[lug] Speakers

Wayde Allen wallen at boulder.nist.gov
Tue Aug 1 09:09:13 MDT 2000

On Mon, 31 Jul 2000, D. Stimits wrote:

> > > I've found a number of speakers have low quality connectors that die.
> > > Many are corroded or tarnished right out of the factory. As far as hum
> > > goes, it could be from a cheap power supply, or it could be picked up
> > > from inside the machine by unshielded cables. If it doesn't work, make
> > > sure the connectors didn't die.
> > 
> > I'll do a double check tomorrow, but if everything seems fine (remember
> > that I verified they worked with another computer), what fundamental
> > reason could there be for them not to work on my machine?
> I suppose that it could be nothing more than one machine having a better
> tolerance on the spring clip placement than the other. Sounds doubtful
> though.

I'd rather doubt that the connectors would be the cause.  Connector
problems should manifest themselves more as intermittent scratchiness, or
popping sounds, and the speakers probably wouldn't respond much to
inter-modulation products due to corrosion in the connectors.

My guess, is that you are getting a 60Hz hum due to inductive coupling
from the computer.  Computer speakers, unlike HiFi speakers are
(supposedly) magnetically shielded to prevent such interference, and to
prevent the speakers themselves from magnetizing your floppy disks, etc..
These days, transformers are manufactured so that the core is right on the
verge of saturation.  The net result is that if your power line is driving
it a bit higher than the design specification, the core saturates and the
excess field radiates.  If you put a high impedance speaker coil in the
vicinity of such a transformer you would likely couple to this field.  You
might simply try moving the speaker away from the computer, or try turning
off peripherals (printers, modems, etc.). 

The other possibility would that you have a ground loop.  That could be
caused by poor connectors, but more likely by an improper connection.

- Wayde
  (wallen at boulder.nist.gov)

More information about the LUG mailing list