I was trying to hook up an external MOSFET driver to run the berd air pump with the F0 fan output, however, I accidentally reversed the pins on the connector for a split second. Instantly, I smell and saw a puff of smoke from the DuetWifi v1.0 board. My heart sank, but everything else seems to function. Note, the berd air pump has a 1N4007 flyback diode across the terminals, but the MOSFET driver board did not have the protection.
I ran some diagnostics and got weird results.
- Multimeter shows 12V on both "always on" connectors at the board and end of the wire harness with the correct polarity.
- I was able to hook up a LED light and verify the F0, F1 and F2 fans was able to vary the voltage using the web control. The LED goes from low brightness to completely off based on the speed control setting. It seems like the control is able to modify the duty cycle of the PWM signal.
- Hooking up additional loads like a fan to the same LED connection instantly drops the current to the point where the LED dims and turns off
- Verifying the F0, F1 and F2 connection using the same step as #2 with a fan would not be able to make the fan blades move. It seems like there is not enough current to drive a fan load but had barely enough current to drive the LED load.
- Switching the jumper from 12V to the 5V setting, the fan runs fine at the 5V level
- It seems like the PMV40UN2 MOSFET transistors are still functioning to a certain level since I was able to control it.
- There is not enough current from the 12V source to the fan drivers but the 5V side have enough current to run the fans.
- I did not see any obvious signs of a part being burned out on the PMV40UN2 transistor, but the smell of burnt plastic is obvious
- Hooking the LED and fans directly to a 12V source shows normal operation and full speed/brightness.
- I just re-examined the board and the U10 (HC125A PSCC) part looks like there is a small ring of discoloration. It might be the culprit, but I don't have a schematic to look at the chip's function. The mouser catalog shows it as a quad buffer line driver. Is this a TSSOP-14 package size? Will the part below be comparable?
- Is there a regulator or component that feeds into all the always on, F0 to F2 fan controllers?
- Could some of the traces on the board be burned out?
- Is there a schematic for me to troubleshoot this more?
I ordered replacement PMV40UN2 MOSFET transistors just in case, but would like to know DC42's comments.
I found the V1.03 board schematic online. Its odd the U10 is a SPI1 buffer on the schematic.
I just did a heart bypass by connecting the middle terminal of the fan voltage select jumper to the 12V coming into the board (V_FAN) and now I have +12V directly to the jumper. I just checked the F0 and was able to control the output on the web control. It seems to point more towards a burned circuit connection from the fan voltage select jumper to the source of the V_IN to that jumper. At least the bypass trick seems to work for now.
On PCB revisions prior to 1.03, if you short out a fan connector then the PCB trace (or more likely a via in it) that connects +VIN to the VIN end of the fan voltage selector jumper block usually fuses. I think that's what happened in your case, although there is enough leakage there still to run an LED but not a fan. I suggest you check it with a multimeter. You can bypass that trace using a wire on the back of the board. Make sure you connect it to the correct end of the jumper block!
Thanks. I think that is what happened in my situation. I am using a jumper to the middle pin directly to the input terminal where the +12V comes in. It works great right now so I won't have to do any major surgery on it yet. That burnt smell and smoke made me fear for the worse.