TC measurements via the PC sound card

["Tesla list" <tesla-at-pupman-dot-com>]

Original poster: "Paul Nicholson" <paul-at-abelian.demon.co.uk> 

Relatively few coilers, it seems, possess test equipment such
as signal generators and 'scopes, leaving them with no easy
way to measure the resonant behaviour of their coils.  However,
surely many more coilers own PCs with sound cards, and this
makes me wonder if there's any way we can make use of the
PC sound card for TC measurements.

One might picture some sort of simple electronic widget which
connects between the sound card and the TC - you hook it up
to your magnifier and run some software which reports back to
you the important three frequencies.

The obvious difficulty is that the sound card typically only
works up to about 50k samples/sec, whereas the important coil
resonances may go up to a few hundred kHz.  Hence the need for
the widget.

Now one might contemplate various techniques involving
heterodyning or frequency multiplication.  The particular
thought I had in mind was to have the widget ping the
coil with a few volts of free running square wave at say
a few tens of pings/sec.   Then the coil current, which
contains signals at all the resonant frequencies, is
picked up and mixed with the sound card output in a
balanced mixer.   The mixer output is returned to the
soundcard input.

In operation, the PC would generate a continuous square
or triangle wave and sweep it across the audio spectrum.
The harmonics of this signal mix with the coil signals
to produce a mess of beat frequencies in the signal
returning to the sound card input.

The PC software then goes to work on analysing the mess.
Each frequency component seen in the sound card input
signal would be intepreted as |nG - F|, where G is
the audio frequency being generated by the PC, F is
a coil resonant frequency, and n is a positive integer
indicating which harmonic of the sound card output is
involved.   By fiddling with G, the PC program can
determine n and decide whether it's nG - F or F - nG.

In this way the PC software can in principle determine
an F for every frequency component visible in the mixer
output.   As the program sweeps G over the range 0-25kHz
it will encounter the same coil frequencies beating
with various harmonics and it could collect these into
a table.   Interfering signals such as pickup of radio
broadcasts can be ignored because they wont have the
pattern of sidebands corresponding to the pinger rate.

Once the heterodynes of the coil signals are identified,
the program could go on to look at their ringdown
envelopes, in the manner of tcma,

  http://www.abelian.demon.co.uk/tssp/tcma/

to determine the Q factor of each resonator mode.

So I wondered if anyone could see any basic difficulty
here?  I think a balanced mixer would be necessary,
so we don't have to worry about harmonics of each F.
Does the widget need to be more than an NE555 and
an MC1496?   I was thinking of have a dabble with this
but wondered if anyone could save me wasting my time.
--
Paul Nicholson
--