Ask questions about your build! Having trouble printing a particular part? Having trouble getting parts to fit? Trouble sourcing non-printed parts? I am here to help
I stumbled upon this build while looking for complicated mechanisms to test out the capabilities of the FormLabs Form 2 in making precision mechanical components/systems. I think there is little question that a Curta fits that build.
However I note that this build is based around using Fused Filament prints. I'm wondering, for those who are already familiar with the 3d geometry and the development of these cad models, what further consideration should be taken in attempting to print the components using an SLA printer. Should the extra "stock on" of the 3d models be reduced, or eliminated? As I understand it the interfacing components in the models are designed to have interference, with the intent of manually grinding the surfaces to match. I'm wondering if it is practical to reduce this stock on.
Additionally, what would be the bottle neck to scaling the build down? Would a 2:1 assembly be practical? Obviously all the non-printed components such as hardware, bearings, springs, etc would have to be re-assessed and new components sized. I imagine the entire assembly would need to be redesigned as opposed to simply scaling the solids and hitting the go button. My question is, would such a thing be practical? What trial components/fits could one start with to substantiate that the redesign of the assembly to a different scale would not be a wasted effort?
Any and all feedback is welcome. Thank you.
I am just starting this amazing project and have a few parts printed. However, I am having a little trouble sourcing the springs. Especially the ones linked on eBay because they are out of stock. Have been doing some searching but can't come up with anything. Anybody get their springs elsewhere?
First, thank you for sharing this with us. It's going to be a long road to getting all done, but I'm really looking forward to the challenge. I've been enamored with the Curta ever since I learned about it at the Computer History Museum.
To that end, I've been going through the build manual to plan out the parts to print. I'm having some trouble matching up the "Parts Required" names with the various file names. Does anyone have a map between the two?
For example, the first instructions call for Support Column. Is that framesupport-_frame_support?
Had a ton of failed prints on my upper housing. Prusa MK2s, Hatchbox PLA. Failed after about 10-20 layers down, the tiny contact area of the sharp edge was not enough to stick. Tried a raft, increasing supports, forcing supports, etc. Played with temps etc, all same result and print came loose. Finally setup the contact area to be .15 vs .2 on the supports, as well as forced supports for 20 layers and it stuck. I did print it threads "down" because that again increased supports to increase contact area. The supports came off just fine with only a touchup of sanding on the threads.
I am trying to minimize painting BTW on my first one...just because I "Like" the printed look.
Just a FYI if anyone is having problems that is something to check.
Working on a build and pretty excited about it! But I find myself with a few questions relating to carry pins on the result dials:
(1) How deep should they be? In the original engineering diagrams, they go nearly to the bottom of the hole and extend maybe a third of the way up the dial's narrower section (on the undivided-gear dials). Is there a rubric for just how far they should stick out?
(2) Apropos of that, is there a good reason not to redesign the result dials to have the carry pins integrated into them? That would require 3 different shapes to print instead of 2 (type-2 + full pin, type-2 + half pin, type-1 + half pin), but it seems it would have the advantage of greater regularity in how far the pins stick out and enforcing the angle on the half pins. Is there a structural problem with doing it this way, or is it just ahistorical (i.e. the pins in the original Curta were separately machined and had a separate part number)?
Just finished printing (well reprinting) all of the parts and on to assembly. Have the selector knob assemblies done, along with the main casting initial setup. Winding the springs was entertaining lol. I am limiting my filling/bondo to essential only, I am intentionally trying to keep my first build finished only to the level that the printer provides, I will be making a second that I will finish/paint.
Only tweak left to figure out is how to stiffen the transmission axles. I tried to print some in carbon PLA with unsatisfactory results so far, but that was my first carbon piece with supports, so I expected some tweaks. If I can sort them out then I may reprint the transmission axles in carbon, but they have to be stiffer and I am not 100% sure they will be! Right now assembly will be with standard PLA printed upright as per the standard build.
My one part that I am not happy with so far is the result dials. Since they have a small flange on them they get built with support, and that prevents the bottom edge from being "crisp" and square (ends up rounding in a touch). I may split it and reprint, then simply glue on the flange as a washer as then the piece will be nice and crisp. Really though the digits cover hides the edge quite well, so it it probably not needed.
Will post some pics and updates to the build as I start assembly, but so far it is really been fun and a super interesting project! Thanks!!!!!
I've mentioned that I've been working on a way to number / letter the Curta that is more accessible to others. It's been a while since I started on that project and I've hit a few roadblocks. I'm writing up this post so you can see what I'm planning, how far along the project is, and where I'm going with it.
The idea was to build a Pantograph for doing the lettering / numbering on the Curta. Stencils and jigs to hold the parts would be printed on a 3D printer. The pantograph would have a mount for a Dremel (or other rotary tool) flex shaft. A small end mill would cut very lightly mostly through just the paint to uncover the plastic underneath (hopefully you're printing in white). This should be easier and more importantly cheaper than buying a Cricut to cutting paint stencils. Hopefully you already have a Dremel (the one I linked is one of the cheaper ones and supports the flex shaft) and a flex shaft. If you have neither, the total between those two links is $75 -- not great, but much better than a Cricut machine. I want to stress here that I have not gotten this working yet, so unless you were already desiring a Dremel and flex shaft, don't go buy them yet!
Unfortunately, the initial plans I had didn't work out (It's a shame too, I was proud of those print-in-place gimbals -- got the tolerances right on the first try, too). I'm attaching some photos of what I had planned. The idea was to use two pantographs. One on top of the other with a Dremel flex shaft and stylus mounted into gimbals so that angle when cutting on a curved surface would be copied from the template to the part (the test template I was experimenting with was just flat, though).
The first problem I ran into was that the double pantograph idea offered too many degrees of freedom. It was impossible to keep the Dremel bit steady. Perhaps if I was able to align the pivot point of the Dremel to the tip of the end mill, it might have worked better.
Next I removed one pantograph and designed / printed a fixed mount for the Dremel and stylus. This eliminated the pivot problem, but the bars over the length of the pantograph arms just weren't rigid enough. Flex in the pantograph arms caused the numbers to still be inaccurate. I'm going to cut the bars down this weekend to reduce the size of the pantograph (It was much larger than necessary anyway) and try again. I also have ideas for easier ways to align the template and the part to the pantograph.
I hope to get much better results soon.
I had a mistake in my BOM. The size of the upper carriage spring was listed as 1.8x28x25. That was incorrect. I have uploaded a fixed copy to the proper size -- 1.8x28x40. I didn't realize it until I ordered more at the wrong size. I hope this post prevents some of you from ordering incorrectly as the only source I know is in China and shipping takes a while.