Attachment 166700

Here is the kind of scrambled video I get when I feed in video from my B&K, my color bar generator, or even my dvd player into the grid of the video output tube. I can't sync anything. I'm just feeding this in with bare leads. Should there be a coupling cap, a diode (since this is past the video detector diode) or anything else needed to properly couple the signal to the tv?

I can't even tell if there is even a video image to pull from this. I just simply go from a flat grey raster to this mess when I feed anything in. Adjusting the output levels on the B&K does nothing but fade to picture to snow.

That looks promising to me - you just need some horizontal sync. It should fade to snow as you drop the level. Have you tried flipping the polarity switch on the B&K ? Do the horizontal clipper, amp, discriminator, etc. circuits check out OK ?

Yeah, I'd use a coupling cap so as not to damage the video source. Maybe a few 100 pF. You don't need a diode if you're feeding in a composite signal.

I'm not familiar with the set, but it looks like the horizontal frequency is way off. Have you tried to adjust the horizontal hold?

John

Okay, after fiddling around with the horiz. sync on the television, it just mustn't be syncing on the Horizontal. The best I can do is get it to "lock" on a multiple of 8 repeated horizontal images of the B&K test pattern. Flipping the B&K's polarity switch makes the image go negative.

I've checked every single component in the horizontal output circuit, and I guess I'll start working my way back to the sync clipper, etc. Tomorrow night.

With the video if circuits working away, could that possible swamp the B&K's signal, considering the contrast control is a really a B+ gain control for the IF strip?

Did you try adjusting the horizontal hold control? I agree with bandersen that all your set's picture needs is sync, and yours is way off. I'd try adjusting the control first, before pulling the chassis. That failing, I'd check the horizontal AFC, as it may be yanking the oscillator so far off frequency that even the hold control won't bring it back.

I found an adjustment for the horizontal on the schematic (L21) and began twisting away. It functions almost as a course horiz. frequency adjustment, so the picture quickly pulled into sync a faint image of my B&K test pattern. Hooking up a dvd player revealed a scrambled mess of a picture. the difference in sources was enough that I had to readjust everything again. Now I have an image to play with!

Attachment 166706

Vertical lock is pretty well nonexistent, but I am more concerned about the faintness of the image. For this tv, since all the contrast control does is ramp up the plate voltages on the IF, I will need to find another way to tighten up this image. retrace lines are everywhere, too, but the lack of vertical hold is probably contributing to that.

Awesome! I had to make the same 'coarse' horizontal adjustment on my last set too.
Are you feeding the signal into the video amp tube or the antenna terminals ? If it's the video amp, perhaps the level is too low ? That could account for the lack of vertical lock too.

It's being fed right into the video amp's grid. I would suspect that the piddling signal from the dvd player just isn't enough to drive a decent picture to the single video amp tube to the crt.

OTOH, the IF strip could be swamping the dvd signal. Should I pull the final video IF tube to drop this potential interference? Also, does the dual diode do anything for a signal added to the video output grid as I am doing, ie do I get dc resto?

Here's a link to a partial schematic for the 802. The DC restorer comes after the video amp so it should be OK. Have you tried doing this using the B&K ?

Set the RF selector to 'IF'
Set the IF frequency to somewhere around 22 MHz
Inject the video signal at the grid of the 3rd video IF (pin 4)

Do you get a picture now ? (you may need to fiddle with the IF frequency). Is it any better ?
Likewise you can work you way back through the 2nd and 1st IF amps.

I replaced some more caps and resistors, focusing on the vertical section this time. Nothing too terribly out of line, except for a mica cap (c136 from Banderson's schematic) was open. I've found a number of open micas in this set, something that I rarely find in other tvs.

I believe that the tuner has at least one open one in its signal path, hence that is why I have no fm or television reception. The service manual describes the tuner as "easily removed from the chassis for servicing" well, it took about two hours to remove the tuner, mark the dozen or so leads that needed to be pulled, and even then it appears to be physically impossible to get to those mica caps.

Anyway, back to inputing composite video to the video ouput grid. Here are some results:

Attachment 166710

Attachment 166711

The distortion in the first picture comes when I install the dual diode (6h6 on the schematic, but mine is a 6al5). Lots of nastiness sweeping from the left to the right.

When the 6al5 is pulled, you have the better looking result in the second picture. Vertical ALMOST wants to lock. I have a pretty good contrast balance, considering that now twisting the contrast knob just throws off the horizontal sync. I do get a bit of distortion-a pretty variable black shadowiness with jagged edges (not unlike a lesser version of what was happening with the dual diode installed) that changes size with the material on screen. I believe this may still be an issue with the horizontal lock, as the horizontal blanking bars have also been creeping in unless the horizontal is dead on.

Weird that the 6AL5 would cause problems. Just curious - what are you testing the mica caps with ? I have a couple vintage capacitor testers that use a bridge and eye tube, but I don't really trust them.

I have an old Heathkit cap tester. There was an online tutorial on how to fix this particular model, and I've never had any issues with it since it was recapped/repaired. Typically, when I test a cap, I test the potential new one going in, and then test the old one going out. In the case of the 240pf cap that was dead in the vertical section, it didn't register as anything on the magic eye tube.

Does your set call for a 6AL5? I mean is it on your schematic or the tube chart? The 6AL5 and 6H6 have different pin-outs.

John

Nice - I should restore mine.

I was wondering about that too. A 6AL5 is a 7-pin miniature and a 6H6 is octal :scratch2:

Indeed those dual diodes do have quite a different pinout. The socket install looks factory, but I'll check the pinouts to make sure the 6al5 is hooked up properly. GE made a fair amount of changes to these sets during their production run, and I could see this being one of them. However, it is on none of the documentation I have been able to find for this set.

Not too much time to work on this tonight, but I did discover that the plate voltage is pretty much missing from the clipper half of the 6sn7-sitting at about .3vDC when it should be at 11vDC. The voltage is supplied by a 330k resistor, the supply voltage is 215vDC on one side of the resistor and .3 on the other. I've tested this resistor a couple of times, and it always tests at about 340k- a little high, but no reason to shut out voltage like it does. But having a freaky clipper certainly would explain the sloppy sync I am getting....

In other news, I emailed the screenshots that I have been getting of the set to the original owner, who only grew up with it in his house and never saw it running in the fifty years he has had it before giving it away. He was pretty pleased to see it working at all.

I'd be intersted in seeing the sync clipper portion of the schematic if someone could post it.

The grid voltage should probably be about -0.5V. It's usually derived by grid leak from the sync pulse. If the sync pulse isn't getting to it the grid will be too positive and that would give you a low plate voltage.

Or you could have a problem with the 6SN7. It could have a leaky grid. But, in this sense it emits electrons, rather than collecting them from the cathode, so it becomes too positive. You need a tube tester that can check for grid leaks or a special tester like the Sencore LC-2 leak tester.

Not having seen the circuit, I might have said something stupid.

John

My schematic comes from a Wallace Telaide with their ledger size pages. Does this scan cover the areas of interest ?

Yeah, it does, thanks. It looks like C40B could also cause the problem he's seeing if it's too leaky or shorted.

John

How about C40B... perhaps leaky/shorted/installed backwards ??
Also R-28... perhaps it is way too low ?

jr

oops! it looks like John beat me to it

Nope...it was that darn 330k resistor. I went back and tested it, only to find that it would suddenly read at 400k, then 300k, then open. I pulled it out of circuit, and it read as an open. So, now I have excellent vertical lock, but still somewhat iffy horizontal lock with that weird blanking bar bleed. I'll photograph what I mean (hopefully) later today.

Okay, I am now focusing more on the horizontal section, as there seems to be poor regulation there. The raster has a bit of a wavy motion to it, and I am getting a black smear coming in from the left side of the screen (it is present in the distorted picture I took a few posts back). Looking at the clipper/sync amp tube (v16) all voltages check out now that that vertical issue has been corrected. Now moving onto the horizontal discriminator tube, I have low plate voltage for v17a-about -6 volts with no video input and -12vDC with signal input. It should be -95vDC. V18a and V17b's voltages check out fine. What has me puzzled is where does this -95vdc come from? Following the schematic, I don't see a plate supply voltage running to this one.

A bit of good news is that the vertical lock is SOLID. Rotating the vert. hold control will make the picture a little jumpy at the extremes, but no rolling. This is a pretty solid circuit now.

Maybe C48 is open?

The grids are tied to the plates so the 6SL7s are acting like diodes. I'm not sure of the level of the output of the horizontal sync amp but it'll be rectifying that signal and the feedback from the flyback.

John

I just picked up one of these sets too :D I'm not going to be working on it anytime soon, but I can take some detailed pictures if you'd like to compare it to yours.

According to my docs, the correct plate voltage for V17A is -6.4 v.
Here's a scan

My set has a 6AL5 too.

What's a factor of ten!

"Service selector channel switch at channel 1." Don't see that too often! Or do I misunderstand?

John

Factor of ten ?

That's right - this set has channel 1 :)

Leadlike said V17a plate is supposed to be -95V. Your scan says -6.4V. Roughly an order of magnitude different. -6.4 sounds more reasonable. I wonder if they left out a decimal on leadlike's manual?

John

Well, I'd chalk that up to a misprint. Especially after tonight when I saw that I am supposed to have 345v on pin 2 of the 6bg6. I'm not an expert, but I do believe that would blow out the tube's 6v rated filament. From the schematic, my plate voltages on the 6sn7 would appear about 20% low, but without some reasonable documentation, that will be difficult. Could anyone provide me with the proper voltages for the tubes in the horizontal section?

Checkout post #65 in this thread :)

Woah! Now that is quite different from my numbers! Your info is telling me that I am off by about 100v on my 6sn7 plates (v21 a&b). Guess I better replace those plate resistors to get that back in line...

Here is a screenshot of where I am at right now. The picture doesn't roll or anything like that, just a bad video picture and some pretty bad flag-waving of the raster.

Attachment 166745

Weird. I downloaded the GE 803 Rider's courtesy of the ETF. http://www.earlytelevision.org/tv_sc..._diagrams.html

It's pretty darn close to the 802 and it has loads of waveform measurments in addition to the socket voltage chart. Maybe you can use a scope to check your waveforms ?

You're right in that the 803 schematic is a close match. In fact, a lot of, if not all of the component numbers match up to my 802 schematic.

I spent another day hammering away at this thing. Primarily I was fooling around with that awful video noise producing that black blob of static on the edge of the screen. My scope traced the issue right back to the video input, and after tacking down another bad ground, I had a cleaner picture.

I also fooled around with v21's plate voltages, which should be about 230 and 245vDC. My set was only around 130 and 145 volts. I had to change the plate resistors quite a few times to up the voltage enough to this "stock" setting. It provided no difference in Horiz. lock, and only seemed to introduce some drive lines, so I went back to the lower voltage. The plate resistors (100 and 150k) seemed to be original, and were within 5% of their values. The 801 has lower plate voltages like mine...so mine is probably running properly. The waveforms are certainly clean.

I corrected a blooming issue by replacing a 470k resistor hidden inside of the 1B3's base (the 1B3 is mounted on 3 ceramic standoffs, the wiring in the base covered with a metal shroud).
So here we are:

Attachment 166841

Attachment 166840

I can now use the 6al5 properly, so DC resto is working pretty well. Retrace lines are visible, but I wouldn't be surprised if this set lacks vertical blanking. The picture still has a bit of waving to it, but it is extremely slight. I included the second picture, as it illustrates the only other issue I have with the picture-those faint vertical bars off to the left side. But, it's nice to have a watchable picture on this set.

Looking good! :yes:

John

Sweet :thmbsp: Maybe you can try this modification to eliminate those retrace lines ?

Yesterday I got a great deal done, along with the usual amount of heartbreak and compromise. But as the day ended, I had a working set in my living room.

First, the bad news: I pulled the crt to clean it, and saw that it has hairline cracks all along the neck. This tube is a rebuild, and there are two large cracks running along the length of the rebuilt portion of the neck. Whether this is a failure of the rebuilder, or as a result of it being moved roughly (though I packed the tube in a fitted box every time it was remove from the chassis) it is tough to say. As it is, the tube is holding its vacuum, and produces and excellent picture, so it will work for awhile. I imagine it will continue to function until the cracks grow around the crt neck, and the whole thing pops apart.

Luckily, I have a 10" Admiral that I am converting to a 12" set, so I had a spare crt. I will use the cracked 10" for a tester until it completely comes apart. The screen isn't aluminized, but it still produces a brilliant picture. I added A/V jacks to the chassis, adding audio at the selector switch (so the radio/phono audio isn't overidden by the tv audio) and video directly into the grid of the video output tube.

I also dialed in the B+ on an adjustable wirewound power resistor. Because I used a dropping resistor to burn off the excess voltages produced by my replacement transformers, this creates an issue when switching from Tv to radio/phono. When switching from tv to any other function, most of the set's tubes are shut off, altering the output current of the transformers, and effectively raising the B+ by about twenty five volts. My compromise was dropping the B+ by about twenty volts when the tv is powered up, so when one switches to radio, the voltage will only climb to about five volts over the rated B+. Filament voltage is not affected by this, only the plate voltage, so I shouldn't have to worry about cathode poisoning or any of that good stuff.

The final modification was adding the vertical blanking circuit.Following the ETF link Banderson provided, I reduced the retrace lines a great deal. It was fun to watch those waveforms on the scope.

I now have the set in the living room on top of our CTC-9. It is hooked up to a temporary speaker and dvd player. I plan to play this thing for as many hours as I can this week to shake out any other issues. It played for about four solid hours last night, and I never once had to adjust anything. These are solid sets, and probably would have rivaled the 630 chassis if GE hadn't designed it to be anything other than a 10" set (IE, no provision for space on the chassis beyond the stock crt, making the radio integral to the chassis, etc.).

Nice job :thmbsp: Glad the ETF mode worked. Too bad about the CRT though - mine's just about dead too.

I was wondering about that mode switch and the TV tubes. I suppose they're cutting off the juice to the filaments ? Must be a hefty switch if they are.

I can't wait to get started on mine :D

....and here it is, playing away. The gentleman in the picture is actually somewhat relevant to this forum, so keep an eye out for him in another posting.

Attachment 166862

LOL - it looks so much smaller outside of the cabinet. Nice picture too - both the TV and photograph that is.