On 22 Jul 99, at 15:24, Willmore, David wrote:
> There are several things in a computer which will have their operational
> parameters vary with CPU activity. The most likely ones are: audio
> subsystem receiving noise coupled either directly (magnetically) into its
> signal lines or via its power feed; or the load on the power regulation unit
> itself.
I don't want to disagree with this. But computer systems almost
invariably have switch-mode power supply units which operate in the
50 KHz - 200 KHz range, and are _most_ unlikely to have residual
output in the audible range. I suppose you could get a "beat"
resonance between a PSU residual and the scan frequency of the
display, though.
In my job (looking after network equipment) I've seen a lot of kit
making "strange noises"; so far I haven't seen anything that wasn't
easily explained by either debris contacting a revolving fan, loose
wires or panels resonating from mechanical vibration, siezed fan
bearings (which don't clear by changing the load a bit), or slipping
drive belts (ditto).
> Yes, coils do can make a buzzing sound. The name of the phenomina
> is 'magnetorestrictance' (who wants to be I spelled that one wrong?) which
> is the property of a material to change its physical dimensions under
> varying magnetic fields--it's the reason large power transformers 'hum' and
> how most high power sonar units generate their signals.
I think you have it spelled right. Interesting about sonar - I just
assumed they used piezoelectric crystals & a suitable diaphragm -
that's the most efficient way I know to convert electrical energy
into acoustic energy. Conventional moving-coil loudspeakers are a
couple of orders of magnitude less efficient in dB/W terms.
Yes, AC coils _do_ hum - at mains frequency - much to the annoyance
of hi-fi enthusiasts, who refuse to sanction the use of switch mode
power supplies because residual (supersonic) high frequencies can
blur transient signals in the (wanted) audio output. Actually, the
material in the coil doesn't change its dimensions, but the coil
tries to ... the change in current creates a changing magnetic field,
which tries to propel the wire in which the current is flowing due to
the motor effect. The wire and its former are elastic to some extent,
so there is some mechanical movement, which can be transmitted as
vibration either through the surrounding fluid (air) or through the
coil mounting.
The way to deal with this (in extreme hi-fi) is to make the coil
very, very large, so that the force per unit length of wire is small,
and to embed the coil in something firm but compliant, like a huge
lump of heat conductive solid grease, to absorb the mechanical
vibration. Or, even more extreme but definitely better, run the kit
from car batteries instead of directly off the mains. You may think
I'm just being wierd, but I did used to know something about this,
twenty years ago.
Why does consumer electronic equipment tend to use switched mode
power supplies? They're smaller, lighter, more efficient and cheaper,
and, for most purposes, they do the job reasonably well. And they
don't hum, though sometimes they can whistle - usually this means
that overload has brought the operating frequency down into the
audible range. The major _practical_ problem with switched mode power
supplies, as opposed to the old-fashioned transformer & rectifier
type, is that they're a lot more susceptible to being killed by
transient overvolt surges on the supply.
Back more or less to topic - when I was a computing neophyte, a
quarter of a century ago, I was told a story by an engineer working
for a major mainframe supplier. For a laugh, the development team
wired up a loudspeaker (presumably through a buffer and amplifier of
some sort) to one of the bits in a CPU shift register. (This was
easily done in the days when everything was discrete components!)
The trick was then to code your program so that it performed to
specification, but, by insertion of extra instructions to tweak this
particular bit at suitable times, played interesting tunes whilst
doing so!
Regards
Brian Beesley
_________________________________________________________________
Unsubscribe & list info -- http://www.scruz.net/~luke/signup.htm
Mersenne Prime FAQ -- http://www.tasam.com/~lrwiman/FAQ-mers
