Zenith System 3 electronic tuning weirdness

[RetroHacker]

So, this is a long, complicated, weird sort of thing going on, so apologies in advance. This is going to get a bit confusing, I'll try to explain the best I can.

This is a 25" Zenith System 3 Avanti console, SM 2535X7. This is my daily watcher, I use this set for a few hours a day, sometimes for many hours a day. I have completely and totally recapped this entire thing. Every single electrolytic has been replaced. It's been adjusted and fine tuned to the best of my ability, and when it's running, the picture is pretty good. Front of set
Inside

This set ran perfectly for a couple of years after having been recapped, and has always been great. It's connected to a lot of different things, all composite video, run through switch boxes and an RF modulator to channel 3. But... the problem I'm having is one that has slowly been getting worse and worse. At first power on, channel 3 is just static, despite having a signal on it from the modulator. If I let the set warm up for a while - after a minute or two, I can change channels around a bit and come back to 3 and get sort of a picture - but it will frequently lose horizontal sync with an unusual static interference, flickering rapidly back and forth between a good picture and static. If I let it warm up for about five or six minutes, and change back to 3 - it works perfectly and will continue to work perfectly for hours and hours, until it's shut off and let to sit for a while.

The *weirdest* thing about this is that it actually cares about how you got to channel 3. It tends to work best if you come at it from the high number channels, clicking down one at a time. If you started at 36 and hit channel up to 3, it tends not to tune in right. But if you came at it from the 30's, went all the way down, 5, 4, 3 - then it'll tune in and work.

Along the way, other weirdness on channel 14. It looks like this if you got to channel 14 from the higher numbered channels, clicking down. But if you got to it by going up from the lower numbers, go from 13 to 14, it looks like this.

If you have no signal going through the modulator, you get nothing on channel 14 but conventional static - regular no-signal static on this set looks different, with smaller, less bright dots, smaller dots - like this.

After warming up for a couple of minutes, it's still not fully "working", but if you tune to channel 3 by way of going down from higher numbered channels, it can come in perfectly clear for a second or so, then will randomly de-sync horizontally and get the weird static pattern, snapping back to clear, then static, then back to clear - but eventually dissolving into noise.
Clear picture.
Noisy picture

Here are several shots I snapped real quick in succession trying to capture what the noise looks like, and that gave me some insight, it's losing horizontal and tearing as the picture dissolves. Not sure how helpful, if at all, these are, but...
Picture 1 Picture 2 Picture 3 Picture 4 Picture 5

If you don't let it warm up quite enough, and try to use channel 3, it'll be clear and perfect, then a random flick of static as a horizontal bar in the picture that's maybe 3" tall, that will gradually get less and less until it stops entirely. Tuning away from 3 and back seem to speed up this process.

And, yes, I did try another RF modulator. It does the same thing regardless of what's supplying the RF signal, be it a standalone modulator, a VCR, or a Nintendo.

Now, here's the frustrating bit. Every time I pull the heavy console out from the wall, take all the stuff off of it, unhook everything, get it out of where it is to pull it apart, and poke around inside... the problem tends to not appear. Like, it'll appear for a short bit, then I'll start poking around, measuring voltages, and it'll work fine. And then I'll turn it off and let it sit, back off, in the middle of the room for a day, come back, turn it on... and it'll work fine. Obviously, this points to a bad connection - I have been all through the set, I have resoldered the entire tuner, checked and cleaned those Zenith edge finger connectors, made sure everything is tight and well connected. The set will work fine for a couple of days, I'll put it back together - it'll work for a few days more, then it'll start with the issues again. Hitting it/shaking it has no effect.

A bad connection sure makes a lot of sense here, but I can't find one, and I can't affect it by hitting it or any of the normal wiggling that should show something. Voltages I've checked have been fine. Other thoughts were a transistor or a chip working better once it's warm, and being close to the wall makes it colder due to the fact that it's an exterior wall made of concrete and there's an air return duct right next to it. My next round of diagnostics will probably involve me hitting things with freeze spray to see if that affects the tuning.

This is where my inexperience with these electronic tuners comes in - this thing is complicated. The schematics for this set are massive. I have spent many hours staring at the tuner, the IF section, video stages... it's just so complex, there are so many variables, there are so many components, and all the chips are custom Zenith parts with no documentation. Honestly, I'd be happy to say screw the tuner entirely and feed in composite video - which is usually pretty easy in most sets, but this one I haven't been able to figure out where even such a point could exist - there's a video amplifier transistor, but it's after all the sync gets picked off. And yes, I am aware this is a hot chassis set, if I were to modify it, I have an isolation transformer right here I can install. This stupid set is so complicated, I haven't even been able to figure out how to adjust the width, or if that's even possible without modifying the circuit.

Unfortunately, I do NOT have a tuner substitution box or an IF video generator.

Sooooo... yeah. If you read this whole mess, I apologize for all the confusion, but this is a weird problem that I've been chasing for some time. I figure I'd ask here in case there's something I should be looking for, if someone has any insight into this, so when I do haul it out from where it is, and take it apart, I have more things to look at when it'll inevitably work OK while open....

At the moment, I've just been turning the set on, going to channel 32, letting it sit there for 5-10 minutes, then click all the way down to 3 - the picture will come in perfectly, and everything will work just fine.

Any suggestions or help would be welcome. Thanks for reading!

-Ian

[old_tv_nut]

Is there an AFT (automatic fine tuning) switch? Does the set act up when the AFT is turned off?

[old_tv_nut]

Note - if there is some sort of automatic color switch, it may activate the AFT also, so both switches should be off to try it without AFT.

[RetroHacker]

Messing with the AFC normal/special switch or the cable tv band selector switch has no effect.

[zeno]

Start easy. CLEAN the AFC & band switches first. Then set to normal & V-U.
( ota settings). NOT HRC or IRC. NOT spcl.
Cold joints are possible. The TUNER itself is on the 9-151 board. Watch
the shield for cold joints. It completes ground paths.
The SELECTOR ( IIRC 9-165 ?) its the board up by the key board. Watch for
cold joints on shield & plugs ( not edge connectors ).
There is a 3 legged IC ( 5V reg ??) on the shield, check VDC both when running normal & acting up on all pins & see if it changes. Had a run of bad ones on the SM1971 sets but different symptoms. We sold a few dozen of them & most failed in a few yrs.
Now to the 9-160 PS/HV board. All voltages come from here. Half of the
fails are cold joints. Go over it real good especially under large parts.
There is a little transistor called 32V current source or something like that.
Be sure its putting out constant 32 VDC. Its the tuning voltage source.
With that said it sounds like the AFT is searching & not finding the horz pulse
to stop it. I should have the SAMS so I can pull it if need be. Keep us
posted. A u-tube would help also if you can.

73 Zeno
LFOD !