The next problem that needed attention was the broken pin on the socket of the 6AT6 first audio tube. Although there was nothing visibly wrong, a continuity check showed that the contact for pin 1 (control grid) was broken. With a little tug, it pulled right out, making the break very clear:
I thought I had an identical socket I could replace this one with, but checking my supplies, I did not. Even if I had had one, it was not entirely clear how I was going to get the old one out and the new one mounted without having the resort to screws to mount the new one.
I noticed that the original socket had two unused pins. By cutting a little bit of the bottom wafer away, I was able to slide out one of the unused pin contacts:
Getting the remnants of the broken contact out was much more challenging, but eventually I got it out. With a similar section of the bottom wafer cut away around pin 1 (already visible above), I was able to slide the good contact into that position:
Before trying to glue anything in place, I soldered the resistor and capacitor to the new pin, to avoid heating after gluing. Finally, after soldering, some 5 minute epoxy was added to keep the replacement pin in place permanently. This was done with a tube in the socket, so everything would be positioned properly. Final view:
With that fixed, I turned my attention to audio IF alignment. Since the 621TS is an early set that does not have intercarrier sound, alignment of the audio IF is critical. Basically every early set without intercarrier sound needs at least a little touch up of the audio IF alignment, both to get good sound quality, and to get the fine tuning to tune in both good picture and good audio at the same point.
This set was a breeze for audio IF alignment, and after just a few minutes, a perfect FM demodulator response curve could be obtained, centered exactly at 21.25 MHz. Here's the equipment I use for that -- the old HP analog sweep generator on top provides the sweep signal; the digital HP RF generator on the bottom is used for markers (it will only sweep up to 20 MHz, so it doesn't make a suitable sweep generator itself, but it will produce markers up to 40 MHz, which is fine), and the little Tek plug-in module scope in X-Y mode is used for the response display:
On this set, I don't think the video IF needs any touch up, since the picture looks good as is. Sharpness is good, and so is the low frequency response and DC restoration. So no need to mess with it.
Picture width is still a little narrow, so I will look into that further tomorrow, as well as checking all resistors for drifted ones to replace.