Have a dead short on the DCBATOUT side of this board. PU4308 became very hot when I applied current to dcbatout pad of PR4002. Removed 4308 (p. 43), short gone, but pins 1-3 on board still short. Removed PL4303 and pins of PU4308 no longer short. So also removed PL4302 and 4303. VCC Core side of all 3 coils are short. Put current to one of the pads, shows short for sure but can't find smoking gun. Is it one of the caps (ST330U2V, 6 in all) right after the coils? Or one of the caps on page 8 (left hand side of page)? Am I on the right track?
-
-
Re: Dell N5050 dead short
Turned out to be one of the large black caps, ST330U2V, PT4308 I think. Short completely gone. Replaced with one off a scrap board, soldered everything back on and board shorted again as soon as plugged in. Removed all 6 caps again but this time in different order and the one I removed last eliminated the short. But it wasn't 4308 this time. Is there something I should know about replacing these caps that I'm missing? Coincidence both times that last cap removed eliminated the short? Because I checked for short after removing 1 cap at a time... -
Re: Dell N5050 dead short
It is usually over voltage that causes capacitors to go short. Usually kills the CPU also. Do the capacitors still measure shorted after removed from the PCB?
PU4308 must have been shorted or on also.Comment
-
Re: Dell N5050 dead short
Yes, PU4308 was dead too. None of the caps measure short after removing from the board. Can I replace PU4308 and try to plug in the board without those 6 caps just to verify they're the problem? Probably not a good idea if they're not. I've got a DC power supply too, maybe I can slowly raise the voltage? Not sure how to proceed, just know I want to eliminate those caps as the problem.Originally posted by rievax_60 View PostComment
-
Re: Dell N5050 dead short
Plugged board in with no caps, no short! Even got a blinking power light and have 3v and 5v at coils. Will go no further until I can figure out this cap issue. Used hot air to remove and install, is that a no-no for this type?Comment
-
Re: Dell N5050 dead short
It is all a puzzle. Capacitors of any type usually don't become good again after going short.
Judging by what I have seen others do, these capacitors tolerate a lot of heating. People including myself used to get misled into thinking a short was present because of the normal 4 to 6 ohms of the CPU. An I5 CPU I measured lately seemed to be over 100 ohms.
How do you define a short?
I can't find a full data PDF for the controller IC. Most of them have crowbar protection where the low side MOSFETs are turned on to keep Vcore clamped when certain faults occur that try to make Vcore too high.But this will not show as a short while all inductors
are removed.
There can also be a pin on the IC connected directly to the Vcore that is able to clamp to ground. So you might measure a few ohms to ground which can remain long after power is removed. Before you know it, the short is gone.
The shorted high side MOSFET is likely to have been the only fault.Last edited by rievax_60; 11-21-2015, 02:32 PM.Comment
-
Re: Dell N5050 dead short
I installed a CPU, plugged the board in and it shorted again, blowing PU4308. Removed PU4308 and PL4303. VCC Core side pad of PL4303 measures about 1 ohm to ground. With my power supply set to 1 volt and 2 amps on that pad, I measure 1 volt and about 1.65 amps. I know a dead short would show no voltage and 2 amps and no problem would show 2 volts and 0 amps. So what is my power supply showing me?Last edited by p1muserfan; 12-13-2015, 08:53 PM.Comment
-
Re: Dell N5050 dead short
1 volt and 1.65 amps indicates about 0.6 ohms and 1.65 watts of heating somewhere.Comment
-
Re: Dell N5050 dead short
Found it, it was PU4307, replaced, so now I'm back to having 3,5v power to the board with nothing installed. With no power to the board and no CPU installed, I'm measuring pretty high resistance to ground on the 3 CPU power coils. With a CPU installed I measure about 16 ohms, which I understand to be about right. So is my chipset ok? Hate to keep installing a CPU and having the board short every time I plug it in, is there something else I can check?Comment
-
Re: Dell N5050 dead short
What I call Vcore, is called VCC_CORE in the schematic. I had trouble lately with a Lenovo.
The Vcore was going to 6v because of a fake ADP3212 controller IC I put in. It made the I5 CPU go short circuit. The Lenovo would turn on the Vcore rail without the CPU fitted. I caught the rail going to 6v for a short time before shutting down with a DSO.
As a precaution I rats nested a crowbar protector that I soldered across the Vcore rail that was set to trip at 1.5v.
When I fitted a good ADP3212, new MOSFETs and CPU, it all just worked, the crowbar did not need to trip.
I cant see any other safe way to test it. Ill say more if I think of something.Comment
-
Re: Dell N5050 dead short
Still in need of some help here. Have standby voltage and white power light on an empty board (no CPU or memory installed) only because PT4301,2,3,4,8 and 9 are removed. If they're installed, board shorts immediately. Why? Because I measure 19v across PL4301,2 and 3. Isn't that voltage only suppose to go as far as the upper fets (PU4302,5, and 8)? This can't be goodComment
Comment