I have been struggling with this for a while now. Its not too bad, but does seem worse than other sets I have of the same era and chassis type (other RCA and clones of the 12-16 chasis).

Normally I can dial out the line pairing with the vert hold and find a sweet spot that give the best resolution. This one is much tougher (very critical and not always possible to dial out).

Its been completely recapped in the vert circuit, and I have tried several 6kA8 (agc/sync sep) tubes. Interestingly when I got the set it had a 8KA8 tube in place of the 6kA8, I have tried both.

I was advised that sometimes too strong a sync pulse can cause this, I am going to go back and read up on vert sync and equalizing pulses etc.. to see if I can get a better understanding of exactly the timing of when the vert osc is triggered.

the RCA color sets have a simple vert integrater at least compared to some of my older sets, its two 100k resistors and two disc caps. I have NOT replaced those (they tested fine with an eye tube type cap tester).

I am going to go back and take a look at the sync signals with a scope to see if the PP is indeed high, but before doing that I did check the SAMs tube voltage charts. the sync sep plate and cathode were right (130v and 25v) but the grid was more like +15 (no signal) and about +3 with signal (per Sams should be read with no signal). It should have been
-15v. The readings were taken with a HP 410 vtvm. The tube bias is cutting off the video as the sync signals are working as far as the horz and vert lock, So not sure if the difference in the tube voltage per sams vs the actual is a problem, but I need to start somewhere.

I have replaced all the filter caps on the boost that is used on the vert mutivibartor. I know that a problem with stray horz rate noise can be an issue, Reflowed all the ground stakes, did not distrube any of the lead dress, routed all the convergence wiring away from the anode lead.

oh a couple more things, I did check the HOT screen decoupler cap, it was fine. and I did scope the B+ it had the normal amount of ripple and a slight amount of horz rate ripple, but I believe this is normal as even the sams shows a FAT line on the power supple B+ which is what I see when using a slow sweep rate to look for 60cycle ripple.

I was checking some of my old post, yes this is very much the same as on my sylvania RCA clone. Looks like I may be pulling that one out and checking its syc sep tube voltages for comparison. I really need a magic bullet to fix this kind of problem. I prob would be willing to write it off "thats the way the all are" but for the fact that I have some that don't to it at all.

I find voltage errors in Sams all the time, I wouldn't be surprized if it's actually supposed to be +15 on the grid. Same thing happened to me recently with the chroma bandpass amp in my 21-CT-55, Sams calls for -15 grid volts whereas I read 0 on both the CTC-2 and CTC-2B chassis. Your cathode is still 10 volts over the grid, so the bias should be about correct. It should be more or less cut off till it sees a vertical pulse, then it should conduct. Scope the plate to confirm. If it's still acting up, you might try replacing the cap coming off the AGC control since the AGC control will have an effect on the bias of that tube. Make sure the AGC pot is clean too.

that is the 40uf, pretty sure its ok as I did a complete sweep of all the can caps. This is a VERY subtle problem as the lines are not completely paired up, more like about 50% paired if that makes since. I will scope the plate tonight and see if any video is getting thru. I also picked up a few more 6KA8's, just curious to see if they make any difference.

there is are 4.7meg and a 22meg resistors that go from the grid of the sep triode, one goes to the 330v and the other to the 265v sources. lastly there is a .0033 cap that couples the video signal from the video amp to the grid of the sep triode.

I don't recall going over all of those. I was wondering if the 8KA8 was put in there to reduce the amplitude of the sync pulse by a tube jockey or if it just is all they had on hand. I will scope the output with that in there as well.

If you have a sound understanding of the fundamentals and the operation of that specific circuit you may be able to design a circuit mod to vary the interlace as needed. That could serve as a magic bullet, if these circuits are so finicky that trouble shooting them is impractical.

assuming the PP on the sync pulse is correct, and its clean, I would think the only thing that could be the problem is the vert integrater. That is where I am going to focus after checking the sync pulse out.

other things, if I adj to the extreme end of downward vert roll (where it just on the verge of breaking lock, and breaks some times) the pairing will go away. I am not sure if this is speeding up or slowing down natural speed of the multivibrator, but perhaps it just takes a bit more of the pulse to kick it off, and that is more precise a starting point?

of course I doubt the design was ever intended to be so critical, which make me think I am just missing something.

Re-solder ALL the board perimeter eyelets. Be SURE the chassis eyelet pins are well tinned as it can be just coated with solder not actually tinned.

Replace the vertical integrator tube socket. This same tin corrosion problem is also manifest in the folded tin contacts of the socket.

Check the pix for heater/cathode leakage. If in doubt use an add-on isolator.

Look at the "front porch" wave form and compare timing from raster one to two, integrator values may have to be hand tweaked.

Check the vertical hold and vertical linearity controls for leakage to the chassis. If these controls have been dosed with WD-40 or other blather, completely degrease them. If these controls are on a remote panel be sure ground return is intact and runs with cabling, do not allow the cabling to be tightly bundled.

If the pix already has a brightener, the extra wiring can cause an issue.

GL

Chas

guilty of using cleaner on the vert hold and vert lin pots. I have some MAF cleaner that I am sure would degrease.

the wave form off the plate of the sync sep seems to have a greater vert pulse compared to the horz pulses when compared to the RCA service manual (NOT the sams, which does not show the wave from at the plate of the sync sep).

on the RCA manual the pic of the wave from show the vert spike being just a bit more then the PP of the horz pulses, where as my scope has them nearly twice the size, I will get a pic and post in a bit.

the sams does show a pic of the integrated sync pulse by pulling the vert out tube, brighness down, so I will take a look a that.

1st pic close up of sec pic, vert rate sync at plate of sync set tube, I dont know if that stuff in the begining is normal?
2nd pic is the vert rate
3rd is horz rate

that was a still DVD blue screen image. That stuff in the beginning of the pulse on the 1st pic was moving constantly, the rest of the wave form was steady.

Looks like a half-cycle of 60Hz running through the sync waveform, you must track this down. It could be either a leaky cap in the power supply (look in the voltage doubler if so equipped), or heater leakage in that tube or any of the IF tubes. Get cracking with the tube checker/tube swapping.

Be sure that the correct probe for the scope is in use. FWIR: It should be a low capacity probe...

You need to see just two vertical scans, initially, the start of of the scans varies by one horizontal line time plus horizontal blanking. It may be that the triggering for the vertical scan is insensitive to the small time difference. This may be difficult to see on a conventional "tv service" scope. Using trigger hold off or other features could present this waveform more precisely.

Ultimately tweaking the cap values around the vertical integrator may be the answer. I cannot recall exactly, what the integrator has for components but if there are peaking coils involved keep them in suspect too.

BTW did you use metalized film caps in the integrator? These have problems with their Dv/Dt rise times.. Only poly metals (real foil) should be used as their rise/fall times are much faster. Good caps will state the rise times in the engineering data. Check the rise times in your circuit with the scope vs the voltage in the trace area (horizontal pulse time most likely) and see if they exceed the specs for the capacitors you chose. That is if you can get the specs..

I would have never though of this to be a problem but I have seen the metal films when used to trigger stroboscopic flash circuits actually get torn apart in a few minutes from trigger pulses of only 250 volts on a 400 volt capacitor. Net result was a dramatic change in capacity as areas of metalization get disconnected... Unrolled the cap to have the metalization fall away in bits... What happens there is insufficient energy transfered to a trigger coil and the flash fails...

Again GL,

Chas

I will check/sub the IF tubes today. the 6KA8 is ok as I have both subbed and checked it already.

I checked all the filter and doubler caps with a cap tester (eye tube type that looks for leakage at rated voltage). Replaced any that tested less than like new.

the integrator has the orig disc caps, I am pretty sure I tested them and they were ok.

I have one of those 1x 10x china probes, bought new recently. Not sure if its a low capacity, but I think it is.

Find some new subs., too simple not to try, as with metalized, if conductive coating migrates into substrate the dv/dt will change and may effect the charge timing slope. You are interested in some device or problem that is at the speed of one horizontal line.

GL

Chas

I will give it a shot.

a few more developments. I replaced the VDR in the feed back circuit. After reading up on this it would appear the VDR is there as a way to bias the tube to compensate for variations in tubes and provide a consistent vert height.

I double checked the 2uf cap that decouples the screen of the video amp, and the 80uf that decouples the vert out transformer.

a new observation, when I 1st turn on the set, the pairing is gone, returns in about 20 seconds after a screen pic is visable. Not sure if its just the focus is better, but you can see the fine pairing starting up.

I am going to scope the vert pulse again and see if that junk in the beginning of the pulse is there when the set is 1st turned on.

oh I checked the VDR but using a cap tester 450vdc leak test feature and a DMM set to read microamps. the old one would read 2uA where as the new VDR would read 38uA.

nothing new, I checked the disc caps on the top hats in the PS, all fine, really just grasping at straws at this point. there were a few other disc decoupling caps, all fine. the ONLY odd thing I found was a 180k 10% resistor that looks like its part of a voltage divider that puts B+ on the filament of the shunt tube. My guess is to get it closer to the 400v sitting on the cathode. I could not test in circuit but when taken out it was 330k not 150k. I dont see how that could effect anything but since I was checking around there I went a head an replaced it. I went thru all the pcb ground stakes again.

I just don't like that middle scope pic of the vert intergrater pulse with that garbage on the beginning of the pulse. I am going to scope the output of the video detector and compare it to the composite video from the box, and take another look. I need to see if that is there before the video amp tube. its like a little bit of video is not being biased off at the very beginning of the pulse, maybe its the close caption or some other nonsense, of course it may have nothing to do with the pairing issue...