Actually I hate to reply between paragraphs but I can't do anything
else this time.
Since one of the most typical feature of amp reflector is that people
mix up things to win over, bear in mind that the topic is actually
chance of VLF oscillations.
We discuss what chances has an SB220 to oscillate at VLF having or not
a variable reactance circuit between grid and ground.
All the rest, like VHF parasitics, is out of topic.
73,
Mauri I4JMY
> ---------- Initial message -----------
>
> From : owner-amps@contesting.com
> To : "Ian White, G3SEK" <g3sek@ifwtech.com>,
<amps@contesting.com>, "Maurizio Panicara" <i4jmy@iol.it>
> Cc :
> Date : Wed, 27 Sep 2000 19:35:28 -0400
> Subject : Re: [AMPS] SB-220 Smoke Test
>
>
> > The resonance of SB220/TL922 grid-ground circuit is given by the
parallel
> > of few mH and few hundred pF, actually that circuit is never
really "open"
> > at any frequency because of the low Q caused by the consistent Ohmic
> > resistance of the long inductor (litz) wire.
>
> The chokes were various types through the life of the amplifier,
In all the several SB220/221 and TL922 I've seen and repaired (may be
more than 50 pieces) the inductor is a 3 section one and uses a wire
that's or appear or is a litz wire. In any case it's not an High Q
inductor, not less than 1 mH and with consistent wire resistance.
> but
> none I'm aware of were ever Litz wire. Even if they were, the grid
> impedance would still have been very high.
The impedance at resonance is relatively high, but the several ohms of
the wire inductor actually dump the resonance that's never an open
circuit like it was stated in the original message that claimed "open".
> > understand why a Gg mounted triode with some series resistance
between
> > grid and ground should have a bigger tendency to oscillate at VLF
(or any
> > other frequency) than when the grid is straightly grounded and gain
> > increased.
>
> 1.) The Pi network is a lowpass that looks almost transparent at
> low frequencies.
PI coil reactance is neglectable, plate and load capacitors add and the
equivalent circuit produces a low pass filter, about -3dB/octave.
All the above is compensated by the hi-pass, almost -3dB/octave,
response of blocking capacitor.
In SB220/TL922 (and all HF amplifiers) the plate choke is neglectable
reactance at VLF but PS capacitor filter, that is there (in case of
full wave rectifier) to remove 100 Hz (120 Hz in N.America) ripple,
perfecly shunts (bypasses) plate to ground at VLF.
Plate choke is too small at VLF and tube output is dynamically closed
on a very low impedance already at 200 Hz and up to the resonance of PS
capacitor filter.
> 2.) The blocking capacitor can be made to present a wide range of
> impedances at the anode with an inductive load (which is
> independent of the SWR on the operating frequency) at the output.
At VLF where the grid circuit composed by the coil and capacitor
resonates (we have seen how) the tube output is dynamically closed on a
low impedance while blocking capacitor is very high reactance virtually
isolating the the output of the tube from rest of the world, therefore
what's out is neglectable.
> > low reactance at VLF (bypass capacitors have neglectable
reactance) so,
> > unless you have a PS with high internal resistance, the tube plate
is
> > dynamically closed on a very low load expecially at DC and VLF.
>
> Not when parallelled by a low reactance Xc caused by series
> tuning the blocking cap with external inductance (high side of
> resonance).
A fractional plate impedance (less than one Ohm) changes because of a
neglectable reactance capacitor in series is (series) resonated by an
external inductance ?
Which external inductance ?
What external inductance can be so big to resonate 2000 pF at some
100/200 Khz ?
....and, when resonance of blocking cap. and huge external inductance
occurs, what happens is that another very low reactance adds to the
existing to better dynamically bypass the tube output at VLF.
> Putting a choke in a grid, especially without a low resistance
> shunting the choke, is poor engineering.
A coil with consistent wire resistance is perfectly symilar to one with
an high Q with a resistor in parallel.
No poor enginering, here, symply series parallel transformations, try
and measure with a Q meter.
73,
Mauri I4JMY
> It commonly creates problems. Look at the history of mods to the
> 30L1 Collins, to see how long it takes people to figure that out.
> 73, Tom W8JI
> w8ji@contesting.com
>
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|