Page 1 of 1

Print Calibration

Posted: Sun Jan 10, 2016 2:42 pm
by brisrascal
Hi,

I am trying to print a cylinder of dimensions 21mm x 16mm x 5mm. However, the printed model is on dimensions are 20mm X 15mm x 4mm. How do I adjust my EEPROM settings to calibrate it? Thanks.

Re: Print Calibration

Posted: Tue Jan 12, 2016 5:23 pm
by Mark
Under CONTROL->MOTION->X/Y/Z steps/mm, increase it or decrease it, save the value then test it agian. You can manually extruder 100mm, compare to the actual extrude length you measured and then set the proper value.

Re: Print Calibration

Posted: Thu Feb 04, 2016 12:36 am
by neilcoggins
The settings under CONTROL -> MOTION don't appear to allow decimal values to be entered.

For example, my values for X/Y are both 80. Printing a calibration object (20mm cube), I find that it ends up being 19.6mm in both X and Y. That means I need to set steps/mm to be (20/19.6) * 80 = 81.63 steps/mm. Since I can't enter that value using the LCD control panel, I'm stuck with either 81 (yielding a cube that is 19.845mm along each side), or 82 (yielding a cube that is 20.09mm along each side).

Is there another way to set the steps/mm more accurately(e.g. in the firmware source)? Or is there something I am doing wrong with setting the values in the LCD?

Re: Print Calibration

Posted: Thu Feb 04, 2016 4:24 pm
by Mark
you can find the value in the tab "configuration.h" of your firmware, then change it to the value you want and upload the firmware.

Re: Print Calibration

Posted: Thu Feb 04, 2016 6:06 pm
by neilcoggins
Ok, so where is the source code for the M201 firmware? I thought that Geeetech were keeping the GTM32 source code closed?

Also, does the M201 (GTM32) support the M92 (http://reprap.org/wiki/G-code#M92:_Set_ ... s_per_unit) gcode command? Because if GTM32 does support M92 (and M500 http://reprap.org/wiki/G-code#M500:_Sto ... _in_EEPROM), then a quick fix is to simply use those two commands to update the steps per mm with a decimal value for each axis and write it to EEPROM that way (rather than using the LCD, which only supports integer values).

Re: Print Calibration

Posted: Fri Feb 05, 2016 1:43 pm
by Mark
the firmware is here:
http://www.geeetech.com/forum/viewtopic ... 20&t=17046
but this firmware cannot be edited by the arduino...and I have tried M92 and seems not working..

Re: Print Calibration

Posted: Fri Feb 05, 2016 4:06 pm
by neilcoggins
So the M201 cannot be accurately calibrated - the resolution is limited to integer number of steps per mm (ie. 80/81/82, etc).

This actually means the M201 is one of the least accurate printers on the market, and probably unusable for detailed printing in its current state.

Can you tell me how to request a firmware update that includes the M92 gcode command or an update to the LCD mechanism so that it us possible to enter decimal numbers in the steps per mm inputs? Perhaps it could work with click to switch from integer part to decimal, eg.

1. Select X steps/mm
2. Turn rotary control to set to 81
3. Press rotary control
4. Turn rotary control to set to 81.63
5. Press rotary control
6. Save settings

Thanks in advance.

Re: Print Calibration

Posted: Fri Feb 05, 2016 4:19 pm
by Mark
Sorry but all the soft engineers are away on our new year holiday so I'll forward your suggestion to them when they back from holiday, we'll improve it asap.

Re: Print Calibration

Posted: Mon Feb 15, 2016 6:20 pm
by neilcoggins
As a possible improvement for the M201 (until the firmware is fixed/improved), it might be worth suggesting to customers that they switch to DRV8825, and reset the steps per mm value to 160. This would mean that any inaccuracy in steps/mm is effectively halved.

The logic is simple enough - on my printer (using A4988) I need 81.63 steps/mm to be "accurate" (whether it's stretching of the belt, or whatever, that's the value that would work). However, I am stuck at present using 82 (with a resulting error of 0.5%). Having changed to DRV8825 (and 1/32 microstepping), I now need 163.26 steps/mm. By using 163 steps/mm, I now get an error of 0.16%, which is a 66% improvement in accuracy, without requiring a firmware change.

Just thought it was worth pointing out as an interim fix before Geeetech can produce a new firmware that accepts M92 gcode...

Re: Print Calibration

Posted: Tue Feb 16, 2016 8:55 pm
by Mark
seems a good idea! have already forward this problem to the engineer, please be patient.