Message: `The resulting switch offset is negative! Either the nozzle is still too far away or something else is wrong...`

[Gil80]

I find this message to be very confusing. I think I'm getting really good results:

HINT: z position_endstop=2.400 - offset=-0.007500 --> possible z position_endstop=2.407
Z-CALIBRATION: probe=5.420 - (switch=7.380 - nozzle=2.403 + switch_offset=0.450) --> new offset=-0.007500

[stepper_z]
step_pin: zboard:PB5
dir_pin: !zboard:PB6
enable_pin: !zboard:PC3
endstop_pin: zboard:PA7 # This is the Z-endstop on the Creality board where the sexbolt is connected.
#endstop_pin: probe:z_virtual_endstop
position_endstop: 2.4 ;2.67 # ROUGHLY MEASURE THE POSITION FROM THE TOP OF YOUR BUILD PLATE TO THE TOP OF THE SEXBOLT. SEXBOLT SHOULD BE NO HIGHER THAN 4MM
position_min: -3 # SOME NUMBER BELOW ZERO

[z_calibration]
nozzle_xy_position: 169,236
#switch_xy_offsets: 145,196
switch_xy_offsets: -24,-42
#bed_xy_position: 117.5,117.5 ; disabled because I'm using bed mesh reference index.
# wiggle_xy_offsets: 0,0
switch_offset: 0.45 #measured 0.41mm #0.45 for textured side, #0.39 for smooth side.
offset_margins: -1.0,1.0

Yet, when I run: CALCULATE_SWITCH_OFFSET I get: The resulting switch offset is negative! Either the nozzle is still too far away or something else is wrong...

Why is that?

[TitusLabs]

Because the current z position is not in the range to be able to calculate a switch offset. Did you move your nozzle down to zero before? This is step 4 in the documentation: https://github.com/protoloft/klipper_z_calibration/wiki/How-To-Use-It#command-calculate_switch_offset

[Gil80]

@TitusLabs I followed this guide: https://github.com/LoganFraser/VoronMods/tree/main/KlickySetup#determine-your-switch_offset-in-klicky-z-calibrationcfg

[Gil80]

Because the current z position is not in the range to be able to calculate a switch offset. Did you move your nozzle down to zero before? This is step 4 in the documentation: https://github.com/protoloft/klipper_z_calibration/wiki/How-To-Use-It#command-calculate_switch_offset

Following the guide you kindly provided, this is what I get:

switch_offset=0.450 - (current_z=-0.045 - z_offset=-0.045) --> new switch_offset=0.450

CALCULATE_SWITCH_OFFSET

HINT: z position_endstop=2.400 - offset=-0.045000 --> possible z position_endstop=2.445

Z-CALIBRATION: probe=4.610 - (switch=6.625 - nozzle=2.420 + switch_offset=0.450) --> new offset=-0.045000

My switch_offset was already set to switch_offset: 0.45 I used a paper and 0.1mm Z decrements until the nozzle slightly grabbed the piece of paper. On klipperscreen, the Z value was 0.0.

So I still don't get why I see The resulting switch offset is negative! Either the nozzle is still too far away or something else is wrong... when I run CALCULATE_SWITCH_OFFSET after CALIBRATE_Z. Or do I need to run CALCULATE_SWITCH_OFFSET only after I manually get the nozzle to touch the center of the bed?

[TitusLabs]

Correct, it needs the current z height where you think it should be 0.0. Then it tries to calculate the switch offset based on the first calibration run and this z position.

The command is intended to help a little in determining the switch offset.

In you case, you already seem to have a good starting point from where you can fine tune it by printing some tests :+1:

[Gil80]

Correct, it needs the current z height where you think it should be 0.0. Then it tries to calculate the switch offset based on the first calibration run and this z position.

The command is intended to help a little in determining the switch offset.

In you case, you already seem to have a good starting point from where you can fine tune it by printing some tests 👍

so I ran a test print and then I got: Offset is greater than allowed: offset=1.117 > max_deviation=1.000 so I did another calibrate_z followed by CALCULATE_SWITCH_OFFSET and I got: switch_offset=0.450 - (current_z=0.497 - z_offset=0.397) --> new switch_offset=0.350 so I set the new value for the switch to be 0.350... I have no idea why this change all of sudden. I didn't change any hardware.

Then a try to print again and I get Offset is greater than allowed: offset=1.070 > max_deviation=1.000 I'm lost here...

[TitusLabs]

Ok, then make some test of your repeatability. And be careful, check z=0 before starting a print.

And don't get confused by the switch offset and the resulting z offset! You can reduce the z offset by changing the z endstop position. A hint is prompted after every calibration with a recalculated z endstop position based on the determined z offset.

So..

1. change your z endstop position
2. check your z offset by running calibrate_z and go with your nozzle step by step down to 0.0
3. change your switch offset, if needed and check the result again
4. do a test print

[Gil80]

I measured the endstop position from the top of the sexbolt to the bed surface, and that's 2.67mm. do you mean I need to lower it or make it taller?

Every time I use CALIBRATE_Z, I copy the suggested possible z position_endstop=xxxx and I use that value and I redo the CALIBRATE_Z and the next time I get a different value. This is very confusing.

Should I use the suggested value? or should I just use the value I measured with my caliper? (from the top of the sexbolt to the bed surface?)

Or is the suggested value is the actuation point of the sexbolt z-endstop? If it's the actuation point itself, then it's not entirely clear. Either way, I'm confused and unsure how to continue from here.

The test print I ran got the nozzle dragging across the bed surface. Thank god it's a tungsten carbide nozzle. But I have no idea why this recent update caused this issue.

Let me explain my confusion and see where do I get it wrong.

1. Given measured Z endstop using calipers from top of sexbolt to bed (not from actuation point of the endstop) = 2.67.

2. Given calculated switch travel distance to actuation is 0.41 (using this guide: https://github.com/LoganFraser/VoronMods/tree/main/KlickySetup#determine-your-switch_offset-in-klicky-z-calibrationcfg).

3. Given calculated Z offset (nozzle to probe switch) is a value derived from running CALIBRATE_Z is switch=6.635 - nozzle=2.680 + switch_offset=0.410 which gives = 4.365 and my config is set to 4.37 example of config

   pin: ^zboard:PA6
   x_offset: 20.5
   y_offset: 41
   z_offset: 4.37

4. The output of Z-CALIBRATION is:

   Z-CALIBRATION: probe=4.805 - (switch=6.635 - nozzle=2.680 + switch_offset=0.410) --> new offset=0.440000

   AND

   HINT: z position_endstop=2.670 - offset=0.440000 --> possible z position_endstop=2.230

   So my questions are:

5. Do I need to change the switch offset from 0.41 to 0.44? If I do and I run this test again, it will give a different value, so what's the point?

6. Do I need to change my Z-Endstop value from the caliper measured 2.67 to the calculated value of 2.23? If I do and I run this test again, it will give a different value, so what's the point?

The result is that I now no longer trust the outputted values by CALIBRATE_Z. And if I run CALCULATE_SWITCH_OFFSET and I get Z=0, I still get the nozzle dragging across the bed. I have to use the klipperscreen to lift Z by 0.05 increments x 10 times to get it off the bed.

I hope I was able to explain my confusion and I look forward for your kind help. Thank you!

[TitusLabs]

Did you already have the previous version running or is it your first installation? The update does not change any core functionality.

First of all, configuring the z position endstop is the "old way" of calibrating the z offset. There are many instructions out there for this. The caliper method may not be accurate enough. You need to push the endstop until you just hear the click and no further.

The more correct this value is, the smaller the calibrating offset will be. It's ok to have a resulting offset of more than 1mm - just increase the max_deviation/offset_margins to allow this. But, it's more safe to keep it small. I added the hint for the z position endstop to adapt it in a fast way. It's ok that these values vary - even the resulting z offset varies. That's fine so long the results are correct.

So, in the beginning, please verify the results before printing!! It is documented here: https://github.com/protoloft/klipper_z_calibration/wiki/How-To-Test-It

If your probe works correctly, calibrate your switch offset first. Test it every time if the nozzle touches the surface with x=0.0. For testing, you can increase your max_deviation. Then calibrate your z position endstop to decrese the resulting z offset. And set a max_deviation that fits your needs.

[Gil80]

Did you already have the previous version running or is it your first installation? The update does not change any core functionality.

I had the previous version running and it ran perfectly.

First of all, configuring the z position endstop is the "old way" of calibrating the z offset. There are many instructions out there for this. The caliper method may not be accurate enough. You need to push the endstop until you just hear the click and no further.

Then I think the documentation should mention the actuation point, i.e., the travel distance until the switch clicks, like in a keyboard. That makes more sense. Or you probably have that in the documentation and I missed it.

As for this guide: https://github.com/protoloft/klipper_z_calibration/wiki/How-To-Use-It#command-calculate_switch_offset I just want to verify if by "Switch Offset" you refer to the travel distance it takes until the actuation point?

p.s. - I updated my last comment and I broke down the steps where I get confused.

[TitusLabs]

Hm, the z endstop position is not part of my documentation. It is the offset between the endstop and the print surface to shift the z=0.0 onto the print surface. More on this here: https://www.klipper3d.org/Manual_Level.html

But, there are many things which can change this value for instance the rising chamber temperature. That's why I build this plugin. You can say that it creates an offset from the offset. This is the correct offset of the current state in the system and can be different over time.

To keep the offset of the offset as small as possible, the first offset which is defined by the z endstop position should be calibrated correctly. The output of the z endstop position value is meant as a hint. And because this value is derived from the current offset, it changes too. So, don't worry. With this plugin, the z endstop offset does not have to be perfectly calibrated, but as accurate as possible so that the offset from CALIBRATE_Z is not too large.

The switch offset is a variable in the equation which I cannot probe. It is the distance from the switch body to the trigger point of the switch. That's why this needs to be determined by data sheet or by the new command CALCULATE_SWITCH_OFFSET. In both cases it needs to be fine tuned. This is the variable which you can use to get the first layer exactly as you want. And you need to test it if your chosen switch offset is correct. Set the value in your config, restart Klipper/Firmware and go with z step by step down to 0 and check everytime the left distance to the surface.

I hope this will clear your confusions. And please read again the documentation and let me know if I can clarify anything else..

[TitusLabs]

I am closing this ticket now. If there is still a problem, please reopen it...

[suifong0728]

Offset -3.133 is outside the configured range of min=-1.000 and max=1.000 Please help me. My vorontrident 350 is always negative.

[TitusLabs]

@suifong0728 You can calibrate your z endstop position to keep the calibration offset as small as possible. If you have a V2 with sacking gantry, home your z after QGL... (with offset_margins, you can change the boundaries of -1.0 and 1.0)