Back in 1991, commercial NICAM products were well and truly out of my price range. Inspired by the Arcam tuners available at the time, I decided that a DIY stand-alone tuner was the best approach.. In the long summer between leaving school and starting university, I produced this:
What made this was possible was a combination of factors. First, at the time, an uncle was working for Ferguson. Under development was a sound panel for an upmarket TV chassis, and I was given a spare pre-production board. Incidentally, these boards became available in the small-ads of Television magazine a few years later. Here's a picture of one, minus the heat sink and a few other components...
The big problem was that I had no diagrams or documentation for the board, so it sat around until Television ran a series of articles on NICAM. These mostly comprised lots of details on how the NICAM transmission scheme itself worked, but the series ended with a brief look at the the chipset used on this board. That was all the data I had, but it was better than nothing!
I spent a while studying the articles and the board and eventually managed to get some sound out of it! The board takes in the baseband signal, and drives loudspeakers via the on-board power-amps - it also deals with external sources, and even has a graphic equaliser. Sadly, most of these functions are controlled by the I2C bus, but as I had no way of generating I2C, I was unable to make use of them. Remember - this was 1991! But, I was able to identify the DAC, and sniff the audio output from that...
The next problem was the baseband signal. I was experimenting with an old tuner/IF panel from a VCR but the SAW filter was, quite reasonably, rolling off the 6.552Mhz NICAM carrier, so operation was somewhat erratic. But at least I was able to get enough of a signal to prove that the NICAM decoder was basically working...
The Maplin NICAM kit was available by now, but even that was rather expensive. However, I noticed that the UHF tuner was available as a separate kit and was reasonably priced as well (around £40). That resulted in perfect NICAM operation from the lashed up decoder.
Having got this far, I was confident enough to take things to the next level. Most of the electronics on the Ferguson panel wasn't required, so I decided to 'modify' it slightly. A cut 'n' shut operation resulted in this:
It looks a bit crude, but it works well enough. It runs off a single 12 volt supply, accepts a baseband 6.552MHz signal, and outputs stereo sound via the 7-way connector (top-left)...
The reconstruction filters were taken from the main PCB and mounted on the Veroboard, along with a simple audio pre-amp. This was originally on the reverse of the PCB, implemented with SMD components - luckily, they were quite large by today's standards, so reverse-engineering the circuit was pretty easy.
A surprising amount of electronics is required to interface this with the real world. A control system is required to enable the unit to be operated, and an interface board is required to provide NICAM/FM switching. Other functions include I/II channel switching (for bi-lingual broadcasts in theory, but never really used in practice), AV input selection, remote control, tuning, etc, etc...
Active view of the NICAM Tuner
A bit of JavaScript interactivity...
Conclusion
The unit was in use until I got $ky Digital. Even when I was able to buy a NICAM televison and a Hi-Fi stere VHS machine, I still preferred to use the tuner because I preferred the sound and picture quality (obviously, I'm extremely biased!). The Maplin tuner is a rather high quality device, and compared well to commercially available devices. Unusually, it uses separate IF processors for sound and vision, which results in excellent performance. Indeed, it is possible to get good NICAM sound when the picture is almost unrecognisable...
I spent quite a lot of time working on the 'user-interface' to get the details right. For example, muting was critical when switching on and changing channel. Some years later, I met someone who'd built the whole Maplin kit, and discovered that they'd not been quite so diligent - it made the most horrible noises when you changed channel!
I'll be installing it somewhere else soon, possibly in the workshop. Before that, I need to fix a couple of minor problems - when I moved from Surrey, I noticed that it suffered from a minor hum-bar: the mains voltage was obviously slightly lower here. I was trying to keep the regulator dissipation to a minimum as this unit sat underneath my Musical Fidelity A1, so I added series resistors to take some of the heat from the voltage regulators (a good technique when current demands are fairly constant). Shorting out the 1.2Ω resistor makes the problem go away. I'm also planning to replace the standard 7812 regulators with low drop-out version and also I'll replace the reservoir capacitors, which have had more-or-less continuous use since 1991.
Another minor problem is caused by the audio stages being powered from a single 12 volt supply. My S-VHS machine is able to generate enormously high signal levels with certain pre-recorded tapes which can cause clipping. The answer would be to generate a negative supply to power the op-amps and analogue switches from it. There's enough space on the board to allow that relatively easily. However, as that VHS machine is no longer connected to it, I suspect I'll never get around to it...
Finally...
Several people have emailed me asking for schematics, parts lists, assembly instructions, etc. Just to be clear, this is not intended to be a project!
The first-generation NICAM chipset, the Maplin UHF tuner kit, the remote control and SAA5012 decoder chipset, and M491 tuning microcontroller are all obsolete. Indeed, NICAM itself is now obsolete in this country thanks to the fast take-up of digital TV, followed by the switching off of the analogue broadcast signals. When I built this - more than 20 years ago now - I honestly never expected it to outlast the broadcast standard!
Additionally, the CMOS logic used in the control board is not worth trying to copy - these days you could replace all 13 ICs with a single PIC!
It's possible that scans of the schematics might find their way onto the site at some point, but they would be for academic interest only - there are much better ways to do things these days...
- Return to:
- Audio Projects »