Effector shifts during probing

bartolomeus

This is my first delta printer, so please be gentle.
My delta has linear rails, haydn's arms, smart effector and duet ethernet, metal corners and metal carriages (converted anycubic linear plus).

During auto calibration I can see the effector shift sideways when the nozzle touches the bed. This happens on a few locations on the bed. Upon contact the effector wil shift a few mm's in x or y direction. Also, the effector will not always travel in a straight line to the next probing point. I setup bed.g with DC's wizard. After calibration deviation is around 0.04mm. I tried with lower sensitivity settings on the effector, but that had no effect.

I have no idea what can cause this behaviour. Right now I have removed the linear carriages and am going over them which is a pita. Ordered some new bearings for them which will take a while to get here.

But what else can cause this strange behaviour?

Some extra info: I first started using this delta with printed carriages and had no problems. I then switched to metal carriages with integrated tensioners, and since then I have problems. I would have expected the more rigid metal carriages to be an upgrade.

T3P3Tony

@bartolomeus I would also expect metal carriages to be an upgrade.

Can you post your bed.g and config.g please

bartolomeus

@t3p3tony said in Effector shifts during probing:

@bartolomeus I would also expect metal carriages to be an upgrade.

Can you post your bed.g and config.g please

I am away from my printer right now. I can post these tonight.

I setup bed.g with DC's wizard: 6 point calibration, 115 radius, maximum# calibration points.
The smart effector is setup according to the wiki, except I changed mode to 8 instead of 5, added multiprobing, heaters off during probing, F1200 (as standard) and T6000 (as standard).

I probe with the nozzle at 0c, so there's no interference from the fan. The probing does make quite a large 'thump' when it hits the bed. Don't know if thats'normal. I have used a precision piezo on a corexy which made less force on contact.

dc42

Perhaps the sensitivity of the effector is too low and it is needing a hard push to trigger it. Does the green LED flash if you tap the underside of the nozzle gently?

bartolomeus

@dc42 said in Effector shifts during probing:

Perhaps the sensitivity of the effector is too low and it is needing a hard push to trigger it. Does the green LED flash if you tap the underside of the nozzle gently?

Not really a gentle push. I have to tap quite hard (compared to to precizion piezo I had). I went lower on the sensitivity (upto 10:245) but that didn't make any difference, although it did introduce some false triggers. Should I have increased the sensitivity? Did I misinterpretate the settings?

bartolomeus

One thing to add, which may influence things: I am also using a Nimble, with zesty's Nimble adapter.

dc42

Also check that you have set the probing speed to the recommended value (1000 or 1200 mm/min). It's the F parameter in the M558 command.

bartolomeus

@dc42 said in Effector shifts during probing:

Also check that you have set the probing speed to the recommended value (1000 or 1200 mm/min). It's the F parameter in the M558 command.

Yes, F is 1200

bartolomeus

Minor shameful update: putting back the rails after cleaning them I discovered the motor pulley on the y motor wasn't tightened down properly. It could rock around the flat spot on the shaft. That probably explains the shifting.

Edit: Calibrated 6 factors using 16 points, deviation before 0.027 after 0.024