prints a little too close to the bed. about klipper_z_calibration

[bmkahl]

I was able to get the right offset by adding SET_GCODE_OFFSET Z_ADJUST=1 into my macro and then adjusting the switch_offset. Note that I do not have a Voron but rather have a custom built machine that has a limit switch on a SG90 servo.

[gcode_macro CALIBRATE_Z]
rename_existing: BASE_CALIBRATE_Z
gcode:
OG28 #only home if needed
SET_SERVO SERVO=my_servo ANGLE=0
BASE_CALIBRATE_Z
SET_SERVO SERVO=my_servo ANGLE=120
SET_GCODE_OFFSET Z_ADJUST=1

[bmkahl]

I am having the same issue. I have calibrated the limit switches, but it keeps adjusting the Z offset too low.

[laserbeak]

I've made my endstop a little taller to be more level with the bed when the switch is activated, and edited position_endstop in printer.cfg and position_min in z_calibration.cfg, but it didn't do much, if anything at all.

[laserbeak]

I'm glad it works, but doesn't that defeat the purpose of kzc?
how close was it to your bed that you had to use 1?

[bmkahl]

It was about .65mm too low consistently even when the switch offset was set to 0.01.

[laserbeak]

I started at .5, now I'm at .38. I'm printing things and testing ever now and then. I'll probably be down to .34 before the week's over.

[eduncan911]

I'm experiencing the same issue with a "load cell" nozzle probe that touches the actual nozzle.

It has a flex of about 0.2mm, that I need to adjust for. I previously was using probe->z_offset set to -0.2 for perfection before adding this script.

Now, I cannot seem to set this offset in any code. The previous z_offset doesn't seem to change regardless. And using z_calibration->switch_offset only seems to change the Z offset being reported...

My problem is whenever I enter "0mm" on Z, it should be a feather touch on the bed without triggering the probe, which would be -0.2mm on the bend.

Instead, the probe is immediately triggered - and I have to manually adjust Z to 0.2 to untrigger it - which should be 0.0 and not 0.2.

[TitusLabs]

If you need to be more away from the bed and the switch_offset is already nearly 0, then there is a problem in your setup/probing. So, the intention is to probe the body of the switch (you MUST do this !!) If this probing is correct (check it), then the trigger point is above this point (positive value). And it needs to be above the body - otherwise I would need a needle to trigger the switch into it's body. So, if you probe directly on the trigger or the endstop sticks to something else, then the endstop would trigger before reaching the actual trigger point of the switch which leads to the demand of an negative switch_offset...

[eduncan911]

I've since bashed away at the config and found perfect. I'll try to find time to post the entire config here.

But note, i've since edited this Klipper plugin to remove one of the Z calibration tests - and forced it to 0.2 or something. that way, not only is it much faster, the offset is correct everywhere.

[TitusLabs]

I added some more explanation on the switch_offset and why someone would demand for a negative value: https://github.com/protoloft/klipper_z_calibration/tree/v0.8.1#how-about-a-negative-switch-offset