Could short-circuit burning and smoke also damage CPU, or only power IC?

Hi,

I would like to ask a question with short-circuit.

We have built a new board which has a power-management IC supplying various voltages of for board components. It has been tested to work fine. Today however, after some soldering work we forget to take the solder wire off the table before plugging power into the board, and I saw and smell a smoke from the board. It was obviously a short-circuit created somewhere due to the extremely well conducitivty of bare solder wire.

Then I found the device has stopped functioning. Of course the CPU doesn't work, and emulation attempts failed with messages saying that there is a CPU power loss. I suspect that the problem was due to the burning of the power management IC so I tested voltage at several output pins of it. Most of them measured zero as opposed to nominal values, and some give incorrect higher or lower non-zero values.

My question:

1. Which are possible to have been damaged? When there is a short-circuit, circuit theory suggests that the current takes the path of the lowest resistance which is the path created between pins by the solder wire touching both, and it shall avoid the path with higher resistance. I therefore think that it is probably most components should be safe? For example, CPU and other ICs, since the theory doesn't suggest that they should have been burned by excessive current due to their non-zero resistance?
2. Could the fix be as simple as just replacing the power management IC? Or there could also be other components that get damaged?

Matt

Provided the power IC didn't put out any significantly higher than usual voltages as it failed, then the damage is likely limited to the power path up to the shorted junctions, as you think.

Exactly what in that path is damaged depends on what is in the path - the power IC, switching MOSFET, diodes, inductors... are the power traces on the PCB intact?

Rick,

The power IC has a internal power path manager which has an out pin, and this output in turn supplies all other voltage regulators.

Normally, given 5V DC input, the output in of of the power path manager gives 3.6V output; but after the burning I measured 4.9V. It looks to me likely that the power path manager has failed.

It is also true that the PMIC module has also MOSFETs and inductors. However, MOSFETs and inductors are all part of sub-modules which are supplied from the power path manager's output, then if the power path manager failed at first there is no way to test these down-stream parts.

Your answer helped a lot. I will make a closer examination.



Matt

Exactly as FoxyRick says: it generally depends on what was shorted to what and by which way. I might happen that PMIC died and because of some spike CPU internal core voltage regulator also died, but nothing else did. It may as well happend that cpu's unresponsive because short current went eg. through input protection diodes of JTAG ports (byebye jtag, permanent codeprotect -> enabled ).

I remember when one time a friend of mine had this pimped out devboard for some MCU (i believe it was worth like $1k) and he forgot to put a cover on the fuse holder in mains supply . IIRC only things that remained functional (as opposed to blowing up or catching on fire) were one BJT and and LM358 .

poorchava and FoxyRick,

It was very fortunate that only the PMIC was damaged. I had it replaced at the solderer and everything went back normal.

Thanks for help!

Bob