Duet Hung Up Mid Print

wilriker

Duet Wifi (HW rev. 1.03) running RRF 2.0

Something I never had before, my Duet Wifi seems to have hung up mid print. I am at work, controlling my printer at home remotely. It is no longer connected to Wifi, so I cannot reach it anymore from here.

Can I do anything to get a M122 when I come home before explicitly resetting it via the reset button so there is a chance finding out what happened? Telnet is disabled but would a USB connection work? Or any other way?

dc42

Yes, please connect USB and try to get a M122 report before you reset the Duet. Also check that it hasn't just paused due to e.g. a temperature fault.

wilriker

@dc42 Would a temperature fault run pause.g as part of it going into pause-mode?
If so, this is not the case because my pause.g does lift the nozzle 5mm and moves X and Y to 0 but I can see on the webcam that the nozzle is just sitting in the middle of a line of the second bottom layer.

dc42

Yes, anything that initiates a pause will run pause.g assuming that the SD card can be read that that point.

wilriker

@dc42 So, in that case I can already say it was neither something that initiated a regular pause nor a power failure (that moves the nozzle up as well - at least it tries with the remaining 11V). Will see what it is tonight and get back with a hopefully successful M122 via USB.

wilriker

Long story short: it was a loss of power after all (with power never being restored to the printer).

It created a resurrect.g file correctly. Also the print is still reasonably stuck to the mirror but due to personal reasons I will only be able to continue this print on Sunday earliest.

Anyway, as it turns out my M911 command did not work as expected since the nozzle was not raised before power was out completely. The command looks as follows

M911 S11 R12 P"M913 X0 Y0 G91 M83 G1 Z3 E-5 F1000"

Apparently 11V are not enough power reserve to move Z up (also extruder does not seem to have retracted the filament). I changed that to S12 (as my PSU is outputting 13.6V in idle and 13V under full load) to hopefully give it a little better chance for the future.

Also instead of using M913 X0 Y0 I set M18 X Y to completely shutoff the motors for X and Y immediately - or is there no advantage to this approach in terms of power saving?

dc42

To stand any chance of having enough reserve power to raise the nozzle a little and retract some filament when the power fails, you need to use 24V power.
M913 is better than M18 for this puropse, because it doesn't wait for motion to stop.

wilriker

@dc42 said in Duet Hung Up Mid Print:

To stand any chance of having enough reserve power to raise the nozzle a little and retract some filament when the power fails, you need to use 24V power.

Yeah, 24V is number 2 on my list right now (but it might move up).

M913 is better than M18 for this puropse, because it doesn't wait for motion to stop.

OK, thanks. I will revert that.

At least I am happy that it successfully wrote the resurrect.g file. That will come in handy.

dc42

Check that the resurrect.g file was created when the power failed, and not when you paused this or another print earlier. With 12V power, depending on what else is powered from the 5V rail, the power could have run out before it was written.

wilriker

@dc42 I did that already but thanks for the reminder. I starts with

; File "Upgrades/DuetEnclosure/Lid.gcode" resume print after power failure at 2018-06-20 10:40

and thereby exactly matches the failure today.

wilriker

Finally got me print to start again. I manually heated bed to temperature first and also hotend to somewhere a little below regular working temp. Then I disabled movement-prevention of unhomed axis to very slowly move Z up (since the nozzle still was touching the printed part). All that went well (except for a little blob being left behind but because that will be on the inside of the print I can live with it).

Then I realized that there seems to be no method to directly run resurrect.g, is there? What I did then was to use M23 and M24. But how is that supposed to be done?

dc42

https://duet3d.dozuki.com/Wiki/Gcode?revisionid=HEAD#Section_M916_Resume_print_after_power_failure

wilriker

@dc42 Thanks