Sorta. The distinction has more to do with whether you're following your
output stage with something that needs to see a specific impedance (like a
filter in a receiver), and whether you're building a power stage. I
couldn't think of a good generalization, so I listed some examples below.
Then I realized that it really boils down to the fact that the impedance
matching rule is a tool for predicing circuit behavior, not a guideline for
life.

RF stage in a receiver: you want to match the output impedance to the
following filter or mixer for maximum power transfer and good filter/mixer
performance. You usually _don't_ want to match input impedance because the
best power transfer impedance is not the same as the best noise figure
impedance.

IF preamplifier (between a passive mixer and a crystal or mechanical
filter): Match both input and output impedances, both to maximize power
transfer and to get good performance from the attached stages.

1st buffer in a VFO: _don't_ match impedances with the oscillator! You want
to purposely refrain from taking much power from the oscillator tank,
because a heavily loaded oscillator is a poor oscillator. You probably _do_
want to match impedances on the output so the following stage won't need so
much gain.

1-transistor transmitter: It's an oscillator, but a power stage too. Load
it (match impedances) enough to get useful power on the antenna, but not so
much that you screw up performance.

Power output (usually): match impedances to the input, don't match impedance
to the output (for all the reasons given earlier). This applies to any
class of stage: A, B, AB, C, D and E (yes, there are class D and E stages).

"Paul Burridge" wrote in message
...
On Mon, 5 Jan 2004 15:35:24 -0800, "Tim Wescott"
wrote:

Well, you're objecting rather strongly to an assertion that I never made.
In theory if you want to get the maximum available power out of a
generator
you _do_ match it's output impedance. In practice if you do this with
most
RF final amplifiers you will reduce your final element (whether
transistor
or tube) to slag -- that's why modern commercial radios have SWR
protection
circuitry.

Please see my reply to your post starting with "Hi Leon" for a full
explanation, and please actually read it before replying. Remember also
that everything I say (including the attempt to match output impedances
with
the wall socket -- I was only 8 but that's no excuse) has been backed up
by
experament.

Okay, so if I understand you correctly you're saying that whilst you
should try to match impedances in Class A; it's inapplicable in Class
C? It's just an operation class distinction?
--

"I expect history will be kind to me, since I intend to write it."
-
Winston Churchill