Cylinders aren't round

[tk1138]

My prints with different speeds (different profile speeds in s3d; the slower was at 800mm/s for print speed and 2000 for XY movement, the fast was at 2200 for printing, and 12k for movement) didn't' show significantly different results from eachother, nor improvement from my last print.
I've used PETG, ABS, and PLA and seen this. For most prints, it's not a deal-breaker, but I'd like to track this issue down as it's the biggest variance and causes prints to be short by ~1% at 100mm in addition to being out of true round. Here's my XY test object: http://www.thingiverse.com/thing:1979526

I did tighten down the Y axis belt clamp bolts before my last print, and the spider-to-Y-Axis bolts were all tight. I don't feel any mechanical slop between the bed and hot end in any direction (though the hot end is cold/off).
It is possible I have a different issue than the OP, though the issue that I signed up on this board for has very similar symptoms.

[steve220]

Thanks to all for your help and thoughts.
Ednisely: The set screw is firm against the flat part of the shaft.
Jules: The rail was cleaned and lubed.

Just as some additional info, I took the build plate and heater plate off to check for loose screws both on the part that mounts to the slide on the rail and the part that holds on to the belt. The belt screws tightened a little, so I was hopeful that was it. Unfortunately, it made no difference in the print.

I've opened a ticket with MakerGear.

[steve220]

OK, I think this is fixed
While messing with this issue, I tried pinching the Y axis belt together with my fingers and moving the print bed forward and backwards and I felt a little play. Even though I had tightened the bolts for the bed and the piece that grabs the belt, there was still evidently play between the belt and the thingymabob that connects the belt to the print bed. I ended up shimming between the back of the belt and that piece to force the teeth to engage tighter. I used two layers of paper as a shim. Once everything was back together, the play was gone.
In the pic below, the top row of parts were all printed before shimming and the bottom pieces were printed after.

I'm pretty happy with these. Hopefully it will stay fixed!
I know I should have taken pictures...

[NavyDT]

Would you mind posting a picture of exactly what you did so others might be able to follow along?

[steve220]

Sorry, I didn't take any pictures. But, as it turns out, I found a site that shows exactly what I did. Part way down the following site, this guy did the same thing, and he took pics!

http://www.rcdon.com/html/makergear_m2_project1.html

[ksevcik]

Oh man, that's a good catch there. Not a place any of us were thinking to look for slop, clearly. I wouldn't think this is a wear issue, but it does sound like something to check for on the X and Y axes out of box and whenever you change belts.

[NavyDT]

holy cow, that dude is meticulous! yikes what have I gotten myself into

[tk1138]

Shimming with paper worked for me as well. I used some heavier paper that was hard to slide in, Y-dimension accuracy is on par with X, and cylinders are coming out well.
Thanks!

[steve220]

Glad it's working!

[tk1138]

My glass is (was) sliding around as well. I just noticed it when watching some support material be printed.
I shimmed the front right clip with some silver anti-static material, but I suspect some either some printed shims or cut silicone rubber shims will be a better fix.