see copyright notice. Page created 17-May-2001 updated 7-Nov-2019. Use the button groups above to navigate quickly around the site.
I decided to build my first TV during my 5th form year at school. Fired with enthusiasm for those new-fangled transistor things (largely thanks to the wonderful Philips Electronic Engineer kit) I nonetheless wasn't going to rest until I'd made a real project - one with valves in it. Practical Television Magazine published its "625-line Receiver for the Constructor" design, by Keith Cummins, between March and July 1970. It was an all-valve circuit except for the tuner and IF strip (both commercial "surplus" items) with a 20-inch monochrome CRT type A50-120W/R (£10.15.0 plus 15/- carriage and insurance!)
The mechanical design comprised a folded aluminium chassis, with most of the wiring on tagstrips underneath the vertically-mounted valves (helps to avoid grid sag, dontcha know) with two vertical side-plates to support the CRT and another for the tuner and speaker. My father had a contact in the sheet metalwork business, who was kind enough to make the large holes for the valve bases, speaker etc. leaving me with only fixing holes to worry about. Even more fortunately, I was learning woodwork at school and had been asked to come up with a suitable project; a TV cabinet was ideal! It was finished in wood-effect Fablon (aka sticky-backed plastic), and later acquired an orange colour-correction filter, as used in all the best production galleries to correct the colour temperature of standard monochrome monitors.
A shrewd observer might notice a large black panel where the tuner buttons
ought to be. That's because in 1979 I decided to rebuild the whole thing as a
solid-state monitor, largely it must be said because I wanted to be able to claim
that I'd designed as well as built it. So the rear view is now
very different to that of 1970, the only surviving original components being the CRT
and the speaker. The latter is in fact redundant, but I left it there to cover the
hole. You can see a couple of the original valve base holes toward the
The "new" circuitry is unique and highly unusual. All the low-level signal processing and timebase generation take place in the metal box on the right. It's crammed full of discrete analogue stuff and CMOS logic, which did a rather more thorough job than strictly necessary, but then I'd been studying broadcast video techniques at the time!
The "HT" power supply is only 24V, which made the line output design quite challenging. To get good switching efficiency, I used two BU208A transistors in parallel. Both the line drive and the line output transformer are home-made. The latter is visible halfway up the left chassis plate; the coupled overwindings are to reduce leakage inductance to obtain the correct harmonic tuning. This is actually my second LOPT design, its predecessor caught fire after a few weeks and had to be rewound with more layers of paper between the EHT windings! The EHT is approx. 20kV from a colour TV tripler module. An 800V output from another pin on the tripler drives a rather hairy two-transistor regulator circuit to provide +450V for the CRT first anode.
Just visible above the enormous capacitor on the left is an antique neon lamp. It serves the dual purpose of a large, diffuse pilot light and a critical part of the switch-off spot suppression circuit. But the strangest area of all is the field output stage, directly under the CRT base, which includes an unmarked thyristor-like device that I found in a junkbox. The circuit was arrived at rather empirically and I was always worried that I didn't quite understand why it worked. I hope it never fails because I could only get it to work properly with that particular component...
It was bound to happen. In April 1972, Television Magazine (having ominously dropped the Practical from its title) started a series of constructional articles for "The Television Colour Receiver". I was looking forward to a few spare months before starting university and my parents wanted a colour TV. So began a mammoth 2-year project which I (and I suspect some of the magazine staff) more than once regretted having undertaken - but at the end of the day we got there. The splendid wooden case (again covered with Fablon) was built by my father while I was away at College. I remember the quizzical expression on the face of a College Porter as he handed me a scribbled telephone message from home: "The Tube Has Arrived!" Said Tube is a Genuine Mullard A56-120X, 22-inch, Made in Britain, Those were the Days.
Following a protracted period of testing, modifying and tweaking,
I installed the finished product in the family lounge, where it performed quite
well for a number of years. I must say though that it needed fairly regular attention
to maintain what I considered a watchable picture. The convergence - always a tricky
business on delta-gun CRTs - drifted, along with the picture dimensions
and black level, according to mains voltage changes. The solution of course would be to
stabilise the power supplies, but by then the set was in regular daily use
and would have been sorely missed for the duration of such an upgrade. So in the
end, when I moved into my own house, I bought my parents a new colour set (they had after
all paid for the home-made one!) and took my baby with me - and it remained in use as my
main TV until 2012.
As with the monochrome set, I've made some rather drastic changes over the years - but there's a lot more of this one still in its original form, including the PCF802, PL509 and PY500A valves in the line timebase. The PCF802 line oscillator is now however just a line drive amplifier, fed from a CMOS PLL in the tobacco tin lower right. All three valves and the CRT survived more than 35 years of daily use, which I hope explodes any myths that valves are unreliable creatures.
The required PSU regulation has been added (big heatsinks on left). What's now missing
is the entire receiver section - tuner, IF, PAL decoder and audio stages - which
originally occupied the left side panel (from the rear). The IF board, in particular,
was in my opinion a bit of a disaster; it just about worked with a reasonably strong
signal, but was hopelessly unstable at high gains. I speak with some authority here,
having spent a great many fruitless evenings trying to improve the damned thing's
behaviour! Not that it was that much worse than many other transistor IF strips
of its day; the innovation that was needed was the Surface Acoustic Wave Filter, and
when these appeared in domestic tuners I happily scrapped the old board, with its
temperamental bunch of stagger-tuned critically-damped amplifier (oscillator?) stages,
and built the external tuner pictured below left.
The black-fronted box houses a UHF tuner, SAW IF strip, detector, and a fairly serious transversal equaliser to give a nicely shaped pulse-and-bar on the video output. The board perched on top is the prototype of a PAL decoder that I designed for commercial production. Determined efforts by broadcasters over the years to fill the Field Blanking Interval with little twinkling dots obliged me to add the small piggyback board sticking up towards the rear; it contains a monostable that re-establishes correct field blanking.
In due course TV broadcasting went digital, and I replaced the tuner with a set-top box. However, the writing was on the wall for 4:3 displays, especially old-style CRT ones. Watching the 42-inch LCD/LED receiver that replaced my home-made one, the thing that impressed me most wasn't the picture size, resolution, or lack of PAL artefacts. It was the fact that the picture stayed the same size regardless of its brightness!