Hi Jon,
Subject: Re: [AMPS] parasitic fantasia
Date sent: Mon, 7 Jun 99 09:15:29 -0500
From: Jon Ogden <jono@enteract.com>
To: <W8JI@contesting.com>, <amps@contesting.com>
>
> >
> >Typical amateur amps are actually rather modestly shielded at
> >VHF. They rely heavily on low impedance on the anode and
> >cathode of the tube to hold direct VHF radiation on TV channels to
> >a minimum.
>
> I thought that most of the TVI reduction came from the low pass action of
> the tank circuit. Or are we saying the same thing?
Not entirely.
Layout and component quality (meaning the self-impedance of the
components) is more critical than anything else in suppressing
high-order harmonics.
Even moving a lead to a different terminal of a variable capacitor
can change high-order harmonic suppression a lot.
Since cabinets are less than perfect (actually some are rather
poor) you need to bypass harmonics with low shunt impedances
near to the source of harmonic distortion. That rounds the
waveform, and reduces harmonic energy.
Then all the lead dress, which can be a mess in bandswitched
PA's, is less susceptible to mutual coupling that "leaks" harmonics
around the filter. There is also much less "through the case"
radiation of harmonics.
> >If the cap was "parallel resonant" with a high impedance, the amp
> >would likely be a nightmare in many ways.
>
> Tom, help me understand something here. One of the things that can cause
> an amp to oscillate is a low impedance path from the anode to ground.
No, just the opposite. In this particular case, the anode requires a
modestly high-Q and somewhat high impedance to become
unstable. That's why the grid is a problem, at some frequency it's
parallel resonance (from shunt C in the grid itself and series L in
the grid to chassis connections) becomes a problem. The grid
effectively has a HIGH impedance, with high Q.
If
> a high impedance is introduced at the resonant frequency it can prevent
> the oscillations from occuring. This is definitely what I believe I've
> done with my amp and I have no nichrome in the anode and no resonant dip
> at 80 MHz (typical for 4-1K) either. So help me reconcile your statement
> and my experience.
You understanding is incorrect, even if the results are what you
wanted.
Here's what can be effective. All you need to do is park the grid
resonance (that you have no or little control over) on a frequency
different than the anode parallel resonance. Then the PA is often
stable. That method is very common, and may be what you
obtained.
In the case where the resonances can not be placed far apart,
there is an optimum value of impedance that can be inserted in
series with the anode to stabilize the tube. If the series suppressor
impedance is too high or too low, the tube will become less stable.
With traditional suppressors, the anode to ground path must
present a very low impedance OUTSIDE of the suppressor. That
allows even a small series impedance to dominate the anode
impedance, and give you control of system Q and resonant
frequency.
If the anode to chassis impedance is high, the impedance values in
the suppressor must be made larger or the suppressor has little or
no effect.
That's why suppressors in tubes with long thin leads, like 811A's or
572's, have more turns and use (or require) higher value resistors
than suppressors in 3CX800's or 8877's.
As a matter of fact, if anode to chassis impedance is low enough
you won't even need a suppressor with the small ceramic tubes in
a good layout. That's why they work on VHF without suppression,
and why they are stable.
Tubes like 811's or 572's are a "bear" to stabilize.
I hope this helps.
73, Tom W8JI
w8ji@contesting.com
--
FAQ on WWW: http://www.contesting.com/ampfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|