Kossel IR ABL height inconsistency

I've been using a large Kossel with a Volcano hot end for some time now, and I just can't seem to figure out a way to get a consistent distance from the glass for my first layer. Fortunately, it doesn't matter all that much to me, for PLA prints, with 0.6-0.8mm nozzle, but it does mean that I have to watch and restart my prints around 0-3 times, even with this big margin for error. It could just be that I'm not cleaning enough, though, my Mendel Max 3, with IR ABL hardly needs it. It could be inconsistencies in the carriages, but the error does not seem to affect tilt, only height.

What would be the typical problem solving steps to resolve this?

Some details:

• 1m tall Kossel, 2020 extrusions
• GT2 belt driven carriages with mini-delrin wheels
• Magnetic ball joints, 10mm thick aluminium arms
• Solid ABS printed effector, no slack or flexibility when motors engaged
• IR sensor (provided with Duet Wifi) firmly mounted in place.
• The variability is around 0.5mm at it's worst, but generally within 0.2mm
• Tilt is almost always fine ... and I can fix it by cleaning the glass
• Glass is thick borosilicate
• Black BBQ paint on the rear side ... same as on my Mendel Max 3
• Mendel Max does not have variability issue with it's IR sensor
• There is no aluminium heat spreader
• The heatbed is a silicone type
• I sometimes notice that water evaporates along certain lines, presumably due to the silicone heater when cleaning the glass

0.5mm RMS deviation is very high, and even 0.2mm is high. Perhaps your wheeled carriage are not moving smoothly? My Kossel got a lot better when I switched to linear rails.

Before printing, do you run auto calibration, or just use G30 to set the Z=0 position at the centre, or just home? If you run auto calibration before printing, then it's best not to home between calibration and printing, in case your homing switches don't always trigger at exactly the same point.

Are you getting a consistent trigger height from the IR sensor?

If you set the nozzle to a measured height at a particular place, then move away and back again, does it go back to exactly the same height?

---

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].

I'm running the G32 auto-calibration routine, which I've modified, to remove the non-delta stuff:

M561						; clear any bed transform, otherwise homing may be at the wrong height
G31 X0 Y0					; don't want any probe offset for this
G28							; home the printer

G30 P0 X-73.6 Y-42.5 Z-99999	; X tower
G30 P1 X0 Y-85 Z-99999			; between X and Y towers
G30 P2 X73.6 Y-42.5 Z-99999		; Y tower
G30 P3 X73.6 Y20 Z-99999		; between Y and Z towers
G30 P4 X0 Y67 Z-99999			; Z tower
G30 P5 X-73.6 Y20 Z-99999		; between Z and X towers
G30 P6 X-36.8 Y-21.25 Z-99999	; half way to X tower
G30 P7 X36.8 Y-21.25 Z-99999	; half way to Y tower
G30 P8 X0 Y42.5 Z-99999			; half way to Z tower
G30 P9 X0 Y0 Z-99999 S6			; centre, and auto-calibrate 6 factors

G1 X0 Y0 Z10 F15000			; get the head out of the way of the bed

I'll edit in a moment, with the answer to the other questions.

I had another look at my carriages, and there's a minute amount of play in them. I'm worried that might be the cause.

Regarding consistent trigger height, I'm not entirely certain I'm measuring that correctly, but I think that the answer is yes. Trigger height is -3.3 / -3.1 Z before the ~465 reading, and -3.7 / -3.8 Z before the >500 reading. At that point, my 0.55mm feeler gauge meets friction under the nozzle, without pressing the bed enough to change the sensor reading. I also tested if using E-stop then measuring produced a different result, but it was consistent.

Regarding triggering, then moving, then triggering in the same measured spot, here's what I did to test that:

Auto-calibrate ... move to >500 trigger height (0.4z), measure with feeler gauge (0.45-0.55mm gap below nozzle) ... Move Z+200, X+100, Y+100 then back to the original trigger spot. Trigger still trips at same 0.1mm, and still measures the same with the feeler gauge.

Ok, so I got the following measurements by using G1 move commands to the calibration locations, after calibration. I feel like this might be a clue:

G30 P0 X-73.6 Y-42.5 Z-99999	; X tower					Z1.05
G30 P1 X0 Y-85 Z-99999		; between X and Y towers	Z0.95
G30 P2 X73.6 Y-42.5 Z-99999	; Y tower					Z1.05
G30 P3 X73.6 Y20 Z-99999		; between Y and Z towers	Z0.95
G30 P4 X0 Y67 Z-99999			; Z tower					Z0.85
G30 P5 X-73.6 Y20 Z-99999		; between Z and X towers	Z0.75 / 0.85 - on downward movement wouldn't trigger, on upward stayed triggered at 0.85
G30 P6 X-36.8 Y-21.25 Z-99999	; half way to X tower		Z0.75
G30 P7 X36.8 Y-21.25 Z-99999	; half way to Y tower		Z0.85
G30 P8 X0 Y42.5 Z-99999		; half way to Z tower		Z0.85

The above was after a failed print, with too much squish.

Afterwards I rehomed, calibrated and got z0.2-0.3 in each position. Repeated the process again, and got the same results. This is quite baffling.

Even more curious. I tried to print again, and my hot end was further away from the bed ... about 0.5mm more I'd say. So I adjusted the config.g file's z offset, and restarted to apply the changes, and I'm back to triggering at ~0.85mm. I do it again and I'm triggering at 0.5mm.