The quarter wave transmission line stub as a TVI filter.

 I recently set up a Diamond DJ-130 discone for local repeater use on VHF/UHF. The discone was 5m off the ground and about 15m and at 90 degrees to the main lobe of the main house TV antenna. The TV antenna is a VHF/UHF combo design with seven elements on UHF with a corner reflector and a six element log periodic on VHF. This feeds a TV masthead amplifier whose output runs to a 4 way splitter in the roof space. All four outlets have a TV connected.

Australian TV is still mostly analog transmissions on VHF channels with digital TV also transmitted on VHF. There are some UHF services and translators but in my location it's nearly all on VHF. Hence our TV masthead amplifiers must have gain between roughly 40MHz to 240MHz as well as 500MHz to 800MHz.

Preliminary tests on UHF with a 30 watt mobile showed no TV interference but VHF 2M transmission was another matter. The two metre radio puts out 30W on high power and around 4 watts on low power. High power transmissions completely wiped out all analog VHF and UHF TV reception as well as significantly affecting all digital reception. Low power transmissions caused significant degradation of analog TV reception very minor digital TV disruption. The neighbors were unaffected by either high or low power transmissions. Until I could resolve the issue I limited my VHF operation to low power and times when there was "nothing good on TV" - on that basis one might ask if there really is a problem at all.

All cabling and connections were checked as was the TV antenna but no faults were found. It was clear that the problem was due to fundamental overload of the TV's as a result of the gain in the masthead amplifier amplifying the 2M transmissions. I thought I would try to resolve the problem by using a 1/4 wave stub in the antenna feedline. Theory tells us that a length of transmission line that is one quarter wave long at the frequency of interest will look like a short circuit when open at the far end and look like an open circuit when shorted at the far end. The plan was to add a 1/4 wave open stub in parallel with the feedline from the antenna that would severely attenuate any 2M signals that might appear.

With the help of a friend with a spectrum analyzer and tracking generator we set about cutting a stub of RG6 coax to the correct length for the centre of the 2M band. We set the analyzer and generator to sweep 20MHz centered on 146MHz. In free space, a quarter wave at 146MHz is around 513 mm. take into account the velocity factor of the cable at say 0.8 and you get 410mm. This was our starting point and after a bit of trimming we came to a final length of 402mm from the tip of the connector to the end of the stub. The trace on the spectrum analyzer showed a pronounced dip at the centre frequency in the order of 30 - 35dB but there was also some attenuation present at either end of the sweep.

Performance was initially very good with only the slightest trace of noise visible on the analog TV channels when transmitting on high power. Low power transmissions showed no effects whatsoever. The MWF (Minister for War and Finance) was suitably impressed with the results and all was good in the world. That was several weeks ago. In the last few days we had significant troppo ducting which all but destroyed all TV reception in my area. Before I realized what was happening I had undertaken a major investigation of the TV antenna system. Of course there was nothing wrong and once the index fell back to lower levels everything came back to normal - well almost everything. During the investigation it became clear that the stub was having some effect on analog TV reception. Although the reception was acceptable, there was some noise evident in the picture with the stub connected.

I decided to replace the stub with a simple series tuned trap built into the masthead amp. Nothing more than a six turns of a smallish gauge enameled copper wire 6mm in diameter in series with a 15pf trimmer. I used a VHF receiver with a short whip to monitor a local 2m beacon which uses FSK modulation. By connecting the trap across the feedline I was able to trim the cap and notch out the beacon all together. The notch was very deep. With my receiver set to FM the S-meter showed the beacon at 40dB over S 9. Once the trap was tuned, the meter needle did not move and I could barely hear the beacon on SSB. Is my meter calibrated correctly? I doubt it very much, but it does show that the the trap was tuned correctly and does have significant impact in the 2m band. The new trap had no effect on TV reception unlike the stub and TVI is non-existent. Again a very happy outcome.

Conclusion.

What happened? I'm not sure, either the weather affected the stub or feedline in some way or I didn't check the picture too carefully after I first added the stub. Maybe I didn't pay enough attention to the amount of attenuation provided by the trap outside the 2m band. We certainly were interested in getting the notch bang on 146MHz when we were cutting the stub, perhaps we should have been a little more careful with the width of the notch too.

Having thought about it for a while I'm not sure that I would try a stub type TVI filter again. Most modern commercial TV masthead amplifiers now include a 2M trap to limit the effect of pager interference. This trap is a coil and a small fixed capacitor and the installation technician squeezes or stretches the coil to tune the trap. From a cost perspective a short length of coax might be cheap but if you add a couple of F-connectors and a tee piece and cable ties plus the hassle of weather proofing etc it certainly exceeds the cost of a short length of wire and a small ceramic capacitor or a trimmer cap that can all fit neatly in the masthead amp or directly to the back of the TV antenna wall plate in the house.