While I had the chassis on the bench, I decided to take another look at the breadboard circuit. It is being powered by the CTC-4 chassis but none of its signal leads are connected. Now I can see that it's oscillating lower than the ideal frequency.
As before, one output of the quadrature transformer looks irregular and the other is a clean sine wave, when they are connected to a probe one at a time:
It was interesting to connect the two scope channels to the two transformer outputs. When both outputs had some loading, the irregular waveform got more regular. With some adjusting of the bottom and top slugs, I was able to bring them 90 degrees out of phase, as I did before with the TV's native circuit:
This little exercise gave me a better grasp of how the transformer works. The bottom adjuster affects (mostly) the amplitude of one output, while the top adjuster affects (mostly) the phase of the second relative to the first. The closer they came to quadrature, the more the irregular wave seemed to smooth out. I was also able to observe an amplitude peak while adjusting the bottom slug, something that you're supposed to observe with a voltmeter in the service manual's alignment procedure.
Tacking a 100-pf cap in parallel with C213 gave me a TV that displays stable color bars (even though all of the colors are wrong), so I guess the oscillator is running in the right ballpark.
Phil Nelson