There seems to be more interest in 3d printing knobs than I expected 
If people want I could start a deep dive thread on it. It’s something I’ve been working on for some time.
Back about 15 years ago I build a Midbox SID, started a midibox FM, and started a midibox sequencer (back when it still used a PIC instead of a STM32 for the main processor) But…all of those projects stalled out when I discovered how much knobs and buttons cost.
The SID I got working - and built a partial interface for it…but to this day it still looks like this:
(And as two years ago suddenly stopped working. Every couple months I poke at it but still haven’t tracked down what the heck suddenly went wrong. Pretty sure it’s an issue with the main core module but I’ve tried swapping in a known working core and even a new known working PIC and neither solved it so still stumped.)
The FM is still just a board. I’ve been lazy about building the bipolar power supply for it (now that I have two on hand I should dust that project off and try to fire it up finally) and honestly it didn’t seem like as much of of a synth to play with. I also recently built an XFM2 (https://www.futur3soundz.com/) and splurged and got a flash FM synth (https://hpi.zentral.zone/flash) because I just couldn’t pass up an entire FM synth that fits inside a midi jack and is entirely powered by midi - even if it was more than I could justify spending. But…I just don’t find FM synthesis to be nearly as much fun.
But I’m digressing - knobs…and buttons! Which brings me to the midibox sequencer. I had let the project languish because without knobs and buttons it just didn’t make any sense…and I just couldn’t bring myself to spend what it would cost to buy the knobs and buttons. But…about 4 years ago I picked up a 3d printer and found it solved all kinds of problems for me. So I got to thinking about the sequencer again…and saw it had updated to using the STM32F4 discovery board and I happened to have one on hand not being used. So for Christmas 2 years ago I ordered myself a set of circuit boards - then for my birthday a few months later ordered all the bits to populate them and built the thing.
Then I dug in on my plan to 3d print knobs and buttons…and quickly realized that I had overestimated the ability of a FDM printer to do that kind of detailed work 
My first attempts at keycaps were underwhelming at best:
Even with the smallest layer height I could pull off there just wasn’t enough definition to make the tops feel right under my fingers. But…I still had a trick up my sleeve…I had some ABS filament and figured I could try vapor smoothing it with acetone. So I tried that next.
They were better. But really inconsistent.
They just looked…lumpy…and unprofessional…but not in an appealing way
They worked…and I’ve had them on there for 2 years still not willing to pay for “real” caps. Not that I even can buy “real” caps for this. I cut the panel on a CNC machine I built (a MPCNC) and the only bit I had was a 3mm endmill so I couldn’t do square holes…and I didn’t want round buttons. I needed oval buttons but just couldn’t find any. So I’ve lived with lumpy caps.
I also printed knobs for it…and those came out ok. I’m still not big on them and every couple months I try printing a new design - but the limits of FDM printing have always left me less than 100% satisified with any of the knob designs I’ve tried.
Incidentally - getting the sequencer working has been a big motivator in finally building a modular. I need for things for this beast to sequence Right now it only has MIDI but I’m waiting on the last few parts to finish an analog out card for it so it can do CV/Gates/Triggers directly - I’m curious which will happen first now, those arriving or me getting enough modules built to be worth having the new abilities!
So…FDM printing can work but you’re just not going to get very professional looking results. I thought about trying resin printing…but the price was too high and it just sounded too messy and hard to deal with from my discussions with people who had resin printers.
Then anycubic announced the photon zero and it got some pretty impressive reviews. But reading deeper it seemed the reviewers really just like the wash/cure station that was released alongside it and the photo zero itself is actually a little underwhelming as a resin printer. All resin printers have fairly small build volumes and the zero’s is even smaller than most, plus it’s lower resolution…and the whole reason I wanted to try resin was for ultra-high resolution. So…I crossed it off myself. And instead talked myself into getting a Phrozen sonic mini which seemed to get rave reviews.
I ordered it from a place here in the states…but then found they just drop ship from the manufacturer. And after 2 weeks I still hasn’t got a shipping notice - apparently it was Taiwanese holiday and the manufacturer was closed for a few weeks. I decided to cancel the order and go with my second choice - an Elegoo Mars Pro which had come back in stock on Amazon. But…while I was doing that Phrozen came back from vacation and shipped my printer…so now I had two printers on the way. Oops.
They both arrived and I dug right in on learning to print knobs. And quickly learned it’s not as straight forward as I had hoped.
Take those keycaps. Print them right on the print bed and you’ll get “elephants foot” which is almost unavoidable with MSLA printing. So the bottom squishes out and changes the dimensions so it won’t fit over a button anymore:
Print it on supports…and…you get deformation from the supports…and if you just raise it up then there are suction issues while it prints that can cause bigger deformities or even cause it to pull off the supports while it prints:
Side note - one other thing I hadn’t counted on which seems obvious in retrospect. But the color of resin matters a lot. Since it’s a light based process and color is a function of light different colored resins print differently and require different exposures. You may have a part printing great in grey…but switch to black and suddenly you’re almost starting over! And there aren’t nearly as many color options as there are with FDM…and you generally get a dull matte finish with resin.
Anyway - I did a bunch of tests to try and figure out the best position and support settings to get good prints:
And…I’m still trying to figure out what’s best. They all have compromises. Angling the caps in two dimensions and using heavy supports seems to work best…but whatever is pointing down still gets deformed. You can see how that happens on these test knobs:
The side that’s up in these photo was down against the build plate while printing.
Printing flat against the build plate I found I could add a chamfer to the part to compensate for the elephants foot - though just how big of a chamfer changes depening on the resin due to the exposure changing. Black resin takes more exposure so needs a bigger chamfer to compensate for more elephants foot caused by the longer exposure. And…you can still run into failures due to the suction effect that happens with resin:
But…for whatever reason I just won’t give up and keep trying. That big spinner on the radio interface I showed earlier came out nice…I did do a little post-processing to it lightly wet sanding the face and giving it a light spray of polyurethane to make it smooth:
But the back of that same print really shows off how mangled it gets from the supports:
And notice how the flat back isn’t flat. Also - note the holes - printing things solid in resin is also troublesome…so it’s best to hollow the parts out and leave a way for uncured resin to escape. Finally notice the shaft and set screw holes - the white is the cured resin being worn away from friction. The threads in the set screw barely work…and the shaft opening wasn’t quite right so its a tight fit even without the set screw…but since the resin is so stiff it doesn’t flex to fit it just rubs away some resin.
So…my experiments continue. I got some flexible resin and have been experimenting with blending the flexible and colored - which seems very promising. The flexible resin is shinier when cured and by itself is almost like rubber - but mixed with traditional resin gives a nice less brittle finish.
I’m also thinking about trying a hybrid approach. Printing most of the knob on my FDM printer and then making a SLA printed cap that fits over it for a polished final look. That way the more flexible FDM materials deal with the issue of fitting on the shaft…but I can still have nice hard detailed outsides on the knobs. Even commercial knobs use this method sometimes…like the ones that came with the pot I used on this first module:
There’s some kind of soft (nylon?) plastic bit inside that does the mating to the pot shaft, and an outside shell made of some kind of aluminum (or more likely pot metal.) I’m thinking this may work best and is something I’m probably going to be trying soon.
I think it’s worth the effort and frustration…nice knobs matter after all. And it’s great being able to make custom ones any shape/texture/size that you can dream up. Not just for aesthetics…but for function and like the button caps on my sequencer to accommodate other aspects of DIY projects.
(And no, this wasn’t the deep dive. I tend to really ramble when I get going…and can obsess over little things. If I really got going on the details of what I’ve learned about trying to print knobs this would be 10 times longer!)
