I am trying for the last 5 months to troubleshoot this specific problem but to no avail. I tried looking throughout the whole forum but no post seems to address the combination of problems I am faced with.
MPCNC built 400x450 dimensions (the standard height of 81).Initially the problem was with the the Z axis leadscrew, on command of 5mm up or down, on way down it was 4.98 then on up 5 then dropping back to 4 or 3 randomly as if the motor couldn’t hold the position and held in position by binding. Today I reprinted the z clamps and even though the binding is almost gone (before the bearings were not all in contact with the tubes no matter how tight or lose I was trying to get them, now they are all in contact).Now the leadscrew is extremely smooth to the point it can’t hold 0.8kw spindle, the whole z assembly keeps dropping like free fall. As Ryan pointed in other posts I assume this is good? This was with motors unpowered
When I tried engaging the motors (actually commanding them to move again 5mm), the z axis was climbing 5mm then it goes back to the bottom until it hits the spoilboard as if the motors are disengaged the moment the move stops. Until now I was quite sure that there is a mechanical problem but is there any chance it is a software issue and fluidnc actually disengages the motors by the time the commanded move stops?
The stepper motor on the Z axis is quite week (they are 17HS4401S rated at 42nM) and I have already ordered a replacement 17HE24 that is rated at 60nM (sorry about the metric, I am not quite sure how it measures in the imperial units)
The control board is TinyBee board running the latest fluidnc and the commands for move were both given through fluidnc interface and NodeJS (tried both).I have the automatic powerdown disabled for the drivers (drivers are TMC2209, with a vref of 1.6 and they are not overheated, they seem cool to me)
For your control board, FluidNC needs to be told holding current, and it has a timeout for how long it holds after commanded moves. The $1 parameter (IIRC) sets it, and $1=255 should set.it to never time out. Check holding current and the value of $1
Edit: I see you have the vRef set, is that with the pot? Does the TinyBee support Serial UART communication to the TMC drivers?
Unfortunately my understanding is that it doesn’t. I tried setting the config with UART but it didn’t work so I opted for the legacy method. I am looking now for your suggestion about $1 parameter in the fluidnc wiki is it something that works only with UART?
Thanks for your valuable help it is greatly appreciated
Actually that did it! The default value was set to 250, changing it to 255 kept the motors energized and now it is able to hold the axis! A small problem that still needs to be dealt with is the random small error on going up or down.
Tried with + and -5 movements and each time was off by max 0.05 but random so I assume it is not going to be fixed with steps calibration. For example gave the command 5mm up and went 4.97, another 5 went 10.1, another five 14.98 so the error was not cumulative. Could it be still binding?
Finally anything else could be done about the assembly crashing down when power is off aside from changing the leadscrew from 4 start to 1 start?
For the random errors, seems like it might be a mechanical issue, but hard to know without seeing it.
For power off crashing look at the Z axis brake threads E-Brake for Belted Z axis Use Z-Brakes!
Those are referring to the belted Z axis on the MP3DP, but could easily apply to a LR3 or Primo.
Tomorrow I am gonna get a photo of the setup and get back to you, after I reset all the bearings, I have a feeling that while trying to set the tension with the bearings on the core (shouldn’t have I know ) they got all out of place
) they got all out of place