend are on. I am 99.999% sure it is not PSU.
I can move everything around using manual controls with the heat bed on, hotend on, homing, 7.2 amps max.

Still, I can disable the heatbed as a test, that should be easy enough.

I have Z probe AND endstop. Funny, it seems all mendel type printers would have Z min end stop. Connecting a Z probe should be no different than any other type printer. (I am new to Z probes)

I think the next thing to do would be to start over with a new marlin folder, new configuration.h, and retest using settings from wiki without Z probe stuff, just follow wiki, disconnect z probe. This is what I will do now.
:)

Hmm, same result with Marlin RC8.

Here is full log -

start
echo:Marlin 1.1.0-RC8

echo: Last Updated: 2016-12-06 12:00 | Author: (none, default config)
Compiled: Jan 11 2017
echo: Free Memory: 3529 PlannerBufferBytes: 1168
echo:Hardcoded Default Settings Loaded
echo:Steps per unit:
echo: M92 X655.36 Y655.36 Z40960.00 E6953.59
echo:Maximum feedrates (mm/s):
echo: M203 X50.00 Y50.00 Z1.00 E5.00
echo:Maximum Acceleration (mm/s2):
echo: M201 X300 Y300 Z1 E17
echo:Accelerations: P=printing, R=retract and T=travel
echo: M204 P150.00 R17.00 T300.00
echo:Advanced variables: S=Min feedrate (mm/s), T=Min travel feedrate (mm/s), B=minimum segment time (ms), X=maximum XY jerk (mm/s), Z=maximum Z jerk (mm/s), E=maximum E jerk (mm/s)
echo: M205 S0.00 T0.00 B20000 X5.00 Y5.00 Z0.01 E0.10
echo:Home offset (mm)
echo: M206 X0.00 Y0.00 Z0.00
echo:Material heatup parameters:
echo: M145 S0 H180 B42 F0
M145 S1 H180 B42 F0
echo:PID settings:
echo: M301 P25.35 I2.22 D72.27
echo: M304 P155.27 I30.36 D198.53
echo:Filament settings: Disabled
echo: M200 D3.00
echo: M200 D0
echo:SD card ok
echo:SD card ok
echo:enqueueing "M23 3dbenc~1.gco"
echo:enqueueing "M24"
echo:Now fresh file: 3dbenc~1.gco
File opened: 3dbenc~1.gco Size: 4697928
File selected
T:141.1 /0.0 B:35.0 /42.0 @:0 B@:0 W:?
T:139.9 /0.0 B:34.8 /42.0 @:0 B@:127 W:?
T:138.8 /0.0 B:35.0 /42.0 @:0 B@:127 W:?
T:137.7 /0.0 B:35.3 /42.0 @:0 B@:127 W:?
T:136.6 /0.0 B:35.7 /42.0 @:0 B@:127 W:?
T:135.4 /0.0 B:36.3 /42.0 @:0 B@:127 W:?
T:209.1 /208.0 B:42.1 /42.0 @:73 B@:33 W:0
echo:busy: processing
X:120.00 Y:120.00 Z:5.00 E:0.00 Count X: 78643 Y:78643 Z:204800
X:120.00 Y:120.00 Z:0.00 E:0.00 Count X: 78643 Y:78643 Z:0
X:120.00 Y:120.00 Z:15.00 E:0.00 Count X: 78643 Y:78643 Z:2
X:120.00 Y:120.00 Z:15.00 E:0.00 Count X: 78643 Y:78643 Z:8
...Arduino locks up...

I just tried again with no heatbed, lockup.
Printing from SD card, printing from Pronterface, same difference.

Try a different arduino? Maybe move to Marlin forum.

Maybe a different firmware?

With Repetier firmware, everything works, except extruder at the moment. Physical problem, not related to firmware.

Strange that Marlin locks things up. (EEPROM is/was disabled by the way)

Cool to see it move around!

Good to hear it was firmware related problem and have a solution. Congrats!

Just out of my curiosity, what was the voltage on PSU when you measured 7.2A with everything running?

11.6v @ 7.44a

Ok, so I spend some time this weekend on printer. After tinkering with Repetier, I took the settings and applied to Marlin, and printer worked! So I am back on Marlin (RC8).

Settings that are working for me:

#define DEFAULT_Kp 25.35
#define DEFAULT_Ki 2.22
#define DEFAULT_Kd 72.27

#define DEFAULT_bedKp 155.27
#define DEFAULT_bedKi 30.36
#define DEFAULT_bedKd 198.53

#define DEFAULT_AXIS_STEPS_PER_UNIT {655.36, 655.36, 40757.73, 7024.20}
#define DEFAULT_MAX_FEEDRATE {30, 30, 0.350, 4.17}
#define DEFAULT_MAX_ACCELERATION {30, 30, 10, 17}
#define DEFAULT_ACCELERATION 30
#define DEFAULT_RETRACT_ACCELERATION 10
#define DEFAULT_TRAVEL_ACCELERATION 30

Changed in RC8 (RC7?)
#define HOMING_FEEDRATE_XY (750)
#define HOMING_FEEDRATE_Z (12)

Still have not printed anything, but I am getting close!

Brother Thorgal,
The printer works!! I printed a 3d benchy and it came out pretty nice! (4 hour print!) The picture has been added to my Made page - http://www.thingiverse.com/make:283877 (last picture)

However, the movement is SO SLOW, I know from tweaking with this thing over the months that the movement is capable of faster moves than what I have defined. Yet, I am not sure what settings to tweak to make things faster.

Here are current settings, which one's should I change to make the movement faster? Note, I am pretty happy with the speed of Z. It is X and Y that I'd like to improve.

#define DEFAULT_AXIS_STEPS_PER_UNIT {655.36, 655.36, 40757.73, 7024.20}

#define DEFAULT_MAX_FEEDRATE {30, 30, 0.350, 4.17}
#define DEFAULT_MAX_ACCELERATION {30, 30, 12, 17}

#define DEFAULT_ACCELERATION 30
#define DEFAULT_RETRACT_ACCELERATION 10
#define DEFAULT_TRAVEL_ACCELERATION 30

What settings have others dialed in?

ToyREP 3D Printer
by Eddiie

Hi Eddiie,
Great job and nice pictures. Thank you for sharing!

Speaking about speeds, did you try to find the highest speed possible for your machine? Check: http://reprap.org/wiki/ToyREP_User_Manual#Speeds_and_units Just please be aware you may burn few motors while trying!

Also your X and Y axis default_max_acceleration, default_acceleration and default_travel_acceleration is quite low (only 30). Do you have a reason for this? I would try something around 150-300 (and if you want higher speeds maybe even up to 900-1000).

Thing is, that although you set 30 mm/s as a maximal speed in X and Y axis, with default_max_acceleration set to 30 in X and Y axis you will never reach more that 5,5mm/s on 100mm distance of such small heatbed (To make it simple, think about maximal reached speed as a square root of maximal acceleration, although in real it is not that simple).

On contrary Z (12) is quite high value and shall be around 1. Your acceleration of Extruder is fain.

Hey - newer versions of Marlin (I am using RC8) have different format for #define HOMING_FEEDRATE (750, 750, 12, 0)

#define HOMING_FEEDRATE_XY (750)
#define HOMING_FEEDRATE_Z (12)

They also change #define DEFAULT_XYJERK to
#define XJERK
#define YJERK

Mine is currently at 30, it move the carriage up/down so much faster. It is a hack because I have not properly calibrated the steps and stuff as you linked above. I will do this today. I've replaced so many motors already on every axis, the threads on the retaining screw holes are stripping out.

I was hoping my settings we too (s)low. That is hopefully the cause of my filament clogging (as you exactly described in another reply).

Will update this when done with calibration.

Question: Do you still use your ToyRep?

Thank you!

ToyREP is able to move and print around 30-35mm/s in X and Y axis with hand picked motors that are capable of high speeds. But with speeds above 12.5mm/s the motors are overload and overheated and in my experience will last only few months of occasional prints.

Honestly, I did not use ToyREP for a year. Reason is that for more than year now, I'm printing at work almost every day. I have to use 4 years old semi-profesional piece of crap-rap. Even my CARP Box is better, so it stays at work too (as a backup). And back at home, where I'm left with ToyREP and Scoobydoo (my first full metal printer I've ever build after Mendel) I don't have a mood to print anything else.

Good stuff.

I begin calibration now. Too much stuff in the way to make dot, move, make new dot, and measure the distance - I printed a bar that was 100mm in X, another 100mm in Y, and 50mm in Z.

My results -
X = 94.72mm
Y = 99.65mm
Z = There is no mention of Z calibration on the wiki?

Question:
When calibrating extruder, am I measuring the output of the extruder or the input? I guess it is the output.

Math -
X steps are currently 655.36, the measurement was 94.72, Wiki says ... 655.36 / 94.72 X 100 = 691.891 (new steps/mm)
Y steps are currently 655.36, the measurement was 99.65, Wiki says.... 655.26 / 99.65 X 100 = 657.6618163572504 (new steps/mm)
Z - I will circle back on this, but it seems OK
E - Waiting to hear back from Lord Throgal

Hi Eddiie,
for Z axis there is mechanical calibration described in chapter: http://reprap.org/wiki/ToyREP_Build_Manual#Print_Surface
It is the very basic way and you can do it differently, like start with measuring the height of X axis rod from table on both ends and than level the heatbed surface by driving the hotend into corners and using the piece of paper method.
In firmware, the calculation of microsteps per mm is in: http://reprap.org/wiki/ToyREP_User_Manual#Speeds_and_units
It is motor_steps_per_turn x gear_box_reduction * microstepping / pitch of the threaded rod. in case of 28BYJ-48 with bipolar hack, with 1/4 microspepping and M5 threaded rod = 32 x 64 x 4 / 0.8 = 10240 microsteps per mm (or as in the example with 1/16 microstepping 32 x 64 x 16 / 0,8 = 40960). You will have to print few layers of known height and measure the overall height of printed model, than use the same formula from Wiki

Your recalculation of X and Y seems correct. Keep in mind, that this is influenced by size of printed pulley and tension of belt, so it is perfectly OK that these two numbers are not the same.

In Z axis where is no tension or printed part involved, the basic calculation is usually correct unless you have for eg. motor that do not have gear box with 1:64 but some odd number like 1:63.68395. Keep in mind that if you print for eg. the 10mm heigh calibration box, and slice it with 0.3mm layer thickness, the height won't be exactly 10mm but rather multiplication of either 33 or 34 layers, means 9.9 or 10.2mm. Pronterface will write the actual hight in the status bar on the bottom of the window. For better results, You will have to print something taller than the calibration box in all four corners of heatbed (it do not have to have infill and top layers to save filament and time). Chose one corner as the reference, calculate the difference between corners and make this correction by the M3 screews in corners. One full turn of M3 screw should do 0.5mm height difference. Than use the reference corner height to change the microsteps per mm in firmware and you will be good to go.

For extruder, you have to measure the length of filament, that is coming INTO the extruder, so mark like 50mm and 100mm on filament above the extruder (for eg. measure from top of the extruder) then extrude 50mm in Pronterface (or other software) and measure the length to the second mark. If you will measure for eg. that 53.5mm is left to second mark, than you extruded only 46.5mm. Put this number into the equation and you will get the right number of microsteps per mm. Make new marks and recheck, It should be better but probably not perfect. Some deviance will be always there due to the precision of printed extruder gears.

Later on you will set extrusion multiplier in Slic3r (or Cura) settings. This is to correct the heat expansion and flow of the plastic when printed. For PLA I use something between 0.9-0.95 (90-95%) and for ABS 0.975-1.0 (97.5-100%). These numbers are very dependent on plastic properties, that are very different for every manufacturer and even colour of the filament. You will have to tweak it depending on the print results you will get with the particular filament. For eg. if holes will be printed smaller then required it is probably due to overflow so the multiplier needs to be a bit lower. On contrary if layers on surface seems to have occasional space in between go a bit higher.

Comments deleted.

Nema 14 would be interesting...great Project

You may be right but in my point of view Nema 14 is not enough smaller nor cheaper than Nema 16 and 17. Nema 11 and 8 are noticably smaller but way more expensive and much weaker (also have problems with microstepping, atleast Nema 8 do, due to inductance).
I would rather use Nema17 20mm tall, than Nema14 28mm tall with the same 1kg-cm. In the end it would make reasonably small, cheaper and more reliable machine (in terms of compatibility with common stepper drivers).

Comments deleted.

I don't think, this will let you use 28BYJ-48 without the bipolar hack as the output have only 4 pins. For unipolar motor you need at least 5 or 6 pins.

I'm having trouble wiring the 28BYJ-48 stepper motor to the RAMPS 1.4. My main concern is, you never put what kind of connector it is your using. The connectors that came on my motor are too thick and in the wrong color pattern. So what is the name of the connector I can put into google/amazon?

Also, I accidently order the 12v motors. I noticed the buyer's guide doesn't mention which way to go. You should put a note on the motors that if you're going with the 12v power not to go with 12v motor. I personally read some of the notes then weeks later started buying everything; so I forgot.

With regards to connectors, I used DuPont connectors, like these here:

http://www.ebay.co.uk/itm/121920271555

What you need are the plugs as well as female connectors. They fit snuggly to the connectors on the RAMPS board. You normally would also need something called a crimping tool to fit the female connectors onto the wires, but you can also just use a small wrench; there are plenty of instructions online of how to do this without said crimping tool.

Thank you. The only thing I could find was a JST connector, but even that was coming back with different products.

hello i'm having layershifting , i'm using a GT2 belt and maybe :) (i dont know anymore) i'm using a GT2.5 Pulley could this be the cause of the problems ? at every layer the printhead is moving about 5 mm to the left, i have increased the Vref and checked for overheating, but everything else seems fine

That's exactly the cause

T2.5 pulley with GT2 belt would do no good, but won't cause regular shift either.
Typically such shift is matter of undercurrent or too high movement speed. In both cases, the motor is not able to stay on microstepping and/or do the required amount of microsteps. But that is mostly the case of Nema17, not 28BYJ-48 due to the internal gearbox.

First, check if the pulley is correct for the belt you are using. Than try to print at half of the speed you are using now. If the problem still remains change the motor. Some teeth may be missing in the gearbox.

Comments deleted.

What power supply are you using????

It was discussed so many times. Please read manual and comments. If you want to embed it into the printer, you need 12V, 40-120W LED PSU of sizes mentioned in build manual. Don't ask me for some particular model or brand. It depends on what is available for eg. in you local electro shop. Otherwise any 12W 40-120W notebook PSU can be used too, but you will have to find a way how to unplug it.

Power depends on the overall consumption of the printer. Measure the resistivity of all your heaters and calculate the consumption by OHM law. Find PSU with wattage that is bigger than the requirement. For eg. my machine have 3.6Ohm heater and 4.8Ohm heatbed. PSU is 12V so the total consumption is 12x12/3.6 + 12x12/4.8 = 70W, you can add 1W per every fan and 1W for electronics, if want to be really sure.

Suggestion is also to use an ATX powersupply, it has 12vdc rail, and a 5vdc rail, even old anemic throw away powersupplies run 250 watts or more, they are larger than the small 70 watter for sure, but they should supply ample power with no problems

You mention using a notebook PSU. Could I plug that directly into the mega board (since it already has the female plug.). I'm hoping that since the RAMPS 1.4 and everything else is attached that I could pull the full 120W. I've never actually dealt with wiring my own electronics before this so my knowledge is ver low.

No. That way you would power Arduino Mega only.

When using 12V notebook PSU you can attach some female plug on RAMPS screw terminals and power up Arduino through it. That was my suggestion. Mind the polarity though.

Hi Thorgal,

first of all, thanks a lot for designing this printer! I have no prior experience to 3D printing, so I have used this printer as an introduction in how 3D printers work and it was really a lot of fun building it (pictures of the build to follow).

I have finished building the printer and got everything working smoothly so far and the first few prints come out quite nicely. I just was wondering about one thing and thought someone from the community might advise on this.

For the default feedrates I have currently the following settings:

define DEFAULT_MAX_FEEDRATE {20, 20, 0.15, 3}

Everything is working fine like this, however, if I increase the feedrate for the Z axis (as I thought that is really was quite slow) even to only 0.2 and test Z axis movement in prontrface, the motors start to jitter and I do not get any reliable movement of the axis anymore, i.e. sometimes it moves, sometimes it doesn't. My Vref for the DRV8825 steppers that I have is set to 0.05V, and I have already replaced the motors (I have the 5V version), as I thought they might be the issue, but to no effect. It does seem to me that even at a max feedrate of 0.2 the motors do not seem to have enough torque anymore to move the Z axis. Is this low feedrate of 0.15 normal and do other people experience the same? If not - does anybody have an idea of what the issue might be?

Thanks a a lot and thanks again very much for all the work Thorgal!

Hi Hubert,
Thanks for kind words and feedback.

1) What microstepping did you set by jumpers? For Z axis I would recommend something like 1/4, 1/2 or even fullstep. It will give you a bit more torque. You will have to change DEFAULT_AXIS_STEPS_PER_UNIT accordingly. (For X and Y keep 1/16 or even 1/32 for better precision)

2) what is the acceleration set in firmware? it should be really low, like
DEFAULT_MAX_ACCELERATION {150, 150, 1, 15}
where 1 is for Z axis. And set also
DEFAULT_ZJERK 0.01
to enforce acceleration whenever the axis moves.

3) lubricate the threaded rods in Z axis really well. That will increase friction and reduce torque required to move in Z. Try to move the Z axis all the way up and down few times. It will take a lot of time time, but will pay off in the future.

4) what threaded rod do you use - M5 or M8? With all this you should be able to get somewhere to 0.35mm/s in Z axis on M5. I have no personal experience with M8, but in theory it should be 2.5 times quicker.

5) also try to buy brass or nylon nuts instead of common nickel plated iron.

6) 5V motors are right choice but try to increase current in Z axis stepper driver. There are two motors so you need 0.18-0.30A in total. Vref on this driver should be 0.09-0.15V. (You can use even slightly more if motors in Z are not in break and if printer do not move in Z during retraction but only when changing layers. But check if they are not heating up too much.)

Hope that will help and please keep me posted about your progress.

Dear Thorgal,

thank you so much for replying so fast!

Concerning 2) I have these settings set as you described and as for 3) and 4) I am using a M5 rod which I have lubricated.

You were spot-on with the 6) Vref setting and the 1) microstepping - I had set the microstepping to 1/8 and the Vref was at 0.05V. I changed the microstepping to 1/4 and set the Vref to 0.12V, which did the trick. I completeley forgot that that the driver is powering two motors and therefore needs to pass on a higher voltage! Thank you very much for pointing this out!

At the moment I have the regular nickel plated iron bolts - I will try to replace them with Nylon nuts at some point!

I have uploaded some photos of my build, which now runs smoothly and a tad bit quicker thanks to your advice.

A really great little printer!

Thanks for posting pictures! I'm really happy we found a solution.
Just be careful with nylon nuts. They will help with friction, but as every plastics parts they will wear off after some time. Don't have personal experience with them, so can't tell how soon it will come and it will also depend on precision and smoothness of the thread on the M5 rods. If normal nuts are working for you now, there is probably no reason to change them any more because you probably hit the limit of your motors. But if you would like to experiment, I will be happy to hear about the results you achieved.

Hi

Can I use Octoprint to control the device?
How do you upload the g-code to the printer?
Is it possible to modify for dual extruder?

Octoprint works well on my printer. It's basically universal to all printers so it's just a matter of settings.
As for uploading gcode, most 3d printers use a computer program such as Pronterface or Octoprint to send the gcode over a USB cable. For beginners I would recommend a software called Repetier Host. Octoprint also does this job well too, but requires much more setup. Cura is also user friendly, but kind of assumes that you have an LCD so it seems to be missing some features.
Dual extruders are not entirely feasible on ToyREP because of its low price point and limited motors. Not to mention that adding a second extruder takes at least 20mm of space off the Y axis. In my opinion, if you want a dual extruder setup you should save up and build a prusa i3.

I really don't know if Octoprint works with RAMPS 1.4 and Marlin/Repetier firmware. That is a question for Octoprint support rather than for me.

Using Slic3r for generating g-code and Pronterface to print directly form PC. But that is mostly a case of running some test prints. I get best prints when printing directly from SD card.

In theory, it is possible to install two extruders on ToyREP but right now there are no printable parts to support this feature and no instructions for working setup.

Does anybody have any information about the electronics ? i blew my arduino by guessing how to put it together, i have everything printed and the whole printer build etc. but the project is stranded, how many jumpers do i use how do i put on the steppendrivers (can be both ways) i see that many people have similair problems, isn't there anybody who can post a few lines how things should be connected ?

Sorry to hear that.

Build manual for ToyRep, including wiring on Arduino Mega with RAMPS 1.4 and DRV8825 stepper drivers is here: http://reprap.org/wiki/ToyREP_Build_Manual

If you use A4899 stepper drivers remember they have to be installed other way around. See this page where is the correct orientation of both of them: http://reprap.org/wiki/A4988_vs_DRV8825_Chinese_Stepper_Driver_Boards
(green are A4899, purple DRV8825)

Correct microstepping set by jumpers under stepperdrivers is on this picture: http://doku.radds.org/wp-content/uploads/2015/02/MicroStepping_Modes_tbl.png
For Z motors 1/4 is just fine. For X and Y axis and extruder anything above 1/8 will give you enough resolution. If you put there all jumpers you will get 1/16 microstepping on A4899 and 1/32 on DRV8825. No problem with this setup, only you will lose a bit of torque in Z axis and the amount of microsteps for 1mm will be ridiculously high.

Complete wiring diagram of RAMPS 1.4 is here: http://reprap.org/wiki/RAMPS_1.4
The position of endstops and thermistors is much clearer there. Mind the polarity of 12V PSU too!!!

Don't forget to do the bipolar hack on 28BYJ-48 motors: http://www.electronicsmayhem.com/?p=13

Set the current on stepper drivers to minimum as described on this page: http://reprap.org/wiki/Pololu_stepper_driver_board
0.09-0.15A should be enough!

If any other query, just ask.

i have flashed the marlin firmware, but something goes wrong, i get a lot of rubish response, i use marlin 1.0.1-2 and the config file from here, am i using the correct marlin version ? i have tried repetier and i got normal response but still no movement.
i have set the Vref on the steppers(A4899 ) on 0.055v that should lead to 0.13a is that ok ?

Do you have at least correct readings from thermistors and extruder heater working? can you test the electronic with some nema 17 motor? Than this may be a problem of motor pinout. Try to measure which two wires belong to one coil of motor by ohmmeter.

i got movement today ! :D i came across the firmware for the cherry printer, and just uploaded it, and things started to happen (al the wrong way of course. so i think the electronics are ok, the motors looked underpowered so i'm going to raise the Vref . is it possible or is there a place where i can download the firmware for the Toyrep, so i know that is ok ? it would make troubleshooting a lot easier :D

Great! Send me content of Cherry printer's configuration.h firmware as a PM so we can walk through it together.

Great info tnx :) i have new steppers, but the ramps does nothing anymore, so i think i have to order a new one by our chinese friends ;) you should paste this info in the instructions, helps a lot. i also had a problem getting the right colourcoding for the stepperdrivers, i found some pictures from the cherry printer, an other 28BYJ-48 printer, where i copied it from, but i think some people would need this info :) thx for the fast reply

Hi.. How much material is used to print this in meters or grams?

You can easily print all the parts and not use anywhere near an entire KG of filament

Unfortunately I don't know. In BoM price includes electricity and I don't recall the printing time either to reverse the calculation.
Sorry. But you can slice it with your typical settings and Pronterface will show estimated printing time and length of filament when file is loaded.

do you have any electronics guide? like a ramps 1.4 wiring guide or similar???

especially how the pot meter on the pololu should go and the jumpers on the board???

I used the regular wiring diagram for RAMPS 1.4 with 1 extruder, 1 heatbed (may not be used) and 1 fan (also not required). Second stepper for extruder is not used. Thermistor for second extruder can be used for monitoring, as fail safe or not be attached at all. Connection is described in build manual: http://reprap.org/wiki/ToyREP_Build_Manual

does anybody know where i can find the right endstops ? the one's i have have the correct holes but the levers are to small, or is it possible to mount them somewhere else ? the one for the printbed seems the most tricky one, on the pictures it looks like the metal part slides against the belt. if i mount the ones that i have the switch is already closed. can i use a heatgun to mount them ? i'm afraid the would come loose

I have the same issue. I hot glued mine in place. It's really sketchy, but it works 95% of the time.

does anyone know if I can make the x y and zed(z) axis bigger?

Yes you can extend the length of the rods. Z is no problem at all, for X and Y you will have to make bigger frog extenders. Rest of the plastic parts will remain the same.

Do you extend them 25mm to get the full 125x125x125 or does it take a little bit longer 8mm smooth rod?
Disregard...found it in BOM...duh
X 2x - 230mm for 100mm, 255mm for 125mm bed
Y 2x - 220mm for 100mm, 245mm for 125mm bed
Z 2x - 295mm
Easy enough :)

yes you can. if you download the BOM there is examples of different lengths for bigger print surface otherwise you could also do a little calculation yourself and theoretically make it as big as you want.

when trying to compile i get the error message: missing binary operator before token "("

what does that mean???

nevermind fixed it

Does anybody know if it is possible to print the extruder in pla ? I read that you have to cool it, where should i place an extra fan ?

As I build this printer, I am running into problems with mounting the X and Y limit switches. They line up with the holes but when tightening the screws, the switch engages so it is always toggled. Need to move them back a bit, any suggestions? Has anyone else run into this?

I am about to build the extruder assembly now. I lose many motors hammering on pulleys.

Does that mean, that the switches operates normally until you screw them to the printer? You can always drill new holes in different position or stick the switch with hot melt glue.

If your printed models have too tight holes try to file them first. everything should fit precisely. No hammering required.

Yeah- some math errors on my mini kossel makes it print smaller.
Reprint on Prusa i3 and so much better. Not wanting to take it all apart and print new parts, I drilled new holes. the new holes work great. Hot glue, another idea I did not think of.

Very close to finishing up this printer.

I am having a hard time finding glass of this size in the US.
Can someone tell me where they got their's?

I go to the local glass shop, the ones that sell glass to fix windows (every town should have one) get a custom glass cut, I got a 8x8 in for my full size Anet A8 for $5, a 125mm one or 5x5 should be even cheaper

I found an eBay seller that sells 120 x 120 heated print bed, it arrived today and looks great, not yet installed.
Check it out -
http://www.ebay.com/itm/262375406045

Also got a 140x140 piece of glass, don't need it now.

You can use plywood and cover it with tape

how hot does the steppers go? will it be possible for me to print the parts in pla or does it have to be abs??? i struggle to print abs... so pla will be preffered...

Good afternoon. I use Cura app ver.15.04.5. Can somebody give me the settings for this app?

Nice work.
As I make my way through the instructions on the wiki I see errors. Some say use smooth rod when you have picture of threaded rod.
You have completed a huge undertaking by making this, publishing this, creating instructables, wiki, answering questions ... dang.

Can I help by fixing errors I find on the Wiki instructions as I work my way through it? or take notes? You have done enough, would like to help. Know what I mean?

Thanks, man.

Don't worry. I understand what you mean and appreciate Your offer.

Just If you mean assembling PSU cover, the smooh rod is actually correct. It is advisable to bend the wires around smooth rod, and than insert the threaded one. But you are right, it is not described well enough.

I think you can just register on reprap.org and that edit the build manual. Eventually I will have to approve your changes. But I have never undergo this, so maybe I got it totally wrong.

Thank You

email sent for an account today.
I see what you mean about smooth rod on power supply.
More later..

Good afternoon
First to thank thorgal for all content.
I mounted the printer and modified the Electronia, but consiguó not work properly.
The limit switches do not work properly and the engines either.
It is possible to file configuaracion to check if I have mistakes, I use 1.0.2 marlin

PS
I accept suggestions, intentndo take two weeks to fix the problem.
And sorry for my English

Greetings and thank you

Hello! very interesting printer.
Could i ask for more information about correct electronics installation, about adjustment of the current driving the 28BYJ-48 Stepper Motors? What a nominal current for hacked steppers for works without overheating or missing steps?
I plan to use DRV8825 Stepper Drivers.

The motors are getting really hot, i have one of these motors for playing with my arduino kit. I think if you can cool them with a fan they will last a lot longer. I will build one maybe i'll try to add a cooler on the motors.

it is already done http://www.thingiverse.com/thing:981794
feel free to use these alternate models.

ToyREP Stepper Motor Cooling
by thorgal

Ok great! I think the extruder should be fitted with an Nema 17 like the Cherry printer has... maybe i can do that, but that is probably done as well but i cannot find it

No this was not done (yet), but there is possibility to use E3D v5 bowden hotend and any bowden extruder including Nema 17 driven ones. There are plenty of those on Thingiverse.

Hello! very cool printer. I have collected 5 pieces. Very satisfied that it is cheap.
Could you add the small screen to control your printer project. It would look very nice.
http://www.thingiverse.com/thing:1005961

TinyOLED V1.0/V1.1 for RAMPS : 1.3" I2C OLED, TF Card Module, Rotary Encoder, Buzzer, Connector Board

Why don't you add it yourself. You could integrate it at the front.

Why don't you add it yourself. You could integrate it at the front.

I currently add their own. I made a proposal to improve the printer

Comments deleted.

If I get you correctly, You can use larger bed, but you would have to prepare bigger frog for it. The rest is matter of longer rods, no other plastic parts have to changed. Of course you need to adjust firmware for bigger print area.

Comments deleted.

Sorry, I cannot print it for you and I've already mentioned, it is not good idea, to have this printer as only one at home. Still the risk is up to you.

I'm writing this just to clarify that you cannot use GT2 aluminum pulleys shipped with belts due to fact that the 28BYJ-48 motors do not have shaft long enough to attach the pulley to the motor.

Hi,
I read it a few times, still I don't have so much money and I like to build and try things...so I will build the toyrep as good as possible, with it I want to get better at 3d modeling and printing and than, if i still think its worth it, i will start buying more nema 17 and start building "better" printer...but for this I need in primis a "reproduction base" :) so thanks for the heads up and the info about the pulleys...I will have to buy gt2.5 belt :)

Hey, i wrote a message to you. Are you German? Have a look at the message center. You can have mine.

A great low coast way to acquire 8mm smooth rods is from old flatbed scanners or printers. The scanner I took apart had two 17" (~420mm) rods, and most HP printers have 8mm rods that are at least 14" (~200mm). Also, the stepper motors used in most scanners could be (theoretically) be used as an upgrade from the 28byj motors, but I have yet to test mine.
I couldn't find a 3mm threaded rod for the Z axis endstop, so I used a wood screw instead. It is harder to adjust, but it works OK.

Hi Thorgal,

First off, very good job! I'm impressed by your hard work and I appreciate that practically 90% of this is 3D printable...

I was just thinking about building one, but first I wanted to ask you: Have you re-calibrated it since your last update? How are the round prints/surfaces going now?

And mostly... why does Marvin come out "jagged" that way, and what can we do to fix/upgrade the situation?

Thank You for kind words.

If by re-calibration you mean uploading new firmware with changed values of steps per milimeter, than the answer is no. There was no need for it as cubic shapes comes out quite well and with reasonably low deviation. Problem with round shapes is probably in backslash of internal gearbox, although it was not proved to be true yet. Somebody also advised me to use Repetier with backslash compensation, but I do not have time for it either. Last thing I've printed on ToyREP was probably Marvin. Since than I made only few upgrades for ToyREP while working hard on second version, that will be mostly printed and with lesser amount of vitamins.

There is not much that can be done on your end, unless You will print one and by using it help me with development. The quality of prints is good enough considering the price of this machine, but as I said already to many other before you, consider this as a technology demonstrator (or a toy) and do not try to build it as your only printer.

Hello my friend,
I like your work and this was very cheap 3D Printer
I'm making one but I have a question about the programing
What firmware shall I upload to the arduino mega ??
where is it's files ??

The instructions are prepared for Marlin 1.0.0, but you can use newer one. The only change is in the way you declare electronics.
If you are using RAMPS 1.4 Instead of MOTHERBOARD 33 use MOTHERBOARD BOARD_RAMPS_13_EFB

You can download it form https://github.com/MarlinFirmware/Marlin/tags

This is so cool. About how much do the parts cost altogether?

In the Buyers guide I've posted in the comment bellow, or even here on Thingiverse is BoM with prices. It may change slightly, but roughly 100USD.

Hey Thorgal, now i had all parts printed but the BOM confuses me a little bit ^^ I want to build the 125 version but cant find a measurement for the Rods. Which lenghts does i need there ?

Tanks !
Moe

Please see the comments at http://reprap.org/wiki/ToyREP_Buyers_Guide - chapter Rods, Screws And Nuts. Hope that will help.

Thanks man !! I forgot this page... :-/ Too many projects running at the same time :-D

Thank you for asking, but unfortunately for You, I am not selling these machines. That would ruin the whole idea of building your self or someone around you a cheap toy :D. If you would buy it you won't learn anything and in case it fails, you are doomed. I really do not recommend this should be a main printer of yours.

Hey Thorgal, really great work !! I´m Printing one for me :-) But im not shure what i should use as smooth rod ? Normal 8mm steel rods ? Have you done something to smooth it ? Or do you use Prezision rods ?

Thanks !
Moe

Sorry, for misleading term. Mine are 8mm dia. stainless steel precision rods.

Thanks ! No problem bro :-) I just thought you found a cheaper way to get working rods :-)

I did some smaller modification (holes for wires in some parts, mounting holes for zip ties etc.) I will upload it if it all works and fits :-D
I also did a google docs tabel for parts with links (the aliexpress Links) If all parts are here i will also publish it.

This is going to be an awesome small printer :-) Thanks a lot for Publishing it !!!

Great, Looking forward to see it!

I heard about using steel rods, polished with high grade sand paper and printed linear bearings as sliding LM8UU would cause scratches on it, but have no personal experience with this.

Hi thorgal,

Currently printing your files and I will try to stick to your detailed manual step by step. Looking forward for the finished machine.

Uhm, can't find the STL file: ToyREP-PSU1. It is mentioned in the build manual.

Disregard!! Found it as a thing on it's own :)

awesome;) bravo! i'll materialize that! it will take me some time but will build it, keep up the excellent work and let's build together:)

kiril

I would suggest this as the coupling for Z axis:
http://www.thingiverse.com/thing:1148245

Hello :-) I have a question regarding the extrusion solution; Would the motors be strong enough if I take two of them, both with a MK8 Driver gear pulley, similar to the "dual pinch wheel" extruder?

Also, would it be advisable to take the 12v motors instead of the 5v motors if the price does not differ? Will it make a huge difference and can they handle a bit more speed, or will they just tend to live longer?

Thank you :-)

One motor is probably sufficient for direct drive extruder. Someone already wrote in comments under http://www.thingiverse.com/thing:200078 that he have dual motor direct drive extruder you described but I did not find the model for it yet.

With 12V version of 28BYJ-48 you will need at least 24V PSU (probably even 35V) due to resistance and inductance of the joined coils! Try to avoid 12V version at any cost.

Delta extruder with 28BYJ-48 stepper

Thank you for taking the time to reply to my questions!

Yes, I have seen the Delta extruder, but as there was little info on if and how good it works, and I did not see that coment! Thanks, I'll check it out.
I thought it may be better to use two 28byj (As they cost ~80 cents and I was considering a simple repstrap variant where two motors just sit next to each other with adjustable tension, and this would be easy to build.

Yikes, thanks for the heads up on the 12v version. I did a lot of reading regarding these motors and saw some tests where the 12v had more torque (https://grahamwideman.wikispaces.com/Motors-+28BYJ-48+Stepper+motor+notes), but it kind of makes sense that driving it at 12v after the uni/bi polar mod won't drive it at full potential.

I'm considering making a mashup of this and an Itopie (mdf based prusa frame) but using two steppers on the x and Y instead of the idlers. I was wondering though, why do you think you are limited to 10mm/s. Is it that the motors don't have enough strength to move the carriages faster than that or that the speed itself is the issue? My hope was if you made the printed GT2 pulley a larger diameter and used 2 steppers instead of one maybe we could get closer to a normal printing speeds goal is 40mm/s (your thoughts?)

Already discussed this with madmike8. You will find it on the bottom of the comments.

These motors can do only 15RPM reliably. With 50mm circumference pulleys that are installed in ToyREP it is 12.5mm/s. Almost all of my 28BYJ-48 steppers can do 30RPM and some hand picked even up to 60RPM when overloaded, but without good cooling they were prone to die very soon. I was running the machine like this at the very beginning, on the first video you can see the machine doing 30mm/s in X and Y axis. The major problem is how to cool down the coils that are hidden inside the motor and barely touching the metal housing ... So trying to make them spin quicker is probably not good way to go.

I was also experimenting with bigger pulley circumference (on different machine as ToyREP is not designed for such pulleys), but 75mm is already dangerously close to the strength limit. With such pulley you can do 18.75mm/s. Two motors with 150mm circumference pulleys can possibly do 37,5mm/s but you have twice as big deviation caused by the internal gearbox and you have to imagine almost 50mm dia pulleys installed in the printer.

Edit: Maybe 35BYJ-46 is better way to go instead of two 28BYJ-48. They can be found fairly cheap too.

But if you use repetier-firmware, couldn't you use the backlash compensation? It works well with my crappy 28byj-48 that have lots of internal backlash.

Probably can, but I have no experience with this function yet. But the backslash is with bigger pulleys is so huge I doubt you can compensate it just by electronics. I am talking about 3-5mm of free movement in X and Y axis.

Hi all

Well things could not be better uploaded some more pics of my printer.
just finished the 20mm test cube and all is good. just a bit of tweaking.
and we are ready to rock and roll.
Thank you for such a nice design

Happy to hear that. I see You've mount optical endstops too ;)
Good luck and thanks for kind words.

Hi!
I've got a problem, the motors 28BYJ-48 work in the same one direction. Who can help?

I think this may be related to endstops and software limits in firmware. If possible try to home first, than try to move motor CW and CCW or switch off movement limits in firmware.

Or you mean they are going the same direction no matter what direction you sent them? That would be truly strange.

endstops not installed, I took another firmware, everything moves, pick up the trial steps

Hi
I have a strange problem mainly to do with marlin I have version V1.0.0 rc2 and it won't accept the Z axis steps greater than 4 digits ie 9999 so when I put in the 40960 as in your documentation in the steps for the Z axis in the printer it just shows up as **9999 the first 2 chars are strange. so what version of Marlin are you using, is anybody else have this problem.
Cheers
Tony

do not use EEPROM option
use EEPROM will limit to 9999

Now I am using Marlin 1.0.2, but also with previous 1.0.0 I never experienced such problem.

Hi
To elaborate on the problem it's actually in marlin in the display code it is set to show only 4 digits in front of decimal point as it has a mask of 9999 so it's only a problem when you try to edit the values on the lcd I have the full graphics display.

The values are stored correctly, it means you have to compile the values in the source or use repitier host to edit on the fly.

Doing my first test prints today all seems good, in general the way it runs. So I am very happy with it
But I seem to have trashed the y stepper it has got very sloppy in the gear train, has a bit of a delay before it does it's move, swap the x and y and it moves to the x axis. So it's in the stepper itself.

It may be because I have no fan on that motor and it gets a bit warm so may have melted something inside.

So job for when I get up tomorrow, put a fan on the y and replace the motor.

will keep you posted.

Hello! give another problem. I have always goes towards the axis Y. tried different settings and change the speed. Changed driver Drv8825 clocked at maximum capacity.

https://drive.google.com/file/d/0B0bCqFasYLxYRGlQS2xlekxSb1U/view?usp=sharing
https://drive.google.com/file/d/0B0bCqFasYLxYVHFERmhMMVhJdTg/view?usp=sharing
https://drive.google.com/file/d/0B0bCqFasYLxYSUlocTJ6U0pVaEU/view?usp=sharing
https://drive.google.com/file/d/0B0bCqFasYLxYUUdlazA5dGo2ZzA/view?usp=sharing

define HOMING_FEEDRATE {1200, 1200, 50, 0} // set the homing speeds (mm/min)
define DEFAULT_AXIS_STEPS_PER_UNIT {42, 42, 2500, 400}(3263.683958)/0.8, (3263.683951636)/(97*3.1415926)
define DEFAULT_MAX_FEEDRATE {22, 22, 0.25, 4.17}
define DEFAULT_MAX_ACCELERATION {300, 300, 1, 17}
define DEFAULT_ACCELERATION 150
define DEFAULT_RETRACT_ACCELERATION 17
define EXTRUDER_OFFSET_X {0.0, 20.00}
define DEFAULT_XYJERK 0.5 // (mm/sec)
define DEFAULT_ZJERK 0.01 // (mm/sec)
define DEFAULT_EJERK 0.1 // (mm/sec)


I think it's because of the backlash compensation on the motor 28BYJ-48. Where is it in the firmware onastraivaetsya?

Did you tried to print something less complicated like the 20mm calibration box? How did it come out? Do it get right dimensions in X and Y axis?

This seems to me like motor missing steps due to overcurrent or overheating.

In the line with DEFAULT_AXIS_STEPS_PER_UNIT you are missing comment // before ( - that may cause some problems too.
Also, how is it possible that you have only 42 micro steps per mm in X, Y axis and 2560 in Z axis? Even without microstepping there should be something around {102, 102, 5120, 30.8}.

Thanks, all earned. I had a bad motor, changed the play and was gone. Who exactly prints !!

You have to stick the hobbed bolt inside extruder the other way around.
http://reprap.org/wiki/File:ToyREP-61-Hobbet_Bolt_With_Gear.jpg
http://reprap.org/wiki/File:ToyREP-68-Installing_Hobbed_Bolt.jpg
http://reprap.org/wiki/File:ToyREP-70-Installing_Extruder.jpg

Pick the right gear for your hobbed bolt from thing files per the distance of groove to hex head. Use washers to align filament and the groove. The filament will lock the position so there is no need for nut on the other side of the extruder and you will avoid this kind of collision.

realized. Thanks for the help!

This printer is really awesome. Thank you for designing it. I will build one over christmas (Or even sooner), but i have a question: Do i need 12 V or 5V motors? Also, where can i find a Bill of Materials (BOM) for a little shopping :) It would be great if someone could help me!

Thank You.
If you are going to use 12V PSU use 5V motors. For 24V PSU use 12V version of 28BYJ-48.
BoM is included here among thing files or on webpage: http://reprap.org/wiki/ToyREP_Buyers_Guide
Good luck.

Thank you very much! When my printer is finsished i will post a make!

UUUUUUUUU, maybe thats the point why i can´t get it to work...

You cannot move motors at all? With 12V PSU you should be able to move them at around 3-6mm/s instead of 12mm/s. If not, the problem is probably somewhere else.

Hi. Can you change the frog to mount heating bed 120x120 mm? I bought here http://www.aliexpress.com/item/Free-Shipping-1pcs-lot-3D-Printer-Aluminum-Plate-Hot-Bed-With-PCB-Heating-Bed-120-120mm/32496687064.html
Heating bed size is 120x120 mm, the distance between holes 112 mm. I can do it myself and put it here, if you share drawings of the mountings bed. Here is photos of the bed https://onedrive.live.com/redir?resid=F1E6EF158260DA0C!70611&authkey=!AF-KypcbMY1ANLg&ithint=folder%2cjpg

Yes no issue, I will prepare it next week.

BTW. Great work!

Edit: Done and uploaded here.

Is it possible to use nema 14 steppers on the mounts??

No, they are way different from these small steppers.

Ahoj, do you have any picture with Bowden Extruder ??
http://reprap.org/wiki/ToyREP_Build_Manual#Bowden_Extruder
Would like to try test this version...

Regards, Ondra

Hi,
Sorry, I am not using bowden extruders, so I did not even bought bowden hotend to make pictures. Technically it is the same mount as for geared extruder. It is going to hold on carriage just the same.
Of course you need your own bowden extruder, probably the best with Nema17 motor to make it run and find some place to attach it to the printer. Be my guest and try it ;)
Regards
Martin

Hello I am very like the design of the printer. Is it possible to add a cooling Print PLA plastic
(small 30h30h10mm FAN)
roughly the
http://www.thingiverse.com/thing:17768
http://www.thingiverse.com/thing:40463

Fan Duct for Ultimaker
by Maurits
Hybrid annular fan duct

With those two you mentioned would be problem with attachment. They have holes for horizontal mount (under the heatsink or carriage), but my hotend cooler is designed for fan and carriage placed vertically. You can still try to glue it on the right place, but I have no personal experience with such setup. Something to cool down printed PLA is still on my list of improvements, but do not expect it being ready and published before Christmas.

Edit: It took more than two months but fan duct is finished - http://www.thingiverse.com/thing:1276315

ToyREP Hotend Cooler
by thorgal

I love this little design. I was wondering if there is a chance for you to upload the orginal cad/3d files you used? I would like like to make a few changes, but it's a real pain to to get these into a proper cad format at times.

Sorry, I won't release full dwg publicly. But I can send you few parts you want to change if we will make mutual agreement. If you are still interested please write me PM with email.

I'm very interested by direct drive extruder, when will you do it ?

Sorry it takes so long. I am still working on direct drive dual motor extruder. The issue is that it is very dependent on size of vitamins. It would be best to make it parametric, but I have no skill in Openscad. Prototype should be ready in February.

ok, i will wait, thank's

You use 40 w power supply heating nozzle to 200 degrees Celsius to how much time?

The same as with any other stronger PSU. The ceramics heat cartridge is designed to draw 40W - 12V, 3.3A. As long as these requirements are covered there is no difference in its operation.

If you want to know how much time it takes to preheat hotend to 200°C (no matter what PSU is used) then I can measure it next time.

So with the longer legs, is it now possible to use an LCD screen?

I made this thing to mount lcd on M8 orizontal bar: https://www.thingiverse.com/thing:2294654

You need to add some M/F jumpers to reach the ramps because by using the original adapter you'll forced to have longer legs

RAMPS "LCD & SD-Card reader" M8 rod mount
by 3D7

Yes, I am able to use a LCD with the leg extension upgrade I made here: http://www.thingiverse.com/thing:1098421
The 50mm legs should be about the same offset as my extensions.

ToyREP Leg Extender

Yes I thing so, but it is not checked yet.

Hey,

once more me. Currently I am fiddling around with the motors:

Could you please tell me (us) what cable color (from the motor) you connected to 1A 2A 1B 2B at the A4988. I did the hack -> turned down the voltage to absolute minimum. When going up with the voltage the humming gets "stronger" but somehow I cannot get the motor to move.

BTW: Maybe someone is interested: The "easydriver" wont move the BYJ28-5V motor at 12V you need to "step" down to 5V. If you use the A4988 it works with 12V (but as said above, currently I can not make the motor to move).

Thanks.

It should be blue-yellow-pink-orange on 2B-2A-1A-1B

Hmmm, strange, I tried your suggestion. But this is what I initially had. I found on the German RepRap Forum following suggestion that worked for me ...

2B => blue
2A => yellow
1A => PINK
1B => ORANGE

Do not ask me why this is working :) but the thing turns like a dream now ...

What you found on the RepRap forum is exactly what i wrote in the comment above?!?

Yes I know. I probably did a mistake while reading --- :) sorry for the confusion. The wheels are spinning now!

Hey,
could you post a better picture of endstop wireing on Ramps? Isn´t it Ymin?

Try the build manual on reprap.org. The pictures are in quite big resolution if you open them separately. And it have to be Ymax no matter what would be shown on the picture.

Could anyone suggest somewhere (in the UK) to get the parts printed cheap? as I don't have a 3D or have access to one and I can't afford the £100+ most places are quoting : (

Also thank you Thorgal for this project I can't wait to make it! : D

It looks like your designed supports are 0.38mm wide which is less than most peoples nozzle diameter. Rather than print all the parts at less than nominal settings, can you provide the STL's without any pre-designed support material? Thanks.

I will, but please hold on for few days. During weekend I've burned motor on extruder. It is second one since ToyREP was published. I have to check up what's wrong.

Thanks. Maybe you have the current set too high?

Motors came in today... Time to Print it. My nephews have been wanting a printer, so I'm going to give this one a shot for them.

Glad to hear that. Good luck and keep us posted. I hope they will be satisfied, but beware as we've got a saying in Czech: With food grows the taste :)

Morning. Could you do a pulley for 2GT belt?

I see You've been busy last night :)

I can, but question is if you can print it as it will have very fine teeth?

I would like to give it a try if its not too much work for you.

Ok. Only it will be prepared after weekend if you don't mind.

How wide is your belt?

all my belts are 6mm wide

The files with wider T2.5 and GT2 pulleys are ready. They are uploaded here. You should find them at the end of thing files.

Next Question: Why did you use so many washers. $Washers/Nuts would also work in step 2 on the PSU!?

Because in between Z2 and PCB1-1, you would not even fit nut. So there are just 3-4 washers. With PCB2-1, there may be a bit more space but still it would be quite hard to manipulate the nut.

On the other side between Z1 and PCBX-2 there can be two nuts and washers instead of so many washers, but I found it easier for tightening when setting the correct distance between Z1 and Z2.

Anyway. Anybody can do it as he likes.

What about the 5VDC motor in your x setup? Are you working with that type?

5VDC motor? What for?

In your pictures there is a 5VDC motor as X axis motor...

AHA. Well that is not DC motor but my 28BYJ-48 stepper motor with 1:64 gearbox rated 5V when unipolar. All of them are the same and have the same stupid misleading sticker. :D

Thanks a lot for all your work. I´ll post my results soon. Have you thought about a fan for filament cooling yet?

I am looking forward to see it!

Fan for printed filament is on my improvement list. But I did not decided yet if I want to use hotend fan also for filament or a separate one.

Do you think it is possible to connect all the fan wires to the same power connector on RAMPS (where the PSU is connected)?

You are reading my mind, right? :D

I had it this way. Respectively, cables from PSU to 5A clamp on RAMPS, then from this 5A clamp to 11A clamp and from 11A clamp to fans. It was working great, but it is not good solution for PCB heatbed. With all fans (I have 4 mounted on the machine) connected to 5V USB charger I was able to get 106-110°C, but when they were connected to 12V from PSU, it was barely 90°C. Although my fans are only 7 CFM, they are moving air way too much around heatbed.

Did it this weekend, because I wanted to get rid of the USB charger. Now I am going to do some measurements, if all of them can be actually connected to 5V for endstops on RAMPS or what resistors are needed to reduce voltage from 12V.

Maybe you can adjust the voltage a little bit higher on the PSU? On my MKS 1.1 it affects the heating temperature and speed.

Hello,

Just a minor question: What micro switches do you use (I am not sure with the 11mm hole spacings): Will KW11 switches match?

Thanks.

Thank you very much for pointing this out. I am sorry, there should be 9.5mm instead of 11mm.

I am unaware of the type of mine but KW11-3Z-3B or KW11-3Z-3C should work with no issue.

Edit: Size corrected in Instructions and BoM.

Hi, I am just wondering about the 28BYJ 48

Am I right that you are using the 5V version? How do you step down from 12V to 5V? I am currently trying to move my 5V stepper with an easydriver module. If I connect it to 5V there is no problem in moving that thing. However, if I connect the easydriver to 12V and turn down the current to lowest as possible it wont budge (not even a glitch). Also no help increasing the current :(

Am I missing something here?

Hi,
I am not going down from 12V with 28BYJ-48. When running bipolar you joined two coils together, so it have twice that big resistance. Means it should run at 10V instead of 5V and 12V is not too far from it.

What you describes may be caused by uninterrupted trace at the PCB of 28BYJ-48 when doing the bipolar hack. At 5V the motor is still somehow running, but at higher voltage, this short start to throw over current fuses of the stepper driver.

Hi, thanks for your answer. I think the problem is the Easydriver itself. At 12 V I can not downregulate the current to a reasonable level. However, as soon as I put a LM317 and go down to somethinh like 5.1V for input voltage I can run the motor at very good rate.

I think your drivers have finer voltage (and current) regulations? A shortage is almost impossible because I nearly cut the PCB at ca. 1mm ( are there any connections at the back of the PCB that could lead to shortages)?

Just for confirmation: After the "hack" I do have 65 Ohms between the wires for each coil - do you red similar values?

Thanks.

No, there are no traces on the back side as far as I know. If we rule out the PCB than I am clueless.

And yes, each one of the four coils should be around 60-65 Ohms.

I've tried this motor on Chinese clone of DRV8825. Its regulation is tragic - especially for low current. EasyDriver should be much more suitable for this small motor.

It is strange that it is so much dependent on voltage as stepper driver is supposed to controlling current independently on voltage. There should be two voltages delivered to the stepper drivers. One Vref for logic of the driver 3.0-5.5V and another Vbb for the motor. Easydriver have its own voltage regulation so it can prepare voltage for logic side from load power supply, but also have pins to connect own 3 or 5V separately and some pads to switch between different scenarios. I think that for some strange reason your driver is feeding load power supply to logic part of A3967 chip, so it won't work above 5.5V as it gets overvolted. Strange it did not get burned yet when hooked to 12V. :D

Hey guys...I was just wondering is there link or download file of a breakdown sheet on how many parts of a certain part? Thanks

Please see the page: http://reprap.org/wiki/ToyREP_Buyers_Guide

And also here, among thing files, is uploaded BoM.

The built-in supports on the X ends and extruder didnt work with cura 15.04.Anyone else have this problem?
Ive got most of the plastic parts printed just waiting on some bits and pieces.love the design.

Yeah, same problem here. I was able to print the carriages fine without support, but the extruder I had to re-design in sketchup.

Sorry to hear that. They are only 0.35mm thick, maybe that is the problem, although Slic3r don't have any issue printing sub nozzle single walls. I can either upload models with thicker supports or models without them. Unfortunately I have no experience with Cura. What you thing will be better way to go?

I removed the supports in sketch up and printed them with cura's support. X carriages came out fine.

Hi, I found the Hobbed bolt, https://goo.gl/photos/fjMrJjG1YH2CtMGY6, and https://goo.gl/photos/ce1FyTHrzu95du4u8, can you tell me which one is best? Thanks.

23mm from centre of filament to inner side of the hexagonal head is the right size.

Thank you!

I found you had updated many parts after I had printed the extruder v1, can I use the extruder V1 with the new X1, X2 parts?

That is a bit more complicated. You need the most recent files if you want to use stepper motors with four mounting holes or 40mm fans to cool the steppers. Other that that there is no difference, you can combine them as you want.

I can't find the 28BYJ-48 stepper motor have 4 mounting holes, and can you show me? Thanks.

Please see the last picture in the gallery. There is drawing showing the difference between common 28BYJ-48 and the one with four mounting holes.

There is also sobo84's comment from 28 August with link to Aliexpress.

If you have such motors you will have to print revised models. Judging by the way you are asking, you do not have this problem so it is only a matter of fan installation - if you want to cool the steppers or not.

Ok, I got it, I will use the 2 mounting holds motor, it is easy to be found here. I have printed some parts, https://goo.gl/photos/YGSpDcByLizAM6iU6, I think it will take about 2-3 weeks to print the all parts. I will keep reporting my progress.

Thank you again for your design.

You are welcome and I am looking forward to see more pictures and hear about your experiences with ToyREP.
Good Luck

I am using your updated file for X1 and X2 that using M8 threaded rod, but it seems the M8 threaded rot can't be installed to Z-joint, do you think that is OK?

It should fit. Inner diameter of Z-joint is exactly 8mm and M8 Threaded Rod is actually a bit smaller than 8mm. Try to open the Z-joint wider or file the hole.

Yes, I printed them today, and I have printed many parts today, you can see them in my MAKES.

Thanks.

That is a very great design, and thank you for sharing. I want to make one following your instructions. What is the design software do you use for this project?

Thank You. It is based messy hand sketches and then drawn in AutoCAD. I am used to this.

Is there any way that a dual extruder could be used with this printer?

Technically it is possible. But without hotend like Cyclops from E3D, you will lose a lot of print area. With two common E3D v5 clones I would suggest 50mm longer rods in Y axis direction and to place two bowden hotends one on a other. Their fan mounts have to be tightened together by M3 threaded rods as such length of M3 screw is not available. Only one 40mm fan will be used, it will be common for both hotends. You will have to figure out where to put extruders and eventually make an insert for bowden fitting that will change 28BYJ-48 short distance extruder into bowden extruder.

How hot does the extruded get?

I don't have any means to measure it. But judging by the fact, that the carriage did not melt so far, even with 40mm fan at 5V it have to be way under 100°C.

I cb at find a PSU the correct voltage anywhere. Where did you get yours?

Mine is coming from one Czech eshop called t-led.cz, but they are not shipping abroad. Please see my reply to bobaxx 6 days ago or Tempest815 from from 8th August. One link is to ebay and other to aliexpress. Hope that will help.

Hey, just one more thing-- The hobbed bolts I'm finding ( http://www.aliexpress.com/item/3D-printer-accessory-Wade-s-extruder-hobbed-bolt-reprap-M8-wire-feed-teeth-space-1mm-for/32387428107.html ) aren't like the one in your pictures. I saw one on eBay, but it was $10. Can you post a link to the one you bought, or will those work?

I know. My hobbed bolt is one of few other things I exploited from previous printers. The design is called Hyena and is quite expensive. I bought it on RepRapWorld few years ago.

But the one you are referring to should be just fine. The only thing that is important is that the hobbing have to make a slight groove. This will prevent the bolt from sliding out of the extruder as it do not have any nut on the other side.
The distance from centre of hobbing to inner side of the hexagonal head should be roughly 21-23mm. Either you will use the head of the hobbed bolt or you cut the hexagonal head off and place nylock hex nut or two hexagonal flat nuts on the threaded side.

Or If you don't want to cut anything, please measure the distance between centre of hobbing to the inner side of the head and I will upload another gear that will use the head without any other adjustments. That is probably the best solution, unfortunately as they are not providing any datasheet I cannot do it right away.

Edit: zhangzq71 provided 3 most common hobbed bolt spacings from centre of the groove to inner side of hexagonal head. They are 23, 26 and 30 mm. New extruder gears are uploaded to ToyREP files, so there is no need to cut anything.

I'm on an issue to make it foldable but i'm stuck with the fixation on the chassis.

ToyREP is not foldable. If you are talking about the article at 3Dprint.com, originally there was this intention, but this feature was removed from final version for very good reasons.

If you would like to redesign ToyREP to become foldable, be my guest. I gave up on this.

I know that it's not foldable with this setup but i'm sure I could make it with only few changes. I did'nt see the article, i gotta take a look on it.
I'm already working on it and I'll let you know as soon as it's done :)

In the article is nothing important, but it was the only place where I mentioned that ToyREP started as a foldable Printrbot. :D

I'm really interested in your solution. My approach with bearings was way too complicated ...

Do you need some support from my side?

I was thinking to bearings too, but I gave up because it was too complicated and it should have too much mechanical slack.
I tried on a 3d model with an alternative of your Z parts divided in two to create a pivot with threated rods. If i'm stuck with it I'll let you know but I guess it will work :) Just need to work a little bit on those alternative Z parts.

Hey ! Here are my first experiments, it works great :D
I don't have all the parts yet, but it seems to be awesome !!! Just have to change a bit to lock the printer in "folded" mode.
No mechanical slack. Here are the photos :
http://bit.ly/1JLEvLh
http://bit.ly/1XfkF0s

Great! Thank you for photos. I'm looking forward to see it working.

Just one question. Is it possible to bend the Z frame to the other side too? Because eventually you will have X motor and whole extruder mounted there, where it lays flat on the table now. I know it is not important while carrying the printer, It is just for the case you would like to store it folded.

Anyway very good job.

Thanks mate. Yes it's possible to do it. The extruder isn't mounted yet but i did some tests on a 3D sketchup model and i'm pretty sure it will work. Maybe it will be necessary to make the legs a bit longer, I'll let you know as soon as I can mount it.
I'll update with some photos later !

Did you ever get that working?

Is there any way to connect an LCD screen to the printer?

Depends on electronics you use. For eg. Arduino Mega with RAMPS can use LCD and SD card reader. But with ToyREP there are few issues. First, you will need at least 5-6mm longer legs to hook up LCD cables. Right now, there is not enough space under the printer. This is on my list already. Second, you would have to make/download/adjust some enclosure. I may prepare box for LCD module with display 12864 (because this one I have at home), but it is not on my priority list.

What about this one:
http://www.aliexpress.com/item/Promotion-3D-Printer-Kit-Reprap-Smart-Parts-Controller-LCD-Module-Display-Monitor-Ramps-1-4-LCD2004/32314567032.html
No idea how it interfaces with the firmware, but it claims to be completely headless including the calibration.

I think it is clone of SainSmart LCD 2004 Controller: http://www.sainsmart.com/sainsmart-smart-lcd2004-controller-with-adapter-for-reprap-ramps-1-4-3d-printer.html. It should be supported by Marlin firmware. It will let you set temperatures, home the machine and chose gcode file from SD card to print.

It have the same connector to RAMPS as my LCD, so the same issue with length of legs applies here. Other than that, just find your self some holder here on Thingiverse and you are all set for autonomous printing. ;)

I've uploaded some schematic drawing. There are actually two types of 28BYJ-48 motors. They are almost identical, except the mounting holes: http://thingiverse-production-new.s3.amazonaws.com/renders/6d/99/73/f6/cf/28BYJ-48_preview_featured.jpg

In blue is the more common motor with only two mounting holes, in black yours with four mounting holes. I can adjust the models so you can mount both types, but it will take me some time, and as I don't have them home, I am not sure about the result. Sorry for inconvenience.

Or you can drill two more holes in the position mentioned in blue. They don't have to be 4.2mm dia. 3mm would be enough. There would be just very little metal around the hole, but it should suffice. You will also have to put some washers under mounting holes of Z and Y motors.

A fix/adjustment would be great. I´ll print the other parts first. Thanks for your commitment.

Fixed parts are ready. See http://www.thingiverse.com/thing:981794 use them instead of the parts uploaded here.

For your motors use Extruder_Rev3 (Rev2 is for common type of motor with fan). This is quick and dirty fix. I will have to remodel the whole extruder body one day, to make it uniform. Other parts are compatible with both types of motors and fans altogether.

Please let me know how it worked.

ToyREP Stepper Motor Cooling
by thorgal

This is great. Its on my list now! Thanks for sharing!

Will an 1/8" polycarbonate plate work instead of a 3mm?

Sure, thickness or even material of the plate does not matter. It just have to be sturdy, flat and in case you want to print directly on it (without any glue), it have to have certain properties so ABS (or PLA) would stick to it.

Thanks! Any idea what kind of material would stick?

This is a question for a separate article. In short. Most of the plastics have two characteristic teperatures. Glass transition temperature (Tg) - when the plastic becomes sort of elastic, but do not flow yet. And melting point temperature (Tm) - when it starts to flow.

Glass transition is used when printing on heat bed - it have to be preheated above this temperature. For PLA this is around 65°C and for ABS 105°C.

Melting point temperature is most commonly used as a hotend temperature. For crystalline PLA it is 175-180°C. With amorphous ABS it is a bit more complicated, depending on the pressure. Temperatures above 200°C have to be used in 3D printers. In both cases the hotend temperature actually have to be a bit higher. You have to heat the previous layer with the actually printed layer above the melting point temperature so they will bound together.

Now when you are printing first layer on plastic cold bed you are using similar principle. The first layer have to be printed at a temperature that will just bound, so the printed model will hold on the plate, but not too high so it would let you remove the model from the plate when print is done! For this reason, Plexiglas (polyacrylate or rather polymethylmethacrylate) with melting point around 160°C or polycarbonate (PC) with melting point at around 155°C are commonly used as cold beds. You can use any other material, with similar melting point (for ABS it may be even a bit higher). Just check its datasheet.

Anyway, you will have to experiment to find the right temperature for the first layer with your printer. As you can see on my videos the cold bed surface get damaged over time.

I can not find stl for PSU.

PSU covers are separate thing: http://www.thingiverse.com/thing:947676 . Please see Thing Info page, it is the first upgrade among Improvements.

ToyREP Power Supply Cover
by thorgal

This design is very interesting to go everywhere with your printer ! I'm already printing the parts with my first machine and I'll probably do some adjustments for a front LCD screen
. I'd like to install a 80W power supply for a hot bed, but after hours spent on internet, I can't find one small enough.
In fact, the minimal size I've found for a 80W PSU is 159x99x38.
By looking at the photos, I see that the Y motor is close to the PSU. Is there any place for those extra 3 cm or do I need to make some adjustments?

Thank You. I took mine for a walk already. Maybe I will add some legs with wheels to the to-do list :D. LCD would be great upgrade too!

It do not have to be strictly 40W or 80W ... Look for eg. here: http://www.ebay.com/itm/LED-Strip-Light-LED-Driver-Power-Supply-Transformer-AC-110-260V-DC-12V-/161769013165?var=&hash=item25aa2f3fad - see the Specifications in the middle of the page. The 60W, 100W and 120W PSU have the supported sizes. And this is not the only eshop that has them.

You can definitely manage, but I was trying to keep it as small as possible. For 159x99x38mm PSU you would have to use PSU2 setup (129x99x38mm version described in Instructions - Frame) and make all rods parallel with X axis 30mm longer. That means threaded rods for legs would have to be M8x210mm 4 pcs, X axis threaded rods M8x295mm 3 pcs and X axis smooth rods dia. 8mmx285mm 2 pcs. For Y axis it will remain the same - threaded rods M8x265mm 2 pcs and smooth rods dia. 8mmx245mm 2 pcs. The distance between Z1 and Z2 would be 160mm. The distance between Z1,2 and front legs will stay 128mm.

My only concern is about Arduino and RAMPS. I am not sure if such long PSU would have mounting holes close enough, so the PCB mounting can still hold electronics on place.

Thank you so much for the link :D
With this kind of supply, I guess I will be able to add easily some LED and a 100x100 hot bed (I already saw that size on an italian website).
Just bought the rods in a local warehouse. Still need some stuff to build this tiny/useful machine.
And btw, wheels and legs would be an awesome upgrade :p

You are welcome. Check the heatbed first. There are some 100x100mm beds that have significantly higher power consumption. I found 50W or even 70W. But with 100W PSU it should not be an issue. Also the mounting holes may be in different position.

Good luck and send pictures! I would like to see your upgrades.

Sorry, I think they just have wrong dimensions in the package description. MK1, MK2A and MK2B are only 200x200mm as fas as I know.

Can you post some test prints? Love to see what sort of quality this can produce.

Trying to put here some examples. My favorite so far is treefrog: http://www.thingiverse.com/make:155032
For more see the Thing Info page.

Treefrog
by thorgal

I will try later this week. Do you have any suggestion what shall I print?

Thorgal, Question for you. If I try a 40 tooth pulley (Diameter slightly less than 32mm) Is there room for it on the X axis? 30 Tooth would be about 22mm if i remember correctly, could be another option. It looks like the Y axis has room...

For 30T or even 40T pulley X1, X2, Y1, Y2, Carriage and Frog would have to be changed. I'm sorry, I left no room for such upgrade because I was trying to make everything in line and compact.

By end of August I will get 28BYJ-48 motors with 1/16 reduction gear instead of 1/64. I am going to test them and I will let you know about the result. That may solve the problem with speed and it is going to be much easier than changing so many parts.

Thorgal

I am building one of these for my classroom, it's a great design. I have a question about the motors. The 28BYJ-48 seems to come in two flavors, 5 volt or 12 volt. I would have guessed that you would use 12 volt, but I see there are other designs that use the 5 volt one. Does it matter? Which do you use?

In the video, you can read 5v on the X axis stepper.

I see. OK, but I still wonder if 12V would be better for perfrmance and longevity. From what I can see they cost the same.

I am using 5V but when changed to bilinear they are actually 10V. 12V 28BYJ-48 motor were quite expensive, when I was looking for them. 12V motors would require more current and are going to be stronger. In theory, they should run cooler, but as I never had chance to test them. I cannot tell for sure.

I might give them a shot. I can get either from Adafruit for the same price. I know I can do better on ebay, but I distrust the sources.

Sobo84 ran into a problem that some 12V 28BYJ-48 have different mount with four holes. If this is also your case, please try use models from here http://www.thingiverse.com/thing:981794

ToyREP Stepper Motor Cooling
by thorgal

The Adafruit steppers are a perfect match for your original model. I have them all mounted and ready to be wired up. The only problems I am having are of my own making. I wanted to use gt2 pulleys, as I have an abundance of gt2 belt. The pulley I made for the Y axis works wonderfully, but not so much for the X axis. I think my efforts to refine the pulleis went to far and I have to try another. I'm also going to try this http://store.cutedigi.com/cnc-kit-for-arduino/ as a controller. I have to build a circut for the heater and thermistor, but that is not difficult.

However, Ischool has started and I have inherited some new responsibilities and my grand daughter is (as we speak) making her way into the world, so I'll have to put this aside for a bit.

Did you have to hack the Adafruit steppers? or did you leave them unipolar?

Hey, huge fan of the printer. I'm working on my own build now, I'm just having trouble sourcing some parts. The power supply most notably. Any tips for those of us stuck in Americaland?

I'm glad you like it ;)

I cannot help you with stuff available in your country, but the PSU I have is similar to this one: http://www.aliexpress.com/item/New-Arrival-AC110V-220V-to-for-DC-12V-3-2A-40W-Voltage-for-Transformer-Switch-Power/32305327778.html

It have different mounting holes, though. That I can adjust in few days. Just send me their position.

If you are not fond of stuff bought directly from China, try to look for 12V 40W LED PSU in US. If you will find anything that would fit somewhere between 111x78x36 and 129x99x38mm, send me the datasheet and I will prepare new cover, no issue.

The last possibility is to use common notebook PSU. For that I can prepare some socket mount. That would be also interesting upgrade in which I'm willing to participate.

This will fit in the printer, but it is smaller than what I used (if the size in description 90x70x35mm is correct). I am not sure about mounting holes and definitely there is no cover designed for such small PSU. But if you can do some adjustments to files that are available, this is possible to use.

You are welcome. This one should work fine. Don't forget to use the longer set of rods.

ok. Thank you.

Comments deleted.

I've got tons of notebook power supplies around. I think you said a theoretical heated bed would need 80W probably max 120? I'll look around, I might have something in that neighborhood.

For clarification:
My calculation was such, that if for 200x200mm headbed 144-166W is required to reach 110°C. Than for 100x100mm roughly 36-42W and for 125x125mm up to 64W should be just fine for possible future PCB heatbed, that is not designed yet.

Hotend is quite sure. It have consumption c. 40W at 12V when preheating (If you will use 6R8 resistor instead then at 12V 22W is be just fine) and just 9-16W when hot end is preheated and machine is printing.
As I would recommend 40mm fans to cool down the X, Y, E motor and hotend, 4W more should be good idea.
The motors have consumption 1.2-1.8W each and I think Arduino is runnig from USB power supply.

So 44W (or 40W PSU with voltage set to 11.5V) is just fine to preheat the hotend while fans are running. If you are impatient and want to move the motors while preheating, 10W more is good idea. Means total 54W in worst case scenario for the machine as it is now. It will consume only c. 30W when printing, so there is really no reason to look for bigger PSU.

And if anybody ever will make PCB heatbed (I'm pushing my friend, but need him for some other project right now), than the up most limit for printer with 100x100mm heatbed is somewhere around 96W and 125x125mm heatbed 120W at 12V. But If you will always preheat hotend first, than the PCB, and then start the print, you can go as low as 70W respectively 95W.

So make up your mind, what is the best set up for you.

You know about this? Might make it easier on gears.

28BYJ-48 - parametric timing pulley

http://www.thingiverse.com/thing:164185

28BYJ48 - timming puley

Thank You for this link. I am aware of the one for NEMA 17 motor not this one. Do You have personal experience with this?

I had tried using it to print gears for GT2 belts a while back, and the teeth didn't grip the belt well enough. T2.5 should work.

That is pity, but with GT2 it is sort of predictable result as the teeth are really small. It would require 0.25-0.30mm nozzle and very fine layers of PLA to print it. That is also the reason, why I did not even bother to prepare it. Mine T2.5 pulley is not parametric, but it is working (for me at least). It is designed with 0.15mm clearance for overflow (as all other parts). Definitely you can try the parametric one if mine would not work for you. Make it T2.5, 20T, 6.5mm for belt, 7.5mm total height with flanges and 2.25mm distance from pulley to motor to fit it into printer.

I was thinking about making this printer parametric, but have no skill in OpenSCAD yet, so rather chose the easier way for me. May be I will give it a try one day as speed upgrade with bigger pulley is on the todo list :D

First going to test stepper motors with 1:16 reduction gearbox. Already ordered them from China. Found source $1.3 each so even cheaper than the 1:64 I'm using now. If they will be strong enough, the printer can run up to 4 times faster as it is.

Keep us posted on the 1:16 gear steppers. How are your current steppers holding up? Extruder? I've been waiting to see if you solved the heat issues.

With 1:16 geared motor I was not successful. They are barely moving the printer. I did not find correct specifications, but what is stated on some web pages is not true. This motor do not have 30mN/m!

So 40T pulley along with all required changes is back in game.

Comments deleted.

Geared extruder is designed so the motor have similar turning speed as X,Y motors when printing (0.5mm nozzle, 0.35mm layer). I sort of solved the heat issue by reducing the X, Y speed to 12.5mm/s (15RPM), adjusting current and reducing acceleration to {150, 150, 1, 17} and rising jerk to 5.0, 0.01, 0.1. With that I was able to finish whistle in 1 hour and 45 minutes but still was not happy about the temperature (worth noting that interior temperature is here close to 30°C now). So I glued additional 40mm fans close to the X, Y and E motor, to be absolutely sure. Seems it is safe to print 15mm/s and move 25 mm/s because when I was looking for these speeds the machine was making calibration boxes for more than 6 hours in a row without an issue. Than I was busy with build manual so last week machine was disassembled and not printing at all.

I will try to rise the acceleration a bit, because it is quite funny to see the machine accelerating, decelerating and retracting now :D. Than l am going to feed it some really long print and make another crappy video of current setup with my cellphone. Any suggestions of a model to print?

1:16 motors should come somewhere around end of August / beginning of September.

Very, very, hot!
I made it !

Pictures please 8-).

This nice but can you upload the ardiuno config file also please

It is there (for some older version of Marlin), but use it with caution. Also made some comments in Instructions, because it is not as easy as just coping the Configuration.h. Each motor is behaving differently and will allow different speeds.

Thank you i know i have to make adjustments, but if i have the basic i can work with it to start. Thank you. Super great project.

I have bought same steppers and didn't succeed. Now with your files i can built my kid small printer. Thank you. !!! :)

You are the most welcome ;). I would like to see the result if possible.

Just sorry for the missing BoM. It will take me a bit more time. But if you are experienced, you can do without it.

Interesting -- I wonder how long the motors will keep up even under low speeds in the long run.

You can probably drop the two top bearings for the threaded rods without any problems. I hear having those generally cause any wobble in the rod to act on the X axis instead of just having the top of the rods wobble a bit.