Warning to Consider When Fans Powered Externally



  • I'm powering the fans from a separate power supply to the general V-IN and 5VDC for the duet. Logic being that when the big red button is hit two 24VDC guided contact relays are dropped and within a couple seconds the steppers are stationary and the hot bits are cooling. This interlock circuit is also broken by the M81 command.

    The logic of powering the fans externally is two fold. They share the same power supply that also feeds the coils of the guided contact relays. If there is a short circuit in the fan it will pop the supply fuse and drop the contactors stopping heat to the hot end and reducing the chance of a seriously jammed hot end. Like wise if someone hits the big red button or the M81 event is raised the heaters are killed but the fans are not, again reducing the chance of a constipated hot end.

    Logical but when I powered up my system for the first time after converting to the DuetEthernet from Duet 0.6 the VIN light was on when there was no supply to VIN. This confused me as nothing was supplying voltage to the VIN. At this stage the only supply to the board was 5VDC.

    So, referenced the wiring schematic:
    https://github.com/T3P3/Duet/blob/master/Duet2/Duet2v1.03/Duet2_Schematic_v1.03.pdf

    On Sheet 4 I can see the fans are supplied by V_FAN, but there is no connection to these pins. Only wire to these is on the FAN0-, but I was not calling for the fan to be on and there should be an open circuit MOSFET between this and the ground.

    It all becomes clear on sheet 6. When I look at "FAN 0 MOSFET CONTROL" I can see almost what I expect, with a MOSFET which sinks the 24V (that I'm supplying to FAN0- from an external supply) when a signal is received from the main processor. What I didn't expect was the diode going from FAN0- to V_FAN.

    This still didn't quite explain the behaviour until I looked at the "FAN Voltage Select & Fuse" section of the same page. My externally supplied 24V was feeding through the FAN0-, not to ground through the MOSFET but back up to the fan voltage select, through the jumper to the V_IN which was not supplied as it was interlocked and then supplying power to the whole board, including the heaters and steppers! No harm done as the board was only live for long enough for me to see an unexpected light (and hear the fans running unexpectedly) and shut it down.

    Maybe this is covered in the wiki, but if you are supplying the fans in the same way as I am completely remove the voltage select jumper from the board and this problem is fixed.

    Took me a while to figure out what this diode was doing. If you power the fans normally then it is essentially running in parallel to the device on the FAN0 connector, but connected in reverse. I think it is trying to be a flyback diode to protect the MOSFET from inductive kick backs from inductive loads that are not suitably protected at the device. Essentially this is idiot proofing, as to be most effective this should be as close to the inductive load as possible. In view of the wide range of skills and the steep learning curve required for some it is however an understandable addition.


  • administrators

    We added the flyback diodes in PCB revision 1.02 after reports that several users had burned out fan MOSFETs by connecting Berd Air and similar air pumps that use brushed DC motors.



  • Yes thanks, appreciate I the logic. I hadn't anticipated their presence as you normally put flyback diodes as close to the inductive loads as possible to minimise the length of wire that the inductive kick back spike runs through. All of the wire with the pulse can act as an aerial and cause another source of noise.

    I'd suggest an addition to the wiring wiki page to warn that is you do power the positive side of the fans separately to the Duet fan headers then pull out the voltage select jumper completely.



  • @doctrucker Maybe I am not getting something in your setup, so I explain what I understood:

    You have a separate PSU for your fans. You connect the positive side of your fans directly to a positive connector of this PSU. The negative pin of your pin goes to the Duet headers - I assume to still let you control the fans.

    Why don't you connect your fans the regular way to Duet's headers and supply the fans from the separate PSU by connecting the positive from this PSU to V_FAN? This way the jumper has to be removed anyway and you should get the same result - and AFAIK this is also the way it is supposed to be done when powering the fans from a different source (I will use it when I switch to 24V together with a buck converter to keep my 12V fans).



  • That would work fine yes. I was moving an exisiting loom over from a Duet Board where I also needed to fuse the fan supply to stop a full VIN current going down the spindly fan wires in a fault before the fuse was added the the 1.03 boards.


Locked
 

Looks like your connection to Duet3D was lost, please wait while we try to reconnect.