[3.4 stable/RC2] Error on PanelDue when initializing

oozeBot

@dc42 @T3P3Tony - An error has been introduced in 3.4-rc2 and remains in 3.4 stable that did not exist in previous betas which is presenting itself on our printers upon initialization. I spent some time whittling config.g down to the minimum code to reproduce.

Reminder our printers use a Duet 6HC and an RPi. They also use a 5v power supply and a 24v power supply that works independently.

To recreate the bug, the 24v power supply must be disabled. Issuing M999 to reset the printer triggers the issue. Strangely - the error is only displayed on the PanelDue (3.4.1-pre2) and is not logged to DWC. The error reads:

Error: Failed to switch off remote heater 1. Response timeout CAN addr 20, req type 6013, RID=XX.

The error is presented immediately upon initialization. It is not waiting for the G4 S1 (or even G4 S10) to delay the script.

Please let us know if there is anything else we can assist with in identifying the issue.

config.g to reproduce

;==========================================================================================
;== Power Settings ========================================================================
M80 C"!pson"      ; Turns on 24v power supply
M950 P0 C"0.out0" ; Creates P0 as GPIO
M42 P0 S1         ; Enables power to Toolboard(s)
G4 S1             ; Wait 1 second

;==========================================================================================
;== PanelDue ==============================================================================
M575 P1 B57600 S1 ; Configures PanelDue

;==========================================================================================
;== Extruder Heater =======================================================================
M308 S1 P"20.temp0" Y"pt1000"                          ; Configures PT1000 for hot end
M950 H1 C"20.out0" T1                                  ; Creates hotend heater on toolboard
M307 H1 R2.702 K0.632:0.000 D5.77 E1.35 S1.00 B0 V24.1 ; Heating process parameters
M143 H1 S500                                           ; Temperature limit for extruder

T3P3Tony

@oozebot does the toolboard have an independent supply of 24V from the mainboard. it looks like it from the config.g start.

oozeBot

@t3p3tony Since our build plate heaters are mains powered, we are using the Out0 screw terminals for the built in fuse and ease of connecting it to the system. That 1s wait has always allowed it enough time to enable in the past. But like I mentioned, even increasing that to 10s still instantly triggers the error. Is it possible G4 isn't properly waiting when run within config.g?

dc42

@oozebot I am surprised that you didn't see that error message. The warning is generated because you have a heater configured on the tool board, and M999 attempts to turn all heaters off. It's warning that it has lost communication with the tool board so the tool board heater may still be on.

A workaround would be to un-configure that heater (by assigning it to port "nil") before you turn off power to the tool board.

oozeBot

@dc42 Thanks for the info. Strange it's never appeared before 3.4-rc2. Perhaps it's always done this but the error just isn't being suppressed on the PanelDue like it is within DWC as no errors are logged in the console when this happens.

oozeBot

@dc42 @T3P3Tony - adding the following M950 before the M81 corrects the issue.

M950 H1 C"nil" 
M81 S1

Like you said, further investigation shows this error is occurring when powering down, but the PanelDue does not display the message until after it resets with the M999. And again, this never presented itself before 3.4-rc2. However, we are good with the fix.

Thank you both for the assistance.

dc42

@oozebot are you running with attached single bard computer? If so, the change is because it used to be the case that when you ran M999 in SBC mode, it didn't take the Duet through the proper shutdown procedure before resetting. This was fixed between 3.4rc2 and 3.4 stable.

oozeBot

@dc42 Yes, we are. That explains it. I'm glad there was a simple fix for this as it's the first thing our users are met with when turning on the machine..! Thanks again.