Unsolved
1 Rookie
•
18 Posts
0
22
January 5th, 2025 22:12
XPS 15 9560 intermittent non-bugcheck power off followed by CPU failure POST code.
My XPS 15 9560 was purchased in 2017 and has been wonderful. Currently it exhibits a CPU failure post code ("2,1" aka 2 amber blinks, 1 white blink).
I recently replaced the keyboard because some keys were not working. I purchased a keyboard from Dell, installed the new keyboard, reapplied CPU/GPU thermal paste (I used "Thermal Grizzly - Hydronaut - Conductive High Performance Thermal Paste"). The new keyboard has been working great. The system has been working fine for about 4 weeks since the repair. The use was been minimal, about 1 hour per week across 4 weeks, so about 4 times, 1 hour each, for the last month, no issues.
The other day, I powered it on, it was working fine. I sat on the couch, and right when I clicked down on the left trackpad button, the unit powered off out of nowhere, meaning no bugcheck, just power off, as if it powering off was related to the trackpad (if not a coincidence).
I would try to power it back on, and at first it would not power on, but after a little bit, it would power on, no problem. During this time, there was one GPU TDR bugcheck failure which I have never ever seen on this laptop... that was the first TDR as far as I recall.
I decided to open up the unit, perform a repasting with thermal paste again. I used the same paste, and I replaced the thermal pads on the chips (I think the 4 VRAM chips, if not mistaken) and the SSD. I used "OwlTree 4 Pack Thermal Pad ... Highly Efficient Thermal Conductivity 6.0 W/mK" pads.
After redoing all thermals, the system seemed fine again. I ran a extraordinary stress operations by performing intensive disk I/O along with constant CPU activity for many hours. I had hwinfo running, it was reporting normal thermals, nothing extraordinary. Things looked great. I left it running.
It was working fine for about 12 hours when I retired for the day. When I came back the next day, the display was off, the system was off. At that point, when trying to power it on, I noticed the 2 amber, 1 white CPU failure POST code. The unit has exhibited that POST code since.
I read other discussions on this forum, which led me to trying the following:
- Holding power for 30+ seconds (with power connected) to reset CMOS. No change.
- Opening unit, removing all peripherals, removing batter, no power connected, holding power button for 30+ seconds, putting back the peripherals, attempting normal power on. No change, same CPU failure POST code.
- I kept SSD out, and only one of the two RAM cards in. No change.
- I tried the other RAM card. No change.
- If I try without any RAM cards installed, I get a RAM failure POST code, so it seems RAM checks occur before CPU checks (assumption), so I'm not sure this last test means anything other than the RAM is probably fine because with RAM cards installed, there is no RAM failure, just CPU failure POST code.
So the CPU failure POST seems legit.
I don't suspect any thermal issues, though, so all of this is puzzling. Coincidental environmental factors, while not out of the question, are not certainties in any way.
This unit has been operating fine in this environment without issue, and it had been fine for weeks after the keyboard repair, so I'd suspect thermal issues would have prevented good operation for those weeks.
Regarding my pressing the left trackpad right at the exact moment when the initial "power off" issue was observed... while that is a potential red herring, its presence as a symptom correlated to the time of the initial problem makes it worth mentioning.
I don't really expect anyone to offer anything more I can do. I really believe the system CPU is gone... but maybe I forgot to check something, or perhaps there are other suggestions.
What is disappointing is that I do not believe this unit is faulty from a "broad" perspective... meaning, I believe it is "mostly" good and functioning, even if the CPU alone is truly gone. I'm therefore bummed out that an overall good asset might require a total mobo change due to CPU failure.
Any tips or thoughts welcome. Thanks.
HedgeFundManager
2 Intern
•
481 Posts
0
January 6th, 2025 01:52
Get in the UEFI and reset it to optimal
Then maybe its time to install Windows fresh
User773
1 Rookie
•
18 Posts
0
January 6th, 2025 20:33
@HedgeFundManager thanks... but an impossibility... no CPU, no UEFI.
HedgeFundManager
2 Intern
•
481 Posts
0
January 6th, 2025 20:54
Guess its time to plink down some cash for a new machine