District Radio News Reporting -
A Low-Cost Approach

Paper presented to the Commonwealth Broadcasting Association East Central and Southern Africa Group Engineering Committee sessions, 1981. Reprinted in "Combroad," it then appeared in various radio magazines around the world, especially the description of the aerial in the appendix. I've added some illustrations to this HTML version.

by D. W. Harris, C.Eng. MIEE
Deputy Director (Engineering), Radio Botswana

Summary

The collection of news material for radio poses problems in many under-developed countries, where communications by road and telephone are frequently unreliable or of poor quality, over great distances.

This paper describes an approach to the problem based on the use of HF radio transceivers in Botswana. The relative merits of this approach are discussed, together with an analysis of the costs involved. A description is given of the type of equipment used, together with necessary modifications.

Background
In common with many other developing countries, Botswana has internal communications difficulties. Roads are long, and frequently poor in the rural areas away from the railway line. The internal telephone and telecommunications network is hardly developed at all; "telephone quality" is at best adequate, at worst unusable for rebroadcasting.
It is hardly surprising, therefore, that a frequent criticism of the national radio station has been its lack of "Botswana news". A recent report (1) quoted a resident of Serowe as asking "Why don't you just call it Radio Gaborone, because that's what it is?" Clearly, the station had to place high priority on the rapid passing of news from the Districts to the capital.

Uher 4000 tape recorder
A Uher tape recorder from the late 1970s Radio Botswana is part of the Department of Information and Broadcasting, and traditionally the Information Officers in the various Districts have had, as part of their brief, the collection of news and information from their areas. Usually, the medium is the ubiquitous Uher tape recorder, upon which interviews are recorded and voice-pieces composed. The tape would subsequently be sent on to the capital for eventual incorporation into bulletins, newsreel programmes or current affairs slots. An alternative was the use of the public telephone network, in which case a prepared script would be read down the telephone and recorded by Radio Botswana. In the unlikely event of its quality being adequate for transmission, it would be aired forthwith; alternatively, a transcript would be prepared and used accordingly. In either case it was virtually impossible to use up-to-date quality material. The delay might be anything from 24 hours to a few weeks!

The Technical Problem
What was needed was an alternative to the public telephone network which was fast, reliable and capable of sustaining reasonable quality. Cost and ease of operation were also very important.
The obvious solution was the use of radio. Unfortunately, the distances involved precluded VHF for all but the most local Information Offices. HF, on the other hand, is subject to all sorts of interference, and if SSB is used, usually of poor quality and limited bandwidth unless highly sophisticated equipment is available. The choice of frequency can be critical throughout the day to optimise signal-to-noise ratio, and this implies a certain amount of frequency agility in a remote installation, which can cause considerable problems with aerials or tuning units. Not least among the technical obstacles is the provision of power for radio transceiver equipment.
All these disadvantages usually restrict the use of HF SSB radio to basic communication. Commercial SSB equipment is normally fairly reliable, but powering, aerials and above all audio quality frequently reduce its usefulness for broadcast purposes to a minimum.
Nevertheless, it was decided that Radio Botswana should investigate the use of HF SSB transceivers, attempting to find ways round the problems mentioned above. To a large extent, this has been successful. In the process, a low-cost aerial has been developed, and some interesting information derived relating to the tuning of SSB signals by non-technical staff.

The Transceiver
From the start, it was realised that the correct choice of SSB transceiver would be crucial to the success of the project. To this end, a loose specification was drawn up and tenders were invited from various commercial manufacturers. [The specification included: 2.0-18.5 MHz coverage; a.m. detector; 100W output; transmit bandwidth to 5 kHz; 12v operation; reception of medium wave; 6 crystal TX channels; VFO receiver; high-quality audio stages; service manuals and spares.]

Upon evaluation, it was realised that the manufacturer most closely approaching the specification was M/s R L Drake Co. The TR-7 transceiver offered a choice of receiver bandwidths, operation over the whole HF band, broadband output circuitry, a general coverage receiver, 12-volt power supply capability and good serviceability.
The Drake TR-7 transceiver

One of the the most difficult requirements in the specification was a "wider" transmit bandwidth. Reaction from conventional manufacturers to the request for a transmitted bandwidth greater than the standard 2.7 kHz or so ranged from complete incredulity to guarded pessimism. Most commercial SSB equipment is rugged and otherwise suitable for communications, but this restricted audio bandwidth seemed to be a stumbling factor for Radio Botswana's purposes.
TR-7 filter board

The TR-7 IF selectivity board has space for four filters
The Drake TR-7 has a 2.3 kHz wide filter fitted as standard. However, it is mechanically possible to fit any of a range of filters (intended for receiving only) into the "transmit" position. M/s Drake specifically warn against doing this (2) so naturally, being good broadcasting engineers, we tried it.

First attempts were made with the Drake SL-6000, a 6 KHz filter intended for broadcast reception. Audio quality was of course greatly improved, but unfortunately the filter's skirt selectivity did not adequately eliminate the unwanted sideband. The net result was a signal which sounded "burbly" and was difficult to tune in. Some experiments were conducted using the transmitter in the "one-sideband-plus-full-carrier" (A3H) mode. This gave good results, using the excellent low-distortion a.m. detector fitted to the TR-7, but signal-to-noise ratio suffered for distant stations. (However, it is intended that further work should be done on this to allow a.m. transmission from OB vehicles working within about 300 km of the capital.)
In the interests of establishing the network, the transceivers were restored to their original transmit bandwidth (approximately 300 Hz - 2.6 kHz) and installed. Even like this, the improvement over the public telephone network was startling.

SSB Generation
At this point, it is interesting to dwell on the method of setting the carrier frequency in the TR-7, relative to the SSB filter, and the effect this has on the quality of transmission.
Most commercial transceivers generate SSB using a fixed intermediate carrier frequency (crystal controlled) which is accurately placed on the side of the selectivity curve of a fixed frequency crystal or mechanical filter. The SSB thus generated is then heterodyned to the required output frequency.
A disadvantage of this method is that, once the manufacturer has decided on the position of the "carrier" relative to the filter, there is very little the user can do to change it, short of purchasing a range of crystals around the nominal carrier frequency and plugging them in one by one until satisfied. This stems, of course, from the very reasons a crystal was chosen in the first place - ie. high frequency stability and resistance to "pulling" by external circuitry. Lest it be thought that no adjustment is necessary, it is worth pointing out that a difference in "carrier" frequency of only tens of Hertz can have a noticeable effect on transmitted quality. (NB: this refers not to accuracy of subsequent resolution of the signal at the receiver, but to how far down the slope of the filter the carrier is placed). Unfortunately, manufacturing tolerances may mean differences of a couple of hundred Hertz between optimum carrier frequencies for nominally identical SSB filters, and this has a marked effect on transmitted audio quality.
In fairness, the average communications user would probably not notice anything amiss. Certainly, there is normally no provision made for adjusting the carrier frequency in commercial transceivers; in any case, it is unlikely that conventional trimmer capacitors would be able to force excursions of two or three hundred Hertz in an otherwise well-designed crystal oscillator operating on its fundamental frequency around 2-8 MHz.

"Passband tuning" control is achieved by simultaneously shifting the second local oscillator and the BFO in one direction, giving the effect of moving the IF filter in the opposite direction.

The TR-7, however, features "passband tuning" - a facility using a VXO which effectively allows the user to "move" the filter across the passband of the receiver. What is more, this feature is used in the transmit mode to set the carrier frequency for the various types of transmission, eg. upper or lower sideband, a.m., c.w. etc. Preset potentiometers allow a wide range of adjustment of the carrier frequency (several kilohertz), and it is a simple matter to set the optimum frequency for a particular filter or mode. Incorrect adjustment is indicated on SSB by the received signal sounding either deficient in bass, or "woolly" and difficult to tune in (caused by the filter passing the other sideband while restricting the upper frequencies of the correct sideband.)

Successful Modification

For about a year, the TR-7s were used with their original, 2.3 kHz-wide, transmit filters. The non-technical operators quickly learnt to adjust both the received frequency and RF gain to obtain good results. A modification was made to each set to allow the playing of taped material from a Uher recorder, and this was highly successful in permitting the passing of actuality material, such as speeches, for "same-day" rebroadcast. Although the frequency response was naturally restricted, general opinion was that the lack of distortion and generally quieter background more than compensated compared with the public telephone network. In many cases, frequency response was actually better than that obtained from a telephone circuit!

Drake then introduced a 4 kHz filter, the SL-4000, intended again for a.m. broadcast reception. Once again, tests were carried out and this time the wider filter proved to have sufficiently sharp selectivity skirts to allow its use in SSB generation. In conducting these tests, the "pass-band tuning" feature was invaluable, permitting careful comparison between the original 2.3 kHz filter and the new 4 kHz one. The resulting passband for a correctly adjusted filter/carrier combination is approximately 350 - 4500 kHz: a great improvement over the usual SSB bandwidth and the public telephone network. Final adjustment of the carrier frequency is usually made by ear - this can be done in the receive mode with a little practice - and a note is made of the resulting carrier frequency for individual filters for future reference in the case of accidental disturbance of the preset trimmers. In the unlikely event of a setting drifting in a field set, a technician can use the TR-7's built-in frequency meter (another feature!) to make the necessary corrections without having to return the set to base.

Aerials

Mention was made earlier of the need for flexibility in the use of HF channels to cover the diurnal and seasonal changes in optimum traffic frequency. With unskilled operators, this requirement can pose severe limitations on the installation's effectiveness, because of problems with single- or at the most two- or three-channel aerials. The classic cheap "multi-frequency" aerial installation for a field SSB transceiver consists of paralleled dipoles, individually resonant at the appropriate frequencies and then fed at a common centre point. This arrangement is notoriously difficult to adjust in the field because of the interaction between the dipoles. In any case, changes in height, surrounding vegetation and so on can markedly affect the tuning and thus the effectiveness of the installation.

What is really required is a broadband aerial, covering at least the range 3 - 10 MHz, which can be simply fabricated and installed, preferably using locally available materials. There are surprisingly few designs which satisfy these requirements. Usually either bandwidth, simplicity or economy has to be sacrificed.

The TR-7, in common with many of the new breed of solid-state designs, is relatively intolerant of high SWR. Full output is obtained only when the aerial is matched to within 2:1 SWR, although the set will continue to operate with reduced output when the matching is much worse than this. Output impedance is specified as 50 ohms, into which over 100 watts (250 watts pep input) is developed.

Of course, an aerial tuning unit will take care of matching. For a skilled operator, manual adjustment is almost second nature and it is perfectly reasonable to expect tune-up procedures to include aerial tuning. The TR-7, however, needs no tuning adjustments - no plate controls, no RF pre-selectors. For unskilled operation, it would be a pity to throw away the ease of operation by fitting an aerial tuning unit. Automatic ATUs are prohibitively expensive and complicated, even if power is available for them, and fixed-channel versions still need lengthy setting-up and checking.

A search through a considerable amount of technical literature eventually turned up what was described as "A Broad-band Travelling Wave Dipole" (3). Developed in the early 1970s, this aerial was claimed to have a relatively constant centre impedance of about 300 ohms over a wide bandwidth. A version was built, using the instructions in the article, which didn't work - perhaps just as well, because the cost of materials, including copper piping, was frightening! However, results were promising enough to merit further development.

Eventually, after a lot of trial and error, the version shown in the Appendix evolved. It differs from the original by the addition of a third "tramline" down the middle. This simple modification improved the performance of the aerial dramatically, and resulted in a highly repeatable, non-critical aerial, which with the addition of a 6.25:1 balun showed less than 2:1 SWR from below 3 MHz to over 20 MHz. An important advantage was that the aerial could be constructed very cheaply (see diagrams) - typical cost including the home-made balun and supporting ropes was below P100 (US$120-odd). Two cheap steel-pipe masts, guyed, at a further P150 or so, completed the installation. An interesting discovery was that aluminium strip, as sold for trimming the fronts of kitchen units and the like, came in 1.8 m lengths at P1.00 per length, and was ideal for forming the tramline spacers in the aerials!

Power Supply

In general, no public electricity supply is available at the Information Offices in the Districts. The TR-7 is essentially a 12v transceiver, requiring about 1.8 amps on receive and up to 25 amps on transmit. An ordinary heavy-duty car battery is used to provide this, with the usual attention to low-resistance contacts and so on, and a rudimentary battery-reversal protective relay. After abortive attempts to get the local Central Transport Organisation depots to charge the batteries on an exchange basis, the Department decided to invest in solar panels to ensure that the transceivers would be available for use at all times. So far (about a year in the longest case) this has been entirely successful. Panels from a number of manufacturers are in use, and there appears to be little to choose between them. A typical installation uses a 33 or 37 watt panel designed for 12 volt battery charging, although no voltage regulation is used. The reasoning behind this is that the cheap car batteries used will almost certainly deteriorate from poor maintenance and the misuse of constant trickle charging long before they deteriorate from overcharging! Expected life for the batteries is around two years.

Accurate Resolution Of SSB Signals

Initially, doubts were expressd concerning the ability of the operating staff to accurately resolve incoming SSB signals. Many SSB transceivers are already in use in Botswana, and a quick scan of the HF bands revealed appalling netting in many cases. It seems that, provided the recipient can make out what is being said, the "Donald Duck" sound is simply accepted as an unfortunate side-effect of SSB.

Naturally, accurate tuning was necessary if the received material was to be rebroadcast. It was a pleasant surprise to find that, after an initial training period, the news staff were able to hear mistuning and correct it using the usual RIT controls fitted to most commercial units. In this, they were helped by the digital display on the TR-7, which they seem to believe more than their own ears in some cases!

When the 4 kHz filters were introduced, the operational staff commented that it had become much easier to determine the correct tuning point with the extended bandwidth. There may be a lesson in this. No such improvement had been remarked upon during the initial trials with a 6 kHz filter, but this was probably because of the intruding unwanted sideband noted earlier. It is also possible that they had not developed "communications ears" at that stage.

Costing

Some approximate prices are given below, converted to US dollars:

Drake TR-7 transceiver, fitted with digital display, extra filters, microphone, external loudspeaker, noise blanker and accessory board permitting operation on fixed channels:	$1 900 (FOB)
Materials for constructing broadband aerial, including masts and guy ropes, balun and coaxial feeder cable:	$ 400
Solar panel, 33/37 watt, 12v	$ 500
12v 80 Ah car battery or similar, plus cables, clamps, battery reversal protection relay etc.	$ 100
Freight on TR-7, etc (15 kg maximum)	$ 300 (max)

Typical installation charge (excl. labour etc.)	$3 200

Conclusions

This paper has described the use of TR-7 SSB transceivers in Botswana for communications and the passing of material for subsequent broadcast. It has been shown that a typical installation may be made for just over US$3000, and the benefits realised in terms of improved quality and news coversage are great. A low-cost broadband aerial has been developed and is described in detail.

References

(1) "Report for the Government of Botswana by the Consultant on Information and Broadcasting Services" - C.N. Lawrence, Commonwealth Fund for Technical Cooperation, December 1978.
(2) TR-7 Service manual; leaflet supplied with accessory filters.
(3) "Amateur Radio" - journal of the Wireless Institute of Australia, PO Box 150, Toorak, Victoria 3142, Australia. (April 1974 edition). Also see appendix to this paper; other references can be found in the ARRL handbook.

Appendix: . . .

Back to Commonwealth Broadcasting Association . .

Botswana index . .

Background In common with many other developing countries, Botswana has internal communications difficulties. Roads are long, and frequently poor in the rural areas away from the railway line. The internal telephone and telecommunications network is hardly developed at all; "telephone quality" is at best adequate, at worst unusable for rebroadcasting. It is hardly surprising, therefore, that a frequent criticism of the national radio station has been its lack of "Botswana news". A recent report (1) quoted a resident of Serowe as asking "Why don't you just call it Radio Gaborone, because that's what it is?" Clearly, the station had to place high priority on the rapid passing of news from the Districts to the capital.
A Uher tape recorder from the late 1970s	Radio Botswana is part of the Department of Information and Broadcasting, and traditionally the Information Officers in the various Districts have had, as part of their brief, the collection of news and information from their areas. Usually, the medium is the ubiquitous Uher tape recorder, upon which interviews are recorded and voice-pieces composed. The tape would subsequently be sent on to the capital for eventual incorporation into bulletins, newsreel programmes or current affairs slots. An alternative was the use of the public telephone network, in which case a prepared script would be read down the telephone and recorded by Radio Botswana. In the unlikely event of its quality being adequate for transmission, it would be aired forthwith; alternatively, a transcript would be prepared and used accordingly. In either case it was virtually impossible to use up-to-date quality material. The delay might be anything from 24 hours to a few weeks!
The Technical Problem What was needed was an alternative to the public telephone network which was fast, reliable and capable of sustaining reasonable quality. Cost and ease of operation were also very important. The obvious solution was the use of radio. Unfortunately, the distances involved precluded VHF for all but the most local Information Offices. HF, on the other hand, is subject to all sorts of interference, and if SSB is used, usually of poor quality and limited bandwidth unless highly sophisticated equipment is available. The choice of frequency can be critical throughout the day to optimise signal-to-noise ratio, and this implies a certain amount of frequency agility in a remote installation, which can cause considerable problems with aerials or tuning units. Not least among the technical obstacles is the provision of power for radio transceiver equipment. All these disadvantages usually restrict the use of HF SSB radio to basic communication. Commercial SSB equipment is normally fairly reliable, but powering, aerials and above all audio quality frequently reduce its usefulness for broadcast purposes to a minimum. Nevertheless, it was decided that Radio Botswana should investigate the use of HF SSB transceivers, attempting to find ways round the problems mentioned above. To a large extent, this has been successful. In the process, a low-cost aerial has been developed, and some interesting information derived relating to the tuning of SSB signals by non-technical staff.
The Transceiver From the start, it was realised that the correct choice of SSB transceiver would be crucial to the success of the project. To this end, a loose specification was drawn up and tenders were invited from various commercial manufacturers. [The specification included: 2.0-18.5 MHz coverage; a.m. detector; 100W output; transmit bandwidth to 5 kHz; 12v operation; reception of medium wave; 6 crystal TX channels; VFO receiver; high-quality audio stages; service manuals and spares.]
Upon evaluation, it was realised that the manufacturer most closely approaching the specification was M/s R L Drake Co. The TR-7 transceiver offered a choice of receiver bandwidths, operation over the whole HF band, broadband output circuitry, a general coverage receiver, 12-volt power supply capability and good serviceability.	The Drake TR-7 transceiver
One of the the most difficult requirements in the specification was a "wider" transmit bandwidth. Reaction from conventional manufacturers to the request for a transmitted bandwidth greater than the standard 2.7 kHz or so ranged from complete incredulity to guarded pessimism. Most commercial SSB equipment is rugged and otherwise suitable for communications, but this restricted audio bandwidth seemed to be a stumbling factor for Radio Botswana's purposes.
The TR-7 IF selectivity board has space for four filters	The Drake TR-7 has a 2.3 kHz wide filter fitted as standard. However, it is mechanically possible to fit any of a range of filters (intended for receiving only) into the "transmit" position. M/s Drake specifically warn against doing this (2) so naturally, being good broadcasting engineers, we tried it.
First attempts were made with the Drake SL-6000, a 6 KHz filter intended for broadcast reception. Audio quality was of course greatly improved, but unfortunately the filter's skirt selectivity did not adequately eliminate the unwanted sideband. The net result was a signal which sounded "burbly" and was difficult to tune in. Some experiments were conducted using the transmitter in the "one-sideband-plus-full-carrier" (A3H) mode. This gave good results, using the excellent low-distortion a.m. detector fitted to the TR-7, but signal-to-noise ratio suffered for distant stations. (However, it is intended that further work should be done on this to allow a.m. transmission from OB vehicles working within about 300 km of the capital.)
In the interests of establishing the network, the transceivers were restored to their original transmit bandwidth (approximately 300 Hz - 2.6 kHz) and installed. Even like this, the improvement over the public telephone network was startling.
SSB Generation At this point, it is interesting to dwell on the method of setting the carrier frequency in the TR-7, relative to the SSB filter, and the effect this has on the quality of transmission. Most commercial transceivers generate SSB using a fixed intermediate carrier frequency (crystal controlled) which is accurately placed on the side of the selectivity curve of a fixed frequency crystal or mechanical filter. The SSB thus generated is then heterodyned to the required output frequency. A disadvantage of this method is that, once the manufacturer has decided on the position of the "carrier" relative to the filter, there is very little the user can do to change it, short of purchasing a range of crystals around the nominal carrier frequency and plugging them in one by one until satisfied. This stems, of course, from the very reasons a crystal was chosen in the first place - ie. high frequency stability and resistance to "pulling" by external circuitry. Lest it be thought that no adjustment is necessary, it is worth pointing out that a difference in "carrier" frequency of only tens of Hertz can have a noticeable effect on transmitted quality. (NB: this refers not to accuracy of subsequent resolution of the signal at the receiver, but to how far down the slope of the filter the carrier is placed). Unfortunately, manufacturing tolerances may mean differences of a couple of hundred Hertz between optimum carrier frequencies for nominally identical SSB filters, and this has a marked effect on transmitted audio quality. In fairness, the average communications user would probably not notice anything amiss. Certainly, there is normally no provision made for adjusting the carrier frequency in commercial transceivers; in any case, it is unlikely that conventional trimmer capacitors would be able to force excursions of two or three hundred Hertz in an otherwise well-designed crystal oscillator operating on its fundamental frequency around 2-8 MHz.

District Radio News Reporting - A Low-Cost Approach

by D. W. Harris, C.Eng. MIEE Deputy Director (Engineering), Radio Botswana