Duet 2 Expansion Board Malfunctioning

Hardware_SortingRobotics

Hello,

I am currently investigating a strange issue that has occurred on several Duet 2 Expansion Boards.

The issue centers around the two heater PWM outputs H6 and H7. When configured properly, these two outputs normally lie at 0V, and when the heaters are active, they output a 5V PWM signal. The issue is such: sometimes, the boards will output something like 2-3VDC without heaters being turned on. We see this voltage as soon as the duet board receives power from the mini USB cable. The presence of 24V at V_IN is not required for these pins to erroneously activate. Additionally, they are not outputting a PWM signal. Its just a straight 2-3V DC with no oscillations.

We believe that the cause has something to do with the U5 chip, which is responsible for inverting the 3.3V active-low signal which is sent from the 50 pin ribbon cable. Something is shorting and causing the chip to un-invert the signal.

Any advice on this? We have not been able to replicate this issue by shorting the ground of the 24V DC power supply.

chrishamm

@injoi9000 What firmware version are you using? AFAIR the last Duex PWM ports are controlled using a port expander and there have been some bug fixes for that in the past.

dc42

@injoi9000 the only reason I can think of for that type of problem is a bad ground, 5V or 3.3V connection from the Duet via the ribbon cable. Have you tried using a different ribbon cable? You could also check that the soldering of the ribbon cable connector pins on the Duet and on the expansion board look good.

Hardware_SortingRobotics

@chrishamm We are using 3.1.1 at the moment.

Hardware_SortingRobotics

@dc42 We can't really reproduce the failure mode yet to this day. I do believe that once we transfer the board to a new ribbon cable, the problem persists. To make matters worse, the symptoms range a lot from one failure to another.

The last time this happened, there was a constant ~3V offset between H6_PWM and ground. Same with H7_PWM. However the triacs connected to this output did not activate despite this voltage difference. Which does lend some traction to the "bad ground" theory.

Could you elaborate more on what you mean by "bad ground"?

dc42

@injoi9000 said in Duet 2 Expansion Board Malfunctioning:

@chrishamm We are using 3.1.1 at the moment.

We have made numerous improvements to the firmware since that version, which is old and no longer supported. Please upgrade your firmware to version 3.4.5.

droftarts

@injoi9000 said in Duet 2 Expansion Board Malfunctioning:

Could you elaborate more on what you mean by "bad ground"?

Proper grounding of the Duex to the Duet is essential for signal integrity. See https://docs.duet3d.com/Duet3D_hardware/Duet_2_family/DueX2_and_DueX5#power-wiring

Ian