Ruined Radio - The AR88

I have just clicked on what my PC chums call a Radio Button. This virtual version has none of the feel of a real button on a real radio but it has sent an e-mail bulletin to about thirty customers. While I wait for them to get back to me, here's a few notes on what atrocities a young man can rain down on a classic AR88D.

It was Clive's. Clive was the least anoraky of the Development Guys. I think it was marriage that meant he and it would never see each other again. Paid around 45 quid? Sounds about right.

The only place for it was on the radiogram cabinet, the best speaker for it was the one inside. Disconnecting it from the record-player, even then, declaring vinyl dead and hooking it into the back of the AR88 resulted in a mere 2.5 to 15 ohm mis-match. It seemed rather quiet, certainly for a rock-loving listener to Radio Northsea International.

On that first evening, the AR88 was slid out of its oak case, on the hunt for easy mods. The output was a 6K6G, a valve I'd never heard of. It was capable of about 2 watts. Quite healthy - but for me, never enough. A 6V6 was an easy solution. If it seemed OK, try a KT66.

This was the one that went into heavy anode current and cooked the output transformer. That caramelised smell of expense. A sorry lesson was well learned.

Was there a fault on my AR88? It seemed to work best with a 2uF capacitor from AVC to ground. This was available by a connector on the back of the thing. It motor-boated without it. Without understanding what I was doing, I must have extended the AVC time constant to about a week. I was happy because it seemed quieter now.

The DIODE OUTPUT was fed to a VORTEXION tape machine to record Kenny Everett from the new Radio 1 on 247 metres. As this was a feed direct from the detector, the recorder neatly shorted out the AVC. I was happy because it sounded louder now.

My AR88, like so many others, had no S Meter. The over-long long-wires so swamped the front-end, the ANT TRIM seemed to have no effect. Metering was needed. At first, this was a traditional bridge circuit sensing changes in the cathode voltage of an IF stage due to the AVC. Not AGC, you will note. The AR88 had no schematic or manual. Every signal-handling valve cathode was tried. Some moved the meter, others didn't. When the meter did move with signal level, it was affected by the RF GAIN setting.

With no documentation, how was I to know which stages were controlled and which, if any, weren't. With no knowledge, what was I doing inside it anyway?

Meter calibration was carried out by finding the strongest station, then inserting a series resistance chosen to just ease the pointer off the end-stop. None of this 50uV for S9 at 14.2 Mc/s stuff.

Although every set I ever saw had this phenomenon of the S Meter gracefully rising as the RF GAIN was reduced, for some reason this could not be tolerated in my AR88. The cure was to replace the double-diode detector (a 6H6?) with one that had a triode in it to drive a right-hand-zero milli-ammeter in the anode circuit. A rare 6SQ7 was chosen, on the engineering premise of having one in the junk room. Not a junk box, you'll note - a junk room. AVC was applied to the grid turning the valve off, causing the meter to rise.

Calibration? This was done by shorting the aerial to deck and padding the anode load with a resistor until FSD (full-scale deflection, sorry) was achieved. FSD on a right-hand-zero meter gives you a traditional left-hand zero. And you can't end-stop it. All the AVC in the world would only cut the triode off, reducing the anode current and the meter to zero - on the right-hand side of the scale. And turning down the RF GAIN had no effect on the reading. This was a valve-voltmeter reading the AVC directly, sort of.

The meter was from the Derby Radio rally. 20p got me one with an uncalibrated scale, the first two-thirds in white, the last third in bright red. In the centre was an authority-lending legend, AM. The military logo used to put the fear of God into visiting listeners. They really thought that prolonged operation in the red would cause some dramatic failure. The sort of blow-out that puts inpromptu serving hatches into bedroom walls.

Thirty years has gone by since then, along with a career in Instrumentation. I have to say this was the most elegant indicator I ever came up with - totally uncalibrated, of course, but really effective.

A victory like this gives you the confidence to really go to town. Front panels are drilled to take switches for product detectors. A cardinal sin. The rectifier valve is changed for silicon diodes with no current limiting. It seems to hum a bit now, so a weepy electrolytic capacitor is slapped across the HT. While I answered the phone, that capacitor failed. A direct short across the power supply burned out the mains transformer, writing off a classic radio.

Why didn't the fuse go? Ah, well, an easy upgrade you can do as a child is to fit higher-value fuses. This stupidity cost me a radio I dearly miss now. And the motto? If you don't understand it - leave it alone. Enjoy it for what it is - rather than what you were hoping for.

There were no radio buttons on an AR88. The one on this PC is clicked with a mouse. A curious second-hand experience. I am doing it now to see if anybody has responded to my last e-mail. Apparently, I have a Fatal System Error at Module E004765007-00014. Never got those on my dear departed AR88.