I’m not aware of any delays like that in the code. If you’re having an issue, best would be to post the full Klipper log from the event, as it’s hard to give advice without knowing the full details of the setup.
Also, if you’re using “sensorless homing” then note it’s generally considered to only work reliably when there is a hard stopping point (ie, not a squishy stopping point). There’s some info on this at: TMC drivers - Klipper documentation
-Kevin
My apologies for forgetting the Klipper log. See attached. Especially the last movement I executed, it seems it travelled more than necessary because there was not much compression left on the nozzle. I’m using the endstop1_pin to detect when to stop it from moving down. I am also getting some warnings in console: “No trigger on manual_stepper left_z after full movement”
I could’ve sworn this was all working fine a few weeks ago before I left for vacation and as soon as I come back it wasn’t working right. I did play around with some config settings since that time, so perhaps I did something to break it.
klippy-7.log (374.6 KB)
I took a quick look at your log. It looks like you’re using the tmc2209 “sensorless homing” to halt the “left_z” and “right_z” movements. Unfortunately, if you’re not getting good signals from these drivers you’ll likely need to tune the drivers to improve the performance (or alter the movement speed, carriage endstop hardware, or similar). The Klipper code isn’t really involved in the “sensorless homing” part - it just moves until the driver signals that a “stall” occurs - if the driver isn’t sending a consistent signal then the only recourse is to alter the driver configuration. Be aware that the driver may be sensitive to temperatures, speeds, motor current, and “sharpness” of physical contact.
There’s some info on configuring the drivers at: TMC drivers - Klipper documentation .
Timofey has also recently worked on a tool to interactively report the tmc driver “stall” values: Sensorless/Stallguard WebUI graph tool
Maybe that helps a little,
-Kevin
@koconnor Thanks for sharing those links, I’ll take a look. However, I’m using sensorless homing when moving the nozzles up to their home position. I’m still experiencing some sort of lag in detecting the limit switch when I move the nozzles down (endstop1_pin). I was pretty sure this was working well before at stopping right on time, but even the console printout seems to say -sometimes-that the motor has reached end of travel downward without detecting the endstop?