Another RCA CT-100 discovery B8001599

[Penthode]

Originally Posted by old_tv_nut

The CTC7 and later chassis used the X-Z demod with output matrix/amp/DC restorer stages. Any direct R-Y/B-Y parent would not apply. Plus, it was more stable with tube drift due to the final DC restoration.

I had assumed the earliest sets would have all included I Q demodulators. Interesting that Westinghouse, purported to be the first color set sold, did not employ it.

[trinescope]

Originally Posted by old_tv_nut

The CTC7 and later chassis used the X-Z demod with output matrix/amp/DC restorer stages. Any direct R-Y/B-Y parent would not apply. Plus, it was more stable with tube drift due to the final DC restoration.

So did the CTC5 Deluxe; that one appears to be the first time RCA used the X-Z scheme. The demodulation angle, according to the RCA service data, is 57.5°. That got played around with in later chassis using the FBP/FJP CRTs due to the all sulfide phosphors those tubes used in an attempt to get subjectively correct color rendition from CRTs that no longer conformed to NTSC specs.

[etype2]

At the risk of not straying too far off topic, actually the Admiral C1617A was the first with more than 1000 copies made and the Westinghouse H840CK15 was second with approximately 3500 copies made. RCA CT-100 was third with about 4400 copies made.

Admiral was in stores in January, 1954, Westinghouse March, 1954 and RCA April, 1954. Availability of all sets were very limited. As an example, the CT-100 wasn’t available to purchase in NYC until mid May, 1954.

I don’t know what demod scheme Admiral used with their C1617A.

EDIT: Clarification, The RCA could be ordered in April, 1954, but not available until about mid May, 1954.

[Electronic M]

Originally Posted by etype2

At the risk of not straying too far off topic, actually the Admiral C1617A was the first with more than 1000 copies made and the Westinghouse H840CK15 was second with approximately 3500 copies made. RCA CT-100 was third with about 4400 copies made.

Admiral was in stores in January, 1954, Westinghouse March, 1954 and RCA April, 1954. Availability of all sets were very limited. As an example, the CT-100 wasn’t available to purchase in NYC until mid May, 1954.

I don’t know what demod scheme Admiral used with their C1617A.

You nerd sniped me on that last line...After a visit to earlytelevision.org I now know it's an I/Q demodulation set.

The GE 15CL100 that IIRC also beat RCA to market is an R-Y/B-Y set with the added weirdness of not having a chroma osc but instead a crystal ringing circuit like a portacolor....The portacolors demodulator comes from a different brand of early color set the 1957-62 Motorola 21CT2.

I now want to read more about what demod circuits the various early color sets had.

A few weeks ago I got out bid on a 21" Raytheon that I happened to have the Sam's for and it's demod was a fairly novel design.

[old_coot88]

At the risk of drifting the thread a bit more, we had a GE roundie come in the shop with no color sync. After subbing tubes to no avail, noticed a little Ne-2 neon bulb lurking down in the burst area, and it was lit normally. Figured, change it anyway. Perfect color sync. Never pulled a schematic, but assumed it must be serving as burst gate into the ringy circuit.

[old_tv_nut]

Precisely.
They also used a simple RC phase shifter to get the desired angle, so no quadrature alignment was needed.
If you compare phase shift components year to year, you can see just when they adjusted the angles.

[etype2]

The NTSC color space was never truly adopted due to the very saturated primaries and low luminesce from television receivers of that period.

[Penthode]

I have been looking forward to the Christmas break to pick up on the CT100. The rest of the family have been sick (not me and my wife) so we have postponed family gatherings this year.

Have completed two run throughs of the IF/RF and Chroma Bandpass alignments and testing. Subsequent run throughs of the procedure yielded some improvements in an attempt to get better frequency responses. I had spent many hours on this starting in late fall bringing me to this culmination point.

To achieve a flat topped wide IF response is the challenge especially with a stagger tuned amplifier with 6 stages! The video IF alignment of the CT100 require you to first sweep the 6th IF to 2nd detector stage followed by the 1st IF stage on the tuner followed by the first 2nd IF stage (first stage on the main chassis. Then the 3rd, 4th and 5th stages are peak aligned to 44.9MHz, 42.9MHz and 41.3MHz respectively. The final overall sweep requires you to adjust the 3rd, 4th and 5th to achieve the full 4.1 MHz bandwidth flat topped response. I could not achieve the response without tweaking the other stages. I feel it important to restrict the final response tweak to the three stages so experimentation was required to understand and determine where the discrepancy occurs. I found the problem with too much gain around the video carrier (45.75MHz) by very slightly altering the 2nd IF tuning higher in frequency to bring the whole respnse in line.

The luminance channel response with all traps and the chroma channel bandpass responses fell niceling in line with the RCA documentation. I had problems however with the I and Q channel responses. Recall all of the white peaking coils were open and I could not repair them.

I managed to wind myself replacements which worked fairly well. Q Channel 6.7mH coil I altered from a ferrite core 33mH coil only to find the the Q response diminished sharply after 100kHz. I ended up winding about 500 turns of AWG44 wire on a 1.0 Megohm 1W brown resistor which gave me the correct 500kHz response.

The I channel with the coils I provided peaked too sharply at 1.0 MHz. I found an old NOS Meissner 1.0 mH coil shunted with a 10K resistor which yielded the response I desired.

A frequency sweep from the 1st Video amplifier input to the kinescope grids did produce the response published.

[Penthode]

Here are a few more photos including the published notes photos.

For for the IF and baseband bandpass alignment, I used the combination HP8601a RF Sweep Generator with HP8600a Digital Marker. The HP 8601a sweep rand is from 100kHz to 110 MHz.

For the complete VHF Tuner alignment from Channel 2 to 13, I used a Wavetek Model 1050 Sweep Generator with the HP8601a as the Frequency Marker. For Channel 13 marking, the HP8601a second harmonic was easily seen.

(The HP 8601a goes up to only110MHz and the second harmonic of 105.6 MHz is 211.25 MHz which is the video carrier on channel 13).

As my scope has a 100MHz bandwidth, I did not feel necessary the demodulator probe suggested by the RCA notes. Hence my snapshots show the un-demodulated or un-rectified carrier frequency sweeps.

[Pete Deksnis]

Originally Posted by Penthode

I managed to wind myself replacements which worked fairly well. Q Channel 6.7mH coil I altered from a ferrite core 33mH coil only to find the the Q response diminished sharly after 100kHz. I ended up winding about 500 turns of AWG44 wire on a 1.0 Megohm 1W brown resistor which gave me the correct 500kHz response.

Thanks for that info.

Found something similar.

When L42 is replaced with ferrite, the Chroma Reference Osc Control is unstable.

A Miller 4624 will work. Even a pair of series connected Miller 5220's works.

[Penthode]

Originally Posted by Pete Deksnis

Thanks for that info.

Found something similar.

When L42 is replaced with ferrite, the Chroma Reference Osc Control is unstable.

A Miller 4624 will work. Even a pair of series connected Miller 5220's works.

Curious. I will have to research this further how perhaps the ferrite affects the "Q".

[Penthode]

With the initial RF-IF-baseband channel alignment done (including setting traps. Last week while checking the I channel I traced a self oscillation to an open 20uF chasis mount capacitor. I had on hand an NOS triple 20uF 450v which I replaced. (The rest of the original chassis mount electrolytics are original after reforming and checking for leakage. I must have missed the one that was totally open).

All testing so far has been minus CRT. Time now to next test the chassis connected to the 15GP22. Using a set top box, I had scoped good video up to the CRT grids. The sweep circuits were good and there was good high voltage available to feed the convergence, focus and second anode connections.

I managed to get a stable picture with lots of brightness and good contrast. I noticed the Keyed AGC is a bit wonky and need to investigate this. After this setting up colour demodulation, purity and convergence.

The 15GP22 seems to have very good emission. The tube is a rebuild: are the bright filaments normal?

[dtvmcdonald]

Originally Posted by Penthode

I have been looking forward to the Christmas break to pick up on the CT100. The rest of the family have been sick (not me and my wife) so we have postponed family gatherings this year.

Have completed two run throughs of the IF/RF and Chroma Bandpass alignments and testing. Subsequent run throughs of the procedure yielded some improvements in an atempt to get better frequency responses. I had spent many hours on this starting in late fall bringing me to this culmination point.

I commend you on those plots. They are truly top notch.

[Penthode]

Originally Posted by dtvmcdonald

I commend you on those plots. They are truly top notch.

Thanks.

Want to tackle the chroma demodulators. But did purity convergence and gray scale next.

I found it difficult to get a consistent grey scale. I find when I advance the brightness and contrast beyond a certain point, a high luminance level (towards white) causes the picture to go out of focus and the raster expands suggest poor HV regulation. Need to look at the HV rectifier...

What is the technique to get a good consistent grayscale?

[dtvmcdonald]

Good grayscale???

Answer: CONSERVATISM and a dark room.

You need a dual probe scope that can subtract. I have a pile
of Tektronixs that do fine.

Attach probes to the CRT red grid and any CRT cathode (i.e. either side
of Sam's R284.) Scope in subtract mode so you measure the grid to cathode voltage.

Adjust for a good picture, not very bright, well below blooming or other "obvious"
problems. The image you are looking at needs lots of both full black and
full white, and preferably full red too.

When you turn up either contrast or brightness you will note double clipping of the red.
and thus compression.

The DC restorer clips one extreme and the CRT itself clips the other. Turn up
the contrast too much and they BOTH clip. You want to avoid this like the plague,
its bad for the CRT and makes the picture turn cyan.

So what I did was to make up little arrows out of red and blue masking tape.
I set the contrast and brightness at the middle of the "usual" range and attached
a blue arrow to both knobs, pointing up. I then turned up the contrast to just
below the double clipping point and attached a red arrow to it.
i returned the contrast to the blue arrow and turned the brightness to the double
clipping point and attached a red arrow to its knob.

So, when adjusting the set, if you DO set one knob above the red arrow, you
need to set the other one below the blue arrow by an appropriate amount.

I also added blue arrows to hue, color, and tone.


You could also attach the scope across R284 and set the arrows so you never
get (dc) current flow (you will of course get spikes at sharp edges.)

One other note: at constant line voltage, my settings for these have remained
correct for three years now! Never Twice the Same Color is simply wrong.