[Elecraft] OT a bit: K3 Ant Tuner

[David Woolley]

I think we have the maximum power transfer theorem fallacy again here. 
The maximum power transfer theorem does not give the right result when 
you want to maximise efficiency, which is what you generally want to do 
in a transmitter, as it always results in worse than 50% efficiency.

One would expect the output impedance of transmitter to not only differ 
from 50 ohms but to radically differ from it.  If I remember correctly 
for a simple transconductance device, the resistive component will be 
many times 50 ohms.  If heavy negative feedback is used, I think it may 
actually be much lower than the "matched" value.  See 
<https://www.edn.com/Pdf/ViewPdf?contentItemId=4124437> for typical real 
reverse termination impedances.

Unless you put a (ferrite) isolator between transmitter and filter your 
filter design needs to assume far from 50 ohm reverse termination.

Putting a matching network in will simply destroy the PA efficiency, or 
even destroy the PA, through overheating.  If one really wanted to do 
it, you would have to adjust the SWR based on a received signal.

That matching network, will, itself, have a frequency response, as will 
the one leading to antenna.

Another reason for not operating anywhere near 50 ohm reverse terminated 
is that the dynamic resistance of output devices is far from linear, and 
a close match could produce a lot of distortion.

-- 
David Woolley
Owner K2 06123

On 05/07/2019 18:55, Jack Brindle wrote:
> Lets step back a bit and look at the system here. The K3, with ATU, drives into the BPF, when then drives into the external ATU and finally the antenna.
> The external ATU takes care of the antenna matching, and should present a 50 ohm load to the BPF. The BPF, because of its design, should present a 50 ohm load to the ATU as well, so everything is matched there.
> Before anyone jumps on this, remember the signals go both ways, outbound for transmit, inbound for receive. Also, we have a fundamental principal thrown at every EE student, that for best transmission of signal, the source and load impedances should match.