Acer Alienware Nitro 5 (GH51G LA-L181P Rev: 1C) & Dell Alienware M15 (GDP50 LA-k452P Rev:1) - Both Shorted PCH (SRKMA)!

[reessi]

Hello:

I have both these laptops that appeared to have the same fault so I thought I would have a go at replacing the PCH as something to learn, they both are using the same PCH.


I started on the ACER:

Up on connecting power no signs of life but current is consumed from the bench power supply and the PCH is hot.
All the caps on the top of the PCH are showing 0 ohms and the 3.3v rail is shorted.

I followed the videos from Sorin, Electronics Repair School on Youtube heating from the bottom to install. I practised on around 10 scrap boards before committing to find the right air/temp settings on the hot air and baked the board and replacement PCH at 120c in the oven to remove moisture.
I took photos before and after to be sure no components fell off, the install appeared to go picture perfect and I was able to nudge it after a time and it sprang back like it was soldered.

The Used chip from Aliexpress came and I measured the caps on top prior to installation, lowest was 180 ohms on a lot of them and the rest higher so hoping it is good.
Measuring the caps on the removed chip indeed showed dead shorts (I cleaned solder off it first to remove all bridges).

After installation the board consumes no power when connected and charges the battery normally when connected.

When turning on it consumes ~200mah for around 10 seconds then the power light turns off!

I am aware that the ME region needs to be cleaned, I haven't done this yet and need to follow some guides..

But my issue is I believe the board should turn on and stay on but no picture prior to ME region cleaning as this is what happened to Sorin and his consumed ~1.5a (His was also a Nitro 5 but not exact model).

When off the only rail on is the 5v Always, when turned on I only get the 3.3v Always but no 1.8v Always and the IC that turns on the 5v Always to 5v On/3.3v Always to 3.3v ON does not turn on!
I tried jumpering it but made no difference.


I checked the 1.8v Always Power supply and it turns on to 0.26v when the laptop is briefly on & was receiving 3.3v enable signal. I checked it wasn't shorted & injected 1.8v here while turning on the laptop but no difference!
But after this it started working on it's own! How is this possible! Now when the laptop is on it shows 0.26v for a split second then turns to 1.8V!

I checked the SUSP# signal from the EC chip that should turn on all the power supplies and this is missing! Should I try injecting it? Would that also be 3.3v like the other enable signals?

So my questions are where to go from here? The EC chip appears to be working as the power light comes on, battery charging is working but it is not sending out the SUSP# signal..

Should the laptop turn on without the ME region being cleared in all cases like Sorins?
If the replacement PCH is faulty would the laptop turn on still or is the EC waiting for communication?

Maybe I didn't solder it properly, I could try reflowing it again but it is added risk so want to exhaust other things first..

I wish I had checked if the board turned on with the PCH removed, does anyone know if it would have?

When checking the caps on the PCH some of them show 3.3v when I turn the laptop on.


Now on the graphics chip it is showing 0 ohms on the coils, is it dead? What are the diode readings of these chips, the mosfets are not shorted, I lifted the coils to eliminate it from the circuit for now and it made no difference to the behaviour since the power supply's are not being turned on anyway.
I have disabled this chip on a previous Alienware that had a shorted one and the laptop ran fine on the integrated one.

The cpu shows ~30 ohms on it's coils which I believe is normal but again the power supply's are not being turned on..

I have removed the bios chip and backed up the bios, while of I turned on the laptop to see the behaviour and is was the same.
I also tried a bios for this model off this forum. I am unsure if it has a cleaned ME region, presumably any bios I download needs to have a cleaned ME region to work..

The bios battery was drained to 0v as it is on the shorted 3.3v line, I replaced it as I had a Dell XPS that refused to turn on without it in the past but same behaviour.


Dell Alienware:

It seams to have identical faults:

Caps on PCH shorted, 3.3v shorted... 0 ohms on the GPU but no shorted mosfets..

I was hoping to get somewhere with the Acer before ordering another PCH..

This one behaves differently sometimes when plugged in, sometimes it shows 500mah and blazing hot PCH and sometimes shows 500mah for a second and drops to 86mah and the PCH is hot on the corner of the silver glass square when checked with my lips..

If I remove the PCH can I expect the board to turn on all be it with no POST, I also lifted one side of the GPU coils on this to make sure the current was only on the PCH and is it.

I know modern GPUs take a lot of current and have low diode readings but what is expected?

I'm not against buying another PCH to test and learn, I could also buy a stencil and remove the one from the Acer and reball for the DELL.




I also have an oscilloscope for testing but not sure what to test on the ACER, maybe for data lines to the PCH?
I have the boardviews & schematics for these downloaded from this forum..

The reviews from the Aliexpress seller are saying the chip worked, assuming they are all real.

I think I have mentioned everything, let me know if any other info needed...

Many thanks...

[reessi]

Actually on the Acer machine the 2 GPU coils next to each other (GPU Power supply) measure 0 ohms and the one on its own measures 50ohms, this is the VRAM power supply.

On the Dell all GPU coils measure 0 ohms.

[reformatt]

A dirty ME wouldn't prevent boot up, just make it quite slow to come on and boot.

My first thought is you got a faulty one (given it's from Aliexpress). These are generally salvaged off junk boards that don't work (they have no way of knowing they work or a means to test them either). I think the MO for these sellers is that you will think its too much hassle to send back and they keep your money, or they will blame the issue of it not working on you. Aliexpress these days make shipping back anything quite easy though and they generally side with the buyer. So I guess this is just your time that you are wasting if you do get a refund. Pretty sure Sorin had a video about how he doesn't do PCH replacements if it's sourced from China for this very reason.

You may wish to check other power rails used by the PCH, like +3VL_RTC, 1.05V, 1.8V etc. In the schematic, PCH supplies PM_SLP_S4# after EC issues power on after power button press so start there.

[reessi]

Hello,

Yeah it may well be faulty, it was always a risk..

+3VL_RTC is working fine when bios battery is plugged and comes out on some resistors next to the PCH.

I hadn't realised that since the 1.8V ALW started working some other voltages have appeared.

But the 1.05v on the schematic is labelled as +VCCIN_AUX_PCH, this is coming on at 1.8v! Can that be a fault with the power supply? There is no short showing on this coil. When I lifted the coil and measured voltage it shows 0.4v when laptop powered on on PCH side.. Can that be coming out of the PCH if faulty..

I tried injecting 1.05v on the coil side of the PCH with coil removed and powered up, on the bench power supply it jumped up to 1.8v!

After that I removed that power supply and now the PCH side of the coil pads is showing 0.5v with the laptop powered off and 0.66v when on!

The 1.8v is working on the PCH.. When I measure the caps on top of the PCH with it on, I am getting 1.8v / 3.3v and 1.5v on one cap.

PM_SLP_S4# comes on for a split second @ 3.3v when I press the power button which is I guess why the rest of the laptop doesn't power up. If I am reading the schematic correctly it looks like that then turns on SYSON signal at Q13.

And according to the power sequence it should be PM_SLP_S4# -> SYSON -> SUSP# hence why all the power supplies are not turning on.


So in hindsight should I have checked that the PCH voltages are correct before soldering on the PCH, so now I'm not sure if the 1.05v is wrong because of the PCH.. The fact that there is now 0.5v showing on the PCH side with the laptop off and jumping to 0.66v when on does that mean is some sort of short in the PCH, seams like that voltage is coming out if the PCH, I have checked on the boardview and with 1.05v PU1101 power supply removed then the only place left connected is the PCH!

Do you have any idea why to start with the 1.8v ALW was not working until I injected 1.8v? When this wasn't working also the 1.05v was off!

I am now getting also 1.8v on one of the CPU coils -> PLG101 but nothing on the other coils when the laptop is turned on.. The schematic says this is +VCCIN_AUX but doesn't state what voltage this is, I was under the impression the CPU would be 1.05v..
Do you know if 1.8v is correct on that coil... May be this one is also affected by the PCH.. It does not show any extra voltage when off like the PCH power supply..

Any ideas, I'm starting to think the PCH is faulty.. If there are no other things to check, I might give it one more reflow and if no change remove it, if the 1.05v power supply comes back to normal that would indicate a PCH fault right?

Many thanks.

[piernov]

No idea what you are doing, 1.8V is fine for core power rail of HM570 PCH since it has FIVR as indicated in the datasheet. Lower voltages can be requested by the PCH when entering lower power states.
Anyway a random PCH is likely to have Boot Guard enabled for a different platform.
But first without proper BGA rework equipment, chances of soldering a large BGA properly are low.

[reessi]

So 1.8v is ok on +VCCIN_AUX_PCH even though it says 1.05v in the schematic? Is it not strange there is 0.66v appearing on the line when the power supply is removed?

Would Boot Guard stop it from turning on though?

Cheers

[dutchgaucho]

Have you ever solved this mistery? I'm having a quite similar one. Or actually 2 of them, one is also an LA-L181P. They're both Nitro's and before I never had any trouble replacing PCH's. Now for both machines I'm experiencing exactly the same problems you have: board starts for a few seconds, got a power light and then switches off again. No amps, no fan spin. Cleared both ME's and did to both a reflow to no avail. These Nitro boards are just junk.. if it ain't the vcore fets that blow up the cpu, it's the pch. What a waste..

[mcplslg123]

I've successfully replaced SRKMA on LA-L181The PCH was cracked as somebody tried to put up a normal screen for test.Obviously i did it with BGA machine.

The most common fault is SUSP# missing. Check resistance to GND on SUSP# and if its not shorted, try injecting 3.3V on SUSP# pin of SIO and see if it boots.

[dutchgaucho]

Thanks for your help! Will try this tomorrow and get back to you with the results!

[dutchgaucho]

I've successfully replaced SRKMA on LA-L181The PCH was cracked as somebody tried to put up a normal screen for test.Obviously i did it with BGA machine.

The most common fault is SUSP# missing. Check resistance to GND on SUSP# and if its not shorted, try injecting 3.3V on SUSP# pin of SIO and see if it boots.

Just checked. No short on SUSP#. Injected 3v but same behaviour: turns on for few seconds and then off. Cpu not getting hot. One thing I noticed is that +5VALW is present all the time and +3VALW is being created as soon as I push the power button. Could this still indicate a bad pch?

[mcplslg123]

+5VALW will be present in standby and +3.3VALW will come up after power button press-so perfectlyy normal behaviour.
Do you get +1.0V/1.8V rails up after power button press?
Share the resistance to GND on all coils and voltages on all coils after power button with 01.SUSP# injected 3.3V/02.Without SUSP# voltage.
Check whether you have any short on the switched rail side like 1.8VS/3.3VS etc.