Here is my writeup. As I stated above, this was a complicated repair. The original complaint was "no stereo" which was essentially correct. The signal level meter did nothing and because the level threshold for stereo depends on the signal level it never was enabled. The IF strip looks like this:
The signal level is derived from a sample of the IF signal on TR202's collector. As this is at the end of the IF chain it is likely to be in limiting from a low input level so the effectivity of the indication is minimal. But what made things worse was the interaction of the amplifier TR204 and TR202. With a small pickup loop held close to TR204 I could see what was happening with my trusty HP 8595E spectrum analyzer. With the IF signal in the middle of the IF filter it is fine:
But if it is off-center TR204 misbehaves:
The signal meter then indicates a higher level and you might think that tuning is optimal. However, due to the oscillation of TR204 (or TR202) the demodulated audio is distorted. Because the signal meter circuit was designed so stupidly I decided not to try to find the cause of its insurrection but to replace it with a circuit of my own design:
The advantage of this circuit is that it also works at higher signal levels as it acts as a log detector. It uses technology from its time and thus is period correct. It works by taking samples at three points in the limiter. As the level increases and each stage goes into limiting the increase will move to the front of the limiter resulting in a more or less log response of the detected signal level. The diodes are placed in series and will add their detected voltages which end up at TP4. A buffer transistor removes the load imposed by the signal meter and VR201/2. The diode compensates for the V(be) voltage of the buffer transistor.
There is plenty of space on the PCB to accommodate it with the TR204 components removed:
The performance of the log detector is pretty good:
There is a bend in the curve at -80 dBm which could be tweaked by changing the coupling capacitors to the diode chain but as a indicator it works well enough. Getting rid of it would have been a huge effort. I could also have made a detector based on an AD8307 had I wanted better linearity. That would not have been period correct.
Stereo threshold, muting level and signal meter were adjusted. I completed the fix by tweaking the front end, which was almost smack on.
That was quite a journey, and all that for a tuner that is now half a century old!
)
