1957 21" RCA TV Restoration"

A customer bought this set at an estate sale and wanted it restored. It's a big chunk of furniture but has 1950's written all over it, including the "V" on the control panel door and the picture tube shaped station indicator at the top. This is a 21" television, made in the 1957 time frame. It has relatively conventional circuitry and by this time, manufacturers were starting to use printed circuit boards (PCB's) to install many components.

The customer indicated the set had been powered up at the estate sale and that there was some "smoke". This is always a bad sign and can cause damage to hard-to-find components and as it turned out, this set was no exception. The following are pictures of the top and underside of the chassis:

I removed the back cover and began a preliminary assessment. First, I checked the 21AMP4 picture tube. It had been replaced in 1964 per a service card found in the set. The tube tested good. I then removed the chassis and put it on the bench. I noticed a pile of charred substance on top of the chassis between the HV cage and horizontal output tube...the source of the smoke? I looked underneath and couldn't find a source of overheating. I put the ohmmeter to the various sections of the flyback and got readings comparable to the schematic. I then used my EICO 944 flyback tester to confirm there were no shorted turns.

With the CRT and flyback looking OK, I began the rebuild. Most of the components were on three separate PCB's: the horizontal oscillator board, the sound board and the video board. Some components had been replaced in the past and this was done by snipping the old components and "tacking in place" the new components. Since I wanted to inspect the boards themselves, I decided to remove the boards one by one and replace the capacitors properly. The boards connect to the chassis via a number of "wire wrap" posts. I ditch this method and use female molex connectors with the appropriate diameter. These connectors simply slip onto the wire wrap posts, making the board easier to remove and install in the future. The following is a close-up of one of the rebuilt boards with it's molex connectors.

After all the capacitors were replaced, I put the chassis back in. It was at this point that I noticed a problem with the yoke. This is an assembly on the CRT neck that consists of four separate windings and provides for horizontal and vertical deflection. It didn't look so good and I quickly realized where the charred material came from. I also realized I should have verified this component first before diving into the capacitor replacement...yokes are not easy to come by. Oh well, I'm at this point, time to deal with it. A quick eBay search yielded few if any yokes for sale, let alone one for this set. I removed the yoke from the CRT and assessed the damage. The overheating appeared to be at the back end of the assembly where the four coils connected to a terminal strip. I carefully disconnected the coil leads and tested each on the 944. They were all OK! But now I had to repair it. I used a piece of fiber reinforced phenolic board to fab a new terminal strip, using the actual terminals from the old strips. It worked well. I also totally disassembled the yoke at this point, doing a little cleaning before re-assembly. The following is a picture of the repaired yoke:

With the yoke repaired, it was time to assemble and power up the set. With hand on plug for a quick power-down if needed, I turned on the set. The tubes lit but that was about all...no hiss of the flyback. But I did start hearing another hiss and shortly after that, the stench of a transformer in melt-down mode. I also noticed some fireworks in the 5U4 rectifier. Not good and the plug was quickly yanked. The power transformer by this time was very hot and the radio room stunk bad. I pulled the 5U4 and re-applied power...same result, and it indicated the transformer itself had serious issues. The chassis came back out for inspection and assessment. Earlier in the rebuild, I measured the HV winding resistances. They were on the low side compared to the schematic but were both approximately equal. I repeated these resistance measurements and now they were even lower. Yep, the HV winding had shorted coils. The 5U4 had a totally disintegrated filament and one of the filament pins on the 5U4 socket (phenolic wafer material) was burned/carburized. This was also the pin that provided B+ so things started to tie together. Who knows what took out what, the bottom line was that the transformer was toast.

Over the years I have accumulated a number of power transformers at auctions and fortunately I had something that was close...proper HV voltage and current, and a 6.3 vac winding. I also needed a 5.0 vac winding for the 5U4 (the plan was to replace the 5U4 and the socket). But for some reason, this transformer had a 3.7 vac (unloaded) winding. Bad 5.0 vac winding? I decided to abandon the winding, thus abandoning the use of a 5U4. 1N5401 diodes to the rescue! I installed a couple of those, plus a roughly calculated B+ voltage dropping resistor since silicon diodes have a much lower voltage drop than a 5U4. I had to drill a few new holes but the new transformer fit nicely! Here is a picture of the new transformer.

Now it was time to power up again. This time, no smoke but no HV either so there was obviously more work to do. After some channel fine tuning, I was greeted with clear sound! This was good news and indicated a good portion of the set was now working. I replaced a few tubes in the horizontal circuit but things never seem to be this easy. I then hauled out the heavy artillery...the B&K 1077B. This device is a great tool for working on old tv's. I drove the plate of the 6DQ6 with it and was greeted with high voltage and a raster. That's good, I now know the flyback is good and the vertical and horizontal deflection circuits are working. Applying grid drive to the 6DQ6...not so good. Very faint illumination of the picture tube and not what I would call a raster. Further troubleshooting was required.

At this point, testing is a bit of a pain. It is very difficult to do detailed testing with the chassis installed and everything pretty well needs to be connected to do any meaningful testing. I ended up putting the chassis vertically, supported by the bottom of the cabinet opening and fastening the top of the chassis to the top of the cabinet with a piece of wire. This allowed everything to be hooked up and exposed the underside of the chassis for testing.

I ended up spending a LOT of time trying to figure out why the B&K horizontal grid drive was not driving the 6DQ6. At the end of it all, it was attributed to "operator error". For some reason I assumed the "horiz-off-vert" switch on the B&K needed to be in the "horiz" position. In reality, it needed to be in the "off" position to drive a tube. What added to my confusion was the fact I was getting the proper grid drive waveform in the "horiz" position. I looked at the B&K schematic and I still can't figure it out but the bottom line is that the horizontal grid drive worked when the switch was in the "off" position.

Now I was getting a raster from the 6DQ6 when it was driven by the B&K but the set would not drive it. This meant it was time to troubleshoot the horizontal oscillator. It is a conventional design utilizing one tube as a "multivibrator". The board had been taken out earlier at which time I replaced one resistor and properly installed some sprague "orange drops". I removed the board again and replaced all the "domino" mica capacitors; there were about 3 or 4. I used the more modern silver mica caps but since I didn't have exact values I used what I had that was closest and for one I paralleled a few to get the proper value. Put the board back in...still no oscillator. I put the scope to a few points and I was getting nothing that resembled the waveforms indicated on the schematic. There was obviously more work to do. The board was removed for a 3rd time and this time I replaced almost every resistor as a few were significantly higher than original and many had been "lifted" at one end and resoldered, presumably a previous tech had tested them. I also took a closer look at the silver mica caps I put in and discovered a few leads did not solder well. I remedied that and put the board back in.

Finally, success! I was now getting the proper waveforms and the set was producing a raster on its own. I tweaked the board to get the desired waveforms. HV was low at 12 kV so I replaced the 1B3 with an NOS one and this brought the HV up to close to 15 kV. The picture was watchable but not as good as it could be. I tested all the tubes in the tuner, IF strip and video and found a few that were on the weak side. Following replacement with NOS tubes, I now had a better picture.

As can be seen in the above picture, the set lacks resolution. An alignment will likely be needed to improve the picture.