[Amps] Re: [Amps] RF choke selection

i4jmy@iol.it i4jmy@iol.it
Tue, 2 Jul 2002 18:15:38 +0200

I was reading all posts but I didn't see one that pointed out that a 
choke working inside an amp could be easily wrong in another.
A "parked" resonance can infact easily move (enough to disturb) when 
the coil itself is "feeling" what's in the around (i.e. metal panels)
and so on.
The best system is to check for resonance with the tube installed and 
loaded with a resistor that equals what will be in the real, tuning the 
PI network as one would check for match on each band.
To do this, a spectrum analyzer with a tracking generator and an SWR 
bridge is definitely the best (in effects one checks for the return 
loss and resonances will appear the same), also very educative.
As already pointed out a thin wire and a long coil with less diameter 
to minimize capacitance between turns would be the best, but real 
currents and coil size make this approach impossible in practice.
Chokes with parked resonances are typical producers of "strange 
effects" like a band with less power out and low efficiency, or cooked 
turns that breaks after months or years without an apparent reason.
The choke switch system is actually the best and more professional 
approach one can imagine, the only one that insures a condition where 
one has enough inductance at lower frequencies and choke resonances (oh 
yes, not only one may occur) well out of the amplifier range.
The switched chokes are a must with big PAs because since the power is 
increased it gets harder to have only one switch (or two coils). 
Infact, because the wire of the choke is also getting bigger it leads 
to inter windings capacitances (and choke size) that become quite 
significative and more chokes are needed.
Since inductance follows a quadratic law, in my opinion the switch you 
have would have been originally closed above 4 or 5 MHz and only below 
this frequency the choke has to be totally used.
Plate choke is definitely not a rocket science but rather an art, 
typical of very skilled RF designers and the "philosophal choke", good 
in any case, doesn't exist.

Mauri I4JMY

> Hello all
> I've just re-subscribed to this list, so here's a question
> for everyone :-
> I'm trying to modify a surplus RF deck (an auto-tune
> 2 - 30 MHz tube amp.) which has a 3-section plate choke.   
> One section is short-circuited by a vacuum switch, and 
> I guess that this was done above a certain frequency, 
> or maybe over a certain range(s) of frequencies to avoid
> choke self-resonances.  I've had to remove all of the
> auto-tune control electronics (large parts of it were missing
> anyway) but I need to figure out which bands 80 - 10m need the
> extra choke inductance, and which don't.
> Maybe it's as simple as e.g. below 10MHz the vacuum switch is
> open, above 10MHz the vacuum switch is closed.  I've tried
> measuring self-resonances using my ancient Heathkit GD1-U
> GDO (that should jog a memory or two!) but I'm not convinced
> by this method.   Any ideas how I can figure out the switching
> in/out of this plate RF choke section ?
> I have not had time to measure the inductance of each section,
> but I will do this, which will help.
> 73
> Chris     GM3WOJ
> www.gm7v.com
> www.qsl.net/gm3woj/amps1.htm