Custom Rich-Text Page

- HOME PAGE- -SITE INFORMATION- -MY PHOTOS- -LEASING A RADIO SITE- -SITE SURVIVAL- -INTERMOD- -CIRCULATORS- -JAMMERS- -SIMULCASTING- -REMOTE RECEIVERS- NEWS PAGE




INTERMOD AND BANDPASS CAVITIES

By Peter Policani, K7PP

I remember breaking into the two way radio business in the early sixties and finding a great amount of concern over intermodulation and site degradation and desense. These are all ingredients for the disruption of a receivers ability to hear a signal. On community sites, it's intermodulation that rears it's head all to often.

So, let's deal with intermodulation first. I will attempt to dispense with all of the technical bull and hopefully offer you an explanation of concepts. You might find it a little more useful.

If you mix two signals together, you will come up with a combination. You will see the two originals, the sum of the two and the difference. Thats not all. If you have a very efficient mixer, you might also see the difference frequency mixed with the sum frequency and its sum and difference mixed with one of the two originals. In other words, depending on the quality of your mixing device and the strength of the signals that are present you have the ability to generate thousands of new frequencies.

A receiver operating anywhere near these mixes can hear them as if they were a valid signal. The fact is, they are a valid signal. Mixes can take place in any device that acts as a diode. You may hear it referred to as a non linear device. A transmitter final, a receiver front end, a water contaminated connector can all present a place to mix. Supply two or more signals, a non linear device and a way to a radiating element and you have a transmitter.

Let's look at it another way.

Suppose you had two pianos. Both were located right next to each other. If you struck a c note and a d note, you might hear the mathematical difference between the two. You might also hear the sum of the two notes. But suppose you rolled one of the pianos out of the room and almost out of earshot. Well, you might hear a strong c note, a weak d note and a very weak sum and difference note. You see, the mix takes place in your ear, a non linear device, and when one of the original tones is attenuated, the sum and difference are as well.

Suppose we could do that in the front end of a radio. Well, we might be able to clean up a few mixes. Several methods are used to accomplish this. The most common is a very high Q or selective resonate cavity filter. A device that looks like a large diameter pipe and equipped with a tuning device on top. There are usually two coax ports used as an input and an output and any signal applied to one port will appear at the other, so long as the cavity filter is tuned to the frequency applied to it. If another frequency is applied to the port as well, the cavity has the ability to decrease its level a great deal. If that interfering signal doesn't make it to the transmitter or receiver it can't very well mix.

There is also different qualities of cavitys. Some allow you to attenuate signals very close to the one you are working with and those cavitys are usually very large in diameter. The larger the internal surface of the cavity and it's center tuning device, the higher the "Q", the more selective. In order to cover the same frequency range however, the length and the ratio of diameter to internal cavity surface area remain the same. If the length dimensions increase the overall resonant frequency decreases. "Q" or selectivity can also be improved by adjusting the loops inside the cavity. The loops are connected to the input and output coax connectors.

By changing the loops or rotating them for minimum coupling, you might also increase the selectivity but will have to sacrifice insertion loss. Nothing for free, even in electronics. You need to design your system to take into account insertion loss from the filters you add. In a receiver you must remember that if you attenuate a signal into the noise floor, you might find it difficult or impossible to preamplify.

Cavities can also be used to attenuate only one frequency. These are called "notch" filters. These cavities are useful in obtaining attenuation on only one interfering signal.

Duplexers

Duplexers are made of resonant cavities. There are duplexers that are made of just notch cavities, bandpass cavities, or combinations of both.

Notch Duplexers

A notch duplexer is made of several notch cavities which provide attenuation of the transmitter signal on the receive frequency and attenuation of the receive frequency on the transmitter port. These types of duplexers work very well, however they don't provide any protection attenuating any out of band signals. Many amateurs use this type of duplexer because it is inexpensive and they don't know any better. The fact is that these types of duplexers are outlawed on most community sites. Some site managers will accept them if a band pass cavity is inserted in both the transmitter and receiver leg. Then you end up with a cost exceeding that of a band pass-band reject duplexer anyway. The easiest way to spot a notch-notch duplexer, is that the cavities only have one coax port. Don't be fooled by the single port type duplexers that claim to be band pass-band reject. The bandpass characteristics are only effective "in band". Out of band signals seem to sail right through will little or no attenuation. Remember, keeping out of band signals from mixing is what we're after. Also, keep in mind that out of band signals that make it through your duplexer might find you circulator or other filters and or preamps great places to mix and re radiate. If an out of band signal can get in, it's mixing products can usually find a way out.

Band Pass-Band Reject Duplexer

There are several flavors of this type of duplexer. If each cavity has an input and output port as well as a notch tuning adjustment. It's an effective bandpass system. If each cavity has a "T" fitting and a notch capacitor, it will not provide adequate protection from out of band signals. The manufacturer says they are band pass-band reject and thats a true statement, but you could pass a 900 mhz signal through one of these devices with almost no attenuation.

Don't think the site owners and managers aren't aware of this. Save your self some grief and make sure you start with the right type of duplexer. If you already have one of these types of duplexers, you might start looking for a single bandpass cavity to install on each of the ports feeding the transmitter and receiver.

I hope some of this makes sense, as implementing just the cavities alone to combat intermod would be quite effective.

There are other devices that will help to attenuate unwanted signals but I will save it for another article.

Pete Policani, K7PP