HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

[Dannyx]

Good day folks. I'm trying to fire up this laptop which I thought was pretty good and had potential, at least for learning stuff if nothing else. I've attached the schematic provided in another thread on our wonderful community.

The problem is as follows: power LED comes on and that's it. Classic failed GPU I thought, so what I did, since it's broken anyway, I blasted it with some hot air to see what happens. It's at this point that other laptops manage to turn on, but this one didn't. I decided to probe around the motherboard with the help of the schematic and most of the voltages are there, except CPU Vcore. The caps under the CPU where this is supposed to be present, don't give any reading. Page 10 shows us that V Core is generated from VBATR CPU, which reads 19v, by Q505 thru to Q508. These are controlled by U506 ISL95831 on page 11, which I suspect is bad because: +5va is present, CPU_PWEN is also present. I don't have a scope sadly, which would've enabled me to probe the gates of those transistors to see if U506 pulses them. Do ISLs like that usually fail, or should I expect the GPU or something else to be defective....still, why is there no CPU Vcore ? Does it come on after the GPU initializes or something tricky like that ? :|

[caspian]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

first check whether there is a short in CPU Vcore circuit. If there is no short, measure voltage on all pins of U506 and report them here to check.

[Dannyx]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

first check whether there is a short in CPU Vcore circuit. If there is no short, measure voltage on all pins of U506 and report them here to check.

I believe there's no short, but I shall check again and report.

[Dannyx]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

Ok, in case anyone is still out there willing to help, I measured one of the caps on CPU V core (say C236 on page 11) and I get around 13 ohms between it and GND with the CPU in place. If I remove the CPU, it goes up to 1 something kilo ohms, but the board doesn't turn on either without it and I believe 13 ohms is normal with the CPU in place.

[caspian]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

yes. now, voltage values on all pins of U506 are important to check.

[Dannyx]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

Ok but what exactly am I looking out for ? Some of them can't be measured by your standard DMM - a scope would be required which I sadly don't have, but I already checked what I consider to be "important" ones like the "enable" and VCC and stuff and they're all there...unless there's something else I should be looking out for. Adding to the difficulty is lack of a board view and the multitude of unmarked components, so it's impossible to follow the schematic past a certain point...

[Th3_uN1Qu3]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

If you have VCC and Enable, it means the controller should turn on the power to the CPU. Since you say you do not have a short on any of the mosfets, replace the controller chip.

[Dannyx]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

Guess it's on to Ali to get some chips....wait a month then report back Meanwhile, what other causes there may be for this behaviour ? Truth be told, the laptop WAS reflown at some point, but shouldn't CPU Vcore still come on ?

[caspian]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

I recommend replacing both the controller chip and the Mosfets.

[Dannyx]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

The FETs just for good measure ? I did not pick up low resistance anywhere on them, plus I believe something else would've popped on the board if they WERE shorted, though I could be wrong and some of the rails could still be "up" even with a shorted FET there....

[Dannyx]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

Hopefully someone's still around for this project. Here's what's happened: I replaced the ISL and tried firing it up. I DID get CPU V-core briefly and I thought I was all done. I put the board loosely on the case to try it with the display, I pressed the power button, the LED came on, then I heard like a hiss (the kind of noise inductors make), then the AC adapter shut down...not I have a short somewhere on CPU-Vcore AND GPU-Vcore...at least that's what the meter's showing me....diagnostics time. I hope the ISL itself didn't blow. I'm sure I soldered it correctly and I double-checked and triple-checked...Probably a MOSFET went...

[Dannyx]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

Q507 and Q508 shorted - removed them (VERY diffcult!) and now the short's gone....I'll try replacing them. Hopefully they don't blow again...can a bad CPU cause this ?

I really had to blast the hell out of those transistors with hot air until they finally came loose and since they're very close to the GPU, I have my doubts about whether this board is still good or not....I'll push on nevertheless...

[caspian]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

that is strange.

[caspian]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

is there any solution to prevent such things from happening?

[piernov]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

I'd always recommend replacing the buck controller and the MOSFETs all in one go for high voltage supplies. One of the component dying will kill another. On low voltage it's not a necessity, but better safe than sorry.
However, if the MOSFET measure perfectly fine and they short right after replacing the controller, I'd guess this is more a case of bad soldering/bad controller or too much heat on the MOSFETs will replacing the controller with hot air. The MOSFET can be bad even if they measure perfectly fine, but it doesn't happen often.
When replacing the MOSFET you have to get them from a reputable source, as there are a lot of remarked components on eBay/Aliexpress from China. Also, if the buck converter has multiple high-side (or low-side) MOSFET it is better to have them matched or the current distribution can get unbalanced. Usually having the same batch number is fine. Even if only have died, all of them should be replaced. Even better if it's a multi-phase buck converter is to have all MOSFET on all the phases matched, but you usually don't have to be that pedantic. If one of the phase is getting hotter (or colder) than the others then the issue is most likely there though. When you buy them from a reputable supplier, you usually get the same batch. When buying on eBay/Aliexpress you get random components.

[Dannyx]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

Bad solder job on the controller...that makes me a bit nervous TBH, because I'm not entirely sure there aren't any pads that touch, at least as far as I can make out through my magnifier. It's a QFN and has no leads you can actually see, so a little luck has to go with it too, since you can't really see under there.

Too much heat: when I replaced the controller I don't think I blasted it THAT much TBH - the transistors themselves were far worse. I had my station at max 450C and even then it took like 5 minutes to remove ONE of them since they have that pad on the bottom, so with two removed, the other 2 still on the board took quite a beating...so did the graphics chip right next to them. THAT's what I'm most worried about ATM....

What I'll try first is to power up without the FETs and see if I at least get back to where I was, since I don't want to bother soldering them in and trash them immediately...if something is messed up, I should see smoke coming out of the IC or something...again, not 100% sure of my soldering there :|

[piernov]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

When soldering a QFN you need to put good solder (not lead-free or Aliexpress specials…) on the pads, a decent amount. If there is a center pad, don't put too much solder as you'll have a hard time aligning the chip, but still some. Decent amount of flux as well.
Then when with the hot air you start melting the solder, align the chip roughly but it should get somehow on the pads, and press down on the chip (trying not to lose the alignment) to make it sit flat on the board. Excess solder (on the pads or the center pad) will come out from the side.
You'll have to run your iron on each side with some flux to get rid of the excess solder and to make the solder stick to the side of the chip (some of the area of the chip's pads is on the side).
Check with a magnifier and good lighting if every pad is soldered and look good (round, shiny, no bridge, stick up a bit to the side of the chip). and that the chip sits flat on the board at the 4 corners.

[Dannyx]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

I'd give it an 8 out of 10, just because it took me a while to align it and it kept slipping off and running away as the solder melted...it turned out decent in the end I DID get Vcore for a short while, so I reckon the controller is fine. I'm worried the CPU may be damaged because when one of the "high-side" FETs shorted, it sent full adapter voltage into it....just a hunch...

[piernov]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

Yes, if the high-side MOSFET shorted it can kill the CPU. If the low-side shorted then it's not an issue.

[Dannyx]

Re: HP Pavilion DM4 6050A2402401 Motherboard no CPU Vcore

Another nail in the coffin sadly...it was an i5 as well sadly Only luck can make this thing turn on now...that or a different CPU which I don't have...