Just aquired an old laptop. Trouble shooting help...

Quote from jkee : My objections are primarily temperature related and I feel like I've already covered them, but I'll try to outline them again one last time. 1. Semiconductors can tolerate higher temperatures when they aren't energized (turned on), when energized a junction temperature in excess of 110-160 C will cause damage (sometimes it's a delayed chemical reaction). Temperatures much higher than what an internal junction can handle are required for reflow of the solder, peaking at 250-260 C. 2. By overheating a device, the internal junction temperatures are much higher than temps on the outside of the packaging where the solder is - practically guaranteeing damage. The solder pads are external to the chip, and by design the outside of the packaging can handle higher temperatures then the silicon inside. The junction is exposed to lower temperatures when heated from the outside in than inside out. 3. Devices are more likely to experience thermal shock when over heated than when heated with pretty much any of the other methods. Thermal shock can damage packaging, the wafer, the wire bonds internal to the chip.

Sounds good, and how could an IQ challenged grade school dropout like myself, refute any of any of that, but if that info is accurate and applicable, how can one explain me being successful every time I did it?
I'm just a lucky guy?

Holly crap, I didn't think my question was going to cause so much discussion. You guys do this often?

Quote from RockySosua : Sounds good, and how could an IQ challenged grade school dropout like myself, refute any of any of that

I honestly can't tell if you're being angry and truthful or passive aggressive and sarcastic here.
I'm sorry if you took anything I said as a personal attack, it wasn't.

Quote from RockySosua : but if that info is accurate and applicable, how can one explain me being successful every time I did it? I'm just a lucky guy?

If you care to brush up on the subject, this pdf [ti.com] has some relevant info. Some of the best references on the subject aren't thinks I can link to sadly (subscription walls or nda).
It depends on how you measure success. Is it a success if you can make it turn on and it functions when you're done but it fails within 6-12 months or less? It may appear to function most of the time but behave erratically under higher loads.

I would say you've been lucky, but it's hard to know if any internal damage resulted from your technique. X-ray is the only way to really evaluate the solder joints. The particulars of a given ic and pcb design and the exact parameters of the overheating you applied, would make a big difference in your results. It would be very hard if not impossible to get the solder hot enough to prevent a cold solder joint from forming while keeping the internals cool enough to prevent damage.

Following the concept of entropy and the second law of thermodynamics, it shouldn't be hard to understand that if you overheat an ic, the internals of the chip will get hotter than the solder pads on the outside of the chip where the solder is. This is not a good thing. Your assumption that the failure resulted from the chip overheating to a point where the solder flowed is flawed. It's more likely that fractured or brittle solder joints, warping, or corrosion are to blame (caused by over heating or thermal shock in most cases but at lower temperatures than would cause solder to reflow).

Personally I use an old convection oven that I modified a bit and it works great on boards where all the components can handle the heat.

Quote from jkee : I honestly can't tell if you're being angry and truthful or passive aggressive and sarcastic here. I'm sorry if you took anything I said as a personal attack, it wasn't. If you care to brush up on the subject, this pdf [ti.com] has some relevant info. Some of the best references on the subject aren't thinks I can link to sadly (subscription walls or nda). It depends on how you measure success. Is it a success if you can make it turn on and function when you're done but it fails within 6-12 months or less? It may appear to function most of the time but behave erratically under higher loads. I would say you've been lucky, but it's hard to know if any internal damage resulted from your technique. X-ray is the only way to really evaluate the solder joints. The particulars of a given ic and pcb design and the exact parameters of the overheating you applied, would make a big difference in your results. It would be very hard if not impossible to get the solder hot enough to prevent a cold solder joint from forming while keeping the internals cool enough to prevent damage. Following the concept of entropy and the second law of thermodynamics, it shouldn't be hard to understand that if you overheat an ic, the internals of the chip will get hotter than the solder pads on the outside of the chip where the solder is. This is not a good thing. Your assumption that the failure resulted from the chip overheating to a point where the solder flowed is flawed. It's more likely that fractured or brittle solder joints, warping, or corrosion are to blame (caused by over heating or thermal shock in most cases but at lower temperatures than would cause solder to reflow). Personally I use an old convection oven that I modified a bit and it works great on boards where all the components can handle the heat.

In the middle of College Football games so can't address the other stuff, but do want to say that I am definitely not angry.
It's rare that I would get that way and it would be evident.
The IQ challenged thing was my way of saying that all that info is above my knowledge and understanding, well, maybe not understanding, but certainly beyond my knowledge level.

Quote from dhc014 : It has a NVIDIA GeForce Go 6150 (UMA). Specs [hp.com] I was given a dead Dell XPS 1330 which was notorious for having the poorly soldered nVidia graphics, but the one I got actually only had integrated Intel graphics. I replaced the whole motherboard with another Intel IGP mobo and the laptop has worked great ever since.

dhc014,

Where did you purchase the motherboard with the Intel GPU? I have a XPS 1530.