Auto Z offset and PROBE_CALIBRATE

[MichaelOToole]

Is PROBE_CALIBRATE primarily intended for printers that don't use TAP?

TAP can determine the z_offset within a micron according to PROBE_ACCURACY but yet it doesn't save this information to .cfg, why not?

#probe accuracy results: maximum -0.475000, minimum -0.476250, range 0.001250, average -0.475250, median -0.475000, standard deviation 0.000500 no paper cold

I'm confused, if we take the average of -0.475250 from code above, and add the paper thickness 0.075 (measured using PROBE_ACCURACY with paper in place and then subtracting the averages), we get -0.400275 pretty close to what the paper method returns, in my case z_offset = -0.405...

I should be able to put the following in my probe.cfg right?

[probe]
pin: !EBBCan:gpio22
x_offset: 0
y_offset: 0
z_offset: -0.400
speed: 5.0
samples: 3 # was 1
samples_result: average # median
sample_retract_dist: 2.0
samples_tolerance: 0.025
samples_tolerance_retries: 3

 

[shiftingtech]

probe accuracy doesn't really determine anything except repeatability. All those values it shows are just relative to the current coordinate system, not to any particular concept of "correct"
Probe_calibrate is the correct way to at least get started calibrating tap (and then maybe fine tune after)

 

[MichaelOToole]

While probe accuracy check repeatably, it still makes an initial measurement and this measurement is within 10 micron of the value I arrive at using probe calibrate, nevertheless I'll stick to probe calibrate until something better comes along.

 

[claudermilk]

Yep, probe_calibrate even with a Tap. Once it's dialed in, you should get perfect first layers every time after that.

 

[shiftingtech]

While probe accuracy check repeatably, it still makes an initial measurement and this measurement is within 10 micron of the value I arrive at using probe calibrate, nevertheless I'll stick to probe calibrate until something better comes along.

yes, but that initial measurement isn't relative to your actual...anything. It's just relative to the current coordinate system, so it doesn't provide any sort of actual CALIBRATION.

 

[MichaelOToole]

Before any movement you are required to home all, this action provides the current coordinate system (relative to the x, y, z end-stops)...
All measurement from that point on are relative, they are entirely dependent on x,y,z end-stops.

No worries, I'm just thinking out loud...

 

[shiftingtech]

Before any movement you are required to home all, this action provides the current coordinate system (relative to the x, y, z end-stops)...
All measurement from that point on are relative, they are entirely dependent on x,y,z end-stops.

No worries, I'm just thinking out loud...

Right, but since tap IS your z endstop, you're referencing back to itself, which (since it hasn't been calibrated yet) brings us back to numbers that don't really have any absolute meaning.

 

[MichaelOToole]

Fair enough...