Power ON diode fades off after pluging USB, polyfuse gets hot

I have got an Arduino UNO. After playing a little with LCD sketch (uploaded and powered from USB without any problems) I have connected 9V battery to move it away from my PC. First I disconnected the USB then plugged DC jack. After about 25 to 35 seconds LCD turned off.

Now my Arduino behaves like this:

  1. After connecting to my PC via USB:
  • the Power ON diode fades off
  • polyfuse (resettable fuse between USB socket and DC jack) gets hot very quickly (5 seconds is enough to feel it)
  • PC does not see anything. USB port on my PC is working fine. On my laptop it is the same.
  1. After connecting 9V battery via DC jack:
  • the Reverse Voltage Protection Diode is getting hot
  • power ON diode on Arduino board does not light up.

I have removed Atmega chip but Arduino behaves the same way. Diode looks fine (desoldered and tested with multimeter), Polyfuse doesn't look burned out (tested connectivity without desoldering)

My questions are:

  1. Why is the polyfuse getting hot? (possible causes)
  2. Is it possible to fix it by replacing the polyfuse?
  3. Can the polyfuse get hot because of another component is damaged?
  4. Can I ignore the Reverse Voltage Protection Diode if I don't intend to use DC jack anymore?
1 Like

Beszczelny:
My questions are:

  1. Why is the polyfuse getting hot? (possible causes)
  2. Is it possible to fix it by replacing the polyfuse?
  3. Can the polyfuse get hot because of another component is damaged?
  4. Can I ignore the Reverse Voltage Protection Diode if I don't intend to use DC jack anymore?
  1. It works "as advertised": There is too much current flowing. Getting hot is normal behaviour in this case
  2. No, the fuse is doing what it is designed to do
  3. Yes, see 1.
  4. The diode is getting hot for the same reason: There is too much current flowing. What do you mean by ignoring the diode? You could bridge it, but that does not solve the problem

Thank you olf2012 for answering all my questions. By ignoring the diode I meant leaving it as it is as (I thought it is broken) because it only gets hot when I plug the DC jack, and I don't intend to do it anymore. If you have any suspicions about what might be drawing this unwanted current, I will be more than happy to know your opinion.

It seems there is short circuit somewhere. Disconnect everything from the Arduino and test it.

I'd say you've blown... something... on the board. Something else is blown and is getting hot. Find it. I'd suspect the regulator or the 16u2 if it's an official board, or the opamp for the power switch circuit.

I have desoldered 5V voltage regulator, checked it and it is giving -1,25V from the output pin. Can I use the Arduino without it if I am using USB only?

Under what conditions are you measuring -1.25v? A linear voltage regulator (like that one) - broken or not - is incapable of generating a voltage higher than the highest external voltage (Vin) or lower than the lowest external voltage (Gnd), so the output couldn't possibly be -1.25v... How does this relate to your having already removed the regulator, too?

The regulator is not needed if it's powered through USB.

Did you confirm that the regulator was the problem before removing it? Whatever part on the board is causing the problem will be getting warm/hot (the polyfuse is doing it's job - something else is drawing a lot of current, and - with nowhere external to dump that power, it must be getting hot.

I have desoldered it from the board, connected 9V battery (according to datasheet) and measured output voltage with multimeter. I don't know how this is possible but that would have explained why reverse voltage protection diode was heating up right?

I still don't know why the polyfuse was heating up (any ideas?) because when I desoldered this voltage regulator everything start working normally, so yes I confirm that it was it.

Sounds like the regulator was blown. I'd wager when you measured the output, you had your probes backwards, and it's actually outputting 1.25v

There isn't a physical way for a blown linear reg to output a negative voltage.

Yeap, I've double checked it and you are right - human error - as always. Anyway, problem solved, thank you for professional help.
Cheers.