OpenY - mendel Y carriage redesign

(I don't want to turn anybody off with the OpenN naming conventions; I'm not saying that other designs aren't open! it started as a description for the x carriage redesign, but i kinda like it so I'll keep using it.)

I'll be fleshing out the wiki page with more instructions and pictures over the next week or two.
Open Y Carriage
Thigiverse

I've printed this out myself and it is working great on my mendel. it is a fully drop in replacement for the old y carriage. (I suppose you'll have to make a new y optoflag, though.)

You probably won't be able to print upside down, which I'm sure will disappoint many.

on the other hand, this design gets rid of the squashed frog!

I would like to hear feedback from people who print this out to try it. I have additional, optional parts waiting in the wings in case people encounter problems. There is some more discussion Here, but i thought that this new part needed it's own topic for visibility.

--

Now i don't know how well this will work with heated beds, as it is almost entirely plastic. What kind of temperatures do people get under heated beds? specifically, how hot does the plastic get under an insulated heated bed? If it's too hot it might necessitate a specific redesign.

I really like your designs so far, please keep up the good work

I'll probably test it sometime soon

---

---

reprapworld.com
PLA, ABS, PETG multiple colors (3 and 1.75 mm) €16,49 / kg
Megatronics 3.0 €79.99 / Minitronics €37,19

Nice one Buback.

Hi Buback,

Really enjoying your OpenX design. I'm just building my first 'non-strapped' mendel, and I'm thinking about running with your open axes entirely.

The only pressing dimension i was wondering about was the length of the 8mm rod on your OpenY. I assume you need this to be specific so that the holes line up with the print bed holes. I think this is the really pressing parameter as far as running OpenY from your documentation.

Take care!

-ED

ah yeah, thanks. i've been laging behind with the documentation this week.

it's 140mm. Now, with wood (or pencil) it's a little floppy until you attach the bed, so don't be alarmed. If you don't mind the weight or cost, some metal solution is probably better/ more rigid. I tried getting 8mm(3/16th5/16ths- Thanks Cefiar) aluminum tube or rod, but couldn't find any for a cost i could justify, but some left over threaded steel rod would work just fine.

and the y axis isn't easy to photograph, but i'll get some pictures up this weekend.

edit: the length is not as critical as you might think, though. the bed mounting screws hold the pieces where they need to be; the rod is mainly there to transfer the force from the motor/belt to the carriage.

Edited 2 time(s). Last edit at 02/24/2011 06:15PM by Buback.

Thanks for that.

Another request: If you happen to get an estimate of the BOM (screw lengths etc.) It would be much appreciated. Just printed out openX this morning, and I'm doing openY now. I think I can follow your images to get the X carriage but I've not visualized the Y yet.

I won't have electronics to run this system for another week or so but once that happens I expect to be giving you feedback.

Thanks again for your efforts!

ok i just added the instructions to the wiki page, along with a BOM. Pictures will be coming soon!

Note: 8mm is just slightly bigger than 5/16th of an inch (~7.97mm), not the value previously mentioned by Buback in the thread. If you're using imperial, you want 5/16th inch rod.

ah yeah woops. i hate imperial...

Got the OpenY assembled this afternoon. I had to use a slightly shorter rod (~125mm), but it was threaded 8mm, and i'm not sure it was willing to go all the way into the mounts for it. Then again, I was aligning it to the print bed that TechZone supplies, and i'm not sure it has the same hole placement as you were designing for. The 6-bearing 'snap-on' side is looking sharp, i'm still tweaking up the 180 side. The smooth rod i'm attaching it to is actually 5/16ths instead of 8mm, but that doesn't seem to be a show stopper.

More information to come - I've got some metal gears stuck to my stepper motors that are keeping me from moving onward....

looks great!

I just left the belt attachment float. in other words, i didn't use M8 nuts to hold it in place.

I'm not sure if i like having only one piece for the belt mount and the optoflag. If i split them into two pieces, it would be easier to line them up separately, and you might only need a straight, unbent flag. Or maybe i could move the endstop to just above or below one of the y rails.

I put my opto endstops on the y-bar clamps, and attach flags underneath the bed to trip them, so I probably won't have any feedback as far as using that piece for the opto. I was only going to use that for the belt.

The M8 nuts aren't really there for any particular reason, I may remove them later. If wobble from that piece is an issue in the future, they'll be nice to have there.

Buback, Do you think you could add an images of the way you've attached your belt to the OpenX? I realize this is somewhat off topic for this thread, but I've been debating a few ways of doing it (wondering if one of them was the origional intent).

the opto flag for my y axis isn't working yet, mainly because i haven't had time to align it, which is probably a good enough justification to make it a separate piece. having it as a separate piece would make it trivial to adjust, if you are keeping your endstop in the default place.

So i'll do that later tonight. I think it's probably even better to just move the endstop placement entirely, like you did.

here's a picture of how the belt is fed


I would recommend creating the loops in the belt ends so that they are VERY close to each other, and not like in the picture. you also want the attachment point to be right on top or adjacent of where the belt joins, or on the opposite side of the carriage, so that the zip ties don't get pulled around any bearings or into the motor.


This is for the regular x motor mount, with the teeth facing away from the carriage. If you use the new motor mount, i think it will work the same, but the teeth of the belt will grab the arms near the bearing, instead of catching on the center protrusion.

Edited 1 time(s). Last edit at 02/28/2011 02:11PM by Buback.

My new Mendel is running OpenY axes and they seem to be working well. I did have to heat/warp the 330 degree bearings so that they would make better contact with their bar. Maybe if you shift their holes up slightly this would be unnecessary?

I haven't added more detail about its OpenY yet, but you can see it running here: [reprap.org]

Wow, very cool!

ok i'll play with the files some. I need to learn openSCAD so i can remake these carriages in a parametric way.

keep me posted about any other issues you have.

Actually, after watching it print a few more open X carriages and whatnot, I notice that the OpenX bearings have a similar issue making contact, causing a bit of wobble as it is pulled back and forth. I'll probably just warp those into place and they should be fine, but again, maybe 'raising' the hole on the 330 bearings would help.

>Actually, after watching it print a few more open X carriages and whatnot, I notice that the OpenX bearings have a similar >issue making contact, causing a bit of wobble as it is pulled back and forth. I'll probably just warp those into place and they >should be fine, but again, maybe 'raising' the hole on the 330 bearings would help.

Just a heads up, I printed the openX a few days ago and had the same problem. Just waiting on a hardware delivery to provide the last few parts so I can test the carriage properly and give more thorough feedback. I'm still a big fan of the design, reminds me of an automobile chassis.

Woodsmoke
R:-8

in my model, i've assumed a 1.5 mm gap between the bearing and the plastic (for the arm bearings) in order to accommodate a washer. i was just guessing on the thickness, and assumed that it might vary wildly anyway. however, i just looked at m4 washer dimensions and the spec sheet says .8mm.

I guess it should have 2 washers between the bearing and the plastic then, if the specs are right. I only have one washer in there now, so i guess i'll try putting two washers on and using longer screws to see if this makes a better fit.

---
I think if i move any of the bearings, it will put some stress on the plastic when it's put on the rails, spring loading it. it will hopefully be light, but just a heads up. i'll upload the modified arm.stl in a bit for you guys to try out.

edit:
Added the files. This moves the bearing on the arm about .2 mm into the rail (i.e. in towards the center of the smooth bar.

I also changed the belt "tooth" a bit because i get some slipping sometimes. i only get it when i move the carriage by hand, or if i'm going way too fast. this has a second smaller tooth that will hopefully fit in between the teeth on the belt. if this isn't adequate, i'll have to change the belt attachment to make it lock down on the belt

Edited 1 time(s). Last edit at 03/30/2011 02:31PM by Buback.

More updates:

The first OpenY/OpenX system that i referenced is still operating well. Both needed a bit of warping to get those angled bearings to contact as I mentioned.

We have another system that is about to start printing - it is a TechZone LC Mendel with another OpenY to replace a mishandled TZ y-stage. God bless the froglessness of your design. Once this is running we may have some more input.

My own Wooden mendel is getting a replacement too. I'll give you some updates after we have them all running.

Can you design a Y-Axis belt harness piece without that optoflag? I chopped it off of the other piece i made, which allows rotation (a good thing for my setup). Those nuts I put on either side actually do wonders for the stability of the belt harness.

Hopefully we will have some feedback regarding that dual extruder OpenX before the end of the semester.

Cheers!

yes i'll work on the belt harness some more. I know it's probably the worst designed part on the OpenY. I'm open for suggestions or drawings of how this could be done better.

-I like the quickness of using zip ties to attach the belts, but worry that it might damage the belts or effect print quality because zip ties don't have the steel wires that the belt has.

-the belt weaving technique works alright, but i can make it slip (not when printing, thank goodness, but when push/pulling by hand for testing). is going back to a secure screw-down system better?

-I also need to find out where, in the z direction, the belt is in relation to the carriage or y rails. I pretty much guessed and made it 10mm or something from the crossbar to the belt. If i get a more accurate distance, the belt will travel in a flat path, and rotation should be eliminated. (Or, i could make it with adjustable height. Just had an idea...)

i've attached some pictures of my basic idea. you slide the part that looks like a gasket onto the crossmember, and then the 'forks' piece attaches to the belt. the two pieces are joined together with some m4 bolts. by loosening the m4 bolts, the y carriage can be popped right off, leaving the forks piece attached to the tensioned belt.

This makes it easy to tension the belt from the top without the carriage being in the way. (on the other hand, you'll still need to turn the reprap over to tighten the m4 nuts/bolts.)

Honestly, I think the zip ties and the harness works well, only suggesting that the opto hole removed. Having nuts on either side of the harness makes it stable, and I haven't noticed any problems due to vertical displacement of the y-belt.

To me, the only critical issue is those 330 degree bearings not making proper contact with their bars. I think the issue would be solved for both openx and openy simply by 'raising' those tilted holes a few mm with respect to the print surface.