Help with custom IDEX machine.

Christockos

I originally started this post as an inquiry to @sebkritikel who has DIY’d an impressive IDEX machine:

https://forum.duet3d.com/topic/13696/new-large-format-idex-printer-project
I had come across his IDEX machine while seeing if anyone had made something like an IDEX Voron 2.4 (which doesn’t seem to exist, yet).

The more I wrote my post, the more I thought I’d ask everybody, so here goes:

BTW, the reason I’m hardcore Duet is because of the wealth of knowledge, and all the smart folks sharing their knowledge and skills on this board. ahem

I'm planning a Printer Farm of 10–20 IDEX machines not only to produce small runs of my product (Kickstarter in the Fall), but also prototype future products that will eventually be produced in much larger runs (Injection Molded). TMI, I know, but I’m looking for advice.

What I think I need:

Duet controlled IDEX printer with a bed size large enough to print two 250mmx250mm(X, Y) prints side by side. Z isn’t as important so 250mm or greater.

What I could do:
A. Buy large (600x600) mass-produced IDEX printers and replace their boards to DUETs, Hotends, etc. Doable, since I’ve done it with all four of my current printers, but wasteful pulling all the boards, parts, heated bed from the stock model to make it something I want (need).

B. Build my own machine. IDEX, Duet, Volcano, tool plate Kenovo bed, 700(?) x 260(?). Direct Drive, easy-to-enclose, reliable, fast, built like a tank and somewhat easy-to-maintain.

Questions:

Am I out of my mind? Am I better off buying a bunch of $4K machines?  So far, I haven’t found what I need. Something about producing our-own custom machines, and tweaking them for what we need in the future seems more proactive in the long run. I foresee adding/modifying/growing the Printer Farm, and having smarter guys than I am working with me, hot-rodding the Farm for years to come.
Is there a rock solid design out there I’m missing, something I should check out?  I only just came across the Voron 2.4 (I wish it were IDEX). I don’t really need/want 600x600 bed size since I will be mostly doing the same print side by side (no more than 250mm x 250mm). The extra Y of a square bed seems to be a waste.
I love @sebkritikel’s large format IDEX printer, but I came across it while looking for an IDEX Voron. The thing I love about the Voron is how fast it prints (https://youtu.be/P11L0uK5a2M). Can a CoreXY machine w/o the stationary bed (Z moves down, unlike a Voron) be able print as fast and well? What makes the Voron so fast?
I want an IDEX machine as fast as a Voron 2.4, with a 600x250 bed and Duet. Any recommendations?

Forgive the forever-noobiness of this post, but you all are the smartest guys I know (3DP, anyway, heh).

Onward!

Chris

o_lampe

@christockos
building an IDEX printer for printing in 'copy mode' is probably the hardest to do. You can't use mesh leveling, so your bed has to be 100% aligned and flat. There are workarounds, like printing a raft 'into' a sacrificial styrofoam? bed surface. But the smell and fumes of 20 printers might not be funny...
Why not build a simple single nozzle CoreXY? Their failure rate would be lower and you only loose 1 part at a time. You'd need twice the number of controllers, but that compensates soon by less maintenance hours.
I'd look for a DIY design with linear rails and ballscrew(s) for Z.
I started with a 'DBOT' and converted it that way, but there are many other OpenSource printers, with all the right features.

martin7404

I second that about Idex. You need IDEX only if you are going to print 2 materials in one part. In your case 250x250 parts are easy to be done in 350x350 bed even with passive enclosure for ABS. You have 2 options, Get formbot troodon or DIY.
I would personally do CoreXY with Alu profile frame linear rods setup for X and Y and 3 Z on kinematically coupled bed with Leadscrews and linear rails.
Last month done it for a specific project with 650x450 X/Y, Took me 3 weeks, and about 1200 EUR

sebkritikel

@christockos Glad my printer inspired you! Despite not being 'complete' its a blast to run and tweak. I personally couldn't see myself running anything less than an IDEX machine (although I think my next design and build would be some sort of toolchanger).

A note on 'speed' - to print 'fast' with decent results, you need to do at least three things:

Ensure your printer can move at desired speeds with minimal resonance/vibrations in the print
Ensure you have a hotend capable of melting material at your targeted rate
Ensure your machine is capable of accelerating your print heads/gantries/build surfaces at high accelerations.

To do one and three, you really need to plan ahead with respect to component selection, and create a document where you seriously calculate the maximum achievable speeds and accelerations you can run. I've got MathCad document where I put in motor specifications, any gearing I have, target speeds, and see what sort of limits I hit. Definitely reference the stepper motor selection page, and use the RRF calculator to spot check your results. I run nicer motors than most, so I'm capable of running X, Y, and U at 300mm/s with 4000mm/s^2 acceleration (as you saw in my thread, I boosted the output torque of my Y axis NEMA23 by 80% as a result of pulley and belt combinations I selected).

I will caution you on relying heavily on part cooling to give you acceptable quality at higher speed, although that may be a controversial opinion.

Regarding mesh compensation and automatic bed leveling, I think the IDEX 'incompatibilities' are exaggerated. While I do have the capability of using mesh compensation (BLTouch mounted to T0) when doing standard printing (single or multi-material), I haven't actually enabled it in a number of months. My T0 and T1 z-axis offset are pretty close (with T1 being a hair higher off the build surface), and with an appropriate bed temperature selection, I don't have any adhesion issues. Properly selected same-material raft parameters would also go a long way in ensuring a nice first layer (personal preference coming from my Stratasys experience, but I much, much, much prefer a nice first layer on a raft than a high gloss first layer directly on the build surface- I think the aesthetics from the raft first layer result in a much more cohesive print. As a side note (at this time) I do not use any surface prep materials (gluesticks, etc) on my printer (no issues with ABS, HIPS, nylons....etc). The only time I think I had an issue was when I was printing some Ultem gears (to repair a paper printer hahahahaha).

Some items I didn't do great on my printer

My bed mounting isn't awesome. Not using a 'real' kinematic mount, most of my printing artifacts are a result of vibrations after z-moves (z-hops), and I don't think I selected an appropriate ATP5 plate thickness (memory is fuzzy on this, check with vendors, but depending on X-Y size, material thickness plays a role in the called out flatness tolerance). No issues with the bed moving up and down the printer, its just not 100% perfect
The last thing I designed/purchased was the toolhead. Ends up I didn't leave much room for them lol.

Some other general suggestions:

Since you're looking at 10-20 printers... design everything with metal parts in mind. Find a nice local machine shop (ok, local isn't a requirement). Make nice print drawings for them (GD&T likely isn't needed.... oversize the holes... .010" on x.xxx" is probably pretty good) and get the bulk rate quotes. Obviously try to design symmetry into your parts (rather than 1x of part 'X' and 1x of part 'Z', do 2x of part 'B', you'll get better volume price breaks). At least verify the design with plastic parts... but go metal in the end
Check beam deflections under various loads - don't try and build the Y axis out of straight 2020 or 3030 profile as an example.

Not a perfect comparison, but everything else being equal, a single tool version of my printer (dropping the U belt, extruder, hotend, Duex5, some pullets, etc) results in an overall price that is still 86% if the IDEX variant... obviously to some degree machine size drives price, but something to consider.

If you're set on the IDEX route... definitely design and build it yourself. You are correct that there aren't many (any commercial) designs out there that meet your requirements.

I might be leaving you with more questions than answers - but good luck!

Edit: a final note - do not under ant circumstances buy a BCN3D printer. Expensive, slow, poor quality prints, interior firmware.

o_lampe

@martin7404 said in Help with custom IDEX machine.:

You need IDEX only if you are going to print 2 materials in one part.

In my case I print with 2 nozzle sizes, but same material: 0.4 for perimeters and 0.8 for infill. Safes some time for bigger parts with lots of infill (print infill every second layer speeds up even more).