Circles not coming out Circles

My experience is:
-alternative X motor bracket design don't give enouth attachment between belt and pulley
-first X motor bracket design is said to brake the belt after some printing time
-if we often have the same belts, we don't often have the same pulleys, and there often is backlash between belt and pulley (various backlashs for all of us)

So:
-I added 2 bearings to the alternative X motor bracket (photo into my August 09 post, those bearings are fixed with 2 of the motor's screws) to get a better belt/pulley attachment, without the too complex belt circulation of the first X motor bracket design
-and I modified firmware to get squares or circles correctly printed, for me (just for me! ):

#define X_STEPS_PER_MM 9.842
#define Y_STEPS_PER_MM 10.047

I am using nophead's (http://www.thingiverse.com/thing:2079) pulleys now, they were printed on a commercial printer here at university and they fit perfectly on my motors and don't have any wobble whatsoever. I will try the extra bearings for the alternative x motor bracket think I have some 623s laying around somewhere. I will also add extra washers to try and remove any friction on the belt. Hope I can figure this out soon, my prints are starting to get real nice, just need some circular circles...

François Delègue Wrote:
-------------------------------------------------------
> My experience is:
> -alternative X motor bracket design don't give
> enouth attachment between belt and pulley
> -first X motor bracket design is said to brake the
> belt after some printing time

I am using the first X motor bracket and yesterday my X axis belt just broke! the metal wires broke! Maybe it was due to me driving RepRap at 725mm/s for testing R2C2.


> -if we often have the same belts, we don't often
> have the same pulleys, and there often is backlash
> between belt and pulley (various backlashs for all
> of us)

Until now I used 2 types of belts, one with metal wires and other without them. I am getting the not rounded circles with both. Hey, but my pulleys were cheap from Ebay/China!!! I hope to exchange the pulleys... thanks for the tip.

---

---
New cutting edge RepRap electronics, ARM 32 bits @ 100MHz runs RepRap @ 725mm/s:

[www.3dprinting-r2c2.com]

Quote

I am using the first X motor bracket and yesterday my X axis belt just broke! the metal wires broke! Maybe it was due to me driving RepRap at 725mm/s for testing R2C2.

No the first x bracket will break the belt after printing 4 Mendels ( ~ 200 hours) at moderate speeds. With the improved X bracket I have not had a belt snap.

If you have the metal pulleys from Hong Kong [www.ebay.co.uk] try holding the pulley in one hand and sliding the belt in the other. It moves a significant fraction of a millimetre because the teeth of the pulley are a lot thinner than the gaps in the belt. Also the holes are not well centred.

I found this type: [www.ebay.co.uk] from Bulgaria doesn't give backlash. The teeth are not the right profile, the valleys between them are semicircles, but the belt teeth wedge into them in a way that does not slip. Out of three though, two had badly centred holes.

I have been unable to find off the shelf pulleys with the correct tooth profile and 5mm bore. I can get 10 tooth T5 and 16 tooth T2.5 but they need drilling and reaming on a lathe and then tapping for the set screw.

Edited 1 time(s). Last edit at 11/04/2011 05:21AM by nophead.

---

[www.hydraraptor.blogspot.com]

nophead Wrote:
-------------------------------------------------------
> I am using the first X motor bracket and yesterday
> my X axis belt just broke! the metal wires broke!
> Maybe it was due to me driving RepRap at 725mm/s
> for testing R2C2.
>
> No the first x bracket will break the belt after
> printing 4 Mendels ( ~ 200 hours) at moderate
> speeds. With the improved X bracket I have not had
> a belt snap.

For now I am back to the belt without metal wires. It will not break and at least the not round circles seems to not be related to that belt.


> If you have the metal pulleys from Hong Kong

I am sorry, I thought "pulleys" meaning the "bearing". I want to sat my bearings were from Ebay/China. Don't know if bearings have any role on this problem or not.

I just recorded a video printing a part with infill made with circles. It's visible this problem: [blip.tv]

Edited 1 time(s). Last edit at 11/04/2011 07:10AM by casainho.

---

---
New cutting edge RepRap electronics, ARM 32 bits @ 100MHz runs RepRap @ 725mm/s:

[www.3dprinting-r2c2.com]

How did you generate that circular infill? Are they definitely supposed to be true circles?

---

[www.hydraraptor.blogspot.com]

nophead Wrote:
-------------------------------------------------------
> How did you generate that circular infill? Are
> they definitely supposed to be true circles?

Yes, I can see true round circles on SFACT!! Skeinforge have that option for infill since a few versions ago! did you missed it? :-)

---

---
New cutting edge RepRap electronics, ARM 32 bits @ 100MHz runs RepRap @ 725mm/s:

[www.3dprinting-r2c2.com]

I use a very old version of SF. I try not to update software as it always gets more and more bugs in it.

---

[www.hydraraptor.blogspot.com]

nophead Wrote:
-------------------------------------------------------
...
> I have been unable to find off the shelf pulleys
> with the correct tooth profile and 5mm bore. I can
> get 10 tooth T5 and 16 tooth T2.5 but they need
> drilling and reaming on a lathe and then tapping
> for the set screw.

This is how I wound up using GT2 belts and pulleys from SDP/SI. 5mm bore and a smaller pitch diameter than I can get with the usual belts (T5, etc.).

Ok, so long story short, my problem was indeed due to friction of the belt and the large penny washers meant to keep the belt in line. There was so much rubbing that the penny washers had been worn on those sections. In order to fix the problem I used longer bolts and placed another 624 bearing under the first. I did this on all four of the x-axis belt guides and also on two of the y-axis belt guides on the non-motor side. Thank you nophead for your contribution to this project and your support on these forums!

A quick fix to belt friction is to hit the fender washers with some teflon/silicone lubricant. The issue with this is that when the lubricant makes its way onto the pulleys, it'll increase whatever backlash potential that they have, so you also need to make sure that you have good pulleys.

The odd thing is I don't have a problem with the Mendel 624 pulleys, I only got this problem on Prusa with the 608 pulleys. The gap between the bearing and the penny washer is much bigger and the bearing has a bigger chamfer so the belt can slip over the edge and jam between the bearing and the washer. I replaced the 608 bearing on the X axis with a 624 and it works fine.

---

[www.hydraraptor.blogspot.com]

casainho Wrote:
>
> [farm6.static.flickr.com]
> d29b09a.jpg
>
> I wish hardware developers can make a machine that
> people as us can assembly and don't have this
> problems, rendering useless the investment made.

I had a similar problem where the circles comes out flat on Y axis similar to your pic.

Solved the problem: the Y carriage was binding at both ends. It moves smoothly in the center area area but the binds near Y = 0 and Y=200. To confirm, I turned off the motors and move the bed with my fingers. And the bed will not zero; moves smoothly in center; will not move to the end.

I solved the issue by realigning the Y smooth rods so that they are parallel horizontally and vertically. Adjusted the bed spring as well so that the smooth rod remains parallel.

korgon Wrote:
-------------------------------------------------------
> Ok, so long story short, my problem was indeed due
> to friction of the belt and the large penny
> washers meant to keep the belt in line. There was
> so much rubbing that the penny washers had been
> worn on those sections. In order to fix the
> problem I used longer bolts and placed another 624
> bearing under the first. I did this on all four
> of the x-axis belt guides and also on two of the
> y-axis belt guides on the non-motor side. Thank
> you nophead for your contribution to this project
> and your support on these forums!

Look at my post on the first page ...4th Post

I've got the sort of the same problem as the topicstarter, but not that drastic

My circles are pretty good but there not circles, they kinda oval (lemon shaped) i have made a test piece to show it:



As you can see here, the 'ovals' are tilted slightly to the top left corner.

These things i have tried to fix my problem:
- Removed the backlash witth belt-tensioners ( like this: [i51.tinypic.com])
- Made sure that the motors had enough current (pololu drivers)
- Lubricated my axes.
- Put small 'guides' on my washers so that the belt run's in the center of the bearings, so its not rubbing towards the borders.
- Calibrated my firmware, on small sizes they are dead on (0.02 ish diverence) and on lager objects like .20 to .40mm
- Used commerical printed pulleys for my X, Y axes (and extruder) High infill and really nice quality.

Has anybody an idea what i'm doing wrong? Or what i should do to elaminate the 'lemon like circles' problem?

Edited 1 time(s). Last edit at 11/23/2011 11:44AM by Bramboterham.

This is similar to the problem I had, but looks like you've tried everything... In the end for me it was all about the belt rubbing.

Well after a day of testing/calibrating i've got an idea what the problem seems to be...
My eSteps where correctly configured (so i thought) but i saw that X has a value of 80.6something en Y had 79.8something... That is not possible because the belts are the same... the pulleys came from the same machine and there is (nearly) no backlash... So i set them to the same value...and presto! got circles again

Bramboterham Wrote:
-------------------------------------------------------
> - Put small 'guides' on my washers so that the
> belt run's in the center of the bearings, so its
> not rubbing towards the borders.

Are your guides fixed or free to rotate? If they are fixed I would suggest the belt is dragging/rubbing enough to introduce lash. I couldn't print bearing guides like [www.thingiverse.com] and you probably won't be able to make accurate ones either, as you need to be able to print smooth, accurate, round circles, or it will introduce other problems. I did this: [forums.reprap.org] As suggested in that thread, you could also try using two bearings instead of one. Remember to make sure both Y- AND X-axis belts are not binding!

some flat-circle fixes that I've tried on my printrbot:

>tightening belts: helpped a bit
>re-installing the hotend: had a decent effect
>raising the y belt to avoid friction: worked miracles