# TAZ5-ish X-Y accuracy issue
**URL:** https://forum.lulzbot.com/t/taz5-ish-x-y-accuracy-issue/27159
**Category:** LulzBot TAZ
**Created:** [December 24, 2024, 11:03am UTC](https://forum.lulzbot.com/t/taz5-ish-x-y-accuracy-issue/27159 "2024-12-24T11:03:36Z")
**Posts on this page:** 5
**Page:** 1
### Author:  [@gossamer](https://forum.lulzbot.com/u/gossamer)
#### Post date: [December 24, 2024, 11:03am UTC](https://forum.lulzbot.com/t/taz5-ish-x-y-accuracy-issue/27159/1 "2024-12-24T11:03:36Z")
I’ve printed [xy-hole-compensation](https://www.printables.com/model/775001-xy-hole-compensation-calculator-for-orca-slicer-et/files) model in PLA to verify and tune to print more “accurate”. What I noticed is that X accuracy is not too bad. I am off by 0.5% (14.92mm vs 15mm) but Y axis is bit worse. Its off by 2% (14.7mm vs 15mm). I would expect similar shrink error (1% or so).
I upgraded X axis to v-slot, and Y axis is mostly original (only added belt tightening and replaced old belt).
I assume thats due to belt tension ? Although it seems tight enough. Cant really “hear” the difference when strum on the belt.
Just wonder what is expected accuracy that ppl are getting ? Should I just take it as it is or can it be fixed somehow ?
---
### Author:  [@Wrathernaut](https://forum.lulzbot.com/u/Wrathernaut)
#### Post date: [December 24, 2024, 10:29pm UTC](https://forum.lulzbot.com/t/taz5-ish-x-y-accuracy-issue/27159/2 "2024-12-24T22:29:48Z")
If it’s printing small, check for a loose grub screw, play in the belt mounts on the bed.
If you print a 10mm cube, does it come out 9.85mm or 9.7mm? with a 30mm cube, does it come out 29.4mm, or 29.7mm? If it’s consistently short the same amount it’s a backlash issue, but correcting it with backlash correction in firmware is not a good idea if it’s something mechanically wrong that can be fixed with tightening things up.
---
### Author:  [@gossamer](https://forum.lulzbot.com/u/gossamer)
#### Post date: [December 28, 2024, 4:35pm UTC](https://forum.lulzbot.com/t/taz5-ish-x-y-accuracy-issue/27159/3 "2024-12-28T16:35:30Z")
Seems to be consistently off by 0.3 mm in Y axis. I printed 30mm (29,7) cube and 40mm (39.7) cube. something must be loose then … I did notice wobble
on the test cubes.
---
### Author:  [@zenotek](https://forum.lulzbot.com/u/zenotek)
#### Post date: [December 28, 2024, 5:50pm UTC](https://forum.lulzbot.com/t/taz5-ish-x-y-accuracy-issue/27159/4 "2024-12-28T17:50:04Z")
You will not derive anything meaningful from printing cubes. Well, you can, if they are wildly off the size requested. But since each of those cubes was off by the exact same amount, it’s probably flow rate or filament as opposed to steps being off or something mechanical.
---
### Author:  [@Wrathernaut](https://forum.lulzbot.com/u/Wrathernaut)
#### Post date: [December 28, 2024, 7:04pm UTC](https://forum.lulzbot.com/t/taz5-ish-x-y-accuracy-issue/27159/5 "2024-12-28T19:04:41Z")
If it were off in the same amount in both the X and Y, it could be flowrate, but where it’s limited to the Y axis and consistent, I’d say it’s a loose pulley, belt mount, or could be down to the mount of the toolhead letting it get pulled slightly forward or backward.
You can use the manual axis controls to do some more testing. Move it in the +Y direction until you can clearly see the pulley at the back. Make a mark on the end of the motor axle and pulley gear with a permanent marker. Now move the bed in the -Y direction for just .2mm. Do the lines still line up perfectly, or have they shifted? If they shift, it’s a loose grub screw on the pulley. If they remain perfectly aligned, but the bed didn’t move, it’s a loose belt, belt mount, or maybe even a loose bed. If everything moved in sync and the proper amount, try moving in the +Y direction for just .1mm and see if the bed moves. It might take some calipers to verify.