Help-What causes the auto calibration 6 factor to vary homed height

The Delta has been running fine for two weeks with auto cal before every print

Today i did some maintenance of the belt joiner/tensioners and something went haywire.

The physically measured Z height is 618.80 with Z=0 measured with a feeler gauge at 0.03mm above the blue tape, This hasn't changed and the end stops still have the same reported offsets.

Due to whatever went wrong the first print didn't quite put layer 1 down correctly and I managed to rip the Mini height sensor of it's mount.... replaced it and glued back in place and went through a full calibration of trigger heights.

Now i'm getting the routine wanting to change the homed height 0.1mm lower than it really is.....

Here's 3 consecutive runs which are fairly consistent, the M665 at the bottom is after a reset

21:06:49M665
Diagonal 674.60, delta radius 301.86, homed height 618.71, bed radius 280.0, X 0.08°, Y -0.39°, Z 0.00°
21:06:42 Calibrated 6 factors using 13 points, deviation before 0.160 after 0.158

21:03:58M665
Diagonal 674.60, delta radius 301.87, homed height 618.70, bed radius 280.0, X 0.07°, Y -0.39°, Z 0.00°
21:03:51 Calibrated 6 factors using 13 points, deviation before 0.156 after 0.155

21:01:43M665
Diagonal 674.60, delta radius 301.93, homed height 618.72, bed radius 280.0, X 0.07°, Y -0.38°, Z 0.00°
21:01:29 Calibrated 6 factors using 13 points, deviation before 0.205 after 0.152

20:59:33M665
Diagonal 674.60, delta radius 301.72, homed height 618.80, bed radius 280.0, X 0.13°, Y -0.25°, Z 0.00°

Sounds like the trigger height of the IR sensor has changed a little, and you need to reduce the G31 Z parameter in config.g by 0.1mm.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

Quote
dc42
Sounds like the trigger height of the IR sensor has changed a little, and you need to reduce the G31 Z parameter in config.g by 0.1mm.

Trigger height had changed, measured it about 20 times for a fairly consistent average around 2.62mm .... had already reduced G31 Z to 2.58 as with 2.62mm each calibration saw the homed height decrease by about 0.4mm

Bed.g has been tweaked to ensure all probe points are centred on each strip of tape as well.

.....mmmm just run 2 calibrations with trigger z reduced to 2.52mm and got.....

22:24:48M665
Diagonal 674.60, delta radius 301.99, homed height 618.68, bed radius 280.0, X 0.19°, Y -0.23°, Z 0.00°
22:24:32Calibrated 6 factors using 13 points, deviation before 0.156 after 0.156

22:22:24M665
Diagonal 674.60, delta radius 302.01, homed height 618.68, bed radius 280.0, X 0.18°, Y -0.24°, Z 0.00°
22:22:10Calibrated 6 factors using 13 points, deviation before 0.241 after 0.155

22:20:14M665
Diagonal 674.60, delta radius 301.72, homed height 618.80, bed radius 280.0, X 0.13°, Y -0.25°, Z 0.00°


Endstops
22:26:48 M666
Endstop adjustments X0.33 Y-0.15 Z-0.19

What are your initial endstop corrections? Auto calibration normalises the homed height and endstop corrections to make the average endstop correction zero.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

From config.g

M665 R301.99 L674.6 B280 H618.80 X0.13 Y-0.25 Z0.00 ; set delta radius, diagonal rod length, printable radius and homed height
M666 X0.33 Y-0.20 Z-0.19 ; put your endstop adjustments here, as given by auto calibration

David,

I took a deep breath, reset the end stop numbers to where they have been for weeks and set G31 Z at 2.66 (went higher) and got 3 out of 4 runs with height = 618.84 (over by .04) and one at 618.76 (under by .04).

I wonder if this falls into end stop repeatability? as the numbers reported by M666 do change.The end stops are quality micro-switches obtained from brand new mice

For reference current Config.g sections
; Movement section

M569 P0 S1 ; X Drive 0 goes forwards
M569 P1 S1 ; Y Drive 1 goes forwards
M569 P2 S1 ; Z Drive 2 goes forwards
M569 P3 S1 ; E0 Drive 3 goes forwards
M569 P4 S1 ; E1 Drive 4 goes forwards
M574 X2 Y2 Z2 S1 ; set endstop configuration (all endstops at high end, active high)
M906 X1700 Y1700 Z1700 E1300:1300:1300:1000:1000:1000 ; Set motor currents (mA); set each extruder explicitly
M665 R301.99 L674.6 B280 H618.80 X0.13 Y-0.25 Z0.00 ; set delta radius, diagonal rod length, printable radius and homed height
M666 X0.38 Y-0.23 Z-0.16 ; put your endstop adjustments here, as given by auto calibration
M201 X5400 Y5400 Z5400 E600 ; Accelerations (mm/s^2) Set all E-motors the same
M203 X24000 Y24000 Z24000 E3600 ; Maximum speeds (mm/min) Set all E-motors the same
M566 X2600 Y2600 Z2600 E600 ; Maximum instant speed changes mm/minute. Set all E-motors the same
M92 X160 Y160 Z160 ; Set axis steps/mm
M92 E402:402 ; Set extruder steps per mm explicitly for all the extruders
G21 ; Work in millimetres
G90 ; Send absolute coordinates...
M83 ; ...but relative extruder moves

; Z probe and compensation definition

; *** If you have a switch instead of an IR zprobe, change P1 to P4 in the following M558 command
M558 P1 X0 Y0 Z0 ; Z probe is a DC42 Differential IR PCB and is not used for homing any axes
G31 X0 Y0 Z2.66 P500 ; Set the zprobe height and threshold (put your own values here)

Past midnight here..... might look at it some more later today.

An interesting find I think.....

After getting everything dialled in on Saturday, did nothing with the printer on Sunday except run some auto calibration sequences to confirm I was getting repeatable results which I was.

Monday (long weekend here) mid morning decided to try printing the print that kept failing and decided to run the calibration independent of the print.
The calibration came in 0.26mm lower in total height than the previous day..... then it dawned on me..... the garage was also 10c colder at just 5c and what i was likely seeing was thermal expansion and contraction of the overall printer.

Now to keep a close eye on how much this changes with different temperatures.

Cheers
Phil