Unreliable homing height.
Need advice/help please.
Having overhauled my Mini Kossel with a Duet 0.85 board, a flex3d drive and a differential height sensor I am having trouble getting reliable trigger heights.
I am using a 5mm ground ali bed with a printbite surface which i sprayed black on the reverse before attaching to ali plate. All works fine apart from the unreliable trigger readings and the homing is unreliable. With G28 homing it sometimes does not trigger all the endstops and I have to resend the command. Why?
I have set up the differential sensor as per Davids web pages and I have also checked the trigger height at all levelling points and entered the H corrections in the bed.g file. With G32 command I get different correction values every time. I can go and recheck the trigger heights and they will all be different again. I can only assume it is reflection but what else can I do. I had the same trouble before i painted the back of the printbite and I thought painting would solve the problem.
Does anyone have any advice. I'm pulling my hair out.
The solution to the G28 homing issue is covered in the FAQ page of the wiki, in the Troubleshooting section.
Regarding the IR sensor triggering at different heights, check that you don't have direct bright sunlight reflecting off the bed. It's more of a problem when the sun is low (as it is now in the Northern hemisphere) than it is in summer. Also, make sure that you have selected the correct Z probe mode (P1) in your M558 command.
I covered the heated Bed with blue tape and run the bed levelling sequence and I get nice consistant results. Did 10 bed levels and all results within 0.015mm which I presume is not too bad. So I guess the problem is reflection even though the printbite is black.
Anyone else had these problems.
Would welcome a response.
Been doing all the sensing in the garage at night so workshop lights may cause a problem. I have set probe mode P1 and have the trigger set at 500. I will try with a lower light level.
Yes the trigger height for the IR sensor against transparent PrintBite can be rather variable, probably because of the texture inside it. The new black PrintBite is better. However, on a delta printer it's better to use a nozzle contact probe if you can, because any sensor that is offset from the nozzle will be affected by effector tilt.