Y-Axis Gantry Belt Tension

[Joseph Adams]

First a little bit of history.  I own and have used a Rep2 for over a year now and I recently (December) purchased a 2X.  I wanted to be able to print PLA in the 2X so I struggled through January with making various new experimental filament guide tubes.  Once I finally solved that problem I focused my attention on making some real parts with the machine, only to realize that it isn't building very accurately.  Specifically, cylindrical features or pretty wonky, and dimensionally out of tolerance.  I've already eliminated all the usual suspects and feel fairly confident that it's a problem with belt tension on the y-axis, and specifically that its too tight.  The problem is, there isn't any reasonable way to adjust the y-axis belt tension (referring to gantry belts, not stepper belt).  I was considering cutting the belt between the attachment points on the carriage and then loosening it by one tooth, but obviously this is a non-reversible change.  I'm wondering if anyone else has done this and if it's advisable or not?  It seems that replacing these belts would be a huge PITA.  Alternatives?  Thanks.

Joe

[Ryan Carlyle]

Can you post a picture of what you mean by wonky cylindrical features? Loosening belts just sounds like a bad idea. Some circle deformation is normal because of Y axis backlash, but that should be the same on your R2 and R2x.

[Joseph Adams]

These were printed at 5 mm/sec, code generated by Makerware.  I’ve also trialed Kisslicer with the same results.  When I move the Y-axis manually by hand, I can feel variability in the amount of force required to move the carriage along the axis. The “loose” and “tight” spots occur consistently in the same locations (I think) but not at consistent intervals.

--
You received this message because you are subscribed to a topic in the Google Groups "MakerBot Operators" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/makerbot/nLCX60BHzIQ/unsubscribe.
To unsubscribe from this group and all its topics, send an email to makerbot+u...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

[Jetguy]

Given the number of folks with a failed 2X extruder carriage in the belt clamp area, you are probably incorrectly diagnosing the issue and assuming Y is the problem when it probably is your X axis.

To unsubscribe from this group and all its topics, send an email to makerbot+...@googlegroups.com.

[Jetguy]

https://groups.google.com/d/msg/makerbot/rIMvkf9RKHs/uHWVEhy6IC8J

"have discovered recently that there’s a play between the carriage and the x belt on my 6 months old replicator 2x.

i don’t have the tools to measure by how much, but my guesstimation would be around the 0.5mm-1mm."

 

Or this thread https://groups.google.com/d/msg/makerbot/4GwROs4dw90/owIJf4ZZiTMJ

"At our office we bought a replicator 2x a few weeks ago and we were having problems printing nice circular objects (see attached picture: "before"). In the end we found a solution to this problem. I thought it might be nice to post here how we solved it so that people who are having the same issue know how to solve it. Besides this I am quite curious how many others are having this problem.

We noticed that we had quite some play (about 0.6mm) between the carriage and the timing belt, which caused quite some backlash on the x-axis. I contacted makerbot support to explain the issue that we were having and sent them a video showing the play between the carriage and the timing belt (see attached video: "play carriage - timing belt")."

 

 

[Joseph Adams]

 

Thanks Jetguy, you might be right.  I just checked the x-axis lag and there is definitely some slop (about .3mm my guess) between the belt and the carriage.  I’ve found a small clamp to test it temporarily and I’m waiting for the machine to print as I type.  I’ll post results shortly.

[Ryan Carlyle]

The slightly flat section on the left and right side of each circle is an indication of X backlash. When the belt reverses travel direction, the carriage keeps moving in a straight line for a moment.

[Joseph Adams]

I was really hopeful when I found the slop in the X-axis but after eliminating the slop, the problem remains...although it’s different.  I’ve attached a picture of a single cylinder I printed.  I’ve double checked the backlash on the y-axis and it seems to be pretty solid.  Ryan, I totally agree that the flats “can” be an indication of X-Backlash, but it’s a two way street.  You could get the same result if the Y-axis suddenly moved faster, which could be supported by my observation that there are areas of reduced resistance that allow the y-axis to move more freely during some parts of its travel.  This is the theory that lead me to believe that the y-axis tension was too tight (had this issue with my Rep2)…or at least that there is a problem with the y-axis.  I’m not saying that I think I’m absolutely right, just explaining my rationale. 

[Joseph Chiu]

Just to be sure, you've turned off any backlash compensation in the slicers, right?  It sure looks like you still have Y backlash.

There's also a possibility that the Y gantry is not square -- it's possible that your left side is "lagging" the right side so that in the initial start of the Y movement, as X is in full swing, you are actually moving the gantry, but the X is travelling along the tilted Y axis line?

On Mon, Feb 17, 2014 at 3:24 PM, Joseph Adams <jad...@zdr.com> wrote:

I was really hopeful when I found the slop in the X-axis but after eliminating the slop, the problem remains...although it’s different.  I’ve attached a picture of a single cylinder I printed.  I’ve double checked the backlash on the y-axis and it seems to be pretty solid.  Ryan, I totally agree that the flats “can” be an indication of X-Backlash, but it’s a two way street.  You could get the same result if the Y-axis suddenly moved faster, which could be supported by my observation that there are areas of reduced resistance that allow the y-axis to move more freely during some parts of its travel.  This is the theory that lead me to believe that the y-axis tension was too tight (had this issue with my Rep2)…or at least that there is a problem with the y-axis.  I’m not saying that I think I’m absolutely right, just explaining my rationale. 

--
You received this message because you are subscribed to the Google Groups "MakerBot Operators" group.
To unsubscribe from this group and stop receiving emails from it, send an email to makerbot+u...@googlegroups.com.

[Joseph Adams]

Just verified no backlash comp in Makerware.  I found it a little interesting however that the default value for “backlashY” is 0.09, but “backlashX” is 0.  Maybe it’s nothing, just kinda weird that both values aren’t 0.      

[Joseph Adams]

Forgot to mention, I have checked the gantry alignment by pushing the y axis all the way to the front and confirming that both left and right sides are touching the stops simultaneously. 

 

[Ryan Carlyle]

Your latest circle looks better. It still shows some Y backlash though. You cannot completely eliminate Y-backlash on a stock Replicator, because of the three-belt Y motion mechanism. There's just too much slop when you add all three belt/pulley linkages together. It's a consequence of a controversial design decision to make the machine envelope smaller at the expense of precision. This is why Makerware includes Y backlash (but not X) in the default backlash compensation settings.

Here's each step of the Y motion mechanism:

1. Y stepper drive pulley to Y drive belt
2. Y drive belt to transfer pulley
3. Transfer pulley to right X-end belt
4. Right X-end belt to Y axis sync shaft pulley
5. Y axis sync shaft pulley to left X-end belt

As you can imagine, by the time you get to step 5, there's some unavoidable accumulation of slop.

A better design would be less compact but have less Y backlash:

1. Y stepper direct-driving the Y axis sync shaft, which has pulleys to drive the left and right X-end belts

So now instead of three belt/pulley connections on the right side and five on the left side, we have one for each side. 

Now that I'm thinking about it, I bet printing on the right side of the bot would have slightly less Y backlash than printing on the left side. Time to experiment!

[Ryan Carlyle]

If there's a difference between left side Y backlash and right side Y backlash, it's imperceptible on my printer. (Look at the small step between the box and hole in the pics -- that's backlash.) 

If anything, the right side looks slightly worse.

[Levia Zeroth]

Hi, just wanted to chime in with my experiences,
Finished setting up a brand new Replicator 2X and got it printing some cones to test, and noticed it was failing to do accurate circles, as many others have found it was down to the X axis clamp not tightly holding the belt, my unit had around 0.8mm worth of play before it would pull in an opposite direction, this was effecting both circles and the uniformity of solid pieces.

I took an old craft knife blade, snapped off the end so it was just a 0.5mm thick rectangle of metal, gave the edges a bit of a rub with a file to smooth them out a bit then used it as a shim between the plastic clamp and the top of the belt, as seen in my attachments.

I have also included a before and after picture of the top and bottom of the cones, the belt no longer has any play and printing is much more consistent and much, much easier.

Joseph, if you're still having problems manually slide the extruders back and forth on the X axis slowly, and watch the top section of the belt closely to see if it moves back and forth along with the rest of the unit, if theres any lag (belt moves after the unit has already moved slightly), then this fix may help you too.