Steps per mm

My prusa is using t5 pulleys with 1.9 degree steppers, and 1/8 microstepping. I'm using marlin with a gen6 board and my x and y steps per mm are both set to 29.09. When I measure the movement on both axises there seems to be a variation between the two.

What causes a difference between actual / calculated steps per mm?

Printed pulleys?

I don't think you have 1.9 degree motors, it has to divide exactly into 360, so I think you mean 1.8 or 0.9

With 1.8 motors and and 1/8 stepping there are 1600 steps per rev. The distance travelled is the number of teeth times the belt pitch. With 8 tooth T5 pulleys you would have 40 steps per mm. I can't see how you would arrive at 29.09. Maybe 12 tooth pulleys and a 0.9 motor giving 30 steps / mm.

The only difference between calculated and actual can be the belt pitch, which must vary a little with tension. How are you measuring it? Is there any backlash?

---

[www.hydraraptor.blogspot.com]

If you measure distances that are a whole number of revolutions of the pulley then that would eliminate inaccuracy due to printed pulleys.

---

[www.hydraraptor.blogspot.com]

Thanks for the responses,
@nophead I followed your advice and with a demanded movement of 100mm I believe I have the accuracy down to +- .4mm on the X and Y.

Ive also ordered a couple aluminum pulleys to replace the printed ones.
[sdp-si.com]