Inspirion 1525

Do you have a question? Post it now! No Registration Necessary.  Now with pictures!

Threaded View
Hi all,

I'm working on an Inspirion 1525 for a friend.  Its got a bad
overheating problem which caused it to power off in the middle of
installing a service pack... Needless to say, that kind've screwed Vista
up for her.

I pulled out the heat sink, blew out the radiator, put new Arctic Silver
5 paste on the CPU (not much I could do with the ridiculous thermal PAD
over the GPU), reassembled, and the thing STILL runs very hot.

Here's the odd thing...  The second time I took it apart, I noticed that
the copper was VERY hot between the GPU and the CPU, but not even warm
to the touch between the CPU and the radiator, which was also cool to
the touch, even with cpu core reporting a temp of 83 degrees.  Obviously
only cool air was being blow out by the fan.

Is it possible there is a break in the copper between the CPU and the
radiator causing the heat to build up over the CPU?  If not, what else
could cause such a radical difference in temperatures from one section
of copper to the other?

Even if the CPU weren't making contact with the copper for some reason,
the heat radiating from the GPU should still be conducted towards the
radiator, correct?

Re: Inspirion 1525

Ryan P. wrote:
Quoted text here. Click to load it

Hi Ryan! Does that unit use a single heatsink module for both the GPU
and the CPU? It sounds like it does anyway from what you are saying. And
yes, I usually leave the thermal pads alone until they don't do the job
anymore. Plus I never bothered to find replacements either. But what
others and I have done in this case is to remove all of the thermal
pad(s) and use Artic Silver 5 instead. If there is too much of a gap,
you need a copper shim in between.

Now let's talk about the radiator. There is a drop or two of coolant
sealed inside of the copper tube. What sometimes happens is that it may
develop a leak. I don't think you will ever see any signs of a leak
though. And when this happens, they don't transfer the heat to the
radiator hardly at all. Thus the chip(s) overheats.

So if thermal paste doesn't help and the radiator isn't dirty and the
fan spins fast enough (and it is cool there). Then you need to replace
it. Often they are like $15 to $30 used on eBay for most models. Get a
tested one and not one of those "as is" or untested ones. As you might
end up with one like what you have already.

Gateway M465e ('06 era) - OE-QuoteFix v1.19.2
Centrino Core Duo T2400 1.83GHz - 2GB - Windows XP SP3

Re: Inspirion 1525

On 9/13/2011 1:16 PM, BillW50 wrote:

Quoted text here. Click to load it

  Yes, its a single module that "starts" at the GPU, goes over the CPU,
and then to the radiator fins.  There is a significant gap over the GPU,
which is why they have a pad there...  I got one for $2 on eBay, so when
it finally shows up, I'll replace it.

Quoted text here. Click to load it

  Now I learned my new thing for the day!  I had no idea that there was
anything inside the copper tube.  I'm guessing whatever is left in there
is between the GPU and the CPU, which is why its hot to the touch there,
but cool to the touch after the CPU?

Quoted text here. Click to load it

  Love eBay for laptop parts!  I see lots listed for $5 - $10, but I'm
assuming they were just pulled out of an old laptop and parted out.
Thanks for the info!

Re: Inspirion 1525

Somewhere on teh intarwebs Ryan P. wrote:
Quoted text here. Click to load it

The really annoying thing about the way some laptop manufacturers use
heatpipes is that heatpipes are inherantly 'A to B' systems. They work
*really* well when the heat source is at one end and the cooling fins are at
the other end. They're nothing short of amazing!

So you've Googled and Wiki'ed heatpipes by now? Damn impressive things. 95%
of the laptops made in the last decade wouldn't have been possible without
them. However, there's something that a few designers just don't seem to
get. That's the 'A to B' thing I mentioned earlier. When you start including
a 'C' on a single heatpipe then you're asking for trouble.

If the primary heat source is in the middle of a heatpipe then the fluid
will evaporate and migrate away in both directions looking for an area where
it can condense, and give up it's heat-load. If there's a good radiator at
one end and nothing at the other then it'll work fine. The vapour that went
to the radiator (fins) will condense and return to the heat-source via
wicking and complete the cycle. Any vapour that went the other way wouldn't
be an issue. It's own vapour-pressure would prevent any more from going that
way, letting the radiator do it's job. No problem.

The problems start to arise when there are two heat-sources on a single
heatpipe. There is only *one* way that this sort of configuration will work
and that's if the radiator is between the two heat sources. The fluid/vapour
will travel away from the heat (either end), find the radiator, condense and
go back to grab another load of heat. Ideally there should be separate pipes
for each source but the system just described will work quite well.

However, the problems arise when designers try to use a heatpipe to transfer
heat away from two sources that *aren't* at opposite ends of the pipe. That
seems to be the case here. (I'm not familiar with the system and haven't
Googled it.)

Heatpipes are freaking awesome, one of the greatest things to happen to
laptops (in particular). However they don't do well in the configuration I
think you've described. Two heatsources that aren't at opposite ends, with
the fins in between?

What happens in this case is that the heat migrates from the hottest source
and can actually heat another [C/P]PU up instead of the working fluid going
to the radiator. For instance, if the GPU is at one end, the CPU in the
middle and the fins at the other end you can end up with a 'heat barrier'.

See, what happens is this; IME with such a system the GPU is at one end, the
condenser (fins and fan) at the other and the CPU in the middle. Also,
usually IME it's the GPU that's the biggest heat-producer, usually being a
generation or two behind the CPU. (The joys of being socketed.)

So, if you're read up on heatpipes and you understand them then you'll
realise that, in such a system, for the GPU to get any cooling whatsoever it
needs to be considerably hotter than the CPU so that the vapour produced can
get past the CPU to the fins. Otherwise it's doing nothing, the heat is
trapped. Therefore such a system is always going to run hot and is
inherently flawed in it's design. Shame really considering the potential of
heatpipes and also of the hardware involved.

How do morons get to be designers? <shrug> All that's needed for the system
to be a very reliable and cool-running system is for there to be two
descrete heatpipes, one for each processor. As I mentioned previously, the
only way for two prpcessors to sare aheatpipe is for them to be at opposite
ends of it, something that's not impossible considering a heatpipe is quite
a small thing and it can take a loop here or there to make things work as
they should (without compromising the efficiency of the system).

Ho hum. I hope that I wasn't talking to myself. I can do that just fine when
the meds run out. ;-)

Happy computing,

"Humans will have advanced a long, long, way when religious belief has a
cozy little classification in the DSM."
David Melville (in r.a.s.f1)

Site Timeline