PWM Fan mosfets
-
If a pwm fan mosfet went bad(for Fan1 in my case), will Fan1 remain on at all times? I have the exact same config.g file on 2 printers, one works just as intended, but the other printer has its fan1 always on. I tried changing temperature ranges to extreme low to turn the fan off (its set as a thermostatic fan) but this Fan1 never turns off. Everything else works perfectly. Could a bad mosfet do this?
-
@markduet said in PWM Fan mosfets:
If a pwm fan mosfet went bad(for Fan1 in my case), will Fan1 remain on at all times? I have the exact same config.g file on 2 printers, one works just as intended, but the other printer has its fan1 always on. I tried changing temperature ranges to extreme low to turn the fan off (its set as a thermostatic fan) but this Fan1 never turns off. Everything else works perfectly. Could a bad mosfet do this?
Mosfets can fail in various ways. One of the failure modes is to go short circuit, which would leave the fan on all the time.
-
@dc42 Thanks for explaining. One more thing I forgot to mention is I connected a buck converter to the fan1 output, since my fan is a Noctua 12v fan, (which I put on the hot end to reduce fan noise). That is to reduce the 24 volt to 12v. I'm wondering if that could have fried the mosfet into the "shorted" mode. My next purchase will be a mosfet to repair this one on the board. I never removed an SMD part yet, so am a bit nervous.
-
@markduet, you might want to post the way you wired up the buck converter. It is a bit counter-intuitive.
Did you try removing the buck converter and see what the fan outputs are doing ? -
See here on connecting fans and using a buck converter.
https://duet3d.dozuki.com/Wiki/Connecting_and_configuring_fans
The Duet switches PWM on the negative side, so if you use a buck converter, either use it to put 12v into the v_fan pin for all fans, or if you only have a single 12v fan, leave the negative lead of the fan connected to the Duet and the positive side to the buck converter for 12v.
-
@jens55 I took the 2 wires from Fan1 on the Duet board, and hooked them up to the buck converter input. I also tried removing it (the buck converter) and connecting a 24 volt fan directly to the Fan1 output on the board. Same thing happened. Interestingly, I did this identical on 2 machines. One works and other has the fan always on.
-
@phaedrux I have a single 12v fan. I will try your recommendation, i.e, using only the -ve lead from fan1 socket on the board and +ve directly from the buck converter. I understood correctly I hope.
-
@markduet said in PWM Fan mosfets:
@dc42 Thanks for explaining. One more thing I forgot to mention is I connected a buck converter to the fan1 output, since my fan is a Noctua 12v fan, (which I put on the hot end to reduce fan noise). That is to reduce the 24 volt to 12v. I'm wondering if that could have fried the mosfet into the "shorted" mode.
Yes, very likely. Buck regulators take a large surge current when powered up because of the input capacitor, and that surge probably exceeded the mosfet peak current rating every time you turned the fan on.
To use a buck converter, its input and ground must be connected directly to VIN and ground (preferably picking them up from the always-on fan connector, to get the benefit of the fan fuse). Then the 12V output of the regulator should be connected to the positive wire of the fan(s), and the negative wire of the fan connected to the FAN- pin of your chosen fan output. One buck regulator can supply several fans. If all the fans are 12V, then the wiki page that @Phaedrux linked to shows how you can route the 12v through the fan voltage select jumper block instead.
-
@markduet said in PWM Fan mosfets:
I also tried removing it (the buck converter) and connecting a 24 volt fan directly to the Fan1 output on the board. Same thing happened.
In that case it sounds like the mosfet has failed in the on state.
-
@markduet, yeah, that is a no no .... but shouldn't have blown up anything. See the other suggestions for proper hookup.
Edit: forgot about the input caps on the converter but didn't think there would have been enough current for long enough to kill the mosfet. I guess they are more sensitive than I imagined. Since the input caps don't have to do AC to DC conversion, they don't need to be that high a value (or so I thought).Edit2: I prefer the theory that you inadvertently shorted the fan output somehow (unfortunately it happens, sometimes more than once <sigh>) but what do I know,
Replacing my fets was fairly painless although not real pretty. I opted to cut the leads so I was able to deal with one pad at a time.