[NLRS] Bi-directional or two uni-directional couplers

[Dr. Gerald N. Johnson]

For sure a set of mismatches and a few REALLY good terminations can let 
you get good calibration and feel for how they are working. If the 
coupler is looking at reverse and has poor directivity it won't show the 
depth of minimum on a matched load as a better coupler. It will show a 
deep dip with an adjustable load but not quite at 50 ohms. And hamfest 
loads suffer like hamfest attenuators and power meter heads, the chances 
of them being for sale because they've been overheated is high in my 
experience. And then they are junk. My best loads are GR, but a 20 or 30 
dB attenuator rated for the frequency makes a good load too. Doesn't 
matter what's beyond it. But a connector cap or short will give a more 
consistent match than an open connector. By the difference between 40 
and 45 dB return loss or something like that in a range you can't be 
sure of reaching from imperfect directivity anyway.

Another factor is that if you use 1 kHz modulation and a 415 family 
indicator its calibrated on being in the square law region of the 
detector, well below where its a good diode and so the dB scale on the 
415 may be wrong at high levels.

73, Jerry, K0CQ

On 9/27/2010 4:01 PM, Tod - ID wrote:
>
>
> The three replies so far to my query support my intuitive feeling about this
> matter.
>
> Donn expressed possible concern about what might happen if the "line length"
> from two serial directional couplers became  significantly different from
> the line length of a single bidirectional coupler.
>
> Clearly the proper thing to do is to measure the value one gets with 'known'
> impedances. It may be that one can use the same thinking as is present in
> the QEX article to 'correct' for errors in the outputs of the two couplers.
>
> Thanks to the people who took the time to comment. I will go off and do the
> measurements and see what I find.
>
> Tod, K0TO
>
>