[Elecraft] OT a bit: K3 Ant Tuner

[K9MA]

On 7/5/2019 12:55, Jack Brindle via Elecraft wrote:
> OK, so let’s look at the K3 side. The K3 antenna port connects directly to the BPF’s radio port. Again, we are presuming the BPF is designed for 50 ohms resistive source/load. The K3 is designed to transmit into a 50 ohm load, but it may not itself be a 50 ohm source. And, the receiver input may not be 50 ohms as well. Adding a tuned ATU does bring this to 50 ohms, providing a proper match into the BPF, so that optimum signal flows both ways.

Bandpass filters are generally designed for 50 Ohm source and load 
resistances, mainly because the actual source and load impedances are 
usually unknown. These filters will only perform as advertised with a 50 
Ohm source and load. With real world impedances, you just have to hope 
for the best. It's easy enough to use a tuner or other matching network 
to get the load impedance close to 50 Ohms in the passband, but even 
then, the impedance outside the passband may be far different, so filter 
performance will not match the ideal case. It might be better, or it 
might be worse. Unless you know the actual impedances, you can't predict it.

The source impedance is another matter. The output impedance of a 
transmitter is almost certainly NOT 50 Ohms. It's probably much lower. 
After all, an ideal voltage source has an output impedance of zero, and 
is 100 percent efficient. The ATU matches the LOAD to 50 Ohms, so the 
transmitter sees a 50 Ohm load, but does NOT necessarily match the 
output impedance to 50 Ohms. This is another reason real world filter 
performance almost certainly differs from the ideal. Again, unless you 
know the source and load impedances over the whole range of frequencies 
of interest, you can't predict the filter performance. While it's 
unlikely to happen by accident, it's possible to find a source impedance 
for any reflective filter which will result in zero (or very little) 
attenuation at any frequency. It's called a conjugate match.

In reality, though, we seem to get pretty good results with these 50 Ohm 
filters. You can actually measure the parameters that are really 
important, like harmonic attenuation, and see whether it is good 
enough.  Even then, though, there's uncertainty, as that measurement is 
usually made with a dummy load. What if the impedance of your antenna at 
a harmonic isn't 50 Ohms?

73,

Scott K9MA

-- 
Scott  K9MA

k9ma at sdellington.us