William Unruh wrote:
On 2020-01-04, James wrote:

Any suggestion appreciated, I really want to get this to work.
Thanks in advance.

That is a hardware fault, not a software problem. If the system it
dumping sufficient current through you speakers to fry them, that has
nothing to do with software.

If you want to get it to work, your first step is:

1) Connect Line-Out of suspect machine, to Line-in
on Technician machine.
2) Set sample rate to max. (On my two computers here,
both support 192KHz, but the input analog filter that
prevents aliasing, is set to 50KHz or so in the design.)

Then, use Audacity to record this "hot speaker" situation
and discover the "energy" being sent.

Some computer audio setups, have "impedance sensing"
as well as "jack sensing". The original Analog Devices
scheme, sent a burst of 25KHz at a decent amplitude,
to allow measuring current flow and thereby , allowing
an impedance (not a resistance) measurement to be done.

RealTek did not use this idea initially, but apparently
may have acquired the right to use it recently.

Impedance sensing this way is necessary, because
computer audio is fully capacitively coupled. There are
caps all over the place. No DC is intended to be flowing
on *any* unamplified interface.

A few computers have a 2W power amp and drive unamplified
speakers. For the rest, the vast majority, the LineOut has
no power drive capabilities, and can only drive 32 ohms
(recommended) at the best of times. (It's 600 ohms rated
with the headphone amp tick box off, 32 ohms with the
headphone amp tick box on, generally no more than two
connectors with headphone amp enabled.)

If the speakers or the external amp get hot, check for *ultrasonic*
output. If a 1V signal at 25KHz is coming out of the
computer (say, because of the chip "impedance sensing"
function has gone nuts), the technician machine sampling
at 192KHz and 50KHz low pass, can capture that for you.

Power amplifiers can also oscillate on their own,
which would be a discussion for sci.electronics
or the like. Circuits can be "closed circuit stable"
or "open circuit stable", and these general classifications
help guide people in terms of "what to expect" from their
gadgets.

As an example, the ATX power supply has a control loop,
and if more than around 5000uF of bypass caps total
is used on a rail, it could affect the stability of
the control loop. It can't drive just any old load.
And that's useful information if using ATX supplies
as replacements for bench supplies (in an electronics
lab).

In any case, a little creative usage of LineIn on your
"good, working" computer, will tell you whatsup.

And as for audio amps that oscillator, just about
every audio power amp I've ever constructed has
oscillated. One (as a teenager), I took it to high
school and put it on the scope, and it was oscillating
at 500KHz. The speakers of course, cannot follow
this signal, so your eardrums don't get punished.
If the signal is ultrasonic but closer to audible,
there could be an acoustic effect from the super dome
tweeters. If you build or even assemble your own
power amp from a kit, it pays to scope
it before getting too excited about your "accomplishment" :-/
A hint of trouble, is a "hollow sound" in the
midrange frequencies. If you need extreme corrections
with your graphic equalizer to balance the sound,
that's a sign there's trouble brewing in there.

Paul