Ian White, G3SEK wrote:


* The output impedance of the transistor doesn't come into the story at
all - not when characterizing RF power devices that are not operating in
class A. Even the device manufacturer doesn't know or care what it is.
Neither need we.



Tubes and transistor power amplifiers quite oftem
use negative feedback to improve SSB linearity.
Improvements of 5 to 10 dB are common. The
negative feedback reduces the internal impedance
of the tube and transistor amplifiers. The
tube/transistor data sheets do not consider this
factor.

Again, we usually don't really know or care much
about the values of the internal impedances.

But there is a special case. Voice/music/data tube
transmitters operating at low frequencies have a
problem called "sideband clipping" where the plate
tank selectivity may be too sharp and reduces the
modulation bandwidth. The internal impedance tends
to broaden the response at resonance. When
designing the tank circuit this effect may have to
be included.

Bill W0IYH

William E. Sabin wrote:
Ian White, G3SEK wrote:
* The output impedance of the transistor doesn't come into the
story at all - not when characterizing RF power devices that are not
operating in class A. Even the device manufacturer doesn't know or
care what it is. Neither need we.


Tubes and transistor power amplifiers quite oftem use negative feedback
to improve SSB linearity. Improvements of 5 to 10 dB are common. The
negative feedback reduces the internal impedance of the tube and
transistor amplifiers. The tube/transistor data sheets do not consider
this factor.

Again, we usually don't really know or care much about the values of
the internal impedances.

Agreed.

But there is a special case. Voice/music/data tube transmitters
operating at low frequencies have a problem called "sideband clipping"
where the plate tank selectivity may be too sharp and reduces the
modulation bandwidth. The internal impedance tends to broaden the
response at resonance. When designing the tank circuit this effect may
have to be included.

Thanks for that information. A related topic would be the effect of
tank circuit Q on the bandwidth of HF amplifiers; I seem to remember
reading something about, but don't recall what it implied about the
magnitude of the tube internal impedance, as compared with the load
impedance.


--
73 from Ian G3SEK 'In Practice' columnist for RadCom (RSGB)
Editor, 'The VHF/UHF DX Book'
http://www.ifwtech.co.uk/g3sek

Ian, G3SEK wrote:
"A related topic would be the effect of tank circuit Q on bandwidth of
HF amplifiers;"

Class A amplifiers are little used as HF finals, so in practical
amplifiers current is only part-time.

Impedance of a parallel resonant circuit is high. Circuit impedance
rises with inductance. Q rises with capacitance.

A Class C plate tank introduces a load on tube or transistor. It should
waste only a small percentage of the power generated. It should have
enough Q to linearize the output of the amplifier.

Terman says it is easy to show that the Class C tank circuit efficiency
is: 1 - Qloaded/Qunloaded.

Loaded Q is the ratio of the circulating volt-amperes to the transmitted
watts. If Q is too high, bandwidth is too narrow. If Q is too low,
harmonics are high.

As Q is ordinarily high, the tank circuit impedance is higher than the
load on the amplifier.

Impedance on the Class C amplifier has little effect on the tube or
transistor loading. Output impedance presented by the transmitter to the
load is determined in many cases by the percentage of the time the
amplifier is switched-off.

Best regards, Richard Harrison, KB5WZI

I left out the word "tank" in the sentence: Tank impedance on the Class
C amplifier has little effect on tube or transistor loading." Sorry.
Sometimes I delete too much when I shuffle things on the screen, I
should write it first.

Best regards, Richard Harrison, KB5WZI