[davemc]

Update. I removed the oscillator coil and reduced the feedback winding to 10 turns (so now I have tried 10, 15, and 22 turns). I also checked capacitor 13: it shows no leakage and measures 110pf, right on the money. I checked all the resistors again. 14 (32K) measures 37.16K. 10 (51K) measures 53K. 9 (300 ohms) measures 394 ohms (I paralleled some resistance to get it to 300 ohms - no difference). Although the B+ is correct and resistors 19 and 24 are close, the converter screen only measures about 55 volts (Riders schematic says 85 volts to cathode).

IF I DISCONNECT FIXED PADDER 11 the set will oscillate over about half the dial (lower half). But of course with no padder, it is about 600KC too high. Putting ANY value of padder in place kills the oscillation.

This is certainly the most difficult restoration I have ever attempted - and such a simple radio!

It has to be the oscillator coil, but it works perfectly on the broadcast band, and only the MIDDLE part is shorted out on Police (plus the fixed padder 11 is added). Maybe the number of turns on the feedback winding is critical in this case.

Getting ready to punt ....