... as always, under construction ...
On the "Main Unit" PCB:
Replace the AM detector diode, designated D63 (a 1K60 Germanium,) with a Hewlett Packard 5082-2835 Schottky detector diode.
Apply DC bias to the detector diode by removing R121 (a 470k resistor connected between the cathode of D63 and ground) and replacing it with a 3.3M 1/4 or 1/8 Watt metal film resistor connected to the -10V supply rail. The -10V supply rail is available at J9 nearby. Use a piece of heatshrink or plastic tubing ("spaghetti") to cover the resistor. Metal film resistors tend to be less noisy than carbon resistors so this is the preferred type.
After the modification I measured a 14dB increase in SINAD across the range of signal strengths from MDS up to -20dBm. The most profound improvement seems to be on weak signals, presumably because they were previously below the detector's threshold. By ear, in actual usage the R71's audio clarity is noticeably improved and modulation on very weak signals is no longer muffled. Before the modification it took a considerably stronger signal before clear sounding audio was rendered.
Adding DC bias to the original 1K60 Germanium diode may also offer a worthwhile improvement. While testing detector circuits on breadboard I observed that the Schottky diode offered about a 4dB improvement in S/N over the Germanium diode. Theoretically it may be possible to see a 10dB improvement by adding DC bias to the Germanium diode. It will however be necessary to adjust the bias current for the Germanium diode, and it's likely that the bias resistor will have to be increased to perhaps 4.3M or 4.7M.
That's all for now... 12-January-2001.