When the Original Color TV Shows Only Black and White,
Who You Gonna Call?

For the 50 year anniversary celebration of the beginning of RCA's compatible color television, Sarnoff Collection Director Alex Magoun demostrated their working CT100 set, RCA's first commercial color TV. It gave a curiously bluish picture to my eyes but certainly showed a range of hues and saturations that suggested a mere tweak of the screen or drive controls could make it right.

For several months after the event, Alex ran it for visitors who were quite amazed to see a working set of that age.

Until the day the color died.

In an instant, during a demonstration for visitors, the color picture turned into a murky blue-green. The screens and drives and convergence were way out of whack anyway, so the picture seemed to have some color. One could make out green and blue shading and edges, but no warm colors at all, leading some to conclude that the red electron gun in this irreplaceable picture tube had failed. But, after adjusting the drives and convergence, it was clear that this color TV showed only Black and White.

Now, there is no use in demonstrating one of the first color TVs if it only shows B&W pictures. When an important event came close, Alex asked me to come by and see what I could do.

I proceeded with great caution on this set because it is such an important historical artifict, and it is quite rare to find one working condition. A slip of a test lead could have caused more damage, which would have been a very bad thing to do to this set.

First job was to attempt a normal picture using the front panel controls, which included the screen and drive settings (bias and amplitude) for each of the primary colors. Normalizing these and setting the convergence resulted in a normal black and white picture, but no adjustment of the fine tuner or the color control made any color appear.

Since there was only one day to do the job, I economized my effort by attempting to diagnose the circuit without pulling the chassis. This was possible because a metal screen under the chassis shelf could be removed and nearly all the circuitry could be accessed, but only by climbing under the set like an auto mechanic on a dolly. After each reconnection of the scope probe, I had to shimmy out to look at the trace. It came out to about five minutes for each test.

It appeared that the local chroma oscillator was not synching up to the incoming signal. The correction voltage read as a sawtooth at the horizontal frequency. This could have prevented phase lock and caused the color killer circuit to stop any color from reaching the screen. A discussion with color TV veterans at the lab led to the likely theory that the chroma oscillator crystal had gone bad. They had a drawer full of them and gave me a new one. The replacement had no effect. Testing and swapping all the tubes in the chroma circuit yielded no clues. By the end of the day, no progress had been made and I abandoned the effort. The set was left idle for the big event.

Several days later, I came back for another round of wrestling with the CT100, but this time there was no time limit. The chassis would come out.

Chassis removal on the CT100 was unexpectedly easy. Removed all knobs, plugs to the speaker, CRT, and deflection yoke, pulled out the bolts, and out it came. It's a gigantic chassis and therefore easy to work on. I also had an advantage I hadn't earlier - the SAMS PHOTOFACT manual with its resistance measurement charts. Now I could search for problems without powering up the set.

I followed the path of the color signal, starting with the chroma amplifier. My probe was a Radio Shack digital multimeter in a pen-sized package I could hold in my hand and poke the tube sockets with - excellent for such tests. A few resistors did turn up out of spec (more that 20% off), and in sensitive circuits like the phase lock loop, this could cause a problem. Sarnoff's parts department is still reasonably stocked with vintage ½ watt resistors, many probably from the 1940's. These tend to drift with age even when unused, so I checked them before installation.

At the I and Q demodulator tubes, a reading came up that was way off, and I mean way off. These 6BY6's are just like the pentagrid converters in the AA5 radio input stages. The second grid actually doubles as a "leaky" plate, which passes on the input chroma carrier to a subsequent grid which modulates and detects the chroma signal with the two phases of the local chroma oscillator. The second grid in this set read infinity to B+, but these two tubes are supposed to be connected to B+ through a 10 watt 12.5 K power resistor. This resistor, which looked perfectly normal with no sign of overheating, was completely open. The effect of an open resistor there would be a perfect B&W picture. The stock room did not have a 10 watt 12.5K resistor, but a pair of 5 watt 25K resistors in parallel made a perfect replacement.

I finished the resistance tests in the chroma signal path, and with complete confidence that the set was repaired, reinstalled it and turned it on. Eureka! The world was a wonderland of color. After adjusting every front panel control for the best picture I could get without the help of test patterns, I buttoned it up and declared it fixed.

There are some remaining issues to visit in the future if we decide to guild this beast. Fine tuning is a bit touchy: one way and the sound makes wiggly herringbone patterns over the picture, the other way and the picture fades to B&W. Convergence could be better. If the screen gets too bright, high voltage gets wacky and the picture blurs and tears. With the bright lighting in the museum, coupled with the tendency of the set to become more likely to do this as it warms up, makes this a disturbing symptom that could spoil a demo for visiting VIP's. It will have to wait for another day.

In the meantime, the color rendition is probably better now than it has been since the set arrived last year, and it has impressed two NJARC meeting groups since it was fixed.