Somke and burning smell after connecting ethernet cable


 

TimCorrin

New member
Smoke and burning smell after connecting ethernet cable

Hi,

I have recently encountered an issue with my heater meter 4.3 raspberry pi zero W setup. It was working fine with my micro damper setup however I wanted to test it with just using a fan so I soldered up an Ethernet cable to a 5.5mm dc power jack. I plugged this in and was trying to test the fan in manual mode however i noticed the ethernet cable getting hot and there was a nasty smell and smoke coming from the heatermeter.

I quickly disconnected the ethernet cable and tested the unit again with my micro damper but noticed that the fan in voltage mode would only come on at 100%. I then decided that i would leave it alone and come back to it later before i made things worse.

When I did come back to it I tried testing the output volatage on the microdamper jack and the fan output was showing 12V on 5% output. I then went to reconfigure the output to pulse mode and found I could no longer access the web interface raspberry pi was getting hot. At this point it seemed everything was going wrong so I stopped again and decided to write up the issue and ask for help here...

I have ordered a new pi zero W but I am wary of plugging it in in case it suffers the same fate as the first.

I have a soldering iron and multi-meter so should be able to solve the issue.

Any help or advice would be much appreciated.

Cheers

Tim
 
Last edited:
I recently tried to use a HeaterMeter to control a high power DC fan (rated at 0.9A, but pulled over 4A when powering the windings) so it could bear some resemblance to your situation. When it died after running for a minute or so, when the blower output was on (even with no load), the display would get all messed up and the device would keep resetting. I pulled components one at a time, the 47u/25V output cap, the inductor, the power mosfet, the mosfet driver, the diode. Each component checked out as working and with their parameters still in spec. Oddly, I popped a new atmega in and everything started to work again so I put all the components back and it worked perfectly again. I still have the atmega sitting on my desk because I'm not sure how it could have been damaged because it isn't driving an current.

I'd check the 5V voltage on the HeaterMeter's Pi header with no Pi connected and see if you've got a solid 5V there or more. Check the RX, TX, MIS, MOS, SCK, G25 lines to make sure none are above 3.3V. Also you could carefully inspect the drive components, the power mosfet, the smaller BS170 driver, the inductor, the diode for signs of being melty. It is hard to say which part would die first in an overcurrent situation.
 
Thanks for the advice. I have checked the 5v to the pi and that looked good. I also can see that power mosfet looks a bit melty so I have ordered a new one and plan to solder it in. Are there any dangers from heat soldering the board as is? Do i need to take the atmega chip out?

Thanks

Tim
 
I did end up taking the atmega chip out and replaced the power mosfet and the L1 inductor. The power mosfet had definately melted as i had a hard time getting it off the board. I added the new pi and everything is back to working now with the fan slowly increasing speed in voltage mode when manually adjusting the output. Thanks for the help with this one Bryan.
 
Great! Thanks for reporting back too. It is interesting to hear what the failure points are so I can take that into consideration when designing new hardware. Glad you're back up and running though.
 

 

Back
Top