Grounded Experimental Delta Printer

Nicholas, would you consider setting up ConceptFORGE's Github Repository on Bountysource?

If this is something you want to do, it is really easy to set up. You can log in to Bountysource with your Github account, then go to Tools. Bountysource automatically lists your GitHub projects, and you can click on the ones you want to configure.

Open bounties could be announced and followed on the new RepRap Bounties forum.

Of course, if this isn't something you want to do, that is totally cool too!

This is awesome. I will make one, in bright red. I ran Slic3r on the parts using Charles Steinkuhler's Wally .ini in the github repo, and the total came out to 1.07 Kg of plastic.

I can't see the assembly files until I upgrade my Inventor version (a 4GB dl), but in the meantime, a few questions for Nicholas:
- What slicer did you use for the parts? What settings do you recommend?
- There's a large square cutout in motor_arm inside the gear that results in a huge bridge (1+"). Is there any particular reason for this?
- What extruder do you recommend to go with a Simpson?
- Do you use supports for the 608 bearing holes? The topmost bearing holes in many parts have large overhangs.

Thanks!

I used slic3r .9 something. I used a .4 nozzle, .3-.35mm layers, 3 perimeters, 3 top, 3 bottom, 15%-25% infill.

I would suggest making the HUB BOTTOM 50+% infill or more perimeters. It holds it all together.

Ubis is what I use.

I want to eliminate that bridge. I will work to minimize it but I have to have access for the tuning peg. Even if you bridge poorly it has many layers before you get to a mechanically critical part.

What parts are you refering to when you talk about 608 overhanges? Everything should be printable but something may have slipped by.

Quote
sgraber
FYI here's the number of parts to print: [dl.dropboxusercontent.com]

I'm going to slice them tonight and report back on filament amounts and time to print like I did before with Wally.

Thanks for pointing out the printable parts with amounts to print. Do you happen to have a count on each of the non-printable items. I would like to put in an order for all of the nuts and bolts and other things like that.

I am assembling a BOM. I will post it tonight when I get back to my laptop.

Quote
Guizmo
Didn't you have interference problems with the center of the herringbones? Once you told me to make a channel there instead.

Quote
nicholas.seward
@Guizmo: You only need a split between the two sides if you are going to use the same arm blank for both sides. The gears are off by 1/2 a tooth from side to side but there is a 10mm gap.

I don't understand this answer--isn't there a tolerance issue with the center of each herringbone? We're not talking about the tooth offset issue here, but mechanical tolerances in the sharp corners, right?

Quote
MattMoses
Nicholas, would you consider setting up ConceptFORGE's Github Repository on Bountysource?

If this is something you want to do, it is really easy to set up. You can log in to Bountysource with your Github account, then go to Tools. Bountysource automatically lists your GitHub projects, and you can click on the ones you want to configure.

Open bounties could be announced and followed on the new RepRap Bounties forum.

Of course, if this isn't something you want to do, that is totally cool too!

I don't understand what you are suggesting, or what this provides, can you explain?

I have to think you used 25% infill on everything but the HUB BOTTOM.stl. I just sliced all the files using 15% infill with your above-mentioned recommendations (50% infill on HUB BOTTOM.stl) and come up with 33 hrs printing and 665 grams of PLA.

Did I miss a part??? I thought I understood that GUS took ~1kg PLA. Here's my slicing results:

[dl.dropboxusercontent.com]

And here is my slicing settings (Slic3r):

[dl.dropboxusercontent.com]

Shane

Edited 3 time(s). Last edit at 11/27/2013 08:54PM by sgraber.

Quote
A2
I don't understand what you are suggesting, or what this provides, can you explain?

Sure. Basically, Bountysource provides a way (one of many many different ways) to pay people for contributing to an Open Source project. It introduces money into a project, so depending on your point of view it might be a great thing or a terrible thing. It is up to the project leader to decide if it is right for them or not. (Disclaimer - I am not affiliated with BountySource or GitHub. Sorry if this whole thing sounds like a sales pitch.)

Bountysource is pretty well integrated with GitHub. You can see the Bountysource pages of some RepRap-related projects here:

Slic3r
Kintel Marlin
Traumflug Teacup Firmware

If (hypothetically) ConceptFORGE was on BountySource, people could raise issues (and feature requests) and open bounties on things that were important to them. Also, if a person was working on something, they could raise an issue and ask others to contribute money to a bounty on that issue. There is a worked-out example of this process on the wiki page that bobc set up.

So, a couple examples relevant to the current thread:

Nicholas could open issues (or feature requests) in GitHub related to what he is currently working on. He could let people know what he thinks is important, and if people wanted to support the work, they could contribute money to bounties associated to those issues/features. When the work is publised (on GitHub, RepRap Wiki, anywhere with suitable Open Source credibility) Nicholas gets the money.

Another (hypothetical) example: Suppose I'm fixated on self-replicating machines. I'm super excited about BOB Simpson (BOltless and Bearingless Simpson) but I don't have time to work on it myself and I know it's very far down on Nicholas's priority list. Let's imagine I get super super excited and I open a bounty on ConceptFORGE for "a maintenance-free Double Lamina Compliant Joint or equivalent, for use in a BOB Simpson". The idea is that the bounty will give some developer incentive to work on it. When it is done the developer publishes the design with an open source license, everyone gets to use it, the developer gets the money, everyone is happy.

That's the idea anyway. Everyone is supposed to get happy...

@ MattMoses

That's an awesome suggestion!
I didn't realize that GitHub and BountySource could be linked, makes good sense!

Edited 1 time(s). Last edit at 11/27/2013 10:56PM by A2.

Here is a link to the BOM.

A few people are already done printing some parts. That is awesome but I haven't assembled the instructions yet. I just uploaded the STLs. One of the biggest gotchas is that all the holes for M8 and M3 bolts will probably have to be drilled with the exception of where you actually want to thread into the plastic. The other gotcha is that you will probably not be able to drop the 608 bearings into their holes. I put the part in an oven on low. I take it out every minute and hit it against a hard surface. When the tap turns to a thud you are ready to insert the bearings. I have heard reports that you can do this also by heating a bearing with a hot plate or heating the part with boiling water.

@MattMoses: I feel silly setting up my own bounties. However, if people wanted to setup bounties for things they would like me to work on (or people in the community at large) then I will do my best to collect on them assuming they align with my overall goals and time restrictions. I really like the ideas of bounties. (BTW, I can't wait to start working on BOB but so many other things are in front of him. I really think I could get all the non-plastic parts down to just the steppers, controller, and the hot end.)

Quote
DaveGadgeteer

Quote
Guizmo
Didn't you have interference problems with the center of the herringbones? Once you told me to make a channel there instead.

Quote
nicholas.seward
@Guizmo: You only need a split between the two sides if you are going to use the same arm blank for both sides. The gears are off by 1/2 a tooth from side to side but there is a 10mm gap.

I don't understand this answer--isn't there a tolerance issue with the center of each herringbone? We're not talking about the tooth offset issue here, but mechanical tolerances in the sharp corners, right?

Yes. I'm talking about this:

@Guizmo: As you have it drawn the groove is not needed. You only need a split if there is a discontinuity in the teeth. To be clear, you need a discontinuity on the arms so you don't have to make two arm blanks. That isn't a requirement but it makes it easier to not mess up on the registration.

Quote
nicholas.seward
I put the part in an oven on low. I take it out every minute and hit it against a hard surface. When the tap turns to a thud you are ready to insert the bearings. I have heard reports that you can do this also by heating a bearing with a hot plate or heating the part with boiling water.

Heating the part until you get a thud I can understand but but "expanding" the the bearing by heating it does not make sense.

Ralph

What is meant is you warm the bearing to the point that when you insert it into the bearing hole that it melts the plastic a bit so it slides into the hole.

Shane

Quote
Ralphxyz

Quote
nicholas.seward
I put the part in an oven on low. I take it out every minute and hit it against a hard surface. When the tap turns to a thud you are ready to insert the bearings. I have heard reports that you can do this also by heating a bearing with a hot plate or heating the part with boiling water.

Heating the part until you get a thud I can understand but but "expanding" the the bearing by heating it does not make sense.

Ralph

Quote
nicholas.seward
@Guizmo: As you have it drawn the groove is not needed. You only need a split if there is a discontinuity in the teeth. To be clear, you need a discontinuity on the arms so you don't have to make two arm blanks. That isn't a requirement but it makes it easier to not mess up on the registration.

Ok, I got it now. You need it when not mirroring. Thks

I would expect crud to accumulate in the v points. It's hard to clean inside-corners like those, and dust etc won't be able to work its way to an open end and fall out. Does that really not matter in practice? In normal gear teeth there is some empty space because of the details of the shapes of the tips and valleys where crud can escape to.

Quote
sgraber
What is meant is you warm the bearing to the point that when you insert it into the bearing hole that it melts the plastic a bit so it slides into the hole.

I haven't tried that trick with bearings, though I've done it in combination with heating the plastic to get the pulleys onto the motor shafts; if the shaft isn't warm it tends to freeze the pulley in place before it's slid on far enough. And the pulleys are sensitive parts, so I don't want to overheat and distort them. I'm not sure it would work the the bearing holes though. What we found was a problem of out-of-round holes, rather than simply too small, and the plastic really needed to move in order to let the bearings fit in. One thing I did was keep an M8 bolt handy so I could run it through both bearings while the arm was still warm, to make sure they were lined up with each other. Also, I strongly recommend only trying to do one bearing at a time, and re-heating in between; you don't want to try to force them into place or the plastic will distort.

On that point, I strongly recommend trying the oven trick with a scrap piece first. You'll want to get an idea of what Nicholas means by the sound changing - it really does sound like a thud as the plastic softens, but it's also very easy to go too far and make the part too flexible. I've had success with the oven but have also done a couple of variations on his original idea. For the gear arms, I wrapped the gear section loosely in aluminum foil to try to keep it a bit cooler, leaving just the end exposed. And for for the most part I've switched to spot heating with a hairdryer rather than the oven. It helps to have a small enclosure around the part; I use a cardboard box that I could stick the end of the arm into and let the hot air swirl around it. I feel like the hairdryer gives better control, but in both cases I was able to get the bearings and bolts in place without warping the arms (as far as I can tell). This was all on the original GUS; we're still printing the new arms, so haven't tried putting any bearings in yet. . .

Bill.

Quote
nicholas.seward
@MattMoses: I feel silly setting up my own bounties.

It's no sillier than selling a product, or starting a crowdfunding campaign. And I think everyone is in agreement that those things are not silly at all. But of course it is totally up to you. Considering the level of interest in your projects, you probably wouldn't have to worry about setting up your own bounties.

I think the whole concept of bounties could really add a new element to open source hardware projects, but it is only going to work if enough people adopt it (which is why I keep pestering you about this ). So, all I'm saying is: It would be easy to put ConceptFORGE on Bountysource, it might work out really well, and if it doesn't, you can go back to the way it was before.

Why is this calle a "grounded" experimental printer?

Must it be attached firmly to the ground?

Must it be electrically grounded?

Is GUS Simpson a type of "grounded" experimental printer?

usually the delta has "arms" on rods. This printer has them fixed.

@MattMoses: We should chat and figure out some bounties that would make sense. I guess I can I can list out all my projects I have on my list. The reason I would feel silly is that they are projects I would do anyways. However, it could allow people to reprioritize my list. It would also allow for me to attempt projects outside my current means like a jumbo Simpson.

@cozmicray: Grounded just means the arms attach close to the ground. I had to distinguish this from the weird upside down deltas like the Rostock and LISA.

@nicholas.seward

Qusetions about the BOM:

1. The stepper driver is listed at quantety of 1, would I actually need one for each stepper motor?
2. Is a forth stepper missing from the BOM for a filament driver? Or I could be blind...

Thanks,
Mike

My mistake 4 steppers and 4 drivers are needed. I will correct it when I get access again.

Thanks for the info, Nick! Got the latest Inventor downloaded and the assembly looks great. I used to do all my designs and Inventor and teach it to FIRST groups back in '02 and '03, and after spending time with OpenSCAD lately I'm tempted to go back to Inventor for the more complex stuff.

Questions here plus more questions inline:

- Is there a place with a bunch of pictures of assembled Simpsons? I recall seeing ones from MF NY but I can't seem to find them now. Trying to understand the wiring process.
- Are any more basalt beds available? If not available pre-cut, do you have any supplier links?
- What is your preferred "1/2" Flatstock" material for the top and bottom?

Quote
nicholas.seward
Ubis is what I use.

Not hot end; was wondering about the extruder. I see the answer in the BOM now - MBE extruder. Can't believe this thing is bearingless - has to add drag.

Quote
nicholas.seward
I want to eliminate that bridge. I will work to minimize it but I have to have access for the tuning peg. Even if you bridge poorly it has many layers before you get to a mechanically critical part.

I see now, the peg drops in. There's definitely room to minimize this (which I'm sure you know, but I'll list some ideas before making the mods):
- The distance from guitar tuner rotation axis to the start of the curved portion on the guitar tuner body that the faces the large bridge on motor_arm is ~7mm; the bridge is 30mm. So you could curve the edges and reduce the bridge size by about half.
- You could further reduce the need for a full bridge by inserting the guitar tuner twisted 90 degrees to its final placement so that the peg rotation axis is parallel to the long side of the motor arm, which would give a bit of extra space.
- Actually, is any of this needed? If you insert the tip of the peg twisted into the counterbored hole in motor_arm at a slight angle, with the half-cylindrical side first, then rotate it, you would need much less height on the motor arm. Anyone with a printed motor_arm, is this feasible? How much extra height does this get you? If it's ~7.5mm, then simple 45 degree edges may be sufficient to eliminate the bridge; the counterbore for the thicker cylinder on the tuning peg is 11mm deep.

Quote
nicholas.seward
What parts are you refering to when you talk about 608 overhanges? Everything should be printable but something may have slipped by.

I now see the sunken chamfer on all 3 hub pieces, so I see how to print those (chamfer down).

But on the shoulder, I don't see a way to print cleanly without support. You could add another sunken chamfer and print it vertically, but printing flat would of course be faster. If I print it with the longest flat side down, then the counterbores for the 608s are likely to deform at the top where the overhangs become extreme, plus on the far edge counterbore that is on the very bottom when assembled. Do you add any supports? Maybe you've got sufficient cooling to exceed 45 degree overhangs? Maybe the heating process to press in the bearing tends to hide any imperfections in this area. Thoughts?

Thanks! All SkyDelta development is on hold until I get a running Simpson... much cooler. More inertia, but not needing careful spring choices or needing to compensate for idler wrap will be nice.

I'm also interested in an basalt bed, any info where to buy would be nice.

regards
nargos

Quote
nicholas.seward
I already have the electronics for the beta kits but all of those will be getting paired with the mechanicals so I can do some consistent beta testing. It won't be until December when I will be doing this for real.

Beta Simpson Electronics
1 - Azteeg X1 $65
4 - SureStep Drivers $10
4 - Steppers $12
1 - 12or24 Power Suppy $36
1 - 200mm Round Heating Pad $14
1 - Hot End $16
1 - Filament Drive $12

Total: $230 (my cost)

These are prices I got from ordering for 40 machines. I imagine I can do much better when I order for 1000.

Hi Nicholas,

where can I buy a hot end for $16?

regards
nargos

Edited 1 time(s). Last edit at 12/01/2013 04:16PM by nargos.

Quote
brandonh
- Is there a place with a bunch of pictures of assembled Simpsons? I recall seeing ones from MF NY but I can't seem to find them now. Trying to understand the wiring process.

I have a few on Flickr. However, that's the original GUS design, and the new arms will be strung differently. . .
Simpson Set

Quote
brandonh
- Are any more basalt beds available? If not available pre-cut, do you have any supplier links?
- What is your preferred "1/2" Flatstock" material for the top and bottom?

Nicholas will have to answer the question about basalt beds, since they're custom cut to the Reuleaux triangle shape. There's no reason why a circular piece of something wouldn't work fine, though. And the prototypes have nice quality plywood for the top and bottom, nothing special.

Quote
brandonh
Not hot end; was wondering about the extruder. I see the answer in the BOM now - MBE extruder. Can't believe this thing is bearingless - has to add drag.

Based on short but painful experience I don't recommend the QU-BD extruder block. Print something else - Airtripper, or whatever suits you.

Quote
brandonh
I see now, the peg drops in. There's definitely room to minimize this (which I'm sure you know, but I'll list some ideas before making the mods):

Testing with a freshly printed arm and the tuner from the prototype, I think it would easily fit if the roof of the cutout were sloped at 45 degrees from the inside to the outside. However, my printer was able to bridge the gap very cleanly, and it would also be possible to just reduce the height of the cutout by about 9 mm and leave it square. It could be a bit narrower too, maybe 2 mm on each side.

As for the 608 holes, everything is printing cleanly, much better than the prototype when we had an unheated bed. Keeping the arm from warping seems to also make the hole stay more circular. The bearings just about fit in without heating, but I know better than to try that (the part will definitely split). If I get some time later I'll put the first arm pair together. . .

Quote
owens

Quote
sgraber
What is meant is you warm the bearing to the point that when you insert it into the bearing hole that it melts the plastic a bit so it slides into the hole.

I haven't tried that trick with bearings, though I've done it in combination with heating the plastic to get the pulleys onto the motor shafts; if the shaft isn't warm it tends to freeze the pulley in place before it's slid on far enough. And the pulleys are sensitive parts, so I don't want to overheat and distort them. I'm not sure it would work the the bearing holes though. What we found was a problem of out-of-round holes, rather than simply too small, and the plastic really needed to move in order to let the bearings fit in. One thing I did was keep an M8 bolt handy so I could run it through both bearings while the arm was still warm, to make sure they were lined up with each other. Also, I strongly recommend only trying to do one bearing at a time, and re-heating in between; you don't want to try to force them into place or the plastic will distort.

On that point, I strongly recommend trying the oven trick with a scrap piece first. You'll want to get an idea of what Nicholas means by the sound changing - it really does sound like a thud as the plastic softens, but it's also very easy to go too far and make the part too flexible. I've had success with the oven but have also done a couple of variations on his original idea. For the gear arms, I wrapped the gear section loosely in aluminum foil to try to keep it a bit cooler, leaving just the end exposed. And for for the most part I've switched to spot heating with a hairdryer rather than the oven. It helps to have a small enclosure around the part; I use a cardboard box that I could stick the end of the arm into and let the hot air swirl around it. I feel like the hairdryer gives better control, but in both cases I was able to get the bearings and bolts in place without warping the arms (as far as I can tell). This was all on the original GUS; we're still printing the new arms, so haven't tried putting any bearings in yet. . .

Bill.

A metals machine shop will Freeze bearings and bushings for there tight fits.

Ralph

@brandonh: I agree with all that owens said. I would try a shoulder print and see how it goes. (If you still have concerns I can upload a modified version for vertical printing.) The angled cylindrical hole is actually better than a horizontal cylindrical hole. I got the beds custom water cut. PM me if you are in love with them I might have a few in my dinning room. (My wife loves me. I know she does.)

I use an MBE but, as owens mention, you can do better.

@nargos: I get the hot ends from QU-BD. The price I got was the bulk/friend discount. They sell them for $30 normally. See about the beds above.

@owens: I could make it narrower but I didn't want it to be so tight that it would be hard to tighten the M2 screw.

For now I am not going to change anything on the motor arm. For fun, here is a pic of how I could remove the bridging. I don't like it enough to put it in the repo. PM be if you want me to send you the 1.5MB file.



Edited 2 time(s). Last edit at 12/01/2013 11:11PM by nicholas.seward.