Mark wrote:
> Mark wrote:
>
>>Hi
>>
>>I recently received my new Dell 640m with Core 2 Duo processor -
>>lovely! Except I got an annoying high pitch noise - barely detectable,
>>but annoying people with good hearing!
>>
>>So, looking on the forums, it seems people are reporting the same
>>problem on various machines, and it seems to be related to the CPU and
>>the frequency it runs at setting up some sort of resonance.
>>
>>The advice was to download RM CPU Clock utility (freeware, and very
>>good!) and disable the "Enable Popup Mode" tickbox in the "Advanced CPU
>>Settings" section.
>>
>>http://forum.tabletpcreview.com/showthread.php?t=3516
>>
>>I have tried this, and it really works! I can clearly detect the noise
>>stopping/starting as I toggle this setting.
>>
>>I cannot immediately detect and down sides to this change of setting -
>>the RM Clock readme states:
>>
>>"Enable Popup mode (auto C3/C4 to C2) - when this option is off, any
>>bus master traffic causes the chipset to return from C3 or C4 state
>>into C0 (working) state. When this option is on and the chipset
>>observes a bus master request, it will take the system from C3 or C4
>>state into C2 state and enable bus masters. This will let snoops and
>>memory access occur. This setting is available only for ICH6-M and
>>ICH7-M."
>>
>>I would really appreciate if anyone translate this into real-world
>>performance/heat/battery life information for me :)
>>
>>Hope this helps people,
>>Mark.
>
>
> Hi - further tests indicate that the CPU core temperature idles at
> 40degrees with Popup mode enabled, but goes up to 44degrees with it
> disabled.
>
> Hmm, this annoying: either take the whine, or a hotter CPU (which
> presumably means less battery life, and more fan noise!)
>
> So, if anyone else has some ideas, I would appreciate them!
>
> Mark.
>
Where is the noise coming from?
Is it coupled into the audio and coming out the speaker?
OR is it rattling some hardware?

I've seen power transformers and inductors where the wire is loose
or the bobbin is loose, causing the whole thing to rattle at multiple
frequencies. Sometimes you can get magnetic coupling to nearby shield
plates that rattle.
Ceramic capacitors can be serioulsy piezoelectric. With surface mount
caps being so tightly coupled to the board, it's gonna be difficult to
determine which, if any, are causing the problem.

They make ultrasonic leak detectors which are basically frequency
translating sound level meters. With the proper mechanical attachments,
you can get very localized detection of high frequency events.

I've been known to put epoxy on transformers to make the insides rigid.
But that will seriously void your warranty.

And, by all means, complain to the vendor. They won't worry about the
problem until the units start coming back. Sounds like a proper
software/firmware fix could be applied if they were sufficiently motivated.
mike

OH, there's another possibility. Sometimes in the bios there's a
setting to use a pseudorandom system clock. Was originally done to
smear out the peaks in the electromagnetic radiation so they could pass
FCC testing. Might also smear out your noise???? Never tried it.

And one way to look at the temperature is that for a wide range of
electronic devices, over the range of interest, adding 10 degrees C to
the temperature doubles the failure rate.

Are we having fun yet?