[AMPS] Pi-Net math

[Rich Measures]

>
>Sorry - for some reason my Outlook Express does not add
>attribution marks to some messages.
>
€  it's not a bug, it's a feature.  

>>> In normal use, you don't have resistors loading the circuit
>>>at both ends - you put a load on one end and use the circuit
>>>to transform the impedance so that looking into the network
>>>you see the desired impedance.
>
>> However, the tube is a time-varying resistance.
>
>
>In which case one or more of resonance, impedance and Q will
>change in sympathy. In practice when tuning an amplifier we
>work with time averaged values.
>
>> If you remove one of the
>>resistors, I think the resonant frequency will decrease.
>
>Sure, but the tube is there.
>
>
>On the input side, (in g-g) the tube is the load on one end
>of the network - we want the other end to look like 50 ohms,
>but we don't (normally) terminate it in 50 ohms. 

€  The antenna tuner is normally adjusted to present 50-ohms to the 
linear amplifier.  

>
>The resonant frequency of the pi network in isolation will be
>different from when it's connected to the tube, even after
>allowing for tube capacitance.
>
>On the anode side, doesn't the tube approximate to a
>constant current source, 

€  Pentodes are a lot more like a current generator than tetrodes.

> which is a high resistance. Ok.
>it's a variable constant current source so it's a variable
>high resistance. In any case, it's my guess that the tube
>resistance is somewhat higher than the load resistance
>presented to it by the pi network so there's limited or
>minimal effect.

€  It seems to me that when a tube is turned fully on, it looks like a 
fairly low R.  For example, an 8877 has about 250v drop when it is 
drawing 3.1A of peak current.  .  Perhaps RL is an average value?.  To 
get a handle on what's what in a Pi-network, one needs to have both 
resistances in place.  
>
>We desperately need an interactive blackboard capability -
>realtime paintbox perhaps?
>
€  perhaps.  I like to occasionally construct models to put a reality 
check on theory.  

- cheers, Steve

-
-  Rich...
R. Measures, 805-386-3734, AG6K, www.vcnet.com/measures  


--
FAQ on WWW:               http://www.contesting.com/ampsfaq.html
Submissions:              amps@contesting.com
Administrative requests:  amps-REQUEST@contesting.com
Problems:                 owner-amps@contesting.com
Search:                   http://www.contesting.com/km9p/search.htm