?? G1MFG?s ATV Modules Reviewed
By Brian Kelly
Anyone reading the last issue of CQ-TV or monitoring chat on the ATV frequencies will be aware of the ?Cheapest ATV transmitters and receivers in Europe? claims of Giles Read, G1MFG. Of course transmitter kits and convertible satellite receivers for 23cm have been around for some time but Giles has also introduced 13cm units, a band that has lacked use because of the lack of ready-made equipment. Not only are modules for both bands now readily available but their prices are incredibly low; in fact with a pair of his boards it is possible to transmit and receive for under £90 per band. Giles recently loaned me a set of modules to ?play? with so that I could assess their quality and usefulness to the ATV community. I have tested both the standard and ?gold? version of the 23cm receiver to see how much improvement is made by applying all the documented modifications. The ?gold? version is an already upgraded version of the standard module. The modifications are all described on the G1MFG web pages for those wanting to upgrade receivers themselves. My apologies that this review didn?t make it in CQ-TV 191 but we (the Editor, Giles and myself) decided that the similarities in the 23cm and 13cm modules made it impractical to review them separately and by the time all the modules arrived, the print deadline had slipped by.
The modules are very similar. The transmitter pair look identical, the only differences are hidden inside the metal can. The receivers look to have almost identical circuitry but the PCB layouts are different and of course, the inner workings of the tinplate cans are different. Each module has five connectors. On the PCB are three PCB phono sockets for attaching the video, left and right sound channels and a 2.1mm DC power connector. On the metal can is an RF connector. The 13cm units use an SMA type while the 23cm ones are available with SMA or ?F? to order. Frequency selection on both transmitter modules and the 23cm receiver is achieved by setting an 8-way bank of switches. The 13cm receiver utilises a single push-button that cycles through 15 pre-set frequencies and a bank of four red LEDs to show which is currently selected. The last setting is remembered when power is removed. All the boards are low profile, the highest component being the can, which sets the overall height at about 20mm (¾ in.) when measured from the bottom side of the PCB. The receiver board outline is approximately 60mm x 125mm (approx. 2.4 x 5 inches), the transmitter approximately 50mm x 160mm (approx. 2 x 6.3 inches). A single 2.5mm mounting hole is provided at the ?socket? side of the board, the opposite side can be secured by the bush on the antenna connector. There is space on the PCB to drill additional mounting holes if needed.
Transmitter RF Performance
Checks on spectral purity were carried out using the 23cm transmitter; unfortunately, my spectrum analyser top-ends at 2.1GHz so similar checks couldn?t be done on the 13cm version. I am assured that the output is clean up to at least 10GHz.
Nothing that shouldn?t have been there was seen and even mixing products from the two sound carriers were below the noise floor of the analyser. The photographs show sweeps of the unmodulated transmitter output and the signal spread when fed with standard colour bars. Sweeping with a wideband receiver, although not able to take measurements, did not reveal any unexpected emissions. The carrier frequency was quite stable and drift was more than acceptably low. With the synthesiser set to 1316 MHz the frequency was measured at 1315.990
MHz when cold and 1315.979 after an hours operation. The sound sub-carriers were measured at 6.005MHz and 6.506MHz on both transmitters. A modification, to be posted on Giles? web site will explain how to disable the 6.5MHz one; alternatively, with care it can be re-tuned to a lower frequency.
Receiver RF performance
Based on ?off-air? tests, both receivers were exceptionally sensitive. The 23cm receiver easily beat my normal receive system, even with its pre-amp attached. Efforts to attenuate the 13cm receiver input to make it appear deaf, failed to stop it picking up a signal generator across my workbench. A weakness in both receivers was direct IF breakthrough. With an antenna directly attached, both picked up severe interference from UHF domestic TV transmissions. Admittedly, my QTH is 650 feet ASL and line of sight to two main regional transmitters! Sweeping a signal generator across the IF frequency (480MHz) also revealed a window of about 30MHz where leakage through the input socket managed to reach the demodulator. In normal use, a filter or tuned pre-amp in line with the antenna would eliminate this problem. I understand a simple modification reduces susceptibility to direct breakthrough.
With audio fed into the transmitter and monitored through the matching receiver, no obvious distortion was observed. Feeding a signal into only one channel and monitoring for breakthrough into the other showed that isolation was very good. The overall audio quality was more than acceptable. Whether having two channels is of any use for ATV applications is debatable but they are there anyway. The line level inputs at the transmitter returned almost the same levels at the receiver output socket. For use with a microphone, it would be necessary to boost the audio level before feeding it to the transmitter, however it matches well to the output level of most camcorders. There is no separate adjustment for audio deviation in the transmitter or gain at the receiver; this can easily be added in-line with the audio feeds if they are needed.
This is where the performance limitations and the differences between standard and ?gold? receivers really came to light. To be fair, for the price, their performance is good and Giles is actively developing ways to make them better. Most of the improvements are already implemented on ?gold? units. All the modules are derived from origins in the remote video link market where a higher modulation index is used than is normal for ATV. Most of the problems stem from using lower deviation than they were intended for. A level control on the PCB allows the video signal to be reduced to normal levels. The control does not change the sound sub-carrier levels, which are approximately ?30dBc, measured with no modulation applied.
Advancing the receiver gain control beyond its optimum setting results in quite bad video distortion with the picture looking very odd, in fact almost negative. On the standard units the control is already near top position but changes made on the ?gold? 23cm receiver make the adjustment far smoother.
The overall video frequency response through the 13cm pair was quite good and more than acceptable for ATV use. The standard 23cm pair however, exhibited a rather poor HF response with received pictures looking noticeably smeary; I imagine that running the video amplifier at nearly maximum gain is limiting its bandwidth. Most of the modifications made on the ?gold? units are for raising the HF end of the baseband spectrum and indeed the ?gold? pictures were dramatically improved. In fact with the recommended TX pre-emphasis applied there was slight ringing on sharp contrast edges, this shows in the photograph where overshoot can be seen on the rising edge of the colour bars. There is some degree of tilt on low frequency video signals, partially from the TX PLL circuit and partly from the receivers coupling components. In practice this isn?t likely to cause a problem on the air.
Communication with other users of these modules has confirmed my findings about the frequency response problems. The diagrams show the overall response of the 13cm, standard 23cm and ?gold? 23cm modules respectively. The top trace is the signal fed into the transmitter and the bottom trace shows the receiver output terminated by a 75-ohm resistor. The video frequency response graphs demonstrate the HF roll off and how it is much improved on the ?gold? modules. Note that the line showing the effect of TX pre-emphasis has been shifted to take into account the 6dB attenuation the network incurred and then normalised at 1MHz.
One potential problem revealed by the response tests was that the sound carriers were present in the video output. The units do not appear to have any subcarrier traps installed. In normal use this would not be a problem and the addition of a de-emphasis network would almost eliminate the breakthrough.
The units all run from DC supplies between 12V and 15V, on-board regulators ensure that the circuitry sees a constant voltage. The regulator on the TX modules is rated for 12V output; it will therefore work best with a higher input voltage than this. Current consumption for the 13cms modules was 120mA for the transmitter and 160mA for the receiver. The 23cm transmitter drew 130mA and the receiver 300mA. The higher current demand of the 23cm receiver is, I am told, because of a different on-board regulator design. All measurements were taken with a 13.8V supply with the units at working temperature.
For the home or portable ATV operator, these units present exceptional value for money. They are not up to BBC broadcast quality, but you would need to spend significantly more to achieve the same results elsewhere. I would seriously recommend users of the 23cm receivers to apply all the known modifications and for potential new purchasers to consider buying the already modified ?gold? version. The modifications do require a steady hand as tiny surface mount components are used on all the modules. Giles is actively improving their performance all the time and reports on his experiments are posted on his web site, http://www.G1MFG.com along with a wealth of other ATV items. As an almost ?turnkey? method of getting on the air, they are undoubtedly a simple and cost effective way forward. Would I buy them? Based on my findings while testing the modules, I have already placed an order!
Giles can be contacted by email at: