LR3 Y-axis accuracy off by 5mm?

[BenP (Ben Perreau)]

The accuracy of my Y-Axis seems to be way off & not sure what the reason could be.

• I marked the 0,0 with a V-bit on the Makita router
• I enter G0 Y1780 (using the Jackpot with FluidNC)
• I marked the supposed 1780mm point but it reads only 1775mm???
• Did the back-and-forth 5 times, always comes back to the same (inaccurate) point ‘accurately’
• I do not believe that the motor is missing any step (no load)
• Belts are tensioned OK (but not sure really what they should be, an educational video might help)
• Grub screws are tight
• The flex tape that I am using to measure is accurate against another measuring tape

What could be the reasons to have such a difference?

[CesarH (cesar)]

@BenP

I marked the supposed 1780mm point but it reads only 1775mm

After you home your machine, what does your Y axis reports? 0.000?
If not: go to your settings and verify your config for the y axis Set_MM_POS: variable to be 0, save the config.

Also, you can zero the axis manually in the main screen of the webui or send: G92 Y0. 000 that should set your homed axe to 0.

@BenP

I enter G0 Y1780

If your y axis is reporting to be at 5mm when homed, you call it to move 1780mm with G0, it will move 1775mm

[CesarH (cesar)]

What tablet support arm is that?

[BenP (Ben Perreau)]

It’s this one

[BenP (Ben Perreau)]

Thanks Cesar,

PS: My XY axis are swapped (did not mention that earlier to avoid confusion, but now I have to come clean…)

When homing, this is what I see (Xw is 0)

After the G0 X1780, here is what it shows (Xw is 1780, so I assumed that it moved that dimension?)

My mpos_mm is 3.000

[CesarH (cesar)]

@BenP

When homing, this is what I see (Xw is 0)

And you are using the pulleys/belt from a v1 hardware pack or did you sourced your own?

[BenP (Ben Perreau)]

Yes, the motor, belts and everything comes from Ryan.

One thing that I did change in the YAML was the pulse_us.
Ryan has it at 2 in his default file

But the wiki says to use 4??

I have it at 4 right now - could that have an effect on the steps??

[CesarH (cesar)]

My understanding is that the pulses affect the speed of rotation (rpm)

You should be looking at belt tension first but dont get too crazy maybe fine tune the steps if thats the case. Maybe someone else could get you more light into the matter.

[BenP (Ben Perreau)]

Yeah, tried changing to 2 and no changes - still 5mm off.
I also tried to super tension the belt but little effect (only about 0.5mm better).
Thanks for your help - Let see if others can think of any other reasons.

[CesarH (cesar)]

@BenP

My mpos_mm is 3.000

Set this to 0 and rehome. Save the config with: $CD=config.yaml

[BenP (Ben Perreau)]

I did set the mpos_mm to 0
The homing stayed the same

And the G0 X1780 is now 3mm closer to 1780 (1778 actually)!!!

So, should I now try mpos_mm as -2??? Will that bring it to the right 1780 distance??

I guess I am not sure to understand the wiki definition of the mpos_mm as
Sets the machine position after homing and limit switch pull-off in millimeters. If you want the machine position to be zero at the limit switch, set this to zero. Keep in mind the homing direction you choose this number.

Solution:

Changing the mpos_mm to -2 increased the distance to 1780!
I guess that my head translated mpos as ‘minimum position’ as the limit switch was mentioned in the wiki definition, but it probably means ‘maximum position’…

Thanks Cesar for your help!!!

Edit: it is not ‘maximum position’ but ‘machine position’…
For education - Help with Machine Space and Homing | Wiki.js

[CesarH (cesar)]

Mm pos should be zero. At least it works for me that way. After i home my machines coordinates are 0,0 ( i reference my sheets at the origin witn a fence on xmin and ymin 1/8" away

[dos (Daryl)]

@BenP

I did set the mpos_mm to 0
The homing stayed the same

And the G0 X1780 is now 3mm closer to 1780 (1778 actually)!!!

Sounds about right. You probaby need to adjust your steps per mm from there…

Houston we have a problem, belts

Well to my wonderful crew mates, let’s not get me walking the plank over this one. [a pirate walking the plank holding some machinists calipers] Best I can tell we have a belt issue. Our 2mm pitch belt does not seem to be 2mm pitch. I was asked about this during the LR4 beta and blew it off, We have been asked about this many times and blew it off. We have a fixed gear system and calibration should not really be needed. That is only true if all the components are as advertised. I have been u…

[BenP (Ben Perreau)]

Thanks Daryl,

Sounds like from the link that you provided that the belt might be slightly off & the solution is to adjust the step/mm instead of messing with the mpos_mm - this makes sense & I will try it.

• Set the mpos_mm to 0
• Adjust the step/mm for X&Y, as I suppose it doesn’t really matter for Z due to the low travel.

[ttraband (Tom Traband)]

A test I’d try is going half the distance you’ve been going. If the error is the same 3mm, there’s 3mm of mechanical or configuration error in the machine. If the error is half, i.e. 1.5 mm, then there’s a problem with the steps per mm.

[vicious1 (Ryan)]

@BenP

Thanks Daryl,

Sounds like from the link that you provided that the belt might be slightly off & the solution is to adjust the step/mm instead of messing with the mpos_mm - this makes sense & I will try it.

• Set the mpos_mm to 0
• Adjust the step/mm for X&Y, as I suppose it doesn’t really matter for Z due to the low travel.

5mm if far far too much to be belts. That is not the normal first issue.

Belts at worst would be 3mm over 8’, 1.5mm over 4’, so be careful to suggest this, unless the numbers are very similar to that.

[CesarH (cesar)]

Thats why i asked about his pulleys

[BenP (Ben Perreau)]

The 5mm error was with mpos_mm at 3.00
Changing it to 0.00 reduced the error to about 2mm (for a length of 1780mm or ~70", which is then in-line with your numbers
Thanks!

[dos (Daryl)]

@vicious1

5mm if far far too much to be belts. That is not the normal first issue.

Belts at worst would be 3mm over 8’, 1.5mm over 4’, so be careful to suggest this, unless the numbers are very similar to that.

Yep, waited until he was within 2mm over 1780mm before suggesting it.

Ideally, though, he would be testing movements with motions that eliminate the possibilty of backlash.

[vicious1 (Ryan)]

@BenP

The 5mm error was with mpos_mm at 3.00
Changing it to 0.00 reduced the error to about 2mm (for a length of 1780mm or ~70", which is then in-line with your numbers
Thanks!

I don’t understand. Home your machine. Move it forward 5mm (to get rid if any slop/play/lash), mark that position, move it to a few mm from your extreme and mark that position. Measure the marks vs the distance moved between them. mpos_mm has nothing to do with this (that is for homing and physical swithc position).