...Too much communication

- Intermods

Intermodulation interference or intermods as it is more commonly known, is a difficult subject that is affecting communications now, and the outlook is that its presence will increase significantly in the future. With the proliferation of mobile networks and the co-location of multiple antenna systems on single sites, the likelihood of this happening sooner rather than later is all too possible. If we are to enjoy high communications standards, intermods is a subject that the communications industry will need to understand and come to grips with...

Intermodulation distortion, commonly known as Intermods, is a type of interference that can occur in any non-linear junction. It exists when more than one signal is present at the same time. These fundamental signals, known as carrier signals, can combine to create new, unwanted frequencies resulting in interference.

Intermods are formed when carrier frequencies meet in a non-linear device. Although they are not generated harmonically, a harmonic relationship is involved. The resultant products are created from the sums and differences of the harmonics of the carrier frequencies, as a result of a mixing effect between the fundamental signals and their harmonics. As the order of harmonics increases, the less power is involved.

IM products are normally classified as odd order (3rd, 5th, 7th…) and even order (2nd, 4th, 6th…). Of these the odd order intermods are usually of more concern as, where the carrier frequencies are from the same band, the odd orders are more likely to fall in-band and so create interference, whereas the even order normally fall out of band. If the frequencies are from different bands, then even order intermods should be looked at as well. Of the odd orders, interference is most commonly caused by the 3rd and 5th orders. Higher orders can also be involved but these are formed at lower power levels and are usually too weak to cause interference.

To calculate the second order, the simplest form, the carrier frequencies, A and B, are simply added together to determine the sum product and subtracted to get the difference product. For example, if A is 151 MHz and B is 150 MHz (as above), the second order intermods will be at 301 MHz (A + B) and 1 MHz (A – B), thus falling out of band.

Third order products are most often involved in creating interference problems. These can arise when the carrier frequency (first harmonic) mixes with the second harmonic of the another carrier, (or when three fundamentals mix together). To find the difference products where two carriers are involved , the carrier A is multiplied by 2 then B is subtracted to give one. To find the other the carrier B is multiplied by 2 then A is subtracted and then the carrier A is added. For the sum products, the carrier A is multiplied by 2 then the carrier B is added, and the carrier B is multiplied by 2 and then the carrier A is added.

So if carrier A is 151 MHz and carrier B is 150 MHz, the IM difference products will be 152 MHz (2A - B) and 149 MHz (2B - A) thus falling in-band. The IM sum products will be 452 MHz (2A + B) and 451 MHz(2B + A), which fall out of band. The process is similar with the 5th order where the 3rd harmonic of one carrier mixes with the 2nd harmonic of the another carrier. In 7th order the 4th harmonic of one carrier mixes with the 3rd harmonic of the other.

For most communications systems it is the third and fifth order difference products that are the most critical, as these are strongest and fall closest to the wanted signal, thus causing the most disruption. Of course if the frequencies concerned are from different bands, even order products, especially the 2nd order, may well be involved. Usually the 7th and 9th orders are too weak to create a problem. In practice harmonics above the 10th or 15th are rarely involved.

It is important to realise that when you are dealing with large signal levels, the mixer does not have to be highly efficient to create serious interference to on-site receivers. With increasing competition for use of the VHF/UHF bands and decreases in the size of guard bands, intermodulation problems are likely to become more serious in the future. Therefore it is important to ensure that these types of equipment will not be subject to intermodulation effects. Tests can be carried out to ensure that this will not occur. Software packages, from commercial to shareware and freeware, are available to assist in calculating IM products to simplify the process. Where the carrier frequencies are from different bands, both the odd and even order products should be checked.

Intermodulation interference can be produced by receivers and transmitters. It can also result from the mixing of transmitter signals in other non-linear junctions, usually metallic. Rusty tower bolts or similar non-linear metallic junctions in the area like guy wire, tower joints, turnbuckles and anchor rods can all be at fault.

Intermodulation effects may also occur due to non-linearity in the transmission system caused by antennas, transmission lines, connectors and the like. In transmitter intermodulation, the transmitters involved obviously need to be on air at the same time. This occurs when one or more external signals get into a transmitter’s non-linear final amplifier stage via the antenna feeder coax. New frequencies can be created in any or all of the transmitters’ power amplifiers, due to the non-linearity of the amplifier gain as a function of input power.

The front end of receivers is also a common mixing point. Receiver intermodulation occurs when two or more high level off channel signals overload the receiver’s RF amplifier or, in the case of simpler radios, the first mixer. As a result the receiver operates in its non-linear region acting as a mixer.

Passive intermods can also arise where two dissimilar metals in a component meet, for example between the base metal and its plating. When this happens, a diode like junction, which has a non-linear input to output voltage relationship, is created where the two metals meet. Analog RF signal multiplexers can often be a source as they have many simultaneous carriers as inputs and large order intermods can result. Rust between sections of roofs, masts, and the like can act as a non-linear diode.

Because IM products are the same no matter where their origin is, it is not always evident where they are coming from. A useful determining factor is that receiver intermods attenuate more quickly than transmitter intermods, which can be ascertained by inserting an attenuator in the receiver front end. For example, a transmitter intermod will be attenuated 6dB by a 6dB attenuator, whereas with a receiver intermod both signals will be reduced by 6dB.

It is also possible for two signals to create intermods in a faulty transmitter without affecting the transmitter frequency. In this case the fault is very hard to trace as the transmitter, to all extents and purposes, is performing normally. Narrowband filters and ferrite isolators to attenuate signals coming back from the feeder can help in this case.

Where rust is involved, effects will vary according to weather conditions, as wind compresses parts of the rusty metal together or wears them apart. Rain also alters the characteristics of the rust. Ultrasonic sniffers can assist greatly in locating non-linear joints in equipment, and good maintenance procedures , such as use of isolators, combiners, cavity filters, etc., is highly recommended.

While minor corrosion in coaxial connectors or antennas may not be enough to impair VSWR and signal strength, it may be enough to act like a very poor diode and cause intermodulation effects.

There are a number of factors that can assist in the prevention of IM interference and these depend on where the IM products are generated. The key to their suppression is in identifying where the signals are mixed. Unfortunately it is possible to have a number of mixing points.

When intermods are generated in an amplifier, the level of the 3rd order IM products at the amplifier output depend not only on the levels of the carrier frequencies but also on the 3rd order intercept point (TOIP) of the amplifier. The TOIP is a theoretical figure used to indicate the level of the amplifier’s ability to exclude intermod products. A high TOIP predicts a high level of exclusion.

If a low quality amplifier has a TOIP of 5dBm (dB per milliwatt) and the input levels of the carrier frequencies is -10dBm, the resultant 3rd order difference levels at the amplifier output are at -40dBm. However, if the input levels of the carrier frequencies remain the same but the amplifier TOIP is increased to 10dBm (an increase of 50%), the amplifier output levels fall to -50dBm. This drop (10dBm) represents double the amount of the increase in the TOIP level (5dBm). Using an amplifier with a TOIP of 20dBm will drop the levels to -70dBm.

Moreover, for every 1dB decrease in the input levels of the carrier frequencies, provided that this occurs prior to the mixing point, there will be a 3dB drop in the 3rd order IM products. Similarly, for 5th order products, there is a five for one reduction.

Transmitter produced IM products are usually generated in the class C ouput stage of the transmitter, which is non-linear, rich in harmonics and connected to the antenna.

Suppression of interfering IM signals can be achieved by use of bandpass cavity filters and isolators in the offending transmitter where the mixing occurs. The amount of suppression required will be determined by the site noise level and the minimum necessary receive level of the affected receiver.

In the case of receiver produced IM products, it is possible to suppress the IM signals by placing an attenuator between the antenna and the receiver input. With 3dB attenuation, a 9dB reduction in 3rd order products can be achieved. However, as the wanted signal will also be reduced by 3dB, the net result is a 6dB reduction in carrier-to-interference ratio.

The more selectivity used prior to the first active receiver stage, the better the receiver is able to reject IM products. For this reason, the use of two or three bandpass cavity filters in cascade placed in front of the receiver can be successful. Notch filters may also be used. These provide more attenuation close to the desired frequency than bandpass cavities, but only work on a single frequencies, whereas bandpass filters can help with multiple frequencies. Crystal filters are also possible for VHF highband.

Passive intermodulation (PIM) products can be a special problem for triband antenna systems used in cellular systems. Because the base station antennas are often used for transmission and reception at the same time, IM products can be generated when two or more transmitter carriers are combined in one antenna. This is complicated by the practice of locating multiple antennas into the one site. For these reasons it is essential to take this into account in the planning stages, paying particular attention to using suitable materials and avoiding poor mechanical joints. Multiband antennas especially need to be designed so that the three bands are sufficiently isolated to minimise the occurrence of IM products.

When you are measuring intermodulation interference, it is important to remember that the RF measuring tool or spectrum analyser has its own 3rd order intercept. This can be overcome by using an external filter that passes only the IM product you are looking for and rejects the strong signals that made it.

At the planning stage, much can be done to avoid IM problems. For example, when co-locating antennas on a single site, it is important to check that the carrier frequencies will not cause problems. Buying cheap equipment, like transceivers, receivers, antennas and especially amplifiers, may mean that poor quality materials and workmanship have been used, leaving you wide open to the occurrence of non-linear junctions and intermods. Initial savings will mean little compared to the annoyance generated by poor performance, not to mention the cost in time and money of rectifying it.