Once again many thanks for your reply. Using M201.1 to reduce the acceleration has enabled me to achieve a sensorless home! I now feel able to adjust the various parameters to get a reliable performance.
I have read the document on stall detection many times; in fact it was the existence of that document which convinced me that sensor less homing was a viable option. On reading the document I hadn't realised that stealthchop needed to be specifically selected with M569, or that M201.1 could be used to control acceleration whilst homing.
I relied heavily on the section titled Configuring Sensorless Homing, and that doesn't mention these commands. However I also realise that describing stall detection and configuring it is a complex area given the range of stepper drivers used by Duet; and that it would not be possible to describe the software configuration for all hardware combinations
When I set out on building this coreXY machine I realised that there would be a number of challenges along the way and this seems to have been one of them, however it is a good and enjoyable experience when supported by folk like yourselves.

Best posts made by ProfChris
-
RE: Duet3 sensorless homing corexy failure
-
RE: Duet3 sensorless homing corexy failure
Thanks for updating the documentation.
I have now got sensorless homing working fine on the four motor Z axis. I was hoping that I might be able to say that I followed a "process" which might help others along the path, but I'm not sure that was the case.
The process was roughly:
1: establish a clear stealthchop mode using M569 D3 V100...
2: establish the minimum % of motor current which would reliably move the axis; here 30%;
3: establish a speed at which a stall was reliably detected at the end of travel circa 60% of max speed.
This worked and I did not need to modify the acceleration - I had that available if the axis stalled on starting to move (which in hindsight is easy to correct).
Even though at the outset the process appeared a bit random, I'm glad that I persevered with this as there is considerable satisfaction in seeing the axes move smoothly and reliably to the end point with no judder or collision bang. It simplifies the wiring as well.