Csdrhrt

I'm trying to help a friend fix their Toshiba Satellite A665 because its been shutting itself off at random times until it got so bad that it doesn't make it past the BIOS screen before shutting off now. I took it apart, reseated all the power cables, cleaned it out, made sure no capacitors were blown, nothing obviously wrong, etc. and that didn't help.

I then noticed that if I applied pressure to the body just over the hard drive (opposite corner from the power button, next to the power indicator lights but not above the motherboard or anything else), I can reliably make the laptop turn on.

This makes me think short circuit of some sort, but then why would it turn itself off when not being touched? And a failure of the CPU failsafe wouldn't explain it turning on on contact. Anyone have any ideas of what could be the problem, or how to go about diagnosing this?

It's impossible to provide any kind of definitive answer for this type of problem. However, the nature of what you describe suggests several things to look for. BTW, the insides are very sensitive to static electricity, so touch something grounded before you touch anything inside.

• A bad connection that is affected by shifting the case. This could be a loose connector or a break or separation in a wire or flat cable. I assume you have already verified that all connectors you can see are tight. Look closely at all ribbon cables and wires for signs like a kink or indentation that might indicate an internal break. Lightly wiggle each connector to see if one has a bad electrical connection even though physically attached. Lightly wiggle each wire and cable close to where it enters a connector and near anything that looks like a kink or dent in the insulation to see is there is an internal break or separation from the connector. You're looking to affect the operation of the laptop not to feel something while you wiggle things. Be delicate. You don't want to stress anything in the process, and putting pressure on the case causes only miniscule movement inside, so it won't take much movement to simulate it.




• A short that is affected by shifting the case. This could be in the nature of a "signal" short--not something that is conducting significant power but something acting to affect the electronic state of something. Look for things like a daughterboard fastened to the system board that has bare circuitry in close proximity to something it could make electrical contact with. The case can be conductive--bare metal inside, or a coating of insulating paint that can be pierced by sharp points on the system board, or the inside of a plastic case may have metallic paint or film that acts like shielding. See if the system board makes physical contact with anything that could be conductive on the inside of the case. This could also be caused by an errant piece of something conductive, like a broken piece of conductive material or a snippet of wire (or a dropped screw). You might have to do some disassembly (with the battery and power disconnected), and carefully blow off the parts and crevices with compressed air.




• A cold joint. These are circuitry connections that are not electrically reliable. They can be very hard to find unless you are lucky because they often are not visible. I wouldn't recommend doing things to test for this unless you are a trained technician because you could end up doing more harm than good. However, two simple things: 1) Look for a crack or white stress line in the system board. 2) If the system board is held in by screws, loosen the screws slightly so that any pressure on the case doesn't translate to stress on the system board. If that makes a difference (either fixes it or prevents pressing on the case from working), it would indicate a problem for the shop.

All of these problems could be triggered by temperature (physical expansion), or vibration associated with use, which could explain why it shuts off during use and would be affected by pressure on the case.