At 10:03 AM 8/20/2006, ON4IQ wrote:
>I got a 2nd hand 4sq box (Comtek) a while ago and decided to put it on the
>The 4 antenne ports are terminated with 50Ohm suhner dummyloads, the load
>port goes to the network analyser and the xmtr port to the analyser aswell.
>I inject a 3.5 to 4MHz signal into the comtek box and see on the load port
>(where your dummy load needs to go) a level -7.5dB down on the xmtr level.
I assume this is an 80m box? Isn't it the "collins" hybrid (2 inductors, 2
capacitors) plus a transformer to split the center elements feeds?
One might want to try checking Tx->port 1,2,3,4 with loads on the other
ports, just to get a feel for it.
One might try disconnecting one load at a time, and seeing if the reflected
power into the load changes by what might be expected?
There's some work out there on measuring N-port devices with a two port
analyzer. While the idealized case is to terminate the N-2 ports in 50
ohms, there are some accuracy issues, particularly if your 50 ohm loads
really aren't 50 ohms. I forgot the authors' names (Tippet? Speciale?),
but there's a technique where you terminate them into a deliberate mismatch
(something like 2:1) and algebraically reduce it that gives better accuracy.
>This seems pretty high to me as I found in on4un's book notion of -21dB as
>a value for the dumped power when terminated in 50ohm.
>Is there reason to suspect the seller sold me a bad unit?
>TowerTalk mailing list
TowerTalk mailing list