While having another completely separate transmitter does complicate the system, there are a number of distinct advantages to be gained by doing so and these relate to the reduction of intermodulation and other effects of nonlinearities in amplifiers:
But how does one get the audio transmitter's signal on the air? There are two methods that are often used:
These types of combiners (often referred to as filter-combiners) are quite large and complex and require considerable skill to design and align: Very careful attention needs to be paid to loss, phasing, and group delay to avoid excessive distortion of the signals. Practically speaking this sort of filter is not likely to be found on an amateur television installation.
The use of separate antennas simplifies matters: The two systems can operate independently and tuning of filters of one will not affect the other. The disadvantage of having separate antennas is that you need two feedlines, two antennas (of course), and the real estate and money to handle both. Added to that, there is no guarantee that, because the two antennas are necessarily mounted in different locations, they will offer precisely the same coverage: If the audio aural antenna is lower on the tower than the visual antenna, there may be those who cannot receive the audio carrier well because of blockage by some features of the local terrain (i.e. buildings, trees, mountains.) Additionally, differences in the patterns of the two antennas may skew the A/V power ratios excessively, causing an apparent degradation of the overall signal quality.
We have chosen a variant of the filter/combiner system: Since we
only a small fraction of the visual power to transmit the audio, and
FM transmitters are so inexpensive (compared to the linear ones
for AM video) we are just combining them in a lossy fashion using a
The directional coupler shown operates like the classic Monimatch circuit. This circuit is the one used in many SWR bridges: In that application it only couples power that is moving in just one direction on the transmission line. We are doing this in reverse: Transmit power is being applied to the coupling line but it is being coupled to the main through-line in one direction. In this case, the direction of coupling is toward the antenna. By it's nature, it does not couple power from (in this case) the aural transmitter into the visual amplifier. This also works in reverse: The forward power from the visual transmitter is not coupled into the aural transmitter.
The "through" line is a piece of 1-5/8" heliax approximately 6.5 inches long. Approximately halfway between the shield and the center conductor, a long, thin screwdriver was used to bore a hole through the dielectric foam and a length of teflon-insulated wire was run as the "coupling" line. A careful look at the picture will reveal that this coupling line is connected to type N connectors via lengths of #12 copper wire. Careful observers might also notice that they seem to zig-zag a bit and appear to be a bit long for something to be used at 430 MHz, but these are actually matching sections to transform the coupling line's impedance (which is not 50 ohms) The return loss of this section is more than 18db over a 10 MHz bandwidth - more than enough for a narrow audio carrier!
The through line is mounted in the diecast aluminum box (which is approx. 7.5" L x 4.5" W x 2.25" H) by two 1/2" wide pieces of copper scraps cut from a discarded piece of the heliax (and hammered flat) soldered to the heliax (two pieces on each end) and mounted two the screws that attach the N connectors. Some anti-oxidant was used between the copper and aluminum to guarantee a long-lasting and stable electrical connection.
The characteristics of the directional coupler were measured as follows:
Those of you familiar with the design of such systems might raise your eyebrows at some of the return loss and isolation figures in the video service. I should point out that we do use isolators on both the visual and aural transmitters. This goes a long way toward preventing the distortions from reflections and nonlinearities that might result were RF power from one transmitter get into another.
A bit of quick math will reveal that, while we are throwing
over 80% of our aural transmitter's power, we only need to inject 1% of
the total power as aural power to keep the receivers happy. So, if you
had 100 watts of visual power you would need only put about 8 watts of
audio transmitter power into the coupler to attain the desired 20db A/V
ratio. Since our aural transmitter is capable of over 25 watts, we
On a site such as this where the ATV transmit antenna will be in close proximity with other amateur and commercial 450 MHz transmitting and receiving equipment, spectral purity is imperative! To this end, we have a 7 pole interdigital filter. This filter is a homebrew unit constructed of aluminum and held together with TIG welds and stainless-steel screws. The filter shown is from the 439.25 MHz transmit chain of the old WB7FID repeater. This filter has since been retuned for the 426.25 transmit frequency. Since this wasn't the original design frequency, its performance has suffered. We are planning to build a new, lower-loss filter to replace it, but we are using this one in the meantime.
The interdigital filter has the advantage of being able to be designed for a wide, flat response with low loss: Characteristics essential in video service. With this many poles of filtering the skirts of the filter are extremely sharp - sharp enough to provide some degree of VSB filtering on its own. The attenuation provided by this filter is well over 100 db by the time you get to 450 MHz, the top of the amateur band.
Are you interested in putting together a filter of your own? Look at the Putting together an Interdigital Filter page.
Return to the New WB7FID ATV Repeater Page...