Dell Vostro 15 VEGAS/TURIS SKL/KBL 15341-1 short circuit

[Retrocomputer]

Hi everybody,
I have a Dell Vostro laptop with the VEGAS/TURIS SKL/KBL 15341-1 motherboard. Without notice, the other day the laptop was dead and as soon as I plug the charger in, the LED on the charger itself shuts off, as if detecting a short circuit in the motherboard. Measuring on the board between + and - of the plug shows no evident shorts (readings are in the hundreds of kOhms). If I plug a bigger charger with a higher current limit, it will stay on and some chips get hot. Using a thermal IR camera, a couple of mosfets get hot (but, testing them, they are not shorted, I think they go up in temp because of the short circuit current drawn flowing through them). A chip though gets very hot and you can hear a repetitive faint "click" anche the power cycles, as if pulsating. The chip I'm talking about is PU4701, a NCP81382, which is responsible for generation of the VCC_CORE rail. However measuring it with a multimeter shows no evident shorts. How can I proceed in finding the fault? I'm starting to fear the CPU is shorted itself. There seem to be big banks of caps on the VCC_CORE line, could they be the culprit? Or am I completely out of track? I'm a bit newbie at diagnosing such difficult circuits.

Here's the schematic PDF:
https://1drv.ms/b/s!Aiwky_BwJ7fkpAzl...h1yFh?e=WXxrir

Thank you so much in advance!
Davide

[mon2]

Re: Dell Vostro 15 short circuit

1) start with page 103 @ Adapter to follow the power supply flow.

2) Remove the battery (for now) and check the voltage of the external power adapter @ +DC_IN connector. Voltage value @ PU4301, pin 1,2,3 ? (page 43)

3) Voltage value @ PU4301, pin 5-6-7-8 (AD+) ?

4) Voltage value @ PU4301, pin 4 ? Measure with care so that you do not short pin 4 to 1-2-3 with the meter probe. This voltage is of interest only to confirm that the mosfet is being told to enable via the gate. If all is well, SOURCE voltage will pass forward to the DRAIN pins.

5) Then check the voltage reading of DCBATOUT (page 44) - voltage value ?

6) If AD+ rail does not match the external adapter voltage then power down and remove this external adapter. Meter in resistance mode (200 ohms) and red lead on AD+ with black lead on ground. Resistance reading ?

You can repeat for each and every (large grey) coil on the board.

[Retrocomputer]

Re: Dell Vostro 15 short circuit

Thank you very much for you contribution!
Power adaptor voltage, no load and disconnected is about 19,45v. When connected to the board (which is completely bare, nothing connected, removed from the laptop) +DC_IN pulsates and goes really low, averaging 0.7-1v. Spike cycle is too fast for the meter to follow, but it seems that for some fraction of a second it goes to adapter voltage. I may try to catch it with the oscilloscope. Anyway, voltage pulsates nearly the same on the source and drain sides of PU4301. Gate driving signal seems present but pulsates according to dc in voltage, as expected. By thermal imaging, PU4301 and PQ4410 are the only other components on the board apart from the aforementioned PU4701 that go really hot really fast (probably because lots of current is drawn by the short itself). DC resistance between AD+ and GND when power disconnected is about 120kOhms. Just messing around, PU5003 and PU5002 show a short circuit from drain to source. Strangely enough though, they stay fresh.

Thank you very much!
Davide

[Sephir0th]

Re: Dell Vostro 15 short circuit

PR4426 (Any Side) Résistance to GND?

[Retrocomputer]

Re: Dell Vostro 15 short circuit

PR4426 (Any Side) Résistance to GND?

Hmmmm. Interesting. PR4426 shows a reading of only 6.8 ohms to GND, both sides (since it's not open ).That doesn't sound right.

Thank you in advance!

[Sephir0th]

Re: Dell Vostro 15 short circuit

Your Main Power Rail is shorted but it isn't a clear Short to GND, which is likely Bad News for new.

Check all coils for similar Low Ohms Reading Like PR4426. And then Check Resistance between PR4426 and the suspicious Coil(s) (NOT TO GND!).

Share If you find a measurements below 1 Ohms.

[Retrocomputer]

Re: Dell Vostro 15 short circuit

Thank you very much!
So, upon checking, only 3 coils showed suspicious measurements: PL4801 measures 4.6 ohms to GND, PL4701 6.8 ohms to GND and PL5001 around 13 ohms. Resistance from PL4801 to PR4426 is 11 ohms, from PL4701 to PR4426 is 0 ohms (short), from PL5001 to PR4426 is 20 ohms.

Thank you!
Davide

[Sephir0th]

Re: Dell Vostro 15 VEGAS/TURIS SKL/KBL 15341-1 short circuit

I know, you've already excluded PU4701 from your list, but Like it thought... PU4701 is Most likely internal shorted. You could confirm by Lifting or removing PL4701 and Checking after whether the Low Ohms Reading at PR4426 disappears.

[Retrocomputer]

Re: Dell Vostro 15 VEGAS/TURIS SKL/KBL 15341-1 short circuit

Hi! Thank you. I didn’t exclude PU4701, actually I thought since the beginning it might be the culprit. It gets pretty hot immediately when plugged. What’s strange is that with a multimeter it shows no shorts whatsoever.
Anyway, I lifted PL4701, when plugged now the original charger doesn’t shut off anymore, the power led comes up but pulsates on and off and the board clicks quite audibly.

19v from the adapter are now there and stable.


Thank you
Davide

[Retrocomputer]

Re: Dell Vostro 15 VEGAS/TURIS SKL/KBL 15341-1 short circuit

Hmmmm! I made some progress... It seems that, as I initially thought and supported by sephir0th, PU4701 is indeed bad. Maybe it's not the only fault in this mobo, but nevertheless it probably shorts internally when supplied with voltage. After lifting the inductor 4701, some voltage rails got there, such as 19V from adapter (and thus DCBATOUT too), 5V_PWR and 3D3V_PWR were showing up too, generated by PU4503. The board, as I said, clicked and the LED blinked with the click.
Since the click seemed to come from around PU4701, I desoldered it from the board, resoldered PL4701 and tried powering it up. Well... no more clicking in any way or form and the led blinks a couple of times randomly, then everything shuts off (imho it's because the absence of PU4701 removes some sensing signals to other power circuits thus the board shuts off).
Ordered a replacement NCP81382 and let's hope for the best. It's not easy to solder too probably, with that package.

What do you think?

Thank you!
Davide

[mcplslg123]

Re: Dell Vostro 15 VEGAS/TURIS SKL/KBL 15341-1 short circuit

Its not that tough also to solder NCP81382. You need good quality flux though.

[Retrocomputer]

Re: Dell Vostro 15 VEGAS/TURIS SKL/KBL 15341-1 short circuit

So first of all thank to all of you!
The 81382 arrived today and I just soldered it. Unfortunately, though, replacing it showed no difference whatsoever (with respect to it unsoldered). As soon as I plug the charger in, the fan starts for a brief moment, the LED lights up for a second, then everything dies. If I press the power button again, the same repeats, the fan spins for 1 second, the LED lights up for around the same interval of time, then it dies. Nothing on screen. Plugging the battery in it shows the same behavior upon pressing the power button, leaving the battery in makes the LED light up steadily white, as if it's completely charged (strange).
How can I proceed in troubleshooting?

thank you!

[Retrocomputer]

Re: Dell Vostro 15 VEGAS/TURIS SKL/KBL 15341-1 short circuit

Little update. When the power button is pressed,for the brief 1/2 seconds that it shows signs of life, VCCGT and vcc_core are completely absent. Vccsa is present and measures about 1.03V. What puzzles me is that on PU4701 and PU4801, the chips responsible for generation of the absent voltage rails, both the enable and the PWM pins seem to receive a signal.
PU4701:
PWM goes from 0 to 5v when button is pressed
DISB# goes from 0 to 5V too.
PU4801:
PWM goes from 0 to 2v when button pressed
DISB# goes from 0 to 5V.

What can make the two rails not come up? Already tried desoldering PL4701 and PL4801 should there be something wrong down the rail but nothing changed. Those two rails always stay at 0v.

[Sephir0th]

Re: Dell Vostro 15 VEGAS/TURIS SKL/KBL 15341-1 short circuit

Unfortunately a shorted High Side of GPU/CPU usually means the dead of the BGA. This is what i have meant with "Bad News" before.

You could Check whether SVID_ALERT# or PROCHOT# is triggered at the VRM Controller to confirm the theory.

[Retrocomputer]

Re: Dell Vostro 15 VEGAS/TURIS SKL/KBL 15341-1 short circuit

Hi, thank you for the answer.
First of all, something happened: now when I plug the charger in, the LED comes up for a second but without the fan spinning, and even if I press the power button nothing happens, not even the LED lights up. I studied the power up sequence from the diagram it seems that all the rails and inputs are good, but nothing happens. All the power-up involved rails (all the S3, S0, VCC_CPU voltages) come up for a tenth of a second and then they die, except for the rails that are always present in the S5 state (5V_S5, 3D3V_S5, pwr_1D0V, 1D8V_S5...). When I press the power button, the ground signal gets to the EC, all the commands to change state (pm_slp_s4#, s3# and so on) are issued from the ec and returned from the CPU. As I said, all the rails go up to their nominal and correct value for a tenth of a second (even the s0 ones like 5v_s0 or 3d3v_s0) and then down to zero until I press the button again.
Since I replaced that IC no short is present anymore on PR4426 to GND. SVID_ALERT# measures 0V but, I think, not because it's issued and grounded but because almost no voltage gets to it. PROCHOT# spikes to 0.9V upon pressing. SVID_ALERT# reaches 0.2V.
I'd really love to challenge myself to make it working again.

Thank you!

[Retrocomputer]

Re: Dell Vostro 15 VEGAS/TURIS SKL/KBL 15341-1 short circuit

SVID ALERT pin is 256 ohms to ground when plugged in, but when the board is unpowered and it should be in open state it actually measures only 500 ohms to gnd. Could it be the NCP81208 at fault? Or maybe something regarding BIOS and EC? I’m starting to think about that too…thanks!