Printer Model: Voron 0.2 r1
MCU / Printerboard: SKR mini E3 v3
Host / SBC: BTT pi
klippy.log: Not currently available
Describe your issue:
Ive been troubleshooting sensorless homing for a couple of days. ever since Ive got the printer up and running, I can get sensorless homing to work for the most part. It will travel all the way to the end and bump the end stop, but sometimes it skips a step, and i can see it jerk in a direction perpendicular to the homing direction. It happens on both X and Y axis. I only skips once, then sets it at that point, so it is functional. The issue that I have with it is I’m trying to used Elis’s method of finding maximum acceleration that uses a homing at the beginning and end and subtracting the values to determine if the test skipped or not. I am not getting accurate results because at least one axis gets a skip. This morning before work i was messing with it and noticed I can get it to home perfectly when I do the test at lower acceleration and it is reliable. but when I increase the acceleration, then it starts happening. when looking through the macro that handles the sensorless homing, I noticed that there is nothing that sets acceleration, indicating that homing is using maximum acceleration. this happens at every sensitivity until it is unable to home at all. I don’t know why acceleration would effect it, but that’s my current theory. I don’t have any logs right now as I’m on my lunch break currently. Ive tried taking my speeds from 40 all the way down to 15 no change. sgthrs from 30 to 60 no change. current is at .5. When I get home, I plan to see if I can edit the homing macro to set acceleration without impacting the test macro. Any thoughts?
Ok, got home and did some tinkering with a fresh mind. I resolved the issue and it was because i was makeing some bad assumptions. No need for me to post the klippy log anymore. Here is what i figured out:
The default speed in the profile i was using was 20 mm/s. This resulted in a range that was small like the klipper guide says. It was also unstable. Increasing the speed DOES make it MORE stable. When I increased the speed for mine it made it LESS stable, but only because I now needed a larger value for sgthrs. I assumed that increasing the speed would widen the usable range at the current range, so I thought increasing was not producing the results I needed. I tried lowering the speed and it seemed more stable, but that was only because it brought the range back down to where I was initially operating at leading me to believe slower was better. I think that the accelleration was randomly coinciding with successful homing attempts. At 20 mm/sec, the range that would work was about 55 to 57 but it would still skip just less often and would even skip at speeds hire than that. It even resulted in very frustrating times where it would sometimes go to the endstop and skip, while other times not making it all the way to the endstop at the same sgthrs value. At 30 mm/s the range was 76 to 82. I set the sgthrs to 78 and its been working every time
Interesting development: previously it was working every time. However I noticed that it is only working when stepper motors are disabled first which I was doing to reset the tool head. If they still have holding torque, it will skip on the Y axis. I increased the curent from 4.9 (default) to 5.8. This seems to be working with stallguard ranges at the same 76 to 82. What differed from the klipper guide is the values did not change by increasing this value, as it says adjusting current requires recalibrating. Speed definitly did, but the klipper guide said nothing about recalibrating for speed. Increasing the current resulted in the same range, but more stable within that range. I first increased it to 5.5, but that resulted in a most stable value of 81, which to me is too close to the maximum of 82. The additional .3 resulted in that same 78 that I had before working every time.