|
Given the rate of wobble I'd reckon that the power supply filtering is insufficient, your AC outlet voltage is fluctuating, or some component is shorting/leaking/arcing intermittently at a regular interval.
I've seen problems with AC utility voltage fluctuations causing picture size to change 1 a minute to ~3 times a sec. There is a business park about 1 mile away with manufacturing operations there, it was worst for the first 2 years I lived there, but still happens occasionally....I knew it was the power line since I could have 2-3 working tube era sets and one new set fed the same signal source on at once, the 3 tube sets would 'wobble' identically, but the new set would not wobble (due to having actively regulated B+ rails)...The cure was to run the tube sets through a SOLA VRT (AC voltage regulating line isolation transformer).
As for contrast check the video amplitude with a scope and compare to the schematic. If it is low there are a number of things that could be affecting it.
Your AGC may be tied to the low contrast issue....For instance the tube IF grids should be biased negative enough relative to the cathodes to prevent conducting more than their max rated plate current....If the IF tubes are weak and not amplifying enough then the AGC could be be pushing the grids as positive as the biasing design wall allows...When it hits the wall it can't make any more gain, but still it trying to.
It might be interesting to connect that external bias supply back to the AGC line, see how much more contrast you get and how the IF and tuner voltages (especially the AGC controlled ones) compare to the schematic with the external AGC voltage.
May also be good to check the sync amplitude entering the vert mult/osc system. Bad vert hold can be weak sync or osc off freq.
If you can get the vertical to roll both ways (up and then down) slowly, even 'float' without rolling but not fully synced for a while, then the problem is in the sync pulse level/cleanliness and or the osc circuits handling of it.
If vertical only rolls one way then it is off freq and the vert osc/mult/output stages are to be suspected (and sync signal can be ignored till float/dual direction roll is achievable).
|