[SOLVED] Can failed transistors cause overheating in a laptop?

Myronazz

Distinguished
Sep 5, 2016
325
12
18,795
Hello...

So about two months ago, my friend's laptop was experiencing overheating and so they took it to a nearby franchise to have them take a look at it. To put it simply, the technician explained how some transistors in it have failed and because of that, other transistors in the laptop need to take the job of the failed ones and because of that they are basically overloaded which causes the laptop to overheat and shutdown. He also suggested changing some fluid in the laptop (by fluid he probably meant the thermal compound in the processor, I mean what else could he mean?) which he would charge them 20 euros for and also pointed out that this isn't really a good solution as it would only last a few months.

Now, I am no expert in computing and circuits but still I have a feeling that this little overview of the problem the technician provided is kind of bull. I mean... if a transistor fails then a part of the circuit becomes open, which either stops the entire laptop from functioning or some other part of the circuit dies such as the USB circuit which effectively means no USB ports anymore. But apperantely some other transistor will step in and take the job... what? I've never heard of such a thing for as long as I've been around computers but I am no expert so I am asking you people, is that actually a thing?

They got a new laptop anyway because they needed an upgrade and I ended up getting the broken one because they were gonna recycle it anyway. It's a Sony Vaio with a Pentium B980 processor. I cranked it open and took a look at the board, really clean and no signs of it being burned by heat such as the PCB discolorating. I ended up changing the thermal paste since it was really old (from 2012) and now it idles at about 45-50 degrees (kinda high for an idle temp, but is it?) and also runs fine and really fast with no issues. I took it a step further and downloaded Prime95 to stress test it and the max temp it reaches is 74 degrees while averaging at about 70-72 degrees. It's been running the stress test for about an hour now and it has yet to surpass 74 degrees. According to a website, 85 degrees is the absolute max temp that the B980 should reach. So, with that in mind, the laptop is fine, is it not? I am intending to pass it onto my mother because her laptop is an ancient garbage Celeron HP that makes her life hell. In comparison, despite being old and having a Pentium, is really fast and heavenly, especially with an SSD I put in.

So... what's going on here? Is what the technician told us a bunch of bull and changing the thermal paste all that was really needed or was it actually all true and 50 degrees is actually quite high for idle temps so there is indeed an issue here? Which is it?
 
Solution
Smells like BS. Power Mosfets can fail either dead shorted or open. As there are multiple phases in the VRM, an open transistor would only stop that phase from working--this results in very choppy/noisy DC and it is very unlikely for a CPU to run stable on such dirty power. Dead short is much more exciting as it usually occurs at full load and can result in exploding parts.

That temperature is perfectly normal. BTW Sandy Bridge CPUs like that one can slow to as low as 200MHz if required to prevent overheating, so shutting down means it must have been really bad--dried TIM or heatsink fins packed with dust.
It would be POSSIBLE, but HIGHLY UNLIKELY that a failed transistor would cause this issue. If a transistor failed, it would likely cause a cascade effect, and cause many other components to fail simultaneously.

A temp of 74 on load is normal. I would suspect that the issue of heat was the TIM (Thermal grease/compound) that you replaced. Any TIM from 2012 that was placed by the OEM would definitely be ready to be changed.
 
Smells like BS. Power Mosfets can fail either dead shorted or open. As there are multiple phases in the VRM, an open transistor would only stop that phase from working--this results in very choppy/noisy DC and it is very unlikely for a CPU to run stable on such dirty power. Dead short is much more exciting as it usually occurs at full load and can result in exploding parts.

That temperature is perfectly normal. BTW Sandy Bridge CPUs like that one can slow to as low as 200MHz if required to prevent overheating, so shutting down means it must have been really bad--dried TIM or heatsink fins packed with dust.
 
Solution
I think your BS detector is working.
You did well by reseating the cooler.
And, perhaps clearing out the airways.

Now, if you will change out the HDD for a ssd, you will have a very useful laptop.
The ssd will absolutely transform the performance.
 

Myronazz

Distinguished
Sep 5, 2016
325
12
18,795
Alright peeps, thanks for the input! I'm glad my theory that it was all BS came true. The laptop was running the stress test for about two hours and it never exceeded 74 degrees and obviously didn't shut down either. When I stopped the stress test it went back to idle temps of 45 degrees so I'm taking this as the laptop is absolutely fine.

I did install an SSD on it and it's lightning fast for basic things that my Mother needs so it's a really nice replacement over the slow really old HP garbage that runs a Celeron. The difference between the two is honestly night and day despite the 'new' one having a Pentium.

Again, thank you all very much for the help. Its well appreciated!
 

TRENDING THREADS