There's a guy on the videokarma.org forums who has a Philco split chassis that is still working on mostly original caps. A lot of Zenith products somehow manage to somewhat work with original bumblebees...

I own Philcos first consumer color TV the 1955 TV-123 chassis, and mine on mostly original caps somehow managed to make a horizontal line when I variaced it. I have it working on new caps but H output current seems excessive and I need to address that.

The nice thing about Philco using tons of caps compared to other brands is once they're recapped a larger portion of the circuit is new parts thus there's less off tolerance stuff to degrade performance.

I almost got this TV Perfect except it has dark shading on left and right side of images but only sometimes. If I apply a video signal directly to the input of the video output tube it looks perfect but going through the tuner, IF, detector it does this. It has a semiconductor diode for the detector maybe it's leaky. I measured DC resistance across all the peaking coils and I can't remember exactly what they read I think 6 to 10 ohms. I wonder if the agc is working

Electronic memory I do agree with you about capacitors. I was going to just try it out first without replacing them all and try to troubleshoot it. It had an aired horizontal output tube. I changed it and I put it on a dim bulb tester and reformed the electrolytics untill I can apply full power. After full power I woke the dead hahahaha. Nothing worked. The seliniums were high which made all the voltages way too low along with electrolytics that were weak and only 4kv at the tube. No vertical deflection because open yoke I had to do surgery on. I could barely see a vertical line I had to close the cabinet doors to see it hahahaha. Every bumblebee cap was cracked and bad on the sencore analyzer. The only one I didn't replace is the wax tone cap. It's a shame I couldn't keep some of the originals in there and the customer said it worked last time it was plugged in??? Hahaha no way. I hope the crt hold up it had almost zero emissions. I disconnected the green grid wire and powered it up then gave it a jolt with my ccfl tester by touching the grid and it came to life and surprisingly good too. I worked to hard on this dog than the estimate I gave my customer. The am radio works. All I have left is to repair the crusty old record changer hahahaha.

Are you still having this problem. While I doubt that the diode would be the problem, people do have trouble with them. Should be cheap to replace. People successfully use silicon small signal diodes, definitely not rectifier diodes.
Have you replaced the capacitors, including mica, around the AGC circuit?

Edit: 1N4154 is a commonly available diode that is known to work as a video detector.

The 1N4154 would work, from the datasheet. The 1N4148 is more common and I know it works in any position in a tube TV that used a small-signal germanium diode. I've used them for video detectors, ratio detectors and horizontal AFC in many TVs.

It was the 1N4148 that I was really thinking about. I just puled the wrong number from my memory.

Thank for the help. I think it does it on strong signals. I need to check agc voltages and check the micas on my sencore and give them a full voltage leakage test