M558 F parameter only applied during approaching of probe

Googliola

I am using a BLTouch as Z-min and for bed leveling. Unfortunately I get bad repeatability due to the fact that the z-speeds for lowering and lifting are different. One is configured by the M558 F parameter (approaching the probe) while the other is defined by M203 Z parameter.

IMHO, this is a bug as the difference in speeds accounts for the bad repeatability (as I tested). I propose that the M558 F parameter is applied to both directions in order to get a more consistent reading, while the M558 T parameter is only applied to XY movements.
Lowering M203 Z is a possible workaround for bed.g auto-level, but not so for mesh grid compensation (G29). And sacrificing overall z-speed by lowering M203 Z in config.g to get more precision for G29 is just a last resort.

Tests performed on firmware 2.02RC2

dc42

Personally I would find it very annoying if I had to use a slow head lifting speed on my printers when probing, because it isn't necessary on my printers and it would slow down bed probing.

If the BLTouch has special needs in this respect, then I guess the head lifting speed could be changed only when Z probe mode 9 is selected. But are you sure that it is the BLTouch that needs a slow head lifting speed, and not the Z axis of your machine?

I'll need to hear the views of other BLTouch users before I make this change.

Why is lowering the M203 Z parameter not a possible workaround when doing G29 mesh bed compensation?

Googliola

I just noticed that I get different readings if I probe 2 or 3 times and have a high z-speed. Lowering z-speed kind of solved the problem most of the times.
Interestingly, first repeatability tests (as described here) with different M203 Z values could not confirm my suspicion above. Will have to do more testing.

Why is lowering the M203 Z parameter not a possible workaround when doing G29 mesh bed compensation?

What I mean is that there is no way to tell the firmware to lower Z speed whenever a G29 is issued - for example from the DWC. Of course I could create another macro for G29...

Phaedrux

@dc42
I agree that reducing the lifting speed has a positive effect on the BLTouch. I don't have a very fast bed to begin with, but I have found that slowing it down during probing is beneficial.

I simply use a macro to do it though. Before the G30 I call a macro that sets z speed, accel, and jerk to lower values. When it's done it calls another macro that resets them to the default values. Which takes care of homing z.

I use a seperate macro to run G29 and it calls the same macro to lower speeds and restore them again at the end. That may not be a practical answer for most people who don't want to mess with macros for everything, especially when the DWC is the main UI.

Whether it would be better to have that a default behaviour option in M558 I can't really say. I don't have a large enough set of data to say if it would help every BLTouch probe in every situation. I will say that it's also possible to use a too low speed for head lift which can cause the pin to strike the bed and go into error.

For what it's worth, here are the speeds I use for probing

M566 Z6 ; Set maximum instantaneous speed changes (mm/min)
M203 Z100 ; Set maximum speeds (mm/min)
M201 Z6 ; Set maximum accelerations (mm/s^2)
M558 P9 H3 F100 T6000 A5 R0.3 S0.005 B1

Normal speeds are 300mm/min with 60mm/min accel and jerk.

I use a lower 3mm dive height to offset the slower probe and lift speed, with a short pause to let the pin settle.

@Googliola What speeds are you actually using, and how do you have the M558 command configured?

Googliola

Funny, I just came to the opposite conclusion: M558 T and F params dont really account for quality of z-probe values

I ran an automated test Test M558.T_F.params.g that calls Repeat10.g (both stored in /macros/BLTouch/ , remove file extension.g) and got these results Test M558.T_F.params results.xlsx

I built a custom coreXY, based on Tronxy X5S.
My settings are

M566 Z120 ; Set maximum instantaneous speed changes (mm/min)
M203 Z8000.0 ; Set maximum speeds (mm/min)
M201 Z500 ; Set accelerations (mm/s^2)
M558 P9 H5 R1 Fx Ty A2 X0 Y0 Z1 S0.08

Note: Since Z max speed is limited to 8000, the max speed never reached 9000 as it says in the results

Questions:
Whats your take on the results?
What factor would you tune next? Acceleration?

@Phaedrux Love your custom D-Bot

Phaedrux

@googliola

Is this an accurate way to measure repeatability?

G30 P0 X150 Y153 Z-9999
G30 P1 X150 Y153 Z-9999
G30 P2 X150 Y153 Z-9999
G30 P3 X150 Y153 Z-9999
G30 P4 X150 Y153 Z-9999
G30 P5 X150 Y153 Z-9999
G30 P6 X150 Y153 Z-9999
G30 P7 X150 Y153 Z-9999
G30 P8 X150 Y153 Z-9999
G30 P9 X150 Y153 Z-9999 S-1

It was my thinking that to measure the trigger height reliably you'd need to use G30 S-1 and manually reset Z=0 between each probe move. No?

Googliola

@phaedrux said in M558 F parameter only applied during approaching of probe:

Is this an accurate way to measure repeatability?

It was my thinking that to measure the trigger height reliably you'd need to use G30 S-1 and manually reset Z=0 between each probe move. No?

Honestly, I did not consider that, just plain copy&paste from this post

Phaedrux

Yeah I saw the reference post. I'm not sure if it matters?

I guess it depends if you know you can trust your trigger height and repeatability to begin with?

G30 Pn Z-9999 S-1 is typically used to measure multiple points on the bed after you've already determined the probe triggers reliably. Maybe it can be used this way? I'm not really sure.

Phaedrux

This is how I measure my trigger height. I manually enter the values into Excel to get the median and mode.

M291 P"Grid bed compensation map and Z Probe trigger height will be cleared. Ok or Cancel?" R"WARNING" S3 ; User must click OK or cancel.
M291 P"Heights will be found in gcode console if logging successful gcode is enabled" R"Did you remember to enabled gcode logging?" S3

; Clear compensation map and Z probe trigger height
;
G28			; home all axis
M561			; Disable any current bed compensation
G29 S2			; Clear mesh bed compensation parameters
G31 Z0			; Reset Z probe trigger height
G90			; Absolute positioning
T0			; Activate first tool
M98 P"ZSpeedsSlow.g"	; Set lower speeds for Z homing and lower Z motor current
M913 Z60		; Drop motor current to prevent damage in case of head crash

; Preheat to probing temps
;
M291 P"Preheating to bed to 60 and nozzle to 210 for accurate probing"
M104 S210		; Set nozzle to 210 and release
M190 S60		; Set bed to 60 and wait
M109 S210		; Set nozzle to 210 and wait

; 1
;
M291 P"Nozzle will now move to center of bed to reset Z0 and calibrate probe" S3
G1 X150 Y130 Z5 F4000	; Move to bed center
G92 Z8 			; Reset z to 8 to allow jogging up to touch bed to nozzle
M291 P"Jog the Z Axis until the bed and nozzle are touching and click OK" R"Setting Z=0" Z1 S3
G92 Z0			; Set z = 0
M291 P"Probe will now move to center of bed and measure trigger height 10 times" R"ZProbe Trigger Height Calibration" S3
G1 Z1			; Drop bed for nozzle clearance
G1 X190 Y90 F4000 	; Move to bed center
G1 Z10
G30 S-1

; 2
;
M291 P"Nozzle will now move to center of bed to reset Z0 and calibrate probe" S3
G1 X150 Y130 Z5 F4000	; Move to bed center
G92 Z8 			; Reset z to 8 to allow jogging up to touch bed to nozzle
M291 P"Jog the Z Axis until the bed and nozzle are touching and click OK" R"Setting Z=0" Z1 S3
G92 Z0			; Set z = 0
M291 P"Probe will now move to center of bed and measure trigger height 10 times" R"ZProbe Trigger Height Calibration" S3
G1 Z1			; Drop bed for nozzle clearance
G1 X190 Y90 F4000 	; Move to bed center
G1 Z10
G30 S-1

; 3
;
M291 P"Nozzle will now move to center of bed to reset Z0 and calibrate probe" S3
G1 X150 Y130 Z5 F4000	; Move to bed center
G92 Z8 			; Reset z to 8 to allow jogging up to touch bed to nozzle
M291 P"Jog the Z Axis until the bed and nozzle are touching and click OK" R"Setting Z=0" Z1 S3
G92 Z0			; Set z = 0
M291 P"Probe will now move to center of bed and measure trigger height 10 times" R"ZProbe Trigger Height Calibration" S3
G1 Z1			; Drop bed for nozzle clearance
G1 X190 Y90 F4000 	; Move to bed center
G1 Z10
G30 S-1

; 4
;
M291 P"Nozzle will now move to center of bed to reset Z0 and calibrate probe" S3
G1 X150 Y130 Z5 F4000	; Move to bed center
G92 Z8 			; Reset z to 8 to allow jogging up to touch bed to nozzle
M291 P"Jog the Z Axis until the bed and nozzle are touching and click OK" R"Setting Z=0" Z1 S3
G92 Z0			; Set z = 0
M291 P"Probe will now move to center of bed and measure trigger height 10 times" R"ZProbe Trigger Height Calibration" S3
G1 Z1			; Drop bed for nozzle clearance
G1 X190 Y90 F4000 	; Move to bed center
G1 Z10
G30 S-1

; 5
;
M291 P"Nozzle will now move to center of bed to reset Z0 and calibrate probe" S3
G1 X150 Y130 Z5 F4000	; Move to bed center
G92 Z8 			; Reset z to 8 to allow jogging up to touch bed to nozzle
M291 P"Jog the Z Axis until the bed and nozzle are touching and click OK" R"Setting Z=0" Z1 S3
G92 Z0			; Set z = 0
M291 P"Probe will now move to center of bed and measure trigger height 10 times" R"ZProbe Trigger Height Calibration" S3
G1 Z1			; Drop bed for nozzle clearance
G1 X190 Y90 F4000 	; Move to bed center
G1 Z10
G30 S-1

; 6
;
M291 P"Nozzle will now move to center of bed to reset Z0 and calibrate probe" S3
G1 X150 Y130 Z5 F4000	; Move to bed center
G92 Z8 			; Reset z to 8 to allow jogging up to touch bed to nozzle
M291 P"Jog the Z Axis until the bed and nozzle are touching and click OK" R"Setting Z=0" Z1 S3
G92 Z0			; Set z = 0
M291 P"Probe will now move to center of bed and measure trigger height 10 times" R"ZProbe Trigger Height Calibration" S3
G1 Z1			; Drop bed for nozzle clearance
G1 X190 Y90 F4000 	; Move to bed center
G1 Z10
G30 S-1

; 7
;
M291 P"Nozzle will now move to center of bed to reset Z0 and calibrate probe" S3
G1 X150 Y130 Z5 F4000	; Move to bed center
G92 Z8 			; Reset z to 8 to allow jogging up to touch bed to nozzle
M291 P"Jog the Z Axis until the bed and nozzle are touching and click OK" R"Setting Z=0" Z1 S3
G92 Z0			; Set z = 0
M291 P"Probe will now move to center of bed and measure trigger height 10 times" R"ZProbe Trigger Height Calibration" S3
G1 Z1			; Drop bed for nozzle clearance
G1 X190 Y90 F4000 	; Move to bed center
G1 Z10
G30 S-1

; 8
;
M291 P"Nozzle will now move to center of bed to reset Z0 and calibrate probe" S3
G1 X150 Y130 Z5 F4000	; Move to bed center
G92 Z8 			; Reset z to 8 to allow jogging up to touch bed to nozzle
M291 P"Jog the Z Axis until the bed and nozzle are touching and click OK" R"Setting Z=0" Z1 S3
G92 Z0			; Set z = 0
M291 P"Probe will now move to center of bed and measure trigger height 10 times" R"ZProbe Trigger Height Calibration" S3
G1 Z1			; Drop bed for nozzle clearance
G1 X190 Y90 F4000 	; Move to bed center
G1 Z10
G30 S-1

; 9
;
M291 P"Nozzle will now move to center of bed to reset Z0 and calibrate probe" S3
G1 X150 Y130 Z5 F4000	; Move to bed center
G92 Z8 			; Reset z to 8 to allow jogging up to touch bed to nozzle
M291 P"Jog the Z Axis until the bed and nozzle are touching and click OK" R"Setting Z=0" Z1 S3
G92 Z0			; Set z = 0
M291 P"Probe will now move to center of bed and measure trigger height 10 times" R"ZProbe Trigger Height Calibration" S3
G1 Z1			; Drop bed for nozzle clearance
G1 X190 Y90 F4000 	; Move to bed center
G1 Z10
G30 S-1

; 10
;
M291 P"Nozzle will now move to center of bed to reset Z0 and calibrate probe" S3
G1 X150 Y130 Z5 F4000	; Move to bed center
G92 Z8 			; Reset z to 8 to allow jogging up to touch bed to nozzle
M291 P"Jog the Z Axis until the bed and nozzle are touching and click OK" R"Setting Z=0" Z1 S3
G92 Z0			; Set z = 0
M291 P"Probe will now move to center of bed and measure trigger height 10 times" R"ZProbe Trigger Height Calibration" S3
G1 Z1			; Drop bed for nozzle clearance
G1 X190 Y90 F4000 	; Move to bed center
G1 Z10
G30 S-1


; Turn off heaters
; 
M104 S0		; Set nozzle to 0 and release
M140 S0		; Set bed to 0 and release
M98 P"ZSpeedsNormal.g"	; return to normal speeds
M913 Z75	; reset z motor current
G28		; Home Z

M291 P"Probing complete. Turning off heaters and homing axis. Check log for trigger heights and enter into config.g" S2


; Tone to get user attention
;
M400			; Clear movement buffer so tones play reliably
M300 S666 P500
G4 P501
M300 S1111 P300
G4 P301

Note that you won't be able to use that on your own system without adapting it.

Googliola

Thanks for the macro. Maybe, repeatability test will become much easier once Forced averaging for M558 is implemented. Meanwhile I will investigate your macro.

dc42

@phaedrux said in M558 F parameter only applied during approaching of probe:

@googliola

Is this an accurate way to measure repeatability?

G30 P0 X150 Y153 Z-9999
G30 P1 X150 Y153 Z-9999
G30 P2 X150 Y153 Z-9999
G30 P3 X150 Y153 Z-9999
G30 P4 X150 Y153 Z-9999
G30 P5 X150 Y153 Z-9999
G30 P6 X150 Y153 Z-9999
G30 P7 X150 Y153 Z-9999
G30 P8 X150 Y153 Z-9999
G30 P9 X150 Y153 Z-9999 S-1

Yes, that's a good method. On delta printers especially, it's also good to run the same test with movements to alternate high-X and low-X between probings, for example:

G30 P0 X0 Y0 Z-9999
G1 X-100 F6000
G30 P1 X0 Y0 Z-9999
G1 X100 F6000
G30 P2 X0 Y0 Z-9999
G1 X-100 F6000
...

Similarly for Y. This tests for backlash in the movement system.

Googliola

@Phaedrux your macro is quite nice. I adapted it to my needs but I am a little confused:
Considering "my" way to measure repeatability (see above) what is the benefit from

moving the probe away by X40 Y-40 at the end of each round of probing and
lowering the bed to Z10 before probing - in addition to dive height?

Phaedrux

The X40 and Y-40 move is basically to negate the nozzle offset of the BLTouch, so that I am probing the same spot that the nozzle actually touches.

The bed move by 10 was basically done to ensure enough clearance and dive height while I was still narrowing down the dive height. It also gives a longer dive height to allow for the probe to settle. That plus the lower speeds is an attempt to give the probe the best chance of getting a repeatable measurement.

I also have a macro for manual bed leveling and then another for mesh grid measurement. I don't run these very often, but they are nice to have when you need them.

1_1538593590722_0_Grid Compensation Assist.g
0_1538593590721_0_Auto Bed Level Assist.g

And there's also this bundle for tuning:
https://forum.duet3d.com/topic/6181/tuning-macros-menus-accel-jerk-retraction-pressure-advance