The debate in my mind between building a duplexer with 6″ copper transmission line vs. a TV broadcast filter ultimately came down to exactly WHAT was inside the filter cabinets. For years I had been walking past the filter boxes, wondering what was inside. Documents I found at the transmitter site indicate that the smaller box (on the right) is the diplexer, and the larger box is the bandpass filter. There were even drawings of the diplexer and filter that showed the version for higher TV channels, with correspondingly shorter cabinets.
You might be thinking, “If that is a diplexer, you’re all set, right?” Not exactly. Duplexers and diplexers are very similar. In fact, similar enough that when I try to describe the difference between them, I really struggle to find words that clearly define the difference. Try it yourself! They both combine and/or separate signals of different frequencies. What makes them different? Hmmm… Normally we think of a duplexer being used to allow a repeater to transmit and receive simultaneously on a single antenna, but a diplexer can do that too with a dual-band rig.
Anyway, while some readers take a moment to ponder that question, let’s move on. An analog TV transmitter generates separate carriers for sound and picture; it’s really like two transmitters in one. The diplexer (4 cavity cabinet) combines those two feeds, then passes it on to the bandpass filter (6 cavity cabinet). I decided to start with the latter, so I found a 4mm hex head driver, and started removing screws. Literally hundreds of them! Finally, the cover came off. Refer to the picture below, and notice that the filter cabinet contains six resonant cavities.
The first thing that caught my eye was the bright, silver-plated resonators. These are very similar in construction to the resonators in duplexer cavities, just a little larger diameter than average. The resonator is tuned by adjusting the length; the larger diameter section at the bottom slides up and down on the smaller upper section. The longer the stub, the lower the resonant frequency. Rather than a threaded rod, the adjustment is made by loosening a compression nut on the top, and sliding the rod up and down. To provide thermal stability, the rod is made of Invar, an alloy of 36% nickel, 64% iron. At this ratio, the metal exhibits virtually no expansion and contraction due to temperature changes, and helps keep the filter in tune.
Notice that the partitions between the cavities don’t go all the way to the top and bottom. That’s because these cavities are directly coupled to each other right through those holes. The amount of coupling is adjusted by rotating the coupling loops. I don’t know for sure, but I believe very few (if any) duplexer designs use directly coupled cavities. It is far more common to couple signals in and out of individual cavities using coupling loops attached to coax through the top of the cavity.
It appears that the TV signal entered via direct coupling to the 2nd resonator, from 1-7/8″ rigid copper feedline that is visible on the back wall of the cabinet. Energy below the filter’s passband was shunted into the 1st cavity (the longest resonator, therefore the lowest frequency), coupled into it by the adjustable loop at the top of the wall between the 1st and 2nd cavities. The 2nd through 5th cavities are tuned to form a bandpass response with a broad “nose,” several megahertz wide, and energy above the passband is shunted into the 6th cavity (the shortest resonator, therefore the highest frequency). The output signal is coupled from the 5th cavity resonator to a coax connector, in a manner similar to the input.
This is all very interesting hardware, with very heavy-duty construction to handle high power levels (about 8kW in our case). But I had to wonder, could this filter with a fairly broad passband be re-tuned to have a sharper response, and function as a two-way radio duplexer? Or would I have to make fundamental changes to the construction? Could I use the direct coupling between cavities, or would I have to fill in the gaps and use traditional coupling loops that are fed from the top? I even contemplated removing the resonators and possibly installing them inside pieces of 6″ copper pipe, but their large diameter would be too tight a fit.
Before deciding what to do, I would need to see inside the smaller diplexer cabinet – coming in the next installment.