 |
The Great Deviation Debate |
|
The Great Deviation Debate |
|
Having an elevated home, such as Snaefell Mountain, is a mixed blessing for a repeater. On the one hand, it gives remarkable coverage and a large audience. On the other hand, it's wide service area means that it must conform as tightly as technology allows to the current technical standards, if interference to and, indeed, from others is to be avoided. For this reason, the IOM repeater group responded very quickly to the RSGB's declaration that it was to adopt IARU standards for NBFM repeaters. We felt that it was important to lead the way in demonstrating that compliance could be achieved - even on a site like ours, where access is very difficult during the winter months. The existing equipment at the time (Philips F494) was modified to adopt the 12.5KHz channel spacing specification. The transmitter's peak deviation was limited to 2.4KHz and the crystal filters in the receiver were replaced with narrower ones. Although this brought the receiver closer to the specified bandwidth, there was still a lot of headroom, allowing stations with high deviation levels to continue to operate successfully through the repeater. The installation of a Philips FX5000 in September 2002 brought with it a whole new set of "problems". The FX's receiver was built for 12.5KHz channelling and, as such, enforces the specifications more rigidly that ever before. Stations who were previously just about "getting away with it" are now finding that they are having difficulties. Beginning with examples of some typical user complaints, this article sets out to explain what is going on...... |
||
| "I
used to be able to access the repeater. Now they've 'fixed' it, I can't
get in any more"
"My Rig's brand-new, state-of-the-art and makes toast, and I can't access the repeater - It can't be my rig." "I can access other repeaters well enough." "I can talk to my mate across town on simplex with no problems at all - I just can't access the repeater. So it can't be my rig?" "The repeater's been playing-up a lot lately. Must be something wrong with it...
|
||
|
These, and many more besides, are the oft-heard laments of radio amateurs who have fallen foul of the great "Deviation trap". The sad, simple and irrefutable truth of the matter, though, is quite simply this: It IS your rig that's at fault, like it or not. In this note, I will attempt to explain what is happening, why the problem lies with your rig and what you can do about it. |
I'm not in the business of slagging-off
someone's pride and joy, especially if it's just out of the box. I'm
just presenting the facts.
I have spent ten years working in the radio communications industry, and I have worked on a number of amateur repeaters in addition to the commercial ones which came under my remit, so I feel that I can speak from a position of at least some authority. |
|
| The Problem | ||
| The
best way towards a solution is to first define the problem. Since people
who have an understanding of FM principles will already know the causes
and cures of this problem, I will attempt to aim this note at those who
have not.
What is FM? A Frequency Modulated transmitter conveys speech patterns by varying the frequency of an oscillator. As the frequency of the oscillator is changed by the modulating signal, it is said to "Deviate" from it's rest, or unmodulated, frequency. Hence the term "Deviation". The rate at which this deviation occurs is governed by the frequency of the modulating signal. The amount by which the carrier deviates is governed by the amplitude, or loudness, of the modulating signal. In the illustration on the right, you can see how the modulating signal at the top affects the unmodulated carrier, at centre, to produce the final, frequency modulated carrier. Notice how the FM carrier frequency increases to a maximum when the modulating signal reaches it's peak, and decreases to a minimum when the modulating signal reaches it's minimum, passing through zero deviation on the way. The RF spectrum which results from frequency modulating a carrier has, in theory, sidebands stretching to infinity. We don't need to go there, but it is essential to understand that every modulated carrier requires room in the RF world. The greater the level of modulation, the greater the amount of space required. |
||
 |
||
| There's
an important point to be made here. It is quite possible to overmodulate
an A.M. (Amplitude Modulated) transmitter, with the resulting spurious
crud that goes with doing so due to break-up of the carrier.
It is not possible to "Overdeviate" an FM transmitter. As the deviation of an FM transmitter is increased, it merely occupies more and more spectrum, the carrier remains unbroken (conveniently side-stepping Bessel Functions). In view of this point, it should be understood that "Overdeviation" is a synonym for "excessive deviation". |
||
|
|
||
|
The F.M. Receiver In order to be able to distinguish between signals on different channels, a receiver needs a well-known feature called "Selectivity". This simply refers to the ability to accept wanted signals and reject nearby, unwanted, ones. In order to achieve this it is necessary to introduce filters into the receiver which allow the wanted frequencies to pass whilst rejecting others. It is at this stage that the perceived problem occurs. The curve on the right shows a typical response curve for an I.F. filter which would be used in a narrow-band F.M. (NBFM) receiver. The curve shows that the further a signal deviates from the centre frequency, the more it is attenuated, or reduced in amplitude. In this case, a deviation of +/-3.75KHz results in the signal being reduced to half of it's maximum (-3dB). When a signal deviates too far from the centre frequency, the first thing which happens is that audible distortion sets-in because of the attenuating effect of the filter, and the severe phase shift which the filter introduces at it's band edges. This manifests itself as a "Wooliness", or "Fuzziness" in the audio. As the signal deviates still further, the level of the signal is progressively attenuated until it disappears altogether. With the filter shown, a signal with a peak-to-peak deviation of about 5KHz (+/- 2.5KHz) would pass relatively unhindered. Anything much above that would suffer the distortion effects mentioned above, followed by a vanishing act. It is for this reason that a station experiencing these problems can effect a temporary solution by simply speaking more quietly, thereby reducing their carrier deviation. Obviously, a more long-term fix is required. |
|
|
|
|
| The
F.M. Transmitter
The solution to everyone's woes is blindingly simple, but it lies in the individual's transmitter and not, by any means, in the repeater. |
|
|
The block diagram on the right shows the layout of a typical NBFM transmitter's audio stages. Notice that there are two controls; Microphone Gain and Deviation. Some transmitters do not have a microphone gain control, relying instead upon a fixed level of gain. |
 |
|
The audio from the microphone pre-amplifier is passed to a limiter which prevents the audio from exceeding a pre-set level. The higher the limiter setting, the higher the resulting carrier deviation. Because a limiter circuit introduces distortion into the audio, it is necessary to pass the signal through a low-pass filter to remove any harmonics. The usual method of adjusting an FM Modulator is to inject a tone at the microphone input and increase it's level until the deviation stops increasing. At this point the limiter has begun operating. The Deviation control is then adjusted to give the correct deviation for the channel spacing in use (+/- 4.8KHz for 25KHz channels and +/- 2.4KHz for 12.5KHz channels). The microphone gain can then be adjusted on speech to give an average deviation of around half of the maximum. If the microphone gain is set too high, then the limiter will be active for much of the time, giving rise to a compressed audio output which will sound rather "intense". Another result of excessive microphone gain is that background noise is amplified unduly, which can make a mobile transmission almost unintelligible due to high levels of road noise. Notice that simply adjusting the microphone gain is only part of the solution. It is no more effective than moving away from the microphone. Whilst it may result in your transmission being acceptable during normal conversation, the limiter threshold may still be too high. If you should start to speak more loudly, your transmitter deviation will still increase - maybe too far - causing the clipping effects described earlier. The only proper solution is to adjust the deviation control so that the limiter can do it's job properly. Toneburst Access The problems associated with excessive deviation do not only apply to speech. They also apply to 1750Hz tonebursts and DTMF keypads. It is extremely common for 1750Hz access tone generators to be set to an unnecessarily high level. Quite often pressing the tone call button on your rig will result in your signal doing an immediate disappearing act. This is at the root of a great many "I can't access the repeater" problems - the repeater doesn't react to your transmission because it can't hear it. The specification for amateur repeaters requires that they should respond to a 1750Hz tone at a level of "Half system deviation". For a 12.5KHz channel, the system peak deviation is 2.4KHz. Therefore, a toneburst should be accepted at a level of not more than 1.2KHz deviation. The present design of GB3GD will accept a toneburst at a deviation of around 0.5KHz - Very faint indeed. What this means is that tonebursts do not have to be deafeningly loud in order to activate a repeater. The truth is that a toneburst which is too loud can be guaranteed to fail every time!! The audio from a toneburst generator is usually injected into the microphone stage. i.e. prior to the limiter. It follows, then, that if your radio is correctly set up, you should have no problems. You should, however, take care to ensure that your toneburst is set to a level which is well below your radio's peak deviation. |
|
|
|
|
|
Laying Down The Law No-one likes to be told that their rig is maladjusted, but experience has shown that, for reasons unknown, the Japanese never make much of an effort with their FM transmitters. Everything else about their radios is wonderful - superb sensitivity, ease of use, lots of output power etc - but they just don't seem to get FM deviation right at all. It is therefore up to the individual to ensure that, not only are they operating a radio which will get through the local repeater, but also that they are operating within the law. Here are a couple of extracts from various books on the subject:
The first four references are deliberately taken from older publications in order to point out that the requirement for low deviation settings is nothing new; It was being specified 26 years ago!!!! The fifth reference should be all that is required. Your licence states that you should occupy as little space as possible. To do otherwise is to contravene it's conditions. |
|
|
Repeater Specifications
The Isle of Man's 2 metre repeater is based upon a Philips FX5000 transceiver. This radio is designed for use in the hostile environment of a multi-user hilltop radio site and, by law, has to conform to stringent specifications which do not necessarily apply to amateur equipment. The performance of both the transmitter and the receiver are tightly controlled and are, once set-up correctly, beyond reproach. Another reason for the introduction of tight receiver specifications on amateur repeaters (aside from the fact that the concept of true NBFM is well over 20 years old), is that the UK repeater management committee and the RA have begun licensing repeaters on 12.5KHz channel slots. In order to facilitate this, they decreed that all amateur repeaters in the UK should be altered to 12.5KHz channelling. One such repeater which I have been involved with is GB3DN. This repeater has an input frequency which is 12.5KHz higher than GB3GD (145.0375MHz). If our filters were left at the old 25KHz specification, it would be quite possible that people using that repeater could cause problems on ours. Tests with "Real-world" signals on GB3GD have shown that the FX5000 will pass a signal with a maximum deviation of 4.1KHz before clipping occurs. Since the specified maximum deviation for signals coming into the receiver is 2.4KHz, this represents a significant margin of error. Before being installed on a hilltop site, a repeater must satisfy those working on it that it will perform correctly, and continue to do so in the long-term. In order to ensure that, it must be carefully set up with the right test equipment. GB3GD was aligned using a Marconi Instruments 2955A Radio Communications Test Set. Because the parameters of the repeater are clearly set during testing, it can be safely assumed that it represents the standard against which all other radios may be judged, particularly with regard to transmitter deviation. Since the receiver will pass a 4.1KHz peak deviated transmission it follows that, if your signal is being "chopped", your transmitter deviation exceeds 4.1KHz. In turn, it follows that your transmitter is not set-up in accordance with the regulations and that, in fact, you are breaking the law. If you wish to complain that your signal is being carved-up by the repeater, you should bear in mind that to do so is to admit that you know that your transmission is breaking the law. |
|
|
In Conclusion
The foregoing will, hopefully, have shed some light on the subject of FM transmissions and the potential problems which can arise when a transmission doesn't match the expectations of the receiver. The simple truth is that, in 99.9% of cases, it is the user's radio which is at fault. Blaming the repeater, or making banal remarks like "the repeater's playing up these days" is to avoid the real issue - or, at least, to admit ignorance. Having spent a great deal of time and effort in getting the repeater to perform exactly as it should, and having used the best equipment to achieve that end, I find it very offensive to hear people suggesting that there's something wrong with my handiwork when, in fact, the root of problem lies very much closer to home. To return to the comments quoted at the top of this page - and to answer them:- "I used to be able to access the repeater. Now they've 'fixed' it, I can't get in any more" |
|
|
Further Reading: RSGB Repeater Management Committee Repeater Report - January 2000 "Repeater Planning In The New Millennium" RSGB Repeater Management Committee Website "Operating On The VHF Amateur Bands" - RSGB "Amendments to IARU Region 1 Recommendations ; Definition of modes" - International Amateur Radio Union "Technical standards for NBFM and NBFM repeaters" - International Amateur Radio Union "12.5kHz Channel Spacing on 144 - 146MHz - Proposal C5.5 from RSGB"
|
|
October 2002 - GD4HOZ
