Re: OPT equalising circuit

John Byrns wrote:

snip,


> > >> Try measuring the capacitance of a loudspeaker.
> > >
> > >What frequency should I measure at, or do you want a complete sweep?
> >
> > Doesn't really matter. The purpose was just to inform Phil there's
> > capacitance there.
> 
> I would think it matters a lot, I thought whether there is capacitance
> or inductance there depends on the frequency?  There are a lot of
> resonance effects involved in the operation of loudspeakers.
> 
> Regards,
> 
> John Byrns

How much the equivalent capacitance in a dynamic comes&dome
speaker with crossover filters affect an amp depends on the speaker
and the model for it in its enclosure.
If you went looking for a pure 0.22uF in most speakers, you would never
find it.
This is just the right C needed to make a tube amp oscillate at HF if
the amp has loop FB and has not been set up for stable operation.

Most dynamic speakers present an amp with inductance and resistance in
series.
As F rises, the reactance of the inductance rises above the series
resistance value
so the impedance of the L&R becomes increasingly inductive.
This means that at say above 50kHz, the speaker has such high impedance
that its as if there is no speaker connected at all, and some tube amps
will oscillate at HF if no load is connected.
This is especially so in pentode amps because the gain of the output
tubes rises 
with load value increase; gain = gm x RL approximately, and so 
effectively there is more NFB applied when gain increases, and 
the margin of stability reduces to allow oscillation because open loop
gain >1 where
phase ****ft has reached 180 degrees.

The speaker inductance does not cause phase ****ft at HF.
That's because its reactance goes higher as F increases, but a
capacitor's 
reactance becomes lower, and there IS a 90 degree phase lag at HF and
perhaps resonant 
effects with leakage inductance which at HF is the dominant reactance of
the 
amp's output impedance.  Amps have inductiove output Z.
NFB is taken to the input *after* this inductance and when just the
right C value id used as a load,
you have a second order LC filter which has a lot of phase ****ft before
much attenuation has occurred.

If the speaker has impedance equalisation Zobel C&R networks across
drivers, then
the voice coil inductance is neutralized and as F rises the speaker
appears to the amp
as inceasingly resistive, which loads the amp and stops the gain rise
and prevents
the instability due to the gain rise.

But in a radio, there is rarely any such fancy thing as impedance eq,
Zobels or any NFB.

They just have a 6V6 connected to the speaker via OPT with no NFB 
and its all mainly crap what you hear.

OK to get the news each other hour or the weather re****ts.

Or the football scores. Fidelity isn't required for really im****tant
stuff.

And if Russian missles are on their way, the bloody radio won't save
your arse.

Patrick Turner.



 
> 
> --
> Surf my web pages at,  http://fmamradios.com/