Raspberry Pi Forums

Okay, so my brother was fiddling with my pi and a 9v power supply for an LED strip at the same time, and I assumed shorted something. I can't get you exact replication information. I did some googling that day and guessed it was the polyfuse. I gave the pi 4 days to recover. Still, no dice.

What I've tried:
8 power supplies ranging from 800 mA to 1.5mA with no luck. The most any of them put out (according to the reading across TP1 and TP2) was 4.25 (the 1.5mA one). The rest averaged between 3.5-3.8V.

A few more symptoms:
- The voltage across the polyfuse drops between .2 -.3 ish volts from one side of the fuse to the other. The resistance reads 0.6 ohms.
- After the apparent short, my charger (purchased with the pi, a Canakit set) made a very high pitched ringing sound, which, according to what I've looked up, means the switch inside is moving slower than it should, likely due to a bad capacitor. Never had any problems with the power supply before, even with several USB devices, non-powered.
- If that was all, I would assume it was just the power supply. But I've tried, literally, 8 different 5V wall warts. I really am starting to doubt that it is just the power supply.
- I've had a few occasions where the pi will begin to boot, running the full boot sequence. I've even seen the desktop image and my mouse pointer. That didn't last more than 30 seconds, if even that. It seemed like each time I tried after that, it got worse. Then, sometimes the green light would begin blinking, but no boot screen. Then only a solid red light, no more green light. Then a few times, the red light didn't even come on at all. I measured the voltage for this one and it was less than 2.5V (note: this was from a pretty old power supply I dug up. It's recognized by my phone as a charger, so I assume its still working.)
- The port for the microUSB charger gets quite hot very quickly, as does the capacitor next to it and a few other things in that area. The one time I tried using an HDMI cable, that port got quite hot just as quickly.

Okay, after typing all that up, I'm starting to feel like everything is still pointing to the polyfuse. If this is the case, what can I do but wait? How will I know when its ready? Will it ever be the same?

I'd appreciate some thoughts/advice/insight before I jump online and start buying new chargers and/or a new pi.

Replace the polyfuse? Looks like it is not recovering.

you could try replacing it or just bypassing it
or powering it via the GPIO
don't plug in more than 5V on it
don't do what you did before

mewpal wrote: - The port for the microUSB charger gets quite hot very quickly, as does the capacitor next to it and a few other things in that area. The one time I tried using an HDMI cable, that port got quite hot just as quickly.

Unlikely that the socket is at fault, so the heat is coming from something else. The overvoltage protection diode D17 is the most likely. Putting 9v onto the 5v pin of the GPIO connector would have caused it to conduct, to protect the Pi. Since you've been able to boot the Pi at least partially, it appears it did protect it.
However, D17 hasn't survived, and is now more-or-less short, drinking lots of current
Removing it may solve your problem.

From what you've said my diagnosis is that the polyfuse is not the cause of your current problems. The Broadcom CPU+GPU chip is damaged and on its way to being dead if not already. 9 volts and GPIO pins is not a healthy combination.

The voltage drop you are seeing across the polyfuse is not abnormal for the sort of current that will make things get hot. The low TP1-TP2 voltages you are seeing are because that abnormally high current into the Broadcom chip is overloading the external power supplies..

You can try as RaTTuS suggested but I fully expect the only cure is the purchase of another RPi.

mewpal wrote:The port for the microUSB charger gets quite hot very quickly, as does the capacitor next to it and a few other things in that area. The one time I tried using an HDMI cable, that port got quite hot just as quickly.

That sounds like something has become significantly damaged and it probably is time to get a new Pi.

Take care when using any damaged equipment as connecting it to other things can cause further damage and even damage what is connected.

My brother insists that the 9V power source he was working with was connected to the relay board and didn't touch any of the gpio pins. This is the relay module we were using: http://www.amazon.com/SunFounder-Channe ... B00E0NTPP4

So pi and power source were separated through the relay. I am really confused now.

A few more details: the whole board is getting really hot. within 15 seconds of plugging in the board, the CPU in the center of the board is too hot to touch.

I tried plugging in power and shorting F3 polyfuse. Red power light. Voltage 3.5 ish
Same thing when shorting diode 17 (which, on my pi, is labelled d16, not 17?). No changes. Red power light.

I cut open my power sources and tested them directly. They're all putting out a little more than 5 volts.

I don't know what else to try. Every time I plug in the pi, it draws less and less voltage. Where am I losing it all?

What Burngate said... D17 is protecting your RPi by shorting the supply to 0V which will cause the polyfuse F3 to "blow".

mewpal wrote:Same thing when shorting diode 17 (which, on my pi, is labelled d16, not 17?). No changes. Red power light.

Not the best idea, that will cause F3 to stay blown, possibly permanently.

Unless....

mewpal wrote: A few more details: the whole board is getting really hot. within 15 seconds of plugging in the board, the CPU in the center of the board is too hot to touch.

Sounds kind of terminal.

mewpal wrote: I've had a few occasions where the pi will begin to boot, running the full boot sequence. I've even seen the desktop image and my mouse pointer.

Not so terminal....

What to do?

If the problem is with the protection system then remove it, remove D17 (D16 should be one of 3, 3pin devices, (post a pic if you can) further along the board) and short F3. BUT! now you have no protection make darned sure you dont need it, use a reliable PSU. If your RPi works OK then replace D17 and F3, if it dosent work then its new RPi time.

Note that as well as voltage measurements, current would also help diagnosis.

aTao wrote:If the problem is with the protection system then remove it, remove D17 (D16 should be one of 3, 3pin devices

It could be a B 1.20. On those the fuse is labelled F1 instead of F3, and D17 is now called D16. The three 3-pin diodes are now D13, D14 and D15.

(post a pic if you can)

Good idea.

You can see my pics of the 1.2 here http://www.raspberrypi.org/forums/viewt ... 79#p527826

rpdom wrote: It could be a B 1.20. On those the fuse is labelled F1 instead of F3, and D17 is now called D16. The three 3-pin diodes are now D13, D14 and D15.

Ah ha, I am behind the times. Looks like R51 0R0 ground link has scarpered too.