This big old car radio came from
my son's Dodge Monaco. He had reported that it worked but the
audio was extremely faint. A search of the Net failed to come
up with a free schematic so I planned to proceed without one
Having worked out the connections (actually quite simple as there were only four wires), I connected it to a 12 volt power supply and a loudspeaker, then plugged in a signal generator to its aerial socket.
The loudspeaker produced the faintest possible hiss with the volume control turned to maximum and, after tuning the set to 1600 KHz then setting the signal generator to the same frequency and nearly half a volt, I heard the faintest signal.
The first step that I had in mind was to align the radio. This meant adjusting the RF coils to match the dial settings and then determine the IF and confirm the IF amplifier was working properly.
The set uses a small printed circuit board, on which are lots of discrete components dating from the 1960s. No valves fortunately, just ancient germanium transistors. The resistors were the old carbon variety that often seem to drift with age and not many electrolytic capacitors visible . just one 50uF that seemed to have leaked onto the circuit board and a large metal-cased type that looked quite respectable.
I bought myself an ESR capacitor tester last year, and this has proved invaluable in detecting bad electrolytics. I wish I'd bought one years ago, but never got round to it. I removed the 50uF and checked it. Capacity wasn't too bad but the resistance was off the scale. I fitted a new capacitor and the signal improved significantly. The old metal-cased capacitor measured OK with an ESR of around 0.05 ohms.
Now that there was enough gain in the set for alignment I looked at the possible adjustments. There were two large ceramic trimmers and a pair of metal-cased coils with iron slugs. Puzzling I'd have expected more coils. After searching for the intermediate frequency and drawing a blank, even with the signal generator dishing up a volt across the range 100 to 500 KHz, the penny dropped. This radio was not a superhet, it was a TRF set.
US receivers over the years were often TRF types using lots of stages of amplification and no reaction, relying on raw RF gain rather than feedback to produce acceptable overall gain. This old car radio made by Bendix followed the old technique. This is the first transistor TRF receiver I've come across.
Alignment requires the iron slugs to be adjusted at the low frequency end of the scale and the capacitor trimmers at the high end. This achieved, the set began to receive stations with a long aerial connected.
The next step was to measure the old resistors and see if any had drifted too high for comfort. This problem will affect transistors which rely on relatively critical bias settings, whereas in an old valve set, resistor values are not too critical. Satisfactory germanium transistor operation is usually more susceptible to drifting resistors than newer silicon varieties.
Unfortunately early germanium transistors tend to have pretty low impedances and measuring associated resistors is not easy in-situ, however I did find a 68kohm that measured 75kohm, and changing this resulted in another 3dB of gain, still nothing like enough.
The layout of the circuit board was pretty confusing and it was next to impossible to identify the component connections let alone trace the circuit. Other than removing each resistor in turn and measuring it out of circuit what could be done?
There was a clue I noticed that disconnecting the power and reconnecting it resulted in a huge instantaneous increase in gain. I hit upon an old method I must have tried countless times donkey's years ago. I picked out a new 220kohm resistor and connected one end to chassis. I then touched the ends of each resistor in turn to see the effect. Every time resulted in nothing or just a small drop of gain.
Next was to connect the test resistor to 12 volts and repeat the performance. This time, when I touched the end of a 120kohm near the good electrolytic I was rewarded with a huge increase in gain. By trial and error I determined that 82kohm produced the optimum result (making it lower resulted in no further improvement). I soldered this in place. Now the set was working well and background hiss was quite apparent.
I disconnected the supply and reconnected it. The volume doubled for a few seconds before dropping back. I continued with my test resistor and was rewarded a second time. This time a 68kohm resistor was the culprit. I connected another 68kohm across the suspect one and the gain shot up for a second time. and this time didn't change when reconnecting the power.
A final tweak to check alignment, then a tweak at the aerial tuning capacitor which needs to be adjusted for the specific car aerial, and stations roared in across the band. In fact with a long aerial I could hear all sorts of low frequency signals over-riding the tuned circuits. With a short aerial these vanished and left excellent results across the whole medium waveband. Of course long waves are not used in the USA so the set has only a single waveband, no doubt one of the reasons for not employing a superhet circuit. Another benefit of using the TRF (tuned radio frequency) technique is that one does not have to rely on the stability of a local oscillator, particularly in this set with push button tuning.
Just to be sure the old set was OK, I left it on all day. Perfect