OK, time to stop screwing around and get out a scope!
I monitored the output of the detector and waited for the image to fade out.
Sure enough after about 10 minutes the video signal dropped to almost nothing.
DC voltages and waveforms in the IF seemed OK so I went after the detector assembly. I took the covered off and hit it with a little compressed computer duster can turned upside down to cool it. Bam! image was restored
Now, I just need to determine if it's one of the resistors or the 1N34 diode. My bet is on the diode. I'm not wild about having to perform microsurgery to replace it, but I'l give it a try if I have to.
