3D printed Rails

[juggle]

So I mentioned this in the mailday thread and it got some interest. Figured I’d start a new topic with a few more details.

I’ve been 3D printing a lot of my panels and have been really happy with the results I’m getting. My case is just “standard” Kosmo specs with wooden rails. But I’m really struggling with the wooden rails. I’ve got a bad wrist and screwing all those little screws into wood is killing me. Plus getting wood cut thin enough and accurately has proven tricky with my cheap table saw. And the wood I get is either all too soft or too hard and splits. Oh - and finding good wood screws in a proper size has also been frustrating…and those I have found use such a small phillips screwdriver I keep messing up the heads with the hard pine I wound up with for my center rail.

So I’ve been looking at options for some fancier rails. Aluminum extrusion is nice…but rather expensive for my taste.

I tried designing something I could 3D print with either my FDM or my MSLA printers, but the only designs I came up with were either too bulky for my taste or too small to print reliably without supports, and removing the supports on something that small is almost impossible.

Then I found this reddit thread: https://www.reddit.com/r/synthdiy/comments/jjsnds/3d_printed_eurorack_rails_that_snap_together_like/?utm_source=share&utm_medium=web2x&context=3

The design this guy came up with looks really promising. He basically took a standard aluminum extrusion and split it in half so it can then print without supports. But he also designed them in various lengths and in positive/negative versions so they snap together like lego. They still need to be bolted together…but you can stagger the various lengths to get longer rails than you can print in one go.

Intrigued I gave them a try. The one issue I ran into is that you need rather thin M3 square nuts for them. Most of the nuts on Amazon are 2.5mm thick. And these are designed for thin DIN 562 nuts which I couldn’t find on Amazon (well, I could but they didn’t have prime and were kind of expensive at $20 for 100 nuts or maybe it was 200.) So I ordered mine from McMaster, which is always a bit of a gamble since shipping is up to them - but I’ve never really been burned with unreasonable shipping prices from them. I wound up getting 200 nuts for $16 shipped and they arrived in one day.

They fit great, and the rails work perfectly:

You do need to make the case a little taller than for a “normal” Kosmo since there’s a lip on one side of these. So for a two row case you wind up with 4 extra lips and the full height for two rows is about 410mm.

The guy who shared these did release STEP files so you can modify them. You could make them work with the cheaper nuts you can get off amazon…but I got some of those and I’m not big on doing that:

Those are the “correct” DIN thin nuts on the left and the cheap Amazon nuts on the right. Not only are the Amazon nuts thicker (which would result in weaker or bulkier rails to accommodate) but they also have rounded corners so they don’t quite have as much surface area to lock in place. They aren’t bad…and if you’re on a super tight budget are definitely an option. I’m going to stick with the DIN nuts though myself.

I’ve printed this much to test with so far:

I think I’m going to go ahead and use these for my next case. Just waiting on some more black filament since I’d rather print the rails in black and I’m almost out.

How much time does it take to print these? Well, on my prusa Mk3 at 0.3mm draft settings I can print 1 pair of the 36hp rails in about 2 hours. And two sets of 36HP with 2 sets of 18HP takes me just under 7 hours. Working it out…enough rails to fill a “standard” Kosmo case would be pretty close to 48 hours of printing. Actually that would give you more than enough rails for a full Kosmo case. I haven’t tried to work out the optimal number of rails to print and instead have just focused on how many of the full plates with 108hp of rails it would take.

The rails I’ve printed so far in the photo above are about 13 hours of print time. One 7 hour print with 2 36hp sets and 2 18hp sets…and 3 2 hour tests of just one pair of 36hp each.

Oh - and for those wondering. The center rail made from 2 rails bolted together is EXTREMELY stiff and should have no problem spanning a full Kosmo sized case unsupported. It seems quite a bit stiffer than the 1/2" pine I’m currently using.

I also picked up some M3 threaded inserts: https://www.amazon.com/gp/product/B00270ZSIU?pf_rd_r=RAXR5S909RDHQH1EFMF2&pf_rd_p=edaba0ee-c2fe-4124-9f5d-b31d6b1bfbee&th=1

I figure I’ll install those in my case and then use the same screws that bolt the rails together to bolt them to the case. The center rail I’ll use the nut traps provided in the ends to bolt them in from the sides of my case.

There are side supports to create a full frame out of printed parts that the designer shared - but those are sized for eurorack (and he said he may do a 5u size) though with the step files it wouldn’t be hard to create a 200mm tall Kosmo version either.

[willow2x]

I have added my Gietec inspired rail design to Thingie, Thought i had added it already TBH.
I havent tested the things yet, but i have printed them ready for my case build (I might pick up the materials this week)
So, it might be worth cutting the Z height in your slicer to 50mm/100mm for a tester.

[BlackDeath]

Very cool!

It may be more stiff than pine, but how brittle is it? Suppose I am a bit rough with my cables, is there risk of snapping or layer-splitting?

[zorch]

Neat! If I had it all to do over again, I would probably go with these, from the look of them.

[juggle]

@willow2x

So, it might be worth cutting the Z height in your slicer to 50mm/100mm for a tester.

The problem I had printing my own designs like that wasn’t so much the Z height…it was the size of my nozzle.

Here’s the paths from the slicer:

Those two perimeters that connect the top left to the rest won’t be very strong or reliable. They may not even print every layer due to their size.

And if you print it in any other orientation it becomes a support nightmare. That’s why I got frustrated with my own design ideas.

It could print on my MSLA vertically…but the height on there is limited to 150mm or so and the taller the print the longer it takes - so you wind up looking at a VERY long print. I just tested slicing it and it’s looking like 12 hours to print in engineering grade resin (which I’d want to use due to that small cross section) Now…you could fit about 10 of them on the build plate and it wouldn’t change the time at all with the MSLA…but…it would affect the print dynamics and increase the chances of all of them failing. And printing long skinny things on end like this on MSLA is troublesome as well.

And a few tests I did printing extrusions like that on the MSLA anyway…that weak spot was just too weak since even though SLA resin is very stiff it’s extremely brittle. It broke before I even tested fitting a nut into it.

[juggle]

@BlackDeath

It may be more stiff than pine, but how brittle is it? Suppose I am a bit rough with my cables, is there risk of snapping or layer-splitting?

Depends on what you print them in. I’m printing these in PLA which is a bit more brittle than ABS/ASA and PETG…but not nearly as brittle as SLA resin. I’m not too worried about snapping or layer splitting. I’m more worried about my 1/2" square dowel I’m currently using than I think I will be of these.

The big downside to PLA is it isn’t as temperature tolerant. So it’s more likely to deform if things get warm.

But you could print them in PETG or ABS/ASA and they’d be a bit less brittle and more temperature resistant. But…even living in the desert I’m not worried about PLA for this. And PLA is much easier to print. If I was going to be taking it out on the road and subjecting it to more abuse and temperature extremes I would probably go with PETG or ASA…but that’s not in my plans.

In other 3D printing synth news…I did something else just now I’ve been meaning to do for awhile. Drew up and printed this:

12mm on one side, 10mm on the other with some ribs won the sides so it’s easy to grip. Fits my jack and pot nuts perfectly - even with a quick and dirty <1 hour test print:

I have a version in Siraya Blu engineering grade resin printing on my MSLA right now…but that will take 4+ hours.

I’ll share the files a little later once I get them cleaned up and uploaded on thingy and Prusaprinters.

Made tightening the nuts on the Steiner VCF in the background a breeze. Though…that panel is about to get it’s own post in the litany thread

[willow2x]

@juggle

Those two perimeters that connect the top left to the rest won’t be very strong

Hmm i see what you mean, Ive got an idea for a quick redesign ill work on tomorrow if i dont end up soldering =D

Id love a small resin printer, but i dont really know an awful lot about them or the materials =)

[juggle]

@willow2x

Id love a small resin printer, but i dont really know an awful lot about them or the materials =)

It’s a whole different world. And really not great for functional parts. I’d say 90-95% of the resin printer users are just interested in printing figurines. So most of the info out there about them is directed at those users.

I just got mine back in July, I wound up with two due to my impatience and a quirk or where/how I bought the first one. I shared my whole learning process on a another forum here: Experimenting with Resin printing (MSLA) for detailed parts | FliteTest Forum

The detail you can get with resin is remarkable. But…it’s very difficult to get accurate parts due to the physics of how MSLA printers work. There are other styles of resin printers that can do better…but they’re a LOT more expensive. And the other big problem is that most resins are really brittle. There are exceptions but the engineering grade resins are at least twice as expensive as the standard resins and are much slower to print.

Here’s my finger wrench I posted earlier but printed on the resin printer:

Looks pretty good…and even looking closer the detail at first glance is pretty amazing:

This is on my Phrozen sonic mini in Siraya Blu engineering grade resin.

It really captures every detail of my design like the .5mm chamfer which my FDM printer just kind of fudges.

But…it’s not perfect. The other side:

Is pretty ugly. Because it was facing down. It’s usable and grips nuts just as well if not better than the FDM version. But it’s just not very clean. The reason is the supports. I printed this on it’s side at a 50 degree angle. And that means it had to be supported. The blemishes are where the supports attached.

Why not just print it upright like I did FDM? Well, a couple of reasons. First (but least important here) due to the physics of SLA printing you generally get more detail with prints at an angle since the z resolution is finer than the X/Y resolution. But I don’t care about that here.

So the other reasons are - elephants foot. With SLA printing you generally have to overexpose the first layers to get the part to adhere to the print bed. And that means you loose accuracy and get bad elephants foot. I’ve found I can generally avoid that with strategically sized chamfers…but it’s hard to get them right.

But it gets worse…you kind of need that elephants foot to keep the part in place sometimes. There’s actually a LOT of forces on the parts trying to pull them off the bed as each layer is added with SLA. The parts stick to the FEP plastic at the bottom of the print vat as each new layer is exposed. Then the print bed raises and the part is “peeled” off. This is actually a surprisingly complex process where the speed and distance that it’s raised has to be carefully tuned for each resin and exposure. If the print doesn’t have enough surface area against the build plate then the peel can easily pull the part off the print bed and then you’ve got a mess in the resin vat to clean up and a failed print.

Also - vertical height equals longer print times. It’s totally different than FDM. The more you put on the plate with FDM the longer it takes. With resin…the time stays the same…but the physics of the peel get more complex. (the plastic at the bottom of the vat actually flexes as the prints are pulled off - so where parts are on the build plate can dramatically change the forces on them!) But - every layer of SLA printing has a fixed time associated with it - the exposure, plus the peel time. So how tall the tallest part on your plate is determines your print time not how much is on the plate. Tilting this part cuts almost 3 hours off the print time…and it was still a 4 hour print vs. 1 hour on my FDM printer.

On top of all that SLA printing is messy and nasty. REALLY messy and nasty. I made a HUGE mistake starting with it in July when paper towels were still hard to find. And it’s still difficult to find isopropyl alcohol at the stores even now and it’s pretty important for cleaning prints. You can use denatured alcohol, but if you live in CA you can’t buy it…and here in AZ it’s more expensive…and the smell bothers me more than the smell of isopropyl. Also, nitrile gloves. You’ll go through a LOT of them - you don’t want to get resin on your skin. Exposure builds with time - you may be fine with it now…but at some point you will probably reach your bodies threshold and then any exposure will set off a reaction. So it’s best to be careful and always use gloves when exposing yourself to uncured resin. Plus there’s odors. The resin stinks - some worse than others but they all smell more when they cure (though they don’t smell once cured - and are generally inert once cured, there are even body safe resins used by dentists and doctors but they’re really expensive.) The alcohol for cleaning stinks. The other options other than alcohol don’t work as well and tend to stink as well.

I’m glad I have my resin printers…but it’s hard to recommend them. FDM is just so much easier and more useful for most things.

As far as synths go…the best use I’ve found for a resin printer is for printing knobs. And even then it took me a lot of experimentation to get something decent. I print my knobs in two parts. An inner core printed in special expensive flexible resin - and an outer shell in cheap resin (which I’m now custom tinting) to get the look/feel I’m after. Even then printing the shells is really tricky due to the peel effects if they’re printed directly on the bed…and the distortion if they’re printed on supports. I’m getting results I’m fairly happy with…but I’m kind of waiting until I have a 100% full case to print knobs for so I can do them all in one batch to make sure they match.

All that said…the price on Resin printers has dropped a lot lately. And the price on wash/cure stations has come down as well and they REALLY help minimize the mess and how many gloves and solvents you go through. (I LOVE my Elegoo Mercury Plus wash/cure station.) I have both the Phrozen sonic mini and the Elegoo Mars Pro. Both are great printers at a great price. But I’m really torn as to which I prefer. The Elegoo is a more polished package, but the Phrozen has a superior LCD and I had more early success with it. They’re both about the same price (and probably just got cheaper because newer higher resolution models just came out) and have nearly identical print capacity. The Phrozen is slightly faster…but I’d really caution against letting that sway your opinion as they’re both quite a bit slower than FDM.

I can go into more detail if anyone wants…but…fear I may have gone into too much already

[willow2x]

Hmmm they look a little out of my patience range haha well, at the moment.
Need to get some stuff organised at home (closet shelves and other storage, then a table in my tech cave with some component boxes, aswell as sort my gf’s craft room with her…) and eventually i can think about where i could fit a resin printer and other basins haha.
I dont know how i gather so much junk lol

I also reworked the Eurorail today, about to upload it to my Thingiverse now. =)

[juggle]

@willow2x

I dont know how i gather so much junk lol

I know that feeling All TOO well

Almost have enough rail printed to do a two row 144HP Kosmo case:

I can fit 3 lengths of 36HP rail on the bed at once (6 pieces total, 3 of each half) And it prints in just over 6 hours. So one plate of that and a plate with one pieceof 36HP and two half pieces of 18HP makes a full 144HP length.

Do that 4 times…and you’ve got enough rail.

The center piece is fully assembled with all the nuts and screws installed. It’s VERY stiff in the in/out direction but still has a bit of bow/give vertically. Though once panels are mounted to it I don’t see that being an issue.

Reprinting my Steiner VCF plate right now…but overnight I’ll do another plate of 3 36HP sections for the top rail and then tomorrow I’ll crank out the last two 36HP pieces and 4 18HP pieces and should have all the rail I need for my next case. And the screws I ordered should show up tomorrow as well so should be good timing to get a new case built by mid-week.

[juggle]

Finished these yesterday morning and forgot to share:

The screws sticking out of the top/bottom rails are how I plan on mounting them in the next case. Unfortunately I messed up and accidentally ordered 4-40 threaded inserts instead of M3 (didn’t notice that Amazon changed the size on me when I switched it from a 10 pack to a 25 pack.) So will be a few more days until I have new inserts. Which is ok because I need to go to the store and get more wood anyway. The only ply I have left is way too warped to make a case out of.

And when I went to go get some on Sunday the battery in my truck was dead. The downsides of not driving nearly as much with COVID. Car battery was a bit sluggish when I did the grocery shopping as well. Truck battery is just barely 1 year old, car battery is just over 2. Since I live in the desert the car battery starting to die doesn’t surprise me. Thankfully it’s a new car with a 5 year bumper to bumper warranty that specifically includes the battery. Truck on the other hand…is over 2 decades old, has about 250k miles, so I’m on my own there

[juggle]

Well, a little over 2 weeks later I finally got these in a case:

Took me a little longer than I’d hoped to get a case together. The leftover ply from my first case was in really ratty shape and I couldn’t find enough good material to pull off a second case. Then went to by a sheet and the battery in my truck was dead. Finally got to the store and got some…but the ply I got this time is a lot softer and more crumbly…and kept warping on me. So pulling it into a square case took more effort and time than I expected.

Then I ordered the wrong size threaded inserts. I used some threaded brass inserts to mount the top/bottom rail and didn’t notice when I changed the package size to 25 instead of 10 per package Amazon also changed the thread size to 4-40

The threaded inserts worked great. I got two of them slightly out of position and had to drill out the hole in the rail to compensate…but other than that they were a joy to work with.

I also kept putting off finishing because I needed power supplies for this case. But - got the parts for those yesterday so it’s time!

One nice thing I found about these 3D printed rails - since they’re in two pieces you can take just the top off and it’s super easy to add/remove nuts without removing the entire rail. The threaded inserts definitely help with this. Bonus!

And I can hear you asking already. That middle one looks little crooked, are they stable? Well, that’s mostly the fisheye on the camera - but I will be honest, the middle rail would make a lousy straightedge.

But…add some panels:

It really stiffens up! In fact - it feels stiffer than the 1/2" hard pine dowel I use in the center of my first rack. And it’s SO much nicer to work with. Being able to slide and reposition modules without worrying about wood splitting or screws stripping. If it wasn’t for the extra height these take (They have a small lip on one side) I’d be printing some to retrofit my first case. Instead I’ll probably make a second case like this one to replace my first case.

Admittedly - the extra quantizer boards and the mega drone eat up a lot of space - but I’m still a little shocked at how quick this is starting to eat up my back room:

Will be nice to finally have all of these modules that have been piling up on top of the old rack waiting for room and power to find a home

[ChristianBloch]

That looks really great! If I had a 3D printer, it would probably be on duty printing rails 24/7

[CTorp]

Case looks great! And I’m really struggling holding off on starting another hobby and going down the 3D printing rabbit hole.

Also, holy quantizer! That’s a lot of 12ways!

[juggle]

@CTorp

Case looks great! And I’m really struggling holding off on starting another hobby and going down the 3D printing rabbit hole.

Thanks. It isn’t so nice in person The top layer of the ply ripped in a lot of places, and on the bottom I drilled the holes for the threaded inserts all the way through - but on the top where I didn’t drill them all the way through they caused the ply layers to separate and bulge. So both look rough but in different ways.

I’ll probably drill the top all the way through when I take it apart for paint then squirt in some glue and clamp the layers to get them flat again…or may just leave them and hope the paint hides most of it.

Also, holy quantizer! That’s a lot of 12ways!

I’ll tell you when I’ve had enough quantizers!

Seriously though I’ll be putting 4 of those up in the BST thread once I finally get mine built. I just ordered the boards from JLC before @analogoutput decided to put some up on Tindie. I’ve also got a bunch of his power bus boards (it was only a buck or two more to get 10 than 5). I’m just horrible about mailing things so I’m not in a big hurry to put them up for sale. But if someone really wanted one and sent me a message I’d gladly part with them. (Same with this Yusynth EMS filter board I etched for discrete transistors which I don’t need because I got the matched pair IC’s to do that version instead.) Once I get tired of shifting this stuff around I’ll finally list it officially for sale

[CTorp]

Ahh I got you, just testing the rails. I always buy 10 copies! On 2.5cm panels it was only 1.50$ or so for 5 extra boards…why not right?

[eric]

So I checked those rails.
I have one question :
Eurorack HP is 5.08mm, right ?
So the 18HP rails should be 91.5mm. But they are 99mm ?!
Same for the 36HP rails.

[analogoutput]

@juggle

If it wasn’t for the extra height these take (They have a small lip on one side)

I found that can be problematic with my homemade panels — the sheet aluminum I get cut is sometimes a couple millimeters too long. I can use them with Vector rails that have no lip (if there’s extra space between the rails, e.g. I have mine mounted on top and bottom of a wood cross piece) but I have a pair of TipTop rails that have lips and I can’t always use those panels there.

[juggle]

@eric

Eurorack HP is 5.08mm, right ?
So the 18HP rails should be 91.5mm. But they are 99mm ?!
Same for the 36HP rails.

Honestly I have no idea what eurorack HP is. Since my case is Kosmo I don’t really care. And if the width is off just a bit I can always fill any gaps with a spacer since I 3D print all my DIY modules.

Heck, this case isn’t even square. I know for sure because the rails were all the same length - but I had to sand down the ends of the bottom one to get it to fit. But it’s good enough for my needs.

@analogoutput

the sheet aluminum I get cut is sometimes a couple millimeters too long.

Can’t you sand them down to size? Aluminum is fairly easy to work. But again since I 3D print my DIY panels and the rest are nice accurately sized PCB style I kind of like the lips. Though now I kind of want to give a .5 or 1mm black band on the sides of my panels (kind of like the bare aluminum strip on 5U modules) to give some visual separation on them horizontally to match the vertical separate the lip gives in the middle. Wouldn’t add any extra width to them - just make the edges have a layer or two of black the same as how I do the lettering. Looks like it wouldn’t interfere on any of them except my mixer which I was thinking about redoing a bit wider anyway because it’s kind of cramped.

[misha]

I also have a rack with mostly printed rails and threaded inserts for M3 screws.

the STL file is here, print horizontally with 100% infill:
https://res.cloudinary.com/maiik/raw/upload/v1607042500/misha/rail_inserts_2.4-4.6.stl