Re-tuning 12-year old coax stub filters


Of late, I’d noticed that the SWR on the 10 and 15 meter stacks on tower 1 seemed to be getting higher and higher. Sometimes our first reaction is to head up the tower to the antenna to find out what is the trouble. This time, however, I didn’t go with that first gut reaction.

Rather, I began with a very methodical testing routine, with my AA-55 ZOOM connected to the coax in place of the transmitter at position A. I then inserted a precision dummy load at the next coax junction. Next, using the AA-55 I swept the coax on all 6 ham contest bands. I followed this pattern until I found SWR readings higher than 1.2:1.

The first thing I found was my ICE 419B bandpass filter. Even in bypass mode, it drove the SWR higher than 1.2:1. I inserted a brand new barrel connector in its place, and continued my testing.

The next problem found was on the output side of my 10-m coax filter stubs. I replace the stub assembly with another barrel connector, and continued the testing. Likewise, I found the 15-m coax filter stub assembly and the 40-m assembly to be causing SWR problems. More barrel connectors and the testing continued until I reached the distal ends of the six runs of hardline at the inputs of the various Stack Matches at the base of the tower.

The first step at remediation and repair was to test each individual coax stub in the 10, 15 and 40 meter assemblies. I found all of them to be considerable low in frequency relative to what I desired them to do. So, I retrimmed each of them to return them to the proper frequency. Then I reassembled everything, and retested. SWR was flat all the way to the base of the tower.

The second step was to replace the position A ICE 419B with a Dunestar Model 600. Following that, I again retested all the way to the tower. SWR was flat!

Things that make you go “Hmmmm”. My stubs reside in a shed, out of the sunlight and out of liquid rain. What caused them to drift over time and appear too long electrically?

Leave a Reply

Your email address will not be published.