The Set: Pete Deksnis's Site about the CT-100

Restoring a Vintage Color Television Set


restoration log

06-26-2005 This new restoration log page introduces a new CT-100 to the restoration spectacular.

I was fortunate to have acquired a second Merrill in August 2004 after nearly a year of negotiations. It is chassis number B8000194, a brother to my original CT-100. This Merrill was rolling down the Bloomington assembly line just hours behind B8000173, the Merrill that spawned this site in 1999. They are virtually identical in terms of production strategy.

With the B8000194 chassis, I initiated a restoration process anew by following the approach set down in the first four restoration logs. Those early logs were meant to repair an ancient CT-100 and return it to operation. This continuation of that process, however, is designed to improve the reliability of a CTC2 chassis as well as to repair it -- a true
restoration .

After following the initial suggestions in the restoration log, the set received an over-the-air signal and successfully reproduced the audio portion of an NTSC signal. I then selected circuit functions associated with individual tubes and restored those 'islands' of functionality by replacing low-reliability components and testing each circuit function in isolation using an external power supply.

Here for example are the
horizontal oscillator and control and the vertical oscillator and output circuits as restored and individually tested.

After ninety dollars worth of modern capacitors and a month of work, B8000194 was revived on May 29, 2005 with a double delight to the senses-- sound and luminance.

Selected issues faced during restoration of B8000194 include:

(1) The fifth and final intermediate-frequency amplifier in the video chain is a 6CL6 -- a power pentode and an unusual IF-amp choice. It was selected because of the healthy signal that drives it.

[Note. On a CTC2 chassis proper, there are five video IF stages, but there are seven intermediate-frequency amplifier stages when you consider the 6U8-based two IF stages in the tuner. Documentation seems to be evenly divided between these two interpretations of IF configuration (some say there are five IF stages, and some say seven). In actuality, there are seven active devices through which the signal from the first detector is passed.]

This seventh IF amplifier, its IF transformer (L27), and a 1N60 second detector diode in an aluminum ‘can’ that includes a sound-carrier filter (M13), are all housed in a shielded case that extends both above and below the main chassis. Inside the shielded case on top of the chassis, the 6CL6 runs hot; there is one documented restoration where this socket was discovered damaged by the heat and required replacement.

Below the chassis, the shielded area is shown with its cover removed.

Perhaps it contributed to the abandonment of this Merrill in 1956. Note the 2.2K, 1/2W, 5% carbon resistor connected to terminal 'E' of M13 and ground, shown close-up below, see arrow. This resistor is damaged and varies in resistance from 10K to over 100K. It is the lower component of a two-element resistor divider that drops the 275-volt bus to about 3 volts and is used to contribute bias to the first video amplifier, or so it should.

Fortunately, the overvoltage that occurs when the 2.2K resistor drifts does not damage M13 circuitry, including its 1N60 second detector, but it did play havoc with the video-amp bias.

The replacement resistor was relocated outside the shield so that I could more readily monitor the drop across the 2.2K.

(2) Smell the ozone…

Two 50-meg ‘barber pole’ resistors in a series string with the 15-meg convergence pot in-between are connected from the 19.5-kV high-voltage to the top of the 1-meg high-voltage adjustment pot, which itself is 1.8 megohms above ground. Phew! The point is:

The metal end-cap of the second 50-meg resistor, R185, was arcing to the phenolic mounting board. If you look carefully, you can probably see two dark smudges caused by the arcing, which occurred where the end-cap hovered before the lead was cut.

A search through the Allied on-line catalog found a 100-megohm that was rated for the high voltage found in this circuit. As seen next in a bird's-eye view, a parallel-connected pair of the flat resistors fit nicely in place of old sparky. No more ozone.

(3) 07-01-2005 Andorian (think Star Trek) Effect. Here's the biggie. Hue, or color, is determined in the NTSC system by the phase of a sideband. The amplitude of that sideband controls the intensity of the color.

An oscillator in an NTSC color receiver generates a signal used by the set to recover hue and intensity from an NTSC color broadcast signal. The oscillator must be extremely stable; a slight change in the frequency of the oscillator will cause the recovered color to be incorrect. In fact, the 'frequency' stability is measured in terms of phase because it defines accuracy to a fraction of one cycle of the oscillator frequency.

Phase is measured in 'degrees.' There are 360 degrees in one cycle.

When the phase of the oscillator in an NTSC color set is not what is correct for a given color, that color on the screen will be wrong. In Merrill B8000194, there seem to be somewhere around 90 degrees of incorrect phase.

The result is the Andorian effect. Check the 'red' shirt on Merrill B8000194 in the lower right.

An NTSC color broadcast signal actually has two sidebands to carry color and intensity; they are called I and Q. When the information in the I and Q sidebands is recovered, or demodulated, and viewed on an oscilloscope, both channels are clearly seen to be operating. (A gated color bar generator signal is applied to the rf input in this example.)

For stability, the oscillator in a CT-100 is crystal controlled. But that alone is inadequate. An NTSC color television transmitter provides receivers with a reference signal that allows the receiver to 'lock' itself to the transmitter reference signal. Which all works together to get the colors on the screen correct.

Some of the components in a CT-100 that are used to lock onto the transmitter reference signal and generate correct hue are the crystal (3579.545 KC) and a control system that includes a CHROMA REFERENCE OSC CONTROL stage and transformer 1107319-1 (L43) shown here disassembled.

At first, the color problem was catastrophic because the transformer that is part of the chain controlling the crystal-controlled-oscillator (V29B) phase was damaged, as is evident by the flying lead from the transformer winding.

Once repaired, the set displayed stable color, but Andorian-effect color. That state exists to the present day even after installation of missing inductor L42. T'shooting continues...

Success on August 7. Our culprit was L41, the burst amp output transformer. The final adjustment to transformer L41 is not part of the original RCA or Sams alignment procedure. Steve McVoy discovered the omission in 2002 when aligning his second CT-100 restoration. In fact, he posted the information on this site for future restorers. Read Steve's updated procedure from April 2002 at the bottom of this link:
click here . Thanks too to John Folsom who also solved the Andorian effect. Thanks also to Steve Kissinger, Kirk Stankiewicz, Walt Love, and from Australia, Ian Robertson.

HDTV CRT on left and 15GP22 in lower right displaying fried stuff.
The CT-100 still requires convergence, gray scale,
and all other picture calibration.
August 7, 2005 photo