Emachines M6805 overheating problem: foil discovery

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

Threaded View
When I first hit this problem I did some internet research and the
consensus was that the problem is caused by dust. I did get results
upon cleaning the dust the first time I encountered the problem, but
this time a quick blow into the ports without disassembly did nothing
for the problem. I just took the machine apart to do the cleaning
thoroughly and discovered that on the heat sink there is reflective
foil contacting the cpu directly. I brought it to a computer shop and
was advised to remove the foil and add a bit of paste before putting
the thing back together. The foil peels off and beneath it is a firm

For instructions please see:
and its links.

I have only just powered up the laptop so don't have enough results yet
to prove that this fixes the problem for months to come. I suppose the
foil is reflecting heat back to the cpu and that any dust in the
circulation system could put the system over the edge due to the
bottleneck of the foil. Hence the positive results of a dust cleaning
are misleading. If this theory is correct then emachines assembly did
not use this thermal tape properly. They simply should have peeled off
the foil in the first place.

I suspect I will never hear my fan whirl up and lose everything I was
working on again.


Re: Emachines M6805 overheating problem: foil discovery

The foil was part of a "thermal pad" which is used in lieu of thermal
compound in some systems.  This is unlikely to have been the cause of
the problem.  Also, in some systems, you cannot substitute thermal paste
for the pad, because the heatsink mounting system presumes the thickness
of the thermal pad, and without the pad, there is (or there may be) a
"gap" or loss of contact (or at least a loss of TIGHT, pressure contact)
between the heatsink and the CPU.

Tim BandTechnology wrote:

Quoted text here. Click to load it

Re: Emachines M6805 overheating problem: wifi card

Barry Watzman wrote:
Quoted text here. Click to load it

Thanks for the advice Barry.
My experiments agree.
The powerdown problem still occurs with the foil removed.
It seems counterintuitive that a piece of reflective material should be
there but I suppose if the primary heat flow is via conduction then
this is sensible. I really thought I had nailed the problem. Onward

I started removing hardware.
Upon removing the miniPCI wifi card the powerdown problem has not come
up again.
How can this cause overheating?
Possibly a hardware interrupt is errantly firing.
This machine has an Askey WLL3010 wireless card.
It looks like the board rev. is 1294V-0
It is normally disabled for me since it does not work well with linux.
Why it should affect my system at all is unknown to me.

I can't blame this card yet because this could be a BIOS problem or a
motherboard flaw. (powerdown happens with windows xp as well as linux
so probably not an OS issue)
There are a number of people experiencing this problem. Many think that
it is a cooling system problem since those are the only signals.
There seems to be a degradation occurring since the problem grows worse
over time.
At this point cleaning the cooling system does not overcome the
shutdown on my machine whereas it did before and so dust was considered
the object of the problem.
Now that I have gotten this far with it hopefully someone else will
verify this.


Re: Emachines M6805 overheating problem: not the wifi card

Tim BandTechnology wrote:
Quoted text here. Click to load it

While removing the wifi card did cure the problem during the hardware
removal process I just had the machine shutdown without the card in so
I still haven't found the problem.
The overheat is still happening. How these diverse actions can solve
the problem temporarily(from dust cleaning to hardware removal) makes
for an interesting challenge.  The machine must be teetering on the
edge of stability. Dropping some clock speeds could help. The PCI bus
might be involved since this last shutdown I was downloading from an
external wifi card.

My apologies for anyone reading my mistakes above here.
This machine has a 64 bit AMD cpu but is being used in 32 bit mode by
my OS's.
Any ideas?


Re: Emachines M6805 overheating problem: foil discovery

My M6805 emachines laptop has an overheating/shutdown problem.
Please read other thread posts for some background.
Also an older thread:

Removing the wifi card seemed to help but I believe that it was the
mechanical disturbance that caused the change rather than a direct
In desperation I studied the mechanics of the heatsink.
I find that the sink has nice contact with the cpu after removing all
I couldn't figure out a way to test the spring screw system.
I have removed all of the old heatsink paste/tape and applied just
To date this seems to be working; the fan only comes up occasionally
and never hits full bore.
I see that emachines is now offering to fix all notebooks with this
Not sure if they would honor mine since I've been under the hood.

To date my best knowledge of this problem is that it is not a dust
It looks like the real problem is the heat transfer from the cpu to the
heatsink itself.
Mechanical jarring may remedy or worsen the problem.

Here is an unverified theory:
The heat of the cpu softens the platic case allowing it to deform
reducing support for the pcb such that the planar match to the heatsink
is lost.

Until the problem strikes again...


Re: Emachines M6805 overheating problem: foil discovery

This machine was fine for a long time.
I took it on the road and experienced the problem again.
With a few slaps on a table screen side down I managed to make it
This problem is very likely a mechanical instability.
Now still my fan comes up pretty high but not to the point of shutdown.
Perhaps the problem is at the inlet ports where the sink contacts the
body of the machine. The spring loaded heat sink looks like a strong
design; probably stronger than what it is mounted to.

Site Timeline