Measure Step Loss between two machine homings?

[dezsoe]

In my last post I've sent you a version of the macro which disables the soft limits while checking home. The home all button has a fixed function, but you can place a new button on the screen (see manual) and give it the button number 20910.

[mr_cg]

Sorry, i have missed the updated in the macro from your last post.
I have added this and it's working excellent. Many thanks!!!

Last topic which is open is what i have described under 2.
The macro works now excellent for X and Z with almost zero step loss even if there is also backlash configured. Wow!
But for the Y axis the difference is always the same amout of configure backlash for the Y axis.
When increasing/decreasing the backlash configuration for Y axis also the deviation with M20910 macro is showing the same amount.
This is 100% reproducable and i think there you be something wrong with backlash step addition or machine homing/zeroing strategy for the Y axis.
Any idea why i see this issue only for the Y axis?

Would be great if we could fix the remaining issue.

[dezsoe]

We'll have to wait for CNCdrive to answer this question: I never used backlash and don't know, how the homing with backlash is implemented.

[Vmax549]

As a test try doubling the backlash value setting you have now for Y. IF that gets it real close it you may have more backlash than you think.

Just a thought, (;-) TP

[mr_cg]

Attached a screenshot about the issue which could be useful for better understanding.
When changing the backlash value for y axis than also the deviation from M20910 is chaning accordingly.

Output from M20910:

[Vmax549]

But I do not see where you have changed it.

First try removing the backlash value , then try doubling the value to see how it reacts.

(;-) TP

[mr_cg]

As already written before. When i disable the backlash than the deviation for y is zero.
When changing the backlash for y axis e.g. to 0.55mm than the deviation with M20910 is also about 0.55mm.
I have not done screenshots for all the tests restuls, so you need to believe what i am writing.

@ dezoe
Will i get feedback from cncdrive in this thread or do i need to open topic in the bug section?
One idea could be that the backlash for Y axis is included in the counting when driving the homing from (down) the switch?
While for X and Z axis the steps for the backlash are not included when driving the homing from the switch?
Just guessing but the backlash step injection while homing could be different.

[dezsoe]

We have already talked about the problem, but could not find out why only the Y has problems. I'll write here if we can find the problem. Could you upload your profile (.pro) file?

[Vmax549]

yes I read your post But if you noticed I did not say disable the Backlash just change the values to double then zero.

IS it possible that you don't actually have that amount of backlash in Y ? Have you double checked ? What was the process of how you checked it ?

Trouble shooting a remote problem is not always easy and questions have to be asked.

(;-) TP

[mr_cg]

Attached you can find the profile which is almost with stock settings as provided from Stepcraft.

I think we should try to find the explanation (root cause) for this to assure there is noting wrong in the program.
But if we can't find it i would adapt the macro and add the missing steps from backlash for Y axis into the calculation.
It's not very nice to do this exception only for the Y axis but in general it's working.
Question: How is it possible to get the value from the Y backlash accessible in a variable to be used in the M20910 macro?

This change in the M20910 macro is neutralizing the issue but finding the root cause is the better option:
if (WorkAxis == "Y") exec.AddStatusmessage(WorkAxis + ": " + Math.Round((Diff + 0.217), 3, MidpointRounding.AwayFromZero).ToString() + "mm, " + Math.Round((Diff* Stepsperunit + 29), 0, MidpointRounding.AwayFromZero).ToString() + " steps");