FLSUN Kossel with MKS Gen-L Marlin 1.1.8 and Marlin-AG fails G33 P2

I have been trying to get my FLSUN with the MKS Gen-L board to work under Marlin firmware. I can successfully move the print head in absolute terms (using Repetier host UI as well as sending G codes) G33 P1 works and reports a correct max height, but when I try to run G33 P2 it probes centre, X tower and Y tower correctly, but it does not probe Z tower at all. It just sort of hovers over to the Z tower and then moves slightly towards the centre and retracts. I have the nozzle height limit switch connected to the Z- header on the board, and since it correctly probes the height, I am fairly sure that is correct. I am wondering if there are any suggestions on where I should start looking to debug the problem.

Thanks!

Probably a stuck probe switch

OK. Will look into that. Thanks. I have tried to get it close to the actuation point. I might have go too close.

I have checked the probe switch. It looks like it is not stuck. It can correctly find the bed during the G33 P1 setting. When I run G33 P2 it probes centre, then X-tower correctly, and the Y-tower correctly, but when it tries to probe the Z-tower it does not find the bed and sort of stutters towards the centre a bit. There is no error generated on either iteration of G33 P2. But when I run P5, on the first iteration it probes around the centre and starts to probe around the edge, but after getting about a third of the way around the perimeter it starts probing in the air for another bunch of points and then goes wandering off into space. I could take a video if that would help.

Probably a stuck probe switch when it probes at the edge - video would help

It does not have edge probe switches. It is a delta. It has top limit switches and a Z-probe. Will video tonight.

OMG "when it probes at the edge "

the probe switch can probe fine at the centre but get stuck when it probes near the towers

Lol. My bad at reading. Sorry!

SOLVED: I had not adjusted depressing the nozzle probe switch enough to overcome the mechanical hysteresis in the switch. For future reference:

"In micro switches, it’s easiest to think of it like this; when the actuator is depressed, there is a point at which the switch activates, connecting the common contact to the normally open contact. As pressure on the actuator decreases, the point at which the switch reverts to its non-activated state, with the common contact falling back onto the normally closed contact, is not the same as the activation point, it’s later. The distance between the actuating point and the release point is called differential movement, or hysteresis."