RCA 21CT55 B8802897 is Alive!

[miniman82]

Originally Posted by Tomcomm

In the meantime retake and post the 3 tv cluster pix you have.

Hopefully I can get the 21-CT-55 up and running soon, then I will have a pic with a modern monitor, 21CYP22, 21AXP22A (grey), 21AXP22 (green), and green face 21FBP22.

[stromberg6]

Originally Posted by andy

I believe RCA made a chassis in the early 80's with I Q processing. I think it was a CTC13X chassis, but I forget the exact chassis number.

My CTC-136B is wide-band, manufactured in 1986, and has the I-Q chipset. I'll look at the SAMS to see about the single-ended I delay line. I'm getting much from the topic of the "glow" that may be due to grid contamination. I'll see if I can find an hour or more later tonight to fire up my 4, and look for the glow. Thanks for all the info!
Kevin

[Tomcomm]

Originally Posted by Tomcomm

To be truly objective, you must include the gray screen 21FBP22A rare-earth. I have been playing with the color balance feature of my Photo Impact graphics program that reads the RBG values 0 to 255 of any spot on the JPG image. This is the hacker's approach to chromatic spectrum analysis but should give quantified data on relative phosphor differences of all your CRTs.

I have now performed the above Photo Impact “spectrum analysis” on some of my 21CT55 screenshots verses ’84 Sony pro monitor and ’97 Sony 27in tv. All pics were shot with the same Canon S40 and were not color enhanced. I used the DVE color bar for all screenshots tests and “normalized” each pic by placing the virtual sampling probe onto each white bar and setting the brightness to produce approximately the same digital 255 max red, green and blue components. I then placed the sampling probe onto the red bar and took digital readings of its red, green and blue components. Ideally, a perfectly pure red bar should produce a red component of 255 while its green and blue components should read zero. Only the color accuracy of the Canon camera sensor is evolved here since the LCD computer monitor’s color accuracy is never an issue. Since relative CRT phosphor chroma response is all we are interested in, any variance of these readings indicates contamination of the ideal red phosphor’s light output, right?

Preliminary estimates of red phosphor purity of a ’88 13in Sony pro monitor, a ’97 27in Sony "modern" tv and of course the 21CT55 with 21FBP22A “rare-earth” crt follows:

21in FBA.……………Red contaminated with 5.6% green and 8.2% blue
13in Sony……………Red contaminated with 12.6% green and 6.3% blue
27in Sony……………Red contaminated with 29.2% green and 21.3% blue

No wonder the “modern” TV crts appear orangish red!

[old_tv_nut]

Originally Posted by Tomcomm

21in FBA.……………Red contaminated with 5.6% green and 8.2% blue
13in Sony……………Red contaminated with 12.6% green and 6.3% blue
27in Sony……………Red contaminated with 64.2% green and 53.3% blue

Those numbers for the 27 inch Sony should correspond to a visually washed out red - something is not right with those numbers, unless you report that visually it appears that bad.

In regards to "the red bar color should normalize to (255,0,0) RGB," I'm afraid it just ain't so, for a whole slew of reasons: the impossibility of getting ideal filters in the camera, the different spectra of the displays that interact with the filters in the camera; the raw to jpg conversion in the camera having a matrix that is designed to give pleasing color, not accurate color; deliberate non-linear S-curve processing in the camera to mimic the pleasing contrast response of film; etc. etc. etc. etc.

The human eye is a much better instrument for these comparisons than a digital camera.

[Tomcomm]

Wayne, thanks for the reply. I reran the 27in Sony by brightening the white bar. It was way too dark before which caused the red bar to saturate and as you said, "wash-out". The new results:
21in FBA.……………Red contaminated with 5.6% green and 8.2% blue
13in Sony……………Red contaminated with 12.6% green and 6.3% blue
27in Sony……………Red contaminated with 29.2% green and 21.3% blue

Somewhat less contaminated but still way orangish red compared to the Sony pro and the 21CT55 with 21FBP22A. The stated 255,0,0 RGB was "ideal" with no hardware limitations. I believe this simple and cheap method of relative phosphor comparisons can be quite useful when comparing the original NTSC rated crts eg 15GP22 and 21AXP22 with the gray screen 21FBP22A. All I need is optimally setup color bar screenshots of the original GP/AXPs. I don't have much hope for a live eyeball shootout that includes the gray screen 21FBP22A, certainly not one I may witness.

[Findm-Keepm]

Originally Posted by Tomcomm

Only the color accuracy of the Canon camera sensor is involved here since the LCD computer monitor’s color accuracy is never an issue.

Wouldn't the kelvin temp of the LCD monitor backlights be an issue, or is that somehow compensated for/already calibrated/or not an issue?

Mind you, I'm a amateur with all of this, but my two LCD monitors have two Kelvin settings that affect the color spectra - 6500K (default) and 9300K.

Interesting thread! I've worked TV's for 30+ years and about all the discussion I've ever heard is from a old, now-dead TV tech that swore by the "old" RCA red versus the "new" RCA red. I guess he too would've been interested in this thread.

I'm learning, keep posting!

Cheers,

[ChrisW6ATV]

Tom-

That is a very interesting test. Are you able to set the camera into a completely "manual" mode, so it doesn't do any type of color balancing on its own? If so, do you also need to pick a starting color-balance mode? I know mine has several settings for modes but I have not studied its operation very much.

[old_tv_nut]

Originally Posted by Tomcomm

Wayne, thanks for the reply. I reran the 27in Sony by brightening the white bar. It was way too dark before which caused the red bar to saturate and as you said, "wash-out". The new results:
21in FBA.……………Red contaminated with 5.6% green and 8.2% blue
13in Sony……………Red contaminated with 12.6% green and 6.3% blue
27in Sony……………Red contaminated with 29.2% green and 21.3% blue

Somewhat less contaminated but still way orangish red compared to the Sony pro and the 21CT55 with 21FBP22A. The stated 255,0,0 RGB was "ideal" with no hardware limitations. I believe this simple and cheap method of relative phosphor comparisons can be quite useful when comparing the original NTSC rated crts eg 15GP22 and 21AXP22 with the gray screen 21FBP22A. All I need is optimally setup color bar screenshots of the original GP/AXPs. I don't have much hope for a live eyeball shootout that includes the gray screen 21FBP22A, certainly not one I may witness.

This is still a little suspicious. Getting purer colors when the drive is turned up could indicate that the black level is a little high or there are reflections of room light. This would be a valid test of how the tube is set up and viewed, but not necessarily the characteristics of the red phosphor color alone. To get the comparison of the pure phosphor colors you need to do two things:
1) take the picture in a completely dark room, making sure there are no reflections (like from your shirt)
2) turn up the color control and adjust the tint if necessary to make sure only the red beam is on and the other two beams are completely cut off.
2b) Since the color demodulator may include color corrections that deliberately make the primary bars impure, you may need to readjust the tint individually for each primary color if you want to measure the green and blue also.

[miniman82]

Why not use one of the TV tools with 'gun killers'? They include a feature which shunts the guns grids through a 100k resistor at the flip of a switch, though you could do it with clips as well.

[Tomcomm]

Some misconceptions regarding this phosphor spectral analysis procedure seems to exist. My objective was to evaluate its potential to determine the relative spectral purity of the original NTSC spec’d dim CRTs versus the later, brighter CRTs e.g. 21FBP22A:
1) Only previous informal archive screenshot JPGs were used. No rigorous colorbar setup was used regarding black level, chroma level or luminance (contrast) level. All screenshots were taken in total darkness. Purity and focus were always good.
2) The playback LCD monitor is used only to direct the location of the virtual pointer over the JPG image. It’s chroma characteristics contributes nothing to the JPG phosphor contamination determination. Only the Canon S40 sensor can influence the results.
What is needed now is member postings to this thread their quality colorbar screenshots of their 15GP22, 21CYP22, gray 21AXP22A , green 21AXP22 , green 21FBP22 and gray 21FBP22A all taken in total darkness. Thanks for your interest and cooperation

[JBL_1]

Originally Posted by stromberg6

My CTC-136B is wide-band, manufactured in 1986, and has the I-Q chipset. I'll look at the SAMS to see about the single-ended I delay line. I'm getting much from the topic of the "glow" that may be due to grid contamination. I'll see if I can find an hour or more later tonight to fire up my 4, and look for the glow. Thanks for all the info!
Kevin

Just poking around and saw this thread. I was designing televisions at RCA in Indianapolis when the CTC-136 was made. While the Sanyo IC has I/Q demodulation it did not have the wide I channel that was only used in the CTC-131 and CTC-132. I seem to remember the short lived CTC-133 used the same RCA designed chip but the wide I feature was not incorporated. The CTC-121 was the first 25" in the space of a 19" but did not have the Wide I processing..

In those days RCA designed all the IC's in their receivers. I designed the deflection processor in the CTC-146. RCA in Sommerville was the first source and Hitachi was the second source. I remember having to tweak some resistors in the 32 h input circuit as hitachi's transistors behaved a little
different than RCA's.

The CTC-131 chip set included a wide I demodulator and a kine drive IC with auto kine bias. Those were the days....

The CTC-135/136 was a low end chassis designed by RCA Taiwan and overseen by Indianapolis. It used a Sanyo chip that was modified to meet RCA's specs.

I wish i could remember why the Wide I was removed from the CTC-133. Seems there was some reason.. I may have some of those IC's around if I can find them!

[stromberg6]

JBL-1, my previous post stated that I had a CTC-136B, but it's a 133B. There's a 135 in my daughter's bedroom, and it's not bad for a little Taiwan TV.
I wasn't aware that the I wasn't wide-band in the 133. The color resolution on my set is really good when compared to the narrow band TV's in the house, and it will reproduce small detail in color, oranges and cyans, etc, but a really good picture. Interesting information, and I thank you for it. I'll look for more info on this and similar chassis where ever I can find it.
Kevin

[old_tv_nut]

JBL_1, it's great to hear from someone involved with those sets.

The wideband IC demods were discussed in at least a couple a couple of threads here:

http://www.videokarma.org/showthread...ideband&page=2
See post #19 where I gave a quick explanation the single ended delay line, which, IMO, is one of the cleverest inventions in analog video IC design I have ever seen.

http://www.videokarma.org/showthread...nd#post2434742
See post #132, where I talked about my analysis of a sample RCA set with the wideband circuit while I was working at Zenith. Basically, it looked like the IF couldn't support the increased bandwidth without quadrature distortion, and the high-frequency I gain seemed to be turned down to prevent visible edge effects.

[Tomcomm]

Hi Members...........I have returned back to my 21TC55 "lab test bed" after about six months of doing other techi stuff. When I turned it on
about two weeks ago it came right up but was totally out focus due to CRT blooming. I checked all the standard operational readings I use
for setup: Room Temp=60deg, AC input=115vac, B++=424V, FBI=200ma, boost=770V, ultor=30KV, focus=5.4kv, all normal for this
"experimental" setup. After one hour running, RT was 70deg, AC=115, B++=415V, FBI=199ma, boost=800V, ultor=32KV, focus=5.8KV.
No CRT blooming, fine picture. I measured the CRT heater voltage direct at socket with a true RMS Fluke DMM. It read 6.16vac. I installed
an old CRT booster that raised the heater to 7.64vac. The neck was much brighter of course, so I left the TV on for only a couple of hours.
The set was running at 2% under spec heater all these years and now it was running 20% over spec for two hours. I turned the set off.

Nest morning at 7am, RT was 54. I powered up with booster off, the 21CT55 came up with no bloom in perfect focus! Readings were AC, 115V, B++, 420V, FBI, 200ma, boost, 860v, ultor, 33KV, focus, 5.6KV, heater read 6.01vac. After about 4 hours running with fans off, RT, 74deg, Ac, 115vac, B++, 422, FBI, 190ma, boost, 780v, ultor adjusted to 30KV, focus, 5.7KV, heater, 5.93vac. No blooming, good picture so I measured the FBX temp with digital thermo meter applied direct to primary and EHV secondary. Max temp was 112deg, barely warm!

Pleasant surprise, the CRT shutoff uniform glow problem was greatly reduced, almost unnoticeable! Could it be the 21FBP22 running low heater all these years allowed contamination to build-up on the cathodes causing loss of emission and/or causing grid emission that produced the glow? In-the-day the CRT boosters were quite poplar for restoring brightness and were left in place running at close to 8vac which shortened CRT life considerably. I always took them off after a few days/weeks, often with no noticeable decrease in brightness or increase in blooming.

I still have this problem with the hybrid combination of original CTC2B horizontal oscillator and 6CB5 HOT driving a CTC16 flyback xformer and
associated components, driving the original CTC2B yoke. This mismatch works fine but since it has no width control and only limited centering control, the screen is over scanned about 2 inches and is off center about 1 inch. This is unnoticeable except on test patterns.The BIG problem is the very noticeable horizontal video "fold-over" on the left screen presumably due to the horizontal sweep flyback starting up before the previous field's
video is blanked? The magnetic horizontal deflection of color CRTs is probably the most heavily engineered, most critical component matching
requirement, highest efficiency consumer electronic system of the 1950s. It obviously does not permit the "hacker" approach to customizing I usually pursue. I've tried all the usual modifications using all combinations of resistance, capacitance and inductance to reduce the horizontal width
and enable horizontal centering with no success and almost tragic results. Soooo, I am now designing a video blanking circuit that cuts off the foldover contaminated video for the first portion of every video line. It starts immediately after the chroma burst and extends three microseconds which
completely removes all visible traces of video foldover from the screen. Once I get this circuitry installed into the CTC2B under-chassis I will do a painful optimal convergence and be ready to take a batch of screen shots with Judy's new Sony DSC-HX9V 16.2 mega pixel pocket camera.

The CTC2B with "gray screen" 21FBP22A CRT combination has an extreme dynamic range that displays to the human eye details into deep shadows that our other cameras failed to capture. Hopefully, the Sony camera has a greater low-light dynamic range?

The video gate module is now installed, works perfectly! I'll probably do a convergence tomorrow and if I survive I'll take some screen shots with the Sony..........Tom

[miniman82]

Sounds like your CRT is going soft. A strong tube will not lose emission, even at 5.5 volts on the heaters.