FW: [TowerTalk] More on measuring cable loss

[Michael Tope]

Dick,

If you have a good 50 ohm termination, put it at the
end of the coax run and measure the input return loss
of the system as a function of frequency. I think you 
will find that the "anomalies" in your cable thru loss 
measurements will more or less correlate with poor terminated
system return loss figures. If on the other hand, the 
"structural" return loss (i.e. terminated return loss) of the 
system is good (e.g. > 30dB) then the loss measurements made 
via the open circuit/short circuit return loss method should 
be pretty accurate.

73 de Mike, W4EF................... 

----------
From: 	Dick Green[SMTP:dick.green@valley.net]
Reply To: 	Dick Green
Sent: 	Tuesday, January 19, 1999 2:37 PM
To: 	Tower
Subject: 	[TowerTalk] More on measuring cable loss


For what it's worth, I installed the two-radio switching system (anyone
interested in just how many months it takes to design and build an automatic
two-radio switching system, and what it takes to install it outdoors in
January in NH should e-mail me.) I replaced the old (10 years+) Ameritron
with TopTen A/B and 1-of-6 switches. I haven't been able to measure the
losses all the way to the A/B antenna ports because I have go to the antenna
farm to open one of them up don't have time to do that today. But I did
measure losses to the input of the new 1-of-6 switches. Total run length is
several feet shorter that to the old Ameritron, and there are no relays in
the way. Here are the losses on the two cable runs (identical connectors,
suppressors, etc.), as measured with the AEA CIA HF Analyst:


Frequency    Measured        Measured       Measured       Computed
             Before (run 1)  After (run 1)  After (run2)   Loss

3.750 MHz      .95 dB          .85           1.10           .70 dB
7.150         1.65            1.55           1.35          1.00
14.150        1.85            1.65           1.35          1.40
21.225        1.60            2.35           1.90          1.70
28.250        2.20            1.90           2.30          2.00

Looks like the anomoly on coax run 1 has shifted to 21 MHz. Run 2 looks more
like the computed loss curve, but 40M and 20M shouldn't be so close. I have
a feeling that the accuracy of the instrument and SWR method of measurement
are partly responsible for the results. Also, I would guess that some minor
impedance variations along the lines must be responsible for the curve
anomolies. The total coax length is now shorter by a few feet and I've seen
small changes like that change SWR readings when the line wasn't perfectly
matched to the load. These are open lines, but do pass through one lightning
suppressor along the way.

Clearly, there's a dispute as to whether the impedance variations could be
introduced by the connectors. Here's another theory: could the variations be
caused by bends in the coax? I'm using LMR400 UF, which has a foam core. The
bends in the coax aren't severe, but the extra coax is in a loop with
several turns at the far end. Perhaps the slight movement of the center
conductor towards the shield can make enough difference in impedance to
account for the results?

73, Dick, WC1M




--
FAQ on WWW:               http://www.contesting.com/towertalkfaq.html
Submissions:              towertalk@contesting.com
Administrative requests:  towertalk-REQUEST@contesting.com
Problems:                 owner-towertalk@contesting.com
Search:                   http://www.contesting.com/km9p/search.htm



--
FAQ on WWW:               http://www.contesting.com/towertalkfaq.html
Submissions:              towertalk@contesting.com
Administrative requests:  towertalk-REQUEST@contesting.com
Problems:                 owner-towertalk@contesting.com
Search:                   http://www.contesting.com/km9p/search.htm