Bed level trigger height advice

Dugee

I am trying to, funnily enough, make my Delta as accurate as possible and this includes bed levelling and I was hoping to try and use the existing code to automate trigger height reporting across the bed. I am also lazy and find probing the trigger heights manually tiresome and if I can change a couple of codes to get my results then I will.

In order to speed up the initial process, would it work if I included an Annn field in my M558 and S1 for calibration to get multiple probes done at each point?

My thought are to use M558 in the config.g file:

M558 P1 H20 I1 F800 T6000 A5 ; set Z probe type to switch and the dive height + speeds

to probe each point specified in my bed.g file 5 times with an S1 calibration purely to report the results to automate the process.

Would this work and if so do I need to put the S1 in each G30 line in the bed.g file or just my X0 Y0 probe point, as I do with S6? Or is my limited knowledge finding a short cut that isn't there?

Once I get the trigger height sorted I would just delete the Annn command and replace the S1 in the bed.g file with S6 to go for full calibration.

nhof

You should be able to automate the trigger height setting by using the G30 S-3 trigger. This will set whatever height the probe was triggered at to the new probe trigger height. You can then run the normal bed probe process, so it might look something like this:

G28 ; home axes
G31 Z0.0 ; reset probe offset (not sure this is necessary)
G1 X0 Y0 Z5 Fnnn ; move to 5mm above center of bed
G30 S-3 ; Probe in place and set probe offset to trigger height

Then you run normal bed probe cycle after that. This assumes that the nozzle is set at exactly Z0 in the center of the glass tho.

Dugee

@nhof

would this mean that once I get my X0Y0Z0 trigger height set I can use the auto calibration to determine the trigger heights at the points used in my bed.g file by using S-3 for each point that the probe goes to?

I was hoping that by using the Annn field in my M558 that I could probe each point 5 times to check the repeatability also.

If I was more confident I would create a macro that would do the probing to give me my trigger height across the bed, but I don't think that there is any function that would automate the trigger height offset calculation for the H parameter in the bed.g file. If there was then this would do this part of the job for me.

At present I have manually probed the bed at each point 5 times to get the variance and averaged this to get my current H parameters, but to rerun this each time it probably takes an hour or so when it would be minutes if automated.

It's not something that needs done very often, to be fair, but until I am 100% happy I would like the convenience of being able to check this more easily.

nhof

Ah, I understand now that you are talking about trigger-height compensation for effector tilt on a delta? Unfortunately I am out of my depth there as I haven't used a delta machine.

I think the G30 S-3 can only be used on a single point, don't think it can be used to generate all the offsets.

If you know your bed is totally flat you could perhaps run a normal height-map generating probe cycle and use those readings as your offsets.

Dugee

@nhof no problem, I should have specified printer type in my opening post.

I'm going to play around a bit tonight and see what data I can get

Phaedrux

M558 A parameter doesn't work quite the way you think. It doesn't set the number of times that a point is probed, it only sets the maximum number of times a point can be probe if the results of 2 consecutive probes are not within the tolerance set by the S value. The default for S is 0.03mm.

For example, M558 A5 will probe a minimum of two times. If both results are within 0.03 of each other, no further probes are done at that point. If they differ by more than 0.03 it will continue to probe until two consecutive results are within 0.03mm up to a total of 5 times.

Hope that makes sense.

https://duet3d.dozuki.com/Wiki/Gcode#Section_M558_Set_Z_probe_type

Dugee

@Phaedrux thanks. I understand that better now.

Tbh, now I'm home and getting stuck in I am much happier.

I had a missing decimal point that threw my autocalibration out. 192mm z offset at one of the probe points instead of 0.192mm

My autocalibration figure is coming out at 0.052, which I'm now more than happy with, so I'm hoping not to do the trigger heights again for a while.

I was expecting much more pain than I actually got tonight.