@dc42
I'm replying to this rather than make a new post, just so future searching is easier for some people.

I've got essentially the same issue as @GoremanX
I've got a Duet 3 and I'm not sure if it'd be better to connect it to out_1-3 or out_0.

The documentation does say that out_1-3 require external flyback diodes but you say here that they're built in. Is the latter true? Can I directly connect an inductive load to those outputs?

For that matter, are inductive loads safe on out0? It'd actually be easier in my case to run the pump off of out_0 because I won't have to move my outputs around.

Hah, yeah I think I was defining F1 on T0 and it was letting M106/M107 kill the thermostatic fan. I did have the hotend fan on F1, and the way cura sets up its files was apparently allowing that to work.

I think it's actually fixed now.

Yeah. I went through various fan changes and they were assigned oddly (but functioning, for months).

I just changed them back to part cooling on 0 and the hotend on 1.
I thought that fixed the weird behavior but it doesn't seem like it actually did.

For reference, this is the entirety of my config.g

; General preferences
G90                                          ; send absolute coordinates...
M83                                          ; ...but relative extruder moves
M550 P"Sk-Go"                                ; set printer name

M667 S1                                      ; select CoreXY mode

; Network
M552 P0.0.0.0 S1                             ; enable network and acquire dynamic address via DHCP
M586 P0 S1                                   ; enable HTTP
M586 P1 S0                                   ; disable FTP
M586 P2 S0                                   ; disable Telnet

; Drives
M569 P0.0 S1                                 ; physical drive 0.0 goes forwards
M569 P0.1 S1                                 ; physical drive 0.1 goes forwards
M569 P0.2 S0                                 ; physical drive 0.2 goes forwards
M569 P0.3 S1                                 ; physical drive 0.3 goes forwards
M584 X0.0 Y0.1 Z0.2 E0.3                     ; set drive mapping
M350 X16 Y16 Z8 E16 I1                       ; configure microstepping with interpolation
M92 X200.00 Y200.00 Z800.00 E409.00          ; set steps per mm
M566 X700.00 Y700.00 Z60.00 E2500.00      	 ; set jerk (mm/min)
M203 X12000.00 Y12000.00 Z900.00 E3600.00    ; set max speed (mm/min)
M201 X1200.00 Y1200.00 Z140.00 E2500.00      ; set acceleration (mm/s^2)
M906 X800 Y800 Z800 E600 I30                 ; set motor currents (mA) and motor idle factor in per cent
M84 S30                                      ; Set idle timeout
M572 D0 S0.06						    	 ; set pressure advance	(0.085 formerly for PLA, 0.4 nozzle)
;M593 F40  									 ; cancel ringing at xHz

; Axis Limits
M208 X0 Y0 Z0 S1                             ; set axis min
M208 X305 Y300 Z300 S0                       ; set axis max

; Endstops
M574 X1 S3                                   ; configure sensorless endstop for low end on X
M574 Y1 S3                                   ; configure sensorless endstop for low end on Y
M574 Z1 S2                                   ; configure Z-probe endstop for low end on Z

; Z-Probe
M950 S0 C"io7.out"                           ; create servo pin 0 for BLTouch
M558 P9 C"io7.in" H5 F120 T9000              ; Z probe type to bltouch and the dive height + speeds
G31 P500 X31 Y10 Z2.28                       ; z probe trigger						 
M557 X31:280 Y15:290 P6                      ; define mesh grid
M376 H5										 ; taper compensation over 5mm

; Bed Heater
M308 S0 P"temp1" Y"thermistor" T100000 B4138 ; configure sensor 0 as thermistor on pin temp1
M950 H0 C"out1" T0                           ; create bed heater output on out1 and map it to sensor 0
M143 H0 S120                                 ; set temperature limit for heater 0 to 120C
M307 H0 B0 S1.00                             ; disable bang-bang mode for the bed heater and set PWM limit
M140 H0                                      ; map heated bed to heater 0
M307 H0 A115.0 C276.6 D1.5 V23.8 B0			 ; autotune result feb 28, 2020

;Extruder Heater
M308 S1 P"temp2" Y"thermistor" T500000 B4723 C1.196220e-7 ; configure sensor 1 as thermistor on pin temp2
M950 H1 C"out2" T1                           ; create nozzle heater output on out2 and map it to sensor 1
M143 H1 S320                                 ; set temperature limit for heater 1 to 320C
M307 H1 B0 S1.00                             ; disable bang-bang mode for heater  and set PWM limit
M307 H1 A422.5 C138.4 D4.4 V23.8 B0		     ; autotune result april 11, 2020

; Extruder/Hotend Fan
M950 F1 C"out4" Q16000                       ; create fan 1 on pin out4 and set its frequency
M106 P1 S1 H1 T45             		         ; set fan 1 name and value. Thermostatic control is turned on

;Partcooling/Tool Fan
M950 F0 C"out9" Q25000                       ; create fan 0 on pin out9 (servo/laser/pwm) and set its frequency
M106 P0 S0 H-1 X0.35 B0.1                    ; set fan 0 name and value. Thermostatic control is turned off

; Tools
M563 P0 D0 H1 F1 S"BMG-M Mosquito"           ; define tool 0
G10 P0 X0 Y0 Z0                              ; set tool 0 axis offsets
G10 P0 R0 S0                                 ; set initial tool 0 active and standby temperatures to 0C

; Miscellaneous
M501                                         ; load saved parameters from non-volatile memory
T0

I don't, but I can put it in there.

Interestingly, while swapping the fan numbers seems to have solved the problem on the file start,
I lied, it's not fixed. Apparently it's an intermittent problem which is just lovely.

I just finished a test print and the M107 that Prusaslicer sticks at the end of the file (before the user end gcode) will still kill the extruder fan before it's below the threshold.

Duet 3, RRF 3.01 RC10

Moving over to Prusaslicer from Cura and my functioning start gcode is misbehaving weirdly.

I can slice the same part and the Cura gcode works as intended, with thermostatic fan turning as the firmware commands at 45c. With the same start gcode, the Prusaslicer file somehow lets the hotend heat all the way up without the fan on.

The only significant difference (that I can see) in the .gcode files between the two is that Prusaslicer inserts a M107 at the very beginning (before the actual user-start gcode). Cura has the same M107, but it's inserted after the start gcode.

Hotend fan config.g

M950 F0 C"out4" Q16000                       ; create fan 0 on pin out4 and set its frequency
M106 P0 S1 H1 T45             		         ; set fan 0 name and value. Thermostatic control is turned on

Prusaslicer

M107
T0
M190 S60 ; wait for bed temp
M104 S210 ; set extruder temp
M109 S210 ; wait for extruder temp
M82 ;absolute extrusion
G28  ;Home
G29 S1  ; load mesh
G1 Y5 ; move to front edge
G1 Z0.28 F100  ; nozzle to extrusion height
G1 E5  ; make a little blob
G1 X150.0 E15 F500.0 ; purge line
G92 E0 ; zero extruder
G1 Z1
G21 ; set units to millimeters
G90 ; use absolute coordinates
M82 ; use absolute distances for extrusion
G92 E0

Cura

T0
M190 S60
M104 S210
M109 S210
M82 ;absolute extrusion mode
G28 ;Home
G29 S1 ; load mesh
G1 Y5  ; move to front edge
G1 Z0.28 F100  ; nozzle to extrusion height
G1 E5  ; make a little blob
G1 X150.0 E15 F500.0 ; purge line
G92 E0 ; zero extruder
G1 Z1
G1 F3000 E-0.8
;LAYER_COUNT:100
;LAYER:0
M107
;MESH:20mm_cube.stl

The weird catch is that I removed the M107 from the Prusaslicer file and it worked as intended, with the hotend fan turning on at 45c during the M109 S210 command.

@OwenD Unfortunately that's not the issue, at least with those of using using slightly odder part cooling.

More information in general for what's happening specifically for me;

When I call a fan name in M106, the slider is in DWC while not printing, but doesn't actually do anything. When printing, the part cooling fan works as it should, but the slider sits at zero unless I touch it (and I do then have control), until the gcode changes the fan setting (at which point the slider stays where I put it and the fan changes to the speed the gcode called for). In this instance, the option for Tool Fan and the same fan as the name I've set in M106 are available.

If I remove the fan name from M106, it is not available while not printing at all (and neither is Tool Fan), but it otherwise functions the same during a print (Tool Fan becomes available).

If I run M106 P1 S0.5 while not in a print, it does work.

edit:
After updating to RC6, the functioning slider is back while not printing, but it's displaying the pre- X value set in M106. In my case, X0.3 and setting the fan slider at 50% will cause it to snap back to 15. Using the + and - buttons is sort of fruitless unless you click really fast.

@Phaedrux
3.0 (2020-01-03b3) and DWC 2.1.1 (though it was doing it before my recent DWC update)

I'm another with a fan thing happening. Using the laser/servo pwm with a blower on a Duet 3 for part cooling and the blower works as intended, but displays as 0 speed in DWC during a print.

M950 F1 C"out9" Q25000
M106 P1 C"Partcooling" S0 H-1 B0 X0.3

I just encountered this recently while lazily tuning z probe trigger height. I'm using a printer with a moving Z axis for reference

During a print, babystepping works as intended, but I noticed that it has a reversed effect on the Z limit when using the jog controls when not printing.
If I set the babystep offset to +0.15mm and then jog the bed positive and then negative back to it's limit, it will actually let the bed come up to -0.15mm as reported in status/tool position.

As everything works during a print as it should, this isn't actually an issue. I encountered it largely by accident while using babystepping on a test print. I had assumed that if I moved the Z axis to it's limit with the jog controls after the print, I would be at my babystepped height.

Really just curious what the interaction of babystepping is with Z0 limit and why it gives a reversed effect.

Hah! The man himself. Your blog is actually what led me to messing around with this in the first place. I had some RC brushless motors around and I was going to work on a printed blower and then got lucky and found an actual CPAP blower very cheap.
You're responsible for my belt drive z axis with a Rino stepper as well, in fact.