@OwenD It is not a defective driver, because if downgrading the firmware to v3.4 everything is working fine again.

Just a quick info:
The picture of the adafruit LIS3DH on https://duet3d.dozuki.com/Wiki/Accelerometers is misleading because the V3.3 from the duet is connected to 3Vo (Output) instead of the VIN (Input) pin. Maybe it is better to remove it.

@gloomyandy
Some of my findings:
Nr. 1: The purge line is not being completed because the extruder runs backwards after printing about half of the line
Nr. 2: Line is not being printed because the extruder runs backwards
Nr. 3: The extruder motor stands still and neither moves forward nor backwards. Therefore no filament is being extruded until the end of the whole print job.

Findings: There is not only an issue with retracting instead of extruding but also with a complete stop of the extruder motor. Looking at Nr. 1 and Nr.2 it seems to me that the extruder runs backwards whenever the y axis is moving from the front to the back (Y increases).

Hopefully this helps.

@gloomyandy Sorry, of course you're right. Please find attached the the config.g and the macro (PrintStart.txt) including the code for printing the purge line:
config.g
PrintStart.txt

@dc42 I can confirm that the new firmware fixes the issues. Many thanks for the quick fix!

@dc42 I can confirm that the new firmware fixes the issues. Many thanks for the quick fix!

I just activated debugging via M111 P3 S1 and it showed the M409 gcodes being sent from the Arduino. But the Arduino gets no response from the duet.

@dc42 I can confirm that with V3.5RC3 PanelDue stopped working on IO1. Very weird.

@dc42 I am using the Arduino Nano in order to get the printer status, print progess, heater status and temperature in order to display it including animations via Neopixels. Therefore I send different M409 (.e.g. M409 K"state.status) to the duet3.
The issue with V3.5.x is that the Arduino Nano not even get a single byte of response via UART after sending the M409.
Is there a way to see if the duet3 receives the M409 at all? Or do you already have an idea which change between V3.4.x and V3.5.x could cause this different behaviour?

@dc42 @jay_s_uk I tested it with 3.5Beta1, 3.5RC1, 3.5RC2 and 3.5RC3. None of these versions did work.
On the Arduino Serial.available() always returns 0 with these firmware versions.

I now did another test with IO0 and it has the same issue (although a panel due connected to IO0 is working fine). I therefore edited the topic. With 3.4.6 IO0 works flawlessly too with the Arduino Nano.

What I do not know yet is, if it is only an issue receiving anything from the duet board or if there is also an issue sending anything to the duet board from the Arduino Nano. Is there any way to see what the duet receives via the UART ports?

Regarding config via M575 I am using:
M575 P1 B57600 S1
M575 P2 B57600 S1

I did also try M575 P2 B57600 S2 but it did not change the behaviour.

@mule Just begin testing and it already stopped working with V3.5-Beta1

@Phaedrux If I remember correctly the first tests I did was with Beta 4. On the coming weekend I am going to find out since which version the issue exists.

I am using a Duet3 6HC V1.0. Since I upgraded from V3.4.6 to V3.5.x I am getting no response via UART using an Arduino Nano connected to IO1.
Going back to V3.4.6 again solves the issue.

@chrishamm It is an 6HC

Maybe it has something to do with the following:
I am using the duet hardware on a BLV mgn printer with an arduino which polls (M409 K"job" and M409 F"d99f") the duet via serial port in order to get the hotend temperature, bed temperature and printer status to show them via neopixels. The polling frequence is set to 3 seconds. Using this since ages now.

@chrishamm Only one session has been open