Topics created by sverreb

@dc42:

Yes, it's entirely possible that an electrical transient or EMI is putting the hardware in a state that is triggering the problem. One possibility is that it is affecting the I2C interface between the Duet and the DueX.

Is there a flyback diode in parallel with the relay coil, and a snubber network either in parallel with the solenoid or in parallel with the relay contacts?

No, I do not have one. I could switch it to an SSR if that would be better and not require extra hardware. Are there any off the shelf solutions for diodes/snubbers, preferably DIN rail mounted? That would be a good option, I just don't have time to make one from scratch.

I have a feeling that this behaviour was deliberately changed in firmware 1.20 because several users complained that babystepping wasn't remembered from one print to the next.

That will happen if you use mesh bed compensation and the bed is sufficiently uneven that the Z speed changes needed to implement it are constrained by the Z jerk setting. Try a higher jerk setting if your Z axis can handle it.

Feel free to add this feature request to the wish list at https://www.duet3d.com/forum/thread.php?id=961. Many of the items on that list have now been implemented, so I will start a new list soon.

I believe you could do this with a virtual heater on firmware 1.19. If you use a silicone heater with an integrated thermistor and a separate thermistor at the top (or the side or whatever), then you calibrate the PID loop to the first thermistor (showing the temp of the heater) but in your prints wait for the 2nd thermistor (the virtual heater) to be at the desired temp.

This way the bottom will never exceed 150C even if the top hasn't reached that temperature yet.

I have not done this personally but I plan to do this if I ever rebuild my silicone heater bed, for exactly the same reason.

Cheers,
Ben.

EDIT: Sorry, I meant to link to the relevant parts of the documentation: https://duet3d.com/wiki/G-code#M305:_Set_temperature_sensor_parameters

That will be ok as long as the switches are break-before-make. I think most microswitches are. However, it's probably not necessarily if you use the normally closed contacts.

I think you are right. I tried moving the leadscrew by hand and only got 20% of the movement on the bed. Likely the bed resists torque around the x axis enough that the motor mounts (which are printed) deflects more than the bed rotates.