Working on my RCA 21D645 KCS97 again

[rca21d645]

Since my work on my Admiral 12X12 went fairly well, I feel like I may be able to tackle this problem again. Back in early 2018 I started working on restoring an RCA 21D645, the first antique television I ever worked on. I had someone who was helping me learn how to work on these sets, but eventually he just didn't have the time to help anymore. I then started working on it again during quarantine back in March of 2020, and I made the stupid mistake of replacing all the paper capacitors in the set at once without regularly checking if the set worked, and I didn't take any photographs either. After looking over the set I noticed one capacitor on the audio board had not been properly soldered, and after making sure it was secure I was able to get the audio back, but still no video signal. There is a raster on the screen, and I noticed that the 6AQ5 video output tube and the 6AW8 video amplification tube were not lighting up when the set turned on, and removing them had no effect on the set, moving them around in their sockets also did nothing. All the other tubes lit up just fine. I tested all the tubes in my tube tester and aside from one bad 6U8 with a short (which I replaced) the rest of them tested well enough that they should at least produce some kind of picture (some had weaker emissions). I checked all the capacitors on the video board and none of them seemed loose or disconnected. I've gone over the schematic several times, and quite honestly this would've been a million times easier if I had taken a reference photo all the way back before I replaced the capacitors, I learned from this mistake when working on the Admiral. I have the photofact and the RCA service manual for this chassis and while they do help quite a bit, the photo of the chassis in the photofact is incredibly grainy and its hard to make out the leads on some of the capacitors that aren't installed on the circuit boards. If there is any way someone could either take or direct me to a clear reference photo of the underside of a KCS97 chassis, I would be incredibly grateful.

[Kevin Kuehn]

Regarding 2 tube filaments not lighting up, have you checked your filament voltage with the 2 tubes removed and your meter probes top side in their socket holes? Assuming those sockets are soldered to the circuit board it's possible the solder connections are cracked or the terminal connections within the sockets are broken similar to what happens with Philco Predicta sockets. One way or another you need to verify that filament voltages are getting to the tube pins.

[Electronic M]

It sounds like your problem is that the video output tube heaters aren't lighting. Since the tubes test good that means the set is not supplying voltage to the heater pins of those tubes.

Caps usually have NOTHING to do with heater supply issues. To troubleshoot a heater supply problem you need 3 things that it sounds like you have. A brain, a working DMM, and a copy of the schematic. You don't need chassis pictures for this.

You need to measure the AC voltage across the tube heaters of the 2 tubes and trace their wiring. If the set uses a power transformer* (as ISTR reading it did) the heaters are wired in parallel. So either the heater feed wire that feeds both tubes failed, the heater ground connection failed (common on tubes mounted on printed circuit boards), or in the least likely case both sockets have open or damaged heater contacts. If you measure 6V across the heater pins at the socket the socket is bad. If you measure significantly less than 6V there take the probe that's on the grounded socket terminal and touch it to chassis ground...if then you get 6V your ground is open, if not keep one probe on chassis and trace the positive heater line to the solder points it connects to and measure voltage till you trace back to a point that gives 6V AC or you reach the heater winding of the power transformer (which ever comes first). Somewhere you will find a bad connection, bad wire, bad trace or open circuit.

How do you find the heater pins on the socket? The schematic will have symbols for the tubes. Locate the tubes on the schematic and locate the heaters of the tube they have this symbol (^) or this one (^^) on the schematic symbol of the tube. Most schematics put the socket pin number next to the wire connecting to the tube element (if not google for the tube datasheet and it will have that info).
The tube pin number on 7 and 9 pin tubes is counted up clockwise from the big gap (in octals it's counted from the bump on the Bakelite pin) between pins as seen from the bottom of the socket....(if you are looking from the top side of the socket where the tube inserts it's counterclockwise).
Sam's won't show you every wire on the schematic...the hot lead of the heater probably will go to an arrow with an X, Y or Z next to it. The power transformer will have a similar arrow with the same letter next to it. Heater wiring is electrically simple so to keep it from cluttering up an already complex schematic Sam's uses those arrows to indicate every arrow of that letter is electrically connected to every other arrow of that letter with some set of wires. (They do similar things with B+ wiring in most schematics except stuff from the late 40s early 50s).

*If the set is a transformerless series string heater set then there's a short bypassing current flow through those 2 heaters. In transformerless sets Sam's doesn't put the tube heaters on the tube symbols and instead clusters the tube heaters near the power cord to show how the series heaters are wired to simplify troubleshooting the heaters.

[rca21d645]

I should probably clarify, when I had only replaced the electrolytics it did produce a picture, although it was tripled as if there was an issue with the horizontal frequency. It does still produce a raster, but no picture.

[Electronic M]

Originally Posted by rca21d645

I should probably clarify, when I had only replaced the electrolytics it did produce a picture, although it was tripled as if there was an issue with the horizontal frequency. It does still produce a raster, but no picture.

The horizontal frequency being wrong is a secondary issue. If you don't have an oscilloscope and experience using it you have to get video back before you can see what you need to see on screen to properly fix the horizontal frequency (which may only need a service control adjustment). There is a way to adjust horizontal frequency with a scope I use sometimes that I could tell about.

Troubleshoot the dead video tube heaters to get video back and once that's done we can properly address the horizontal frequency.

It sounds like all lytics and paper caps have been changed...once a recap has been done you need to fix remaining problems a stage at a time and prioritize issues that make other issues harder to fix. If you try to solve 2 issues at once both become harder. Divide and conquer.

[rca21d645]

I removed the video board and I found there was a crack right on the trace to the heater of the video output tube. I bridged the crack and now the tv functions fairly well.

[rca21d645]

From the way the picture looks I'm guessing the yoke needs to be adjusted because the picture looks slightly crooked. I'm not sure about the distortion at the top, I think it might be related to the horizontal but I'm not sure.

[Electronic M]

There's probably a clamp that holds the yoke to the neck or a wing nut that holds the yoke to a bracket. Loosen that and rotate the yoke. May want to wear rubber gloves if rotating the yoke while on...it's the best way to get the adjustment right, but if there is any uninsulated part of the yoke windings, terminals or windings it can give you one hell of a zap (ask me how I know).

That distortion is odd. Part of me wonders if there's misadjusted pincushion magnets like some 50s Zeniths had, or if the CRT hardware is magnetized.

[JohnCT]

Originally Posted by Electronic M

That distortion is odd.

Really odd. The vertical non-linearity isn't uncommon, but the linearity changes from one side of the screen to the other. Has the generator been verified on a working TV? I've seen generators produce weird test patterns when *they* have issues.

Other than that, it seems there is some interaction between the horiz and the vertical sweep. This is where a scope really comes in handy.

John

[old_tv_nut]

It could be simply a B+ filter cap issue. If it's B+ and the source vertical is locked to the power line, the distortion will be stationary, but if it's B+ and the source is crystal locked to 59.94 Hz, like a DVD player, the distortion should move slowly.

[old_tv_nut]

If the source is a B&K Analyst, try it on another set to be sure it's not the B&K.

[rca21d645]

Originally Posted by JohnCT

Really odd. The vertical non-linearity isn't uncommon, but the linearity changes from one side of the screen to the other. Has the generator been verified on a working TV? I've seen generators produce weird test patterns when *they* have issues.

Other than that, it seems there is some interaction between the horiz and the vertical sweep. This is where a scope really comes in handy.

John

Yes it has, its a tape that I've used in the same VCR seen on this Admiral.

[Kevin Kuehn]

Could it be the keyed AGC control is set too high and overloading? No harm in tweaking that control just to get a feel for what it does to the picture. But many TV's simply don't work well with a Video tape signal.

[rca21d645]

Originally Posted by Kevin Kuehn

Could it be the keyed AGC control is set too high and overloading? No harm in tweaking that control just to get a feel for what it does to the picture. But many TV's simply don't work well with a Video tape signal.

That's as good as I could get it tweaking the agc, any farther and the picture would either fade to mostly gray or would become more jittery.

[old_tv_nut]

Is this the first time you've looked at the VCR signal on this set? It could be the recovery from the head switching transient that many pre-VCR sets suffered from.

Is it steady, or does it wave from side to side? Waving = head transient. Perfectly steady = some coupling from vertical sweep into horizontal, possibly a bad power supply decoupling cap.