Kossel Mini Auto Calibrate not working

Quote
Algie99
When running auto-calibrate, the effector heads for the z tower ( x0, y,0 z100 not y30 if deploying probe) and slowly descends and then errors before reaching the bed.

Quote
Algie99
Occasionally the effector probes the centre of the bed a few times and then climbs in a spiral until it fouls a tower and I have to pull the power.

Quote
thehankinator
Allen key probe should work fine for auto calibrate.

That is good to know.

Quote
thehankinator
What's the error it gives? I've had this happen to me when the DELTA_HEIGHT was too small or NOZZLE_TO_PROBE_OFFSET is way off (z value in particular).

The error is just 'probe error' on the LCD and in Pronterprint. I have the G29 working on Marlin 1.1.0 with a configuration.h from T3P3. I also have all the measurements from dialing this in. The Z height is with 0.05mm and the tower check points were there as well.

Quote
thehankinator
I've never used an allen key probe but maybe allen key(or another part of the probe) is getting stuck and holding the microswitch in the closed position?

I think the probe is fine. I can hear it the switch clicking and if I reload Marlin 1.1.0 then it works immediately.

Quote
thehankinator
Are you using the config included in Marlin 1.1.9/2.X for the Kossel Mini or did you start with your own?

I have tried several routes to acheiving a config. Spent about 4 days banging my head against this. Several threw up compiling errors that I could not work out. It does not help that the difficult stuff moves around between Marlin releases eg the probe deployment rountine configs are in marlin_main.cpp in 1.1.0 and then they move to configuration.h for 1.1.9 and 2.0 as just the configs; I have no idea where the routine moved to and what variables it may be using. Problem A is the probe does not deploy correctly so is there another constant defined elsewhere but not mentioned? Searching through each document is too time consumimg and they rename things so you cannot always search. Bum!

The refusal to work is a constant in that the machine misbehaves in the same way for 1.1.9, 2.0, 2.0.6.1 and 2.0 bugfix. Something is not configured properly but where is it hiding? Could be a processing bottleneck so I have ordered an SKR 1.3 to see if this helps. I may be entering a whole new world of pain LOL. If that doesn't fix it my Tevo Tarantula can go 32bit. Also looking at Repetier firmware which suggest an Arduino Due and a RADDS board.

Thanks for your help. I will soldier on.

Cheers,

Algie99

Quote
dc42
There are reasons why T3P3 switched to Duet electronics and RepRapFirmware for their version 3 Mini Kossel kit!

I did the original conversion of a version 2 kit to Duet for them. It was my first experience of a Delta printer. It was when I realised how difficult it was to calibrate a delta printer whose towers were not quite at 120 degree intervals that I added least squares delta auto calibration to RepRapFirmware. That type of calibration has become the gold standard.

The Allen key probe on that kit mostly worked well, although I recall adjusting and lubricating it to stop it sticking in the wrong position. Z probes that are offset from the nozzle are not ideal on delta printers because they are affected by varying effector tilt; however the probe offset from the nozzle of that design was small enough to give acceptable results.

Quote
Algie99
These are the configs I am using at the moment. These came from Da Hai Zhu's calibration tutorial on Youtube.

[www.youtube.com]

They were matched up with Marlin 1.1.9 from Chris's Basement as every other version I downloaded was missing menu items used in the above tutorial.

[www.youtube.com]