[Top] [All Lists]

Re: [TowerTalk] M2 20M5 SWR measurements

Subject: Re: [TowerTalk] M2 20M5 SWR measurements
From: Jim Brown <>
Date: Sat, 31 Dec 2011 10:46:29 -0800
List-post: <">>
On 12/31/2011 6:41 AM, Jorge Diez CX6VM wrote:
> I have a 20M5 and 15M6 antennas from M2
> After 3 years using this antennas, I noticed a change in the SWR of the 20M5
> when I added a High power BPF, for what I need to add some coaxial jumpers.

I suggest that you look at the SWR for the new filters into a 50 ohm 
dummy load, and also carefully measure the loss through the filter. The 
filter could be responsible for some of what you're seeing. ANY filter 
can add some phase shift and mismatch. A good one should add much less. 
Look inside the filter -- is there coax from each connector to the first 
circuit elements, or are they depending on the chassis to carry the 
signal return? Using the chassis as a return adds stray inductance to 
the signal path, which means unintended phase shift, can degrade the 
performance of the filter, and increases SWR. There's also the question 
of whether the filter was properly tuned during final test.  An 
inductive or capacitive component that is out of tolerance can degrade 
the filter response and increase the SWR.  Some filter builders do 
serious alignment during final test, but some do not.  Nearly all the 
filters I bought from ICE came to me out of alignment, and none of them 
used coax inside the filter to minimize stray inductance

There's also the matter of how the small impedance mismatch of your 
antenna looks after it is transformed by the length of the feedline. NO 
antenna is a perfect match, and filters are designed to be connected to 
a pure 50 ohm load, and that's how they are tested. But if the load on 
the filter is not exactly 50 ohms, the response of the filter will be 
different, and so will the impedance at the input of the filter. What 
the feedline does will random depending on its length, its loss, the 
frequency, and what the impedance is at the antenna (not exactly 50 
ohms, with some added XL or XC. This means that the feedline at the TX 
end will not be exactly 50 ohms, and there will be some added XL or XC.  
The filter will transform that impedance, and the filter can either 
cancel some of the XL or XC, or it can add to it; it can add to the R 
component (most likely) or it can reduce it.

As for your test with only the jumper added to your long coax run --  
what are you using to splice those cables together?  Is it an Amphenol 
barrel connector, or is it a no-name product?  No-name connectors are 
well known to be cheap junk, and could cause that mismatch. Bad 
installation of the connectors on each of the coax cables, both the main 
long run and the jumper could cause a problem like this. Are they 
soldered or crimped?  Are you sure that the shield connection is very 
very solid for each connector?

73, Jim K9YC

TowerTalk mailing list

<Prev in Thread] Current Thread [Next in Thread>