FIXED: HannStar J-MV1 94V-0/EAL51 turns itself off after a few seconds

[jbdk]

I recently fixed an old laptop (white label/private label) based on the once common HannStar J-MV1 94V-0/EAL51 motherboard. The barebones Pentium M laptop was made in China in 2005, labeled AL51, and was then customized by the reseller with specific choices of RAM, drives etc. As seen in the first few photos below, this particular motherboard is labeled “HannStar J-MV1 94V-0” in one place, “EAL51 LA-2362 Rev. 1A 2005-04-22” in another. Various labels and stickers say “EAL51” or “AL51”. The BIOS is from Insyde. (I am mentioning all these numbers to help people googling for the same problem).



The symptoms

When turned on, the laptop seemed to run normally for a few seconds, then suddenly turned itself off with no warning. Problem occurred both when booting the OS and when staying in BIOS menus. Problem occurred both with battery only, battery plus charger and charger only as power source. Nothing felt warm even the CPU heat sink was cold to the touch. Googling turned up a few similar cases but no solutions. Laptop is 5 years old and out-of-warranty, user has a newer laptop, but WMA DRM has tied the user's music collection to this piece of old hardware.

The big clue

Look at this photo of the DRAM compartment on the back of the laptop:


The circled hole looks like a missing screw (all the other screw holes in view were originally filled, just not when making the photo), but it was something much, much worse as the next photo of the motherboard shows:



The standoffs that held the heat sink and heat pipe assembly on to the CPU had separated from the motherboard and was no longer holding it tight. So the heat didn't get from the CPU to the heat sink (which thus felt cold) and the CPU quickly overheated no matter how fast the FAN ran, so emergency power off was triggered.

To test that this simple mechanical failure was the cause, I held down the standoffs with my fingers and powered the laptop on, it ran OK for more than 5 minutes with the heat pipe assembly getting warm. Moments after I took my hand off, it powered itself down, thus confirming my diagnosis.

Mitigating factors

1. The power and temperature management works and protected the CPU and other components from being permanently destroyed by this failure, although the repeated overheating may have reduced its lifetime.
2. The mechanical (not solder) rivet-like attachment of the standoffs to the motherboard failed in a way that destroyed neither the motherboard nor the standoffs. This makes repair easier and was good engineering by the HannStar people. The joint held itself for 5 years despite tremendous pull from the springs, and then failed in a way that did not destroy the motherboard.

The fix
I simply replaced all 4 heat sink screws from the original M2.5x6mm wafer head to M2.5x10mm wafer head so they would penetrate behind the motherboard, and then attached 3.2mm wafers and M2.5 nuts on each to hold them down. This fits nicely within available space in the laptop (9mm would be too short, 11mm too long) and worked like a charm. Because the heat sink was removed (by itself) and reattached, I also replaced the thermal paste between CPU and heat sink.