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


Cooling Duct that works!

by doctek Jan 13, 2014
Download All Files

Thing Apps Enabled

Please Login to Comment

I have a 24 volt machine. Any thoughts about possible fans? I know I can't find one I think I want to use on Amazon. I had tried a different design, got it all hooked up, used an axial 30mm fan and cooled the hot block 25 degrees in a few seconds, and the software decided that I probably had a thermal runaway. It aborted every print until I turned the fan off. I have given up on one piece designs and now believe that my only hope is a two piece design - one piece to cool the area above the heat break, and a different fan mount, perhaps a long screw to attach to one end of my pseudo-wade mount on my mini-kossel. Then some kind of ring that is around and even (just above) with the nozzle.

But your ideas make sense, and kisslicer is able to deal with throttling the fan for part of the print so you can get good bonding on infill and still do plenty of cooling on the loops. So all I need is the right 24 volt fan. Or I need a light 24 volt fan and I need to modify your inlet to fit it.

I just purchased two of these and paid for the tracking and <20 day delivery. All of that came at half to one quarter the eBay price and not their modern 30-45 day shipping nonsense. I have turned away from eBay this year except for rare stuff because I am tired of 30-45 days of waiting OR I have to find it from the USA and pay 2 to 5 times the price. Aliexpress fills the void eBay left.


That is what I bought and they are 1 dollar a piece and the two I purchased, with shipping and tracking, was $3.58 which is about what one on eBay would cost me plus waiting for less than to equal 20 days (normally from them is 12 days to me) instead of 6 weeks on eBay.

I had the same problem, machine was originally supplied using two 12V fans in series, my quick and dirty solution when I needed to convert it to a single 12V fan was to connect it to the 24V output but limit the maximum PWM to 50% which it is surviving just fine.

Axial vs. centrifugal argument makes sense. Printed and now seeing the difference. Thank you sir.

Going to try this on my I3, looks great, a strip of SMD LEDs would be the SHIZNET.
Only question is how I'll mount it.. LOL.. Ordering the fan today!

My fan runs between 35% and 100%. I use a Ubis Nozzle, and have used a J-nozzle as well with no problems. Since I don't know exactly what fan you have, I'm not sure if those settings will work for you or not.Using Slic3r, it's easy to limit the range of the fan speed, so you could try that.

I gather that your cooling was OK once, then started acting up. If that's true, I have a few other thoughts:

  • My fan requires frequent adjustment to keep if just slightly above the print. I usually check the adjustment before each print.
  • If the nozzle or heat block gets close to the duct, it could cause some distortion that might interfere with the air flow.
  • The duct may have moved and not be centered about the nozzle opening.


What percent do you have the fan on with this duct? Im using this one for quite some time and im noticing now that when the nozzle is close to the bed or i print big areas lots of cold air is hitting the nozzle by bouncing back and cooling the nozzle quite significantly.

I used to have the fan set to 80-100% but not that im testing it manually I cant get higher then 40-50% without reaching the capacity of the heater in the nozzle.

Just wondering what you guys' settings are?

PS. No wonder I was always hitting thermal runaway :D

I have now remixed your great design to be used with dual extruders,

I separated the parts to make for easier printing, probably need some file work before super glueing the parts together

part cooler for use dual extruder

Could you provide a noun? Seriously, I'm not sure what you mean. The fan I use is indicated in the Instructions.

What kind of is required for this?

I love this design and had hoped to make a dual extruder version to fit the Flashforge Creator pro. I just couldn't get the clearance right and gave up, but I was trying to kludge ideas together in Tinkercad. What did you use to design it?

The design was done using FreeCAD. The design files are included in the Files area. Basically, I created the profile of the duct than swept it using a circle as the sweep path. You could change the sweep path to be an oval to make room for dual extruders.

Does it have an even airflow all around?

That's the idea, but I have only empirical evidence. Before I put this duct and fan on, the backsides (away from the fan) of my objects had drooping, blobs, and other flaws that the other three sides did not have. With this duct and fan, the backsides look just as good as the front sides, so I think the design works. From a theoretical standpoint, the flow should be fairly even all around, so I seem to get the expected performance..

That said, I'm working on a device to quantify flow from printer fan duct designs so I can get a better picture of the performance. Stay tuned, that project is down a few in my project stack.

Could suspend something like a string of yarn in the center to try it? Needs to have a high surface are to weight ratio and be fairly uniform, a strip of paper or plastic only reacts in one axis due to being flat, or do it with something flat twice, turning it 90 degrees between tests? If it stays near the center (with some sway allowed obviously) it's balanced, if it gets pushed to one side and stays there it's not.

My instincts tell me the straight path from the fan out the duct on one side lets a larger fraction of air out on that side than the opposite side, have yet to print it and try it though, looking for an affordable place to get a centrifugal fan.

I'll probably also combine it with a LED ring to illuminate the print.

Edit: heh read around about centrifugal fans and saw "usages: bla bla bla laptops bla" remembered I have 2 busted laptops in storage, sure enough, they had one each, after some violence I have both in front of me, they're both 5v one is 0.5A 55mm fan 22mm hub 13mm thick, other is 0.38A 70mm diam fan 32mm diam hub 17mm thick. Both have a open slot on the side running top to bottom though, tried taping off 2/3 of it leaving the side the prop pushes towards if that makes sense clear and the airflow seems good, smaller higher amperage one seems better suited, experimenting with running it at higher voltage and settled on 6.5v as a happy medium, will have to use a resistor, regulator or voltage divider to reduce the 12v from the ramps, reducing the fan speed just gives you PWM:ed 12v

Edit2: Man looking at the main duct model in my 3d editor of choice modo is kinda painful, a lot of really wonky stuff like random polygons, wonky blending of shapes and elements, going to take a lot of time to clean up, kinda tempted to start over fresh.

I plan to make one of these - did you use ABS or PLA? (I guess ABS would be required so close to the hot end).

Also - did you use support for the main duct part? If so, how did you get it out afterwards?!

I used PLA. I print with PLA at temps up to 220C and have used both the J-type nozzle and the Ubis nozzle. The only problem I had happened before I had the clearance right. I got a blob ob plastic caught on one edge of the duct and it distorted it a bit. Still works fine. I've seen no other heat-related damage.

I printed without support. Had almost no sagging or otherwise out of place strands. I would think that truing to clear support material would be very difficult.

Thanks for your interest! I hope it works as well for you as it has worked for me.

I had to laugh when I saw the title of your thing. :-D As you say, there's lots of rubbish fan designs. As soon as you get a bit of back pressure the air forms a donut shaped vortex on the intake and the fan massively looses efficiency. A centrifugal fan spins faster and builds more pressure if you constrict the outlet. I struggled for a while to find a good fan outlet for my rostock and had to make one up in the end... Never looked back! People can also improve pressure on their axial flow designs by increasing cross sectional flow area at the minimum and by creating flow control vanes on the intake and outlet.

The duct would have to be run from the center of the fan to suck fumes. This would be a pretty major re-design!

I have seen no evidence of any melting. But the fact that air is being blown out has a lot to do with that. If you were to suck in air, especially hot air, that could easily change. If you try this, please be sure to post your results.

Perhaps this same technique could be use to suck out fumes? I have seen hoses running down to suck out fumes before.. but this is a neat idea! Does it melt the plastic near the nozzle?