Accuracy problems in Y
Hey Everyone!
I have some trouble with the XY accuracy of the Prusa i3MK2. I use the printer in our company to produce prototype jigs and gauges and accuracy is very important in this case.
I printed two test pieces in PLA with standard Prusa settings in Slic3er:
150 x 150 came out with 149.6 in x and 148.9 Y.
I don’t want to focus on the X axis deviation right now, but over 1mm deviation in Y on a 150mm piece is too much for the intended use. Maybe I expect too much (I’m relatively new in 3D printing), but we have a second Prusa i3 in an associate company, and they don’t have such big accuracy problems.
Is there a possibility to calibrate the printer / stepper motors for more accuracy? I already did the normal calibration, without getting better results. I’m thinking of determining a correction factor or something like that.
Additional info that may be useful:
-The pulley on the Y-motor seems to spin slightly eccentric. The drill hole is 0.05 mm bigger that the motor shaft, maybe it gets some tension, when it gets screwed in place. The motor shaft itself spins perfectly round!
I would be very thankful for any advice, detailed calibration information and tips.
Have a nice day!
Tobi
Re: Accuracy problems in Y
Tobi
1% is very good for this type of printer, however, as long as it is consistent, it is possible to adjust.
1. Belt tension - it is possible that the belt is too loose, causing the Y axis to move slightly less that it should
2. You can include an M92 command in the start-up GCode to adjust the steps: http://reprap.org/wiki/G-code#M92:_Set_axis_steps_per_unit (default is 100 for both X and Y)
Peter
Please note: I do not have any affiliation with Prusa Research. Any advices given are offered in good faith. It is your responsibility to ensure that by following my advice you do not suffer or cause injury, damage…
Re: Accuracy problems in Y
That’s it! Thank you very much, Peter! Right now I’m able to print 150 x 150 x 5 square with just ±0,1 tolerance. That’s pretty good so far! I will observe the results in the next days and weeks. Maybe I’ll try a fine tuning, supported by a professional 3D measuring machine, for the final nerdy kick 😉
Here are some more detailed information for everyone with similar problems:
Because of the 100 steps split it’s an easy percentage calculation and a simple additional line of code in the gcode file.
I choose the line after the “G80; mesh bed leveling”
…
G80; mesh bed leveling
M92 X100.267 Y100.631
...
...
Fortunately this code can be added and saved in Slic3er under the Tab “Printer Settings -> Custom G-Code -> Start G-Code”, so you only have to do this once. Don’t forget to save!
Once again, thank you very much for your support!
Have a nice day! 😀
Tobi
Re: Accuracy problems in Y
Excellent discussion.
Are the values for Z in the M92 command on the same scale as X&Y, or much larger? When I attempted Z100.1, my live Z jumped from -0.420 to -1.5 (or something egregiously large), the layers were printed with the nozzle rammed into the bed. When I stopped the print and raised the Z, I told it to raise Z to 210 but it only went up about 40-50? I wasn't paying close attention as I was trying to avoid panic.
Re: Accuracy problems in Y
Default number of steps per mm for Z is 400, not 100.
Re: Accuracy problems in Y
Curtis
I think that I should suggest that you leave things alone in future unless you absolutely know what you are doing 😉
You can always ask us ...
Peter
Please note: I do not have any affiliation with Prusa Research. Any advices given are offered in good faith. It is your responsibility to ensure that by following my advice you do not suffer or cause injury, damage…
Re: Accuracy problems in Y
I think that I should suggest that you leave things alone in future unless you absolutely know what you are doing 😉
Where is the fun in that? 🙂 If you don't fail you aren't trying hard enough.
Default number of steps per mm for Z is 400, not 100.
and that's the ringer that I was looking for. The RepRap wiki (and all the other sources online) has examples for the M92 command with Z steps of 100, so I assumed it would be the same as X&Y.
See, PJR, I didn't do it ENTIRELY blind! 😉
Re: Accuracy problems in Y
OK, for future reference, the go-to place for the majority of the default settings is the firmware configuration file., the latest version (3.0.8) can be found here: https://github.com/prusa3d/Prusa-Firmware/blob/MK2/Firmware/variants/1_75mm_MK2-RAMBo13a-E3Dv6full.h
Line 27 tells you the steps (3200/8) as steps per rotation divided by the thread pitch.
Peter
Please note: I do not have any affiliation with Prusa Research. Any advices given are offered in good faith. It is your responsibility to ensure that by following my advice you do not suffer or cause injury, damage…
Re: Accuracy problems in Y
OK, for future reference, the go-to place for the majority of the default settings is the firmware configuration file., the latest version (3.0.8) can be found here: https://github.com/prusa3d/Prusa-Firmware/blob/MK2/Firmware/variants/1_75mm_MK2-RAMBo13a-E3Dv6full.h
Line 27 tells you the steps (3200/8) as steps per rotation divided by the thread pitch.
Peter
Can I "Like" or upvote this post?
😎
Re: Accuracy problems in Y
You may want to have a look at Tom Sanladerer's video on why he believes this approach is NOT a good idea.