PWM fan controller possibly destroyed - advise requested
-
Hi all.
I was having trouble with cooling on a PLA print, where it seemed that the fan was not putting out very much air at all running at 100%, and was so quiet it was nearly silent. It's a Noctua NF-A4x10 FLX fan.
While trying to ensure that the fan was indeed getting the correct voltage (the jumper was between the VIN and ground, not between 5V and ground), I was touching the probes of my multimeter to the PWM fan 0 header pins when the probes momentarily contacted each other, and there was a tiny spark.
Now PWM fan 0 doesn't come on at all. After this I sheepishly crimped a red and black wire into a 2-wire Molex KK connector and, keeping the stripped ends carefully separate from each other, measured the voltage on those two pins. It reads 12V, but the fan simply won't come on at all, either by running the fan slider to 100%, or by sending M106 P0 S255.
I already know that the PWM fan 1 on the hotend works fine, since I've tested it by setting the hotend over the thermostatically controlled threshold temperature and waiting for it to kick on. It does.
I switched the fan that was connected to fan 0 over to fan 2 and sent M106 P2 S255 and it comes on, though again it's very quiet and with just barely detectable airflow.
I tested this with a brand new out of the box Noctua NF-A4x10 FLX fan (12V) and it too comes up nearly silent and with extremely weak airflow.
It's possible this model of Noctua fan is just incredibly weaker than I thought it was.
So, some questions:
-
Did I fry a PWM controller or something on fan 0 when I shorted the pins with my multimeter probe? Is fan 0 now dead forever?
-
Why does fan 0 header pins still read 12V when properly probed using a much safer wiring harness, an idea which hadn't occurred to me before the OOPS! moment?
-
Does anyone know of a way either to remap fan 2 to be used in place of fan 0 inside the Duet Wifi configuration? I just checked in Simplify3D for a place where I could specify the fan number to be used for all cooling fan commands, and found nothing. If I can't get my slicer to generate M106 P2 commands in place of M106 P0 commands, and I can't map P2 to P0 in the Duet Wifi config, am I just screwed?
-
Does the company offer some sort of sympathy discount on a replacement board? If not, this is a pretty expensive oopsy.
-
Back to the original problem, is that Noctua fan just really that weak? I was unable to print Baby Groot's head with it running 100% at the very top where it's got some smallish peaks where the layer time is simply low enough that without adequate cooling, the previous layer was still soft enough that it all gooped up into a liquid ball of fail.
-
-
I be able to answer 3) myself, after finding the fan document https://duet3d.com/wiki/Connecting_and_configuring_fans. It looks like I may be able to use an M563 command that specifies F2 for both T0 and T1, and simply use the fan on fan 2 slot instead of the apparently toasted fan 0 slot. Haven't tried it yet because the printer's in the middle of a PETG print that needed no fan, but I'll try it shortly.
It'll be regrettable if I really lost the PWM slot for fan 0 because I was thinking of using an empty PWM fan slot for controlling something else like LEDs, but I can live without that if it saves me $160.
I'm still curious on the other questions. I did look up specs for the Noctua fan, and it does look pretty weak, at only something like 4.8 cfm. I just ordered a pair of Sunon fans that will be much louder, but will do something like 10 cfm, which should be much more effective. If they're too loud maybe running them at 80 or 90% would cut down a lot of noise without cutting too much effectiveness.
-
Hi Sethipus, I'm sorry to hear of your misfortune. It sounds likely that you have blown the FAN0 mosfet. This is not difficult to replace. Do you know anyone who has hot air desoldering equipment?
Yes you can map fan 0 to fan 2 in the M563 commands.
-
Hi Sethipus, I'm sorry to hear of your misfortune. It sounds likely that you have blown the FAN0 mosfet. This is not difficult to replace. Do you know anyone who has hot air desoldering equipment?
Yes you can map fan 0 to fan 2 in the M563 commands.
Thanks for your response David.
I have a soldering station that has a hot air gun for doing surface mount stuff, but I've never done any. I even have a tube of solder paste I bought once when I thought I might do some surface mount stuff, but it never happened. I've only used the normal soldering gun part of the station. I wouldn't mind having a look at the mosfet that would need to be replaced if you don't mind telling me what it is, and I'll make an informed decision on whether it's worth the risk having me actually try to do that.
In the meantime, after my fanless PETG print finished, I edited config.g and added F2 to the T0 and T1 definitions. I've got a print going right now, and the fan kicked in at the appropriate time. The fan slider in the web interface no longer works, but that's not that big a deal. I can make a macro to turn it on if I need to, for testing purposes or whatever.
So, honestly, I might well be better advised to just leave well enough alone. I'm down a PWM header, but I've got a fully functioning printer at the moment with what's actually on it right now. I'll still have a look at the mosfet though.
I had no idea how weak that Noctua fan was for cooling PLA. Naturally I don't use much cooling for PETG, but I've been trying out Hatchbox wood-filled PLA lately, and it's runny as hell, and apparently needs all the cooling it can get. The Sunon fans will be here in a couple days. It will be annoyingly loud, but hopefully will get the job done.
-
See https://duet3d.com/wiki/Connector_and_spare_part_numbers#Fan_mosfet for info on replacing the PWM fan mosfets.
-
Thanks David! I just ordered some mosfets. We'll see if I end up actually doing it, or whether I decide to cut my losses and just the the Duet Wifi as presently configured, since it is in fact working now.
-
I just wanted to say that I just replaced my fan1 mosfet as I must have shorted some wires during installation. I don't have a hot air station and I'd never done any surface mount soldering before, but I did get it replaced and with no collateral damage. Had I known the package with the mosfets was getting delivered that day, I would have waited until afterwords for my normal coffee break as the hands were a bit jittery after that caffeine injection.
-
Fabulous, had the same problem, ordered the part, question though.
I still measure 24v coming from the wire .
Would you still measure voltage if its the mofset blown?
-
Im not bad at soldering with an iron, but dont have a hot air gun solder station is it going to be possible to change the mosfet with a temp controlled iron, magnifyer, a steady hand and nerves of steel?
-
The tricky part is removing the old mosfet without lifting the PCB traces. I've never tried that without using hot air to heat all 3 joints at the same time. I found this video on removing SOIC ICs without hot air: https://www.youtube.com/watch?v=tlSY1uaw0GA. Maybe it would work for small mosfets too.
The new mosfet can be soldered in place using tweezers and a soldering iron with a fine-tipped bit.
-
that video makes it look too easy, We will see I guess possibly this week end then I get get my Printer to be quiet again, Thanks Admin Great board.
-
Did it in 20 min with my intermediate soldering skills. Fan started working right away. Thanks guys for the clear info on what , where , and making possible the components to be replaced.
Cheers
Stav -
The tricky part is removing the old mosfet without lifting the PCB traces. I've never tried that without using hot air to heat all 3 joints at the same time. I found this video on removing SOIC ICs without hot air: https://www.youtube.com/watch?v=tlSY1uaw0GA. Maybe it would work for small mosfets too.
The new mosfet can be soldered in place using tweezers and a soldering iron with a fine-tipped bit.
I cut the legs and de-soldered each one. I was not comfortable using my hot air station there as it's pretty dense.
-
My worry has always been that if I try to cut the legs, the PCB pads would lift. I guess it's OK if your cutters are sharp enough!
