Today's Mystery Guest (RCA CTC-4)

[Phil Nelson]

I am injecting the video signal downstream of the video detector diode -- at the grid of the 1st video amplifier.

Phil Nelson

[Username1]

go back more.... detector and before.... Compare what you get with your reference ctc11....
You got a signal source of any kind you need - pick it out of the 11 and feed it to your 4.

That's a low pass network after the diode, look at that buncha coils, are they checking
reasonable ohms, etc..... Check all that stuff around that diode, back to the tube too....
Voltages in those stages ok....? Compare scope readings to the 11......
That big screen grid cap ok....?

.

[old_coot88]

Originally Posted by Phil Nelson

I am injecting the video signal downstream of the video detector diode -- at the grid of the 1st video amplifier.

Understood. But the reason for asking was: since you're injecting at the grid, and the grid is DC-coupled back to the diode (i.e., no capacitive isolation), a shorted or partially shorted condition in the diode could be lugging down the genny's signal. Like I say, it's a shot in the dark. But easy to test by just lifting one end of the diode.

[old_coot88]

Originally Posted by Username1

go back more.... detector and before.... Compare what you get with your reference ctc11....
You got a signal source of any kind you need - pick it out of the 11 and feed it to your 4.

Good point. Just be sure the two chasses are grounded securely together, and monitor the display on the good set at all times.

[Username1]

Have you swapped out the tubes for the video IF stages,
and the first video amp...? With known good ones....

Have you looked at the signal going into the first video amp
with the tube out...? At the grid....

I can't imagine enough room to set up 2 chassis and make
checks on them and go on for weeks at a time......
Divorce must be that thing over in the corner that
never seems to feel right..... Is it growing....
que - twilight zone music......

.

[DaveWM]

how about just driving with a solid color screen see if you can scope the grid and plates of the Red and Green demod tubes.

[old_coot88]

Will the genny he's using make solid-color blank displays?

[Phil Nelson]

Originally Posted by old_coot88

test by just lifting one end of the diode.

Well, that led me stumbling down a path that eventually led to a real color picture. After I lifted one end of the diode, I checked out the output network downstream of the video detector (which all looked OK). I experimented with injecting the video at a couple of spots other than the video amp grid, without getting anywhere. I was really tired of messing with injected video at that point, so I decided to replace the old diode and send a signal in through the antenna, just to see how bad things really were.

It didn't occur to me to test the diode until after I removed it. Maybe it has been bad all along, or maybe the heat of removal sent it south. It's definitely bad now, anyhow. I wonder if this 1N64 was a replacement, since the schematic calls for a 1N60:



I replaced it with a 1N34A, since that's what I happened to have on hand, and boom -- there was a color picture. Not a correct color picture, but better than I had seen before.

I noticed a persistent vertical jitter that hadn't been obvious when looking at color bars, and spent a long time getting rid of that. Replacing resistors on the 2nd sync amp as well as the vertical circuits finally took care of it.

Some colors are still wrong, and there's a lot more work to do, but at least you can recognize pictures of humans, with some colors in the right spots.



Tomorrow I'll try the color AFC alignment once again, put the scope on some interesting places, and go from there. The resistor hunt is not over -- it has turned up a lot of bad resistors in this set, and I suspect there are more to find.

Thanks again for the advice & encouragement.

Phil Nelson

[miniman82]

Huzzah!

[Username1]

Well that's an improvement.............
A nice rich hi contrast picture....
cool.......
Alright... Next tv......

.

[old_coot88]

Phil,
How does the old diode actually test in forward/reverse resistance measurement? Just curious.
Suspect maybe high reverse leakage(?).

[Phil Nelson]

The old diode looks open (infinite resistance both ways), using either the ohmmeter or the diode check function on my meter. Don't ask me how it was able to pass any signal at all in that condition, but perhaps I torched it during removal, and it worked in some half-baked way before.

Maybe someone can help me make sense of the demodulator adjustment procedure in the RCA manual, which I posted here:

http://antiqueradio.org/art/temp/RCA...dProcedure.jpg

Question 1. It seems like the writer got some colors jumbled:

-- Step 1 tells you to short the blue and green control grids to ground through 100K resistors.
-- Step 3 tells you to remove the 100K resistor from the red grid -- but no resistor was ever placed on that grid -- and then place it the green grid -- which already has a 100K resistor from Step 1.
-- Step 5 tells you to remove the resistors from the red and green grids.

Huh??

One way to make sense of the instructions might be to amend Step 3 to say, "Remove 100K resistor from BLUE control grid and place it between RED control grid and ground." That would leave you with resistors on the red and green grids, which allows Step 5 to make sense.

Question 2. Unfortunately, I don't have a pattern generator that can make the same color bars as their example, so I'm not sure what is meant by an instruction like, "Adjust L125 until the sixth color bar (blue) blends with the background." What color is the "background" of that pattern generator? Black? White? Other?

Adjusting L125 and T115 (top) per those instructions changes the bar hues, but the question is how to do this with my newer pattern generators that produce bars with the usual sequence of white-yellow-cyan-green-magenta-red-blue-black.

An alternate phasing adjustment procedure with oscilloscope is described on page 49-51 of the service clinic manual, but the same question applies -- what waveforms to look for when using the different (i.e., modern) color bar pattern. Perhaps if I stare at Fig. 96 on page 49 long enough, I can draw my own diagram and model waveforms . . . .

Phil Nelson

[Username1]

Well if the diode was open, then removing the new one from the circuit
should produce the symptoms it had before...... There are other
components that most likely would have passed what you were seeing....
Try removing it, see what you get....? Whaddya got to loose....?
It must have worked a little.....
Scope it with some high frequency waves, see if it waffles at some point....

.

[old_coot88]

If t'were me (which it ain't), I would go ahead and get some good red purity on it, set the grey scale, and see what a regular vid with known color content actually looks like.
Then if need be, try some eyeball tweeking of a few of those adjustments (with the exception of the reactance adj, since color sync is already good.

Others' opinions may vary.

[Phil Nelson]

After more eyeballing and tweaking, here's some slightly improved color bars:



As usual, my camera isn't very accurate. The bar that looks white is actually quite yellow, and the dark one between green and red is a dark magenta. I see now from the leftmost bar that the should-be white background was actually a little green. Yes, I should go back and do purity & gray scale, and then return to this level of tweaking.

Here's what a DVD scene looks like:



Again, the real screen is warmer and less washed out than the photo. You can see from the PAUSE letters that edge convergence needs attention, too.

The color sync is still not as robust as it should be. If you fine-tune a little bit off-signal, or even move the Hue control too far at the wrong time, faint rolling bars start to reappear. I'm tempted to reinstall the original L42 inductor on the oscillator. It looked scorched when I pulled it out, but it worked as well as the new ones in the breadboard circuit. That whole area needs to be neatened up, anyway.

The audio's not bad. Now that the rest is sort-of working, I can give that alignment a try. Still more stuff for the to-do list!

Phil Nelson