Gorgeous, Pete!

Hey Pete,

Very nice set up on your CT-100. The color looks great and the convergence looks spot on. Congratulations.

-Steve D.

<<My gnawing itch wonders if WB11 actually transmitted the video with wide-enough gamut to make the effect I saw on the CT-100 vs. the HDTV. I wonder if the cartoon makers had their software set to 1953 NTSC. I wonder if the NTSC generator in my VCR is spitting out a color bar signal good enough to calibrate the CT-100. I wonder, as I wander…!>>

yup unless it is calibrated pretty good as far as color, you could always be seeing just more green due to off-value resistors, coil adjsutment issues, variations in drive of the guns etc. in the color section of the set. The potential for that old tube to display true NTSC is one thing... how accurately that old chassis is going to present it as it currently sits is another. All my old sets seem to have tendency to play up certain colors over others (mainly red) but not enough for me to go in and perfectly realign the color section, plus I don't really have the equipment necessary.

On the subject of pictures on working CT-100s... Steve D. demonstrated his CT-100 for me one day back in '84 or '85, and that was the first time I ever saw a CT-100 turned on. Well, the show that we happend to get was a live news feed from the McMassacre (in progress at that moment) in Chula Vista, CA. A rather macabre way to remember my first CT-100 viewing, but it is a sad and real part of history in California. We sat glued to the little vivid image, maybe for a moment overlooking the fact that it was a CT-100 at all. The CT-100 was, for an instant, restored to its original purpose as a medium of communication of sorts. I wonder if Steve remembers that day??? Deep thoughts :sigh:

Charles

What finally made me go public… ‘er, AudioKarma, was (1) achieving ‘standard’ HDTV-hues that are eyeball-duplicate on the CT-100 and (2) finally accepting the modified chroma-AFC alignment-procedure as okay (where final tweaking of L41 is needed because the circuit needs it, not because a component has drifted).

It should be remembered that all other hues on the CT-100 screen eyeball-matched those on the ‘standard’ HDTV screen. Only the green stood out and only on that character’s hat.

Here’s another CT-100 screen shot taken the same night. There are three different areas of green; two on the hat and one weird snow-capped mountain whose base is green seen through the now-open door. That green matches the HDTV green.

So, I’m speculating for now that the striking green on the character’s hat is a result of wider gamut and not some unusual peaking or weird spurious oscillation somewhere in the CTC2 chassis.

But I could be wrong. And so my next step is to rearrange my living room over the next couple of days, time permitting, with the CT-100 to the right of the HDTV. Then develop some photographic evidence supporting or debunking current speculation.

Was chatting with John Folsom yesterday and recalled that in September I reported the same effect as here. That time it was the eyes of a cartoon character that were vivid on the CT-100 but not in control images.

Here it is: “There's a commercial playing for a Lilo&StitchII (sp?) DVD. The bug-eyed character in the commercial on the CT-100 has bright, intense, large lime-green eyes that pop off the screen at you. Interestingly, in an A-B comparison between the CT-100 and my 38-in. widescreen CRT HD set, there's no similar effect on the HD set, although there may be setup issues to consider. Just came back from seeing the DVD at Target. The eyes in the print version were pale blue and didn't even appear to have been designed to draw attention --nothing like the TV ad to my eyes anyway.”

That's still a truly GORGEOUS picture any way you cut it-Man, I wish I'd scored a CT-100 back about 20, 25 yrs ago when they were just junky old TVs....-Sandy G.

I have been in the back bedroom doing final adjustments of my CTC-10 and was flipping around and noticed that South Park also (now it's on free tv), I think I'll watch it again and see what the dude's hat looks like! (ok it's not a 100 but oh well I'll take what I can get ; )
You should send these pics to the creators of SP and tell them this is what their cartoon ACTUALLY looks like : )
Frenchy

Most digital still cameras use the sRGB gamut (same as HDTV), and computer monitor primaries are similar, so showing what the sets actually look like to the eye will be a problem (although you should be able to see a difference of some kind in photos of the two screens).

Years back, a SMPTE monitor committee had two identical tubes made up (by Sylvania, IIRC) except for the green phosphor, one sulfide green and one NTSC green.. They were mounted in identical monitors and set up in the engineering lab studio at NBC in New York. They had some brightly colored objects in front of the camera including something "Kelly green". With this setup you could clearly see the difference, not only in the greens, but in the cyan saturation and the relative brightness of cyans and reds. They eventually demonstrated that you could get reasonable/acceptable (though not perfect) rendition of colors within the sulfide gamut by changing the matrix in the monitor. As a result, future monitors (Like the Tektronix Trinitron-based series) had a switch to turn the matrix adjustment on and off. You would set up the hue and chroma on color bars with the switch off, then turn it on to get reasonable flesh tones.

SMPTE changes
 

Wayne: Did the capability of the broadcast industry to transmit NTSC gamut shrink with the advent of SMPTE standards?

Or, is it possible — maybe even likely — that the difference in green between the HDTV and CT-100 I saw is a lucky confluence of NTSC-gamut-capable broadcasting plus an overly enthusiastic graphic artist? Rather than some Pete eye-brain quirk!


Also, I won’t be content making CT-100 matrix adjustments using a DVD and a VCR modulator. I suspect that my best chance at a calibrated NTSC-modulated color bar signal is an over-the-air signal.

At one time, broadcasters around here transmitted color bars after they signed off. Nobody seems to sign off anymore. But I found Ch. 13, the PBS station, did so at three a.m. last weekend and transmitted over-the-air color bars.

But I slept right through it. So this weekend if they repeat the bars I’ll try again and use the R and G and B filters that come with Video Essentials to check the two sets. Which brings up another point. I didn’t see any traceable calibration stickers on the color filters… :rolleyes:

Pete,
Try to give the engineering staff at the PBS station a call. I’m sure there’s a good chance once you tell them what you are doing and why they may help you in some way. At the very least they may tell you when they are going to have color bars up.

Great Picture

Good luck
Steve

Pete,

The question of what you should see on today's NTSC transmission is almost unanswerable. The taking characteristics (spectral response) of cameras is not an ideal match to the human eye to begin with (just too difficult to do). Major design criteria are a proper red/green crossover to get flesh hues right, plus a matrix to match the expected monitor primaries generally. Camera designers can do a mathematical calculation of the camera's performance based on a set of color chips, like the Macbeth ColorChecker chart, and adjust the camera matrix to match NTSC. But they would be foolish to not check it on a modern monitor. So, do current cameras and film scanners match NTSC best, or are they tweaked towards current monitors? I don't know, and I don't know that camera makers release this info - they sell cameras based on price, features, and how good they look in a display studio at the NAB convention. Beore blamng the camera guys for tweaking, consider that consumer receivers wiht non-NTSC phosphors have had a wide variety of matrix fiddles to make their rendition closer to NTSC (like the professional monitor matrix I wrote about). So, the system in the US has essentially been out of control.

Now, if you ask about PAL, the answer is simple: PAL standardized on the non-NTSC phosphors and specified their colors precisely, so they cannot transmit colors out to the NTSC gamut. HDTV is the same way - smaller gamut, but precisely defined.

Older material made with image orthicon cameras is intended for NTSC phosphors, but those cameras had no matrixing, so they are the best you can get with optical filtering only. Like most color film, they increased saturation by using narrower than theoretical R,G, and B filters, reducing the need for the matrix.

I was going to say that using the DVD and modulator should not be a problem, but it depends on the modulator having the proper R-Y and B-Y gains, and I suppose they might be wrong (although I think it's unlikely). A video tape of the color bars off your local station should work, since the tape shouldn't affect the R-Y and B-Y differently.

What local station is that? If you got through to the chief engineer and told him what you're doing, you likely could talk him into putting up the signals you want -especially if you invited him over to see the result.

The main requirement for the video essentials filters is that each one suppresses the other two primaries well enough that they don't affect your judgement - so traceable calibration is not needed, just a deep enough color. You should be able to tell if the red filter is letting any green and blue through by looking at the green and blue bars, and so on.

Thanks guys. You've given me plenty to consider.

On another note, recall the label on the CT-100 shows 475 watts at 115 Vac as the power used by a Merrill.

Just measured it. With a 115 Vac line, the unit draws 415 watts with a PF of .85, which measures/calculates out at 485 VA / 488 VA. So in conventional terms the draw is more like 415 watts.

thank god AK is back ... was gettin' the shakes.
 

Checked o-t-a colorbars early this morning. Couldn't see a difference between over-the-air bars or the same station received via cable.

Checked also the HDTV versus the CT-100 after setting both using R, G, and B filters. The green test looks better on the CT-100 and was able to demonstrate it for the site; the other colors test pretty much alike between the two sets. Here's a link:
http://home.att.net/~pldexnis/input/...ntresults.html

To my eyes the yellow on your hdtv looks more 'yellow' than on the ct-100, which looks lime-greenish. Anybody else? Funny that's like the same color area that South Park hat is in. But I'm not looking at them directly, I'm just looking at pics of them on my monitor so....

---

Both NTSC. When receiving analog broadcasts, ATSC sets I'm familiar with (not many) don't seem to equal a modern analog set such as the CTC169 series. I'm guessing the HD set matrix is responding the way it was designed. Jeff Landaro designed the video amp in that set; maybe he's available for an inside expose :D

The lime-green on the CT-100 Frenchy mentioned had been a problem for me too, although since the matrix balance was improved for red, the yellow no longer seems 'off' when viewing the screen. Probably attributable to the strong/wide-gamut green of the 15G.

1) your mono picture of the big set shows less green on the yellow bar, so you would expect its yellow to be less green
2) The red looks darker and the blue looks brighter on the CT100 than on the large screen. The yellow looks greener and the magenta looks bluer. I would check the white point (blue and green drive settings) on the CT-100. If your white is lacking in red, that could be part of the difference. Ideally, you should have a gray background wall in the room lit by daylight or artificial daylight to compare to, or at least be able to look outside at daylight for comparison. Anyway, in turning idown the green and blue drives, it helps to go too far (slightly pink whites) if you can get there, then return to a neutral white, rocking the controls back and forth while you zero in on the neutral point.

It might take a few days, but I'll readjust the B and G drives as per your suggestion; saw a D6500 fluorescent work light at Sears; bought it and have it behind the HDTV a la Digital Video Essentials. BTW, when a 100-percent red screen is applied to the CT-100, it swamps something and shows a dark mustard-yellow, although the 75 percent red is red and at least 'looks' okay. It's all about red in a CT-100 I guess. Back in the '60 when I last fiddled with a Merrill, never had this much fun nor learned so much new stuff about it.

Fantastic reading.You've done a marvelous job in resurecting this set.

My comment that follows is mearly a suggestion.If you're unfamiliar with adjusting gray scale,I thought I might pass on some basics to help.

From some of the pictures you've attached do show that your gray scale is off on both ends.This may be just the pictures,but if what is shown is what you're seeing on the set when looking at the stairstep pattern,it looks to me as if both ends are off.That pattern should show shades of gray from black to white when properly calibrated.It looks bluish on one end and greenish on the other.If you're trying to approach the NTSC color temperature of D65 (6500K),what will help is a colorimeter.Doing it by eyeball is impossible.Prior to any adjustments to these areas,it's imperative that the Black and White (brightness and contrast) levels be adjusted properly.

There are two different adjustment points,the lower end using a 20IRE window pattern and the upper end using a 100IRE window pattern.Most sets have these controls,whether it be by analog pots pots or by service menu access in the newer digital sets.The lower end controls are called cut or bias or even screen by some manufacture while the upper end controls are called drive.If you have R G B for both end,great.Some sets only have two,just depends on the manufacture.

A bias light behind the tv or screen is not a reference for D65 adjustment puposes.It's there for overall viewing reference.

I've had the opportunity to calibrate a few CRT direct view sets and projectors over a span of about 6 years,using a Philips color analyzer and test pattern generator.Have also used patterns from A Video Standard and Video Essentials.The patterns from VE are very good.Getting the gray scale to track properly isn't always easy nor do sets hold them well.But,if this can be done,then the set's color will be as true as it can be,without being shaded.

Please do not take any of my comments or advice as criticism.On the contrary,I am just ecstatic to see life being brought back to these sets.I remember them fondly and have always been fascinated by television.I don't have the repair background that you obviously posses,but my curiosity as to why they all looked different pushed my desire to learn as much as a hobbiest/part time member of ISF could with respect to picture quality.

I agree with what you say about using instrumentation to set grey scale. Actually, I have never had the advantage of being a TV repair technician. For most of my work life I was an editor of one type or another. But I’ve also had opportunity to make a living troubleshooting to component level and have spent a few years working with instrumentation including initial power-on thru final calibration. All this is in support of my first sentence.

I’m going to check eBay; perhaps a reasonably priced colorimeter still in acceptable calibration will turn up.

Some things I’ve grown to acknowledge and appreciate recently are issues that effect stability of this restored Merrill. My feeling is that it’s a result of the original design rather than some artifact of age or the restoration process. In discussing this with John Folsom, who of course is a premier collector of 15-in. vintage color, it seems that most vintage sets are fickle, but the Westinghouse H840CK15 is far more stable than the CT-100, particularly in the area of high-voltage regulation, which I find, in the CT-100, is apparently an issue in an attempt to achieve brightness suitable for a desired, wide, grey scale.

Another area that’s been challenging when adjusting the matrix is to achieve a toned-down green component in the color bar display. It’s well known and fully appreciated that the red phosphor in a 15GP22 is by far the least efficient. I expect to work more on these challenges this weekend, but perhaps there’s just not enough latitude in the red to overcome the brightness and more importantly perhaps, the very wide green gamut, when adjusting grey scale (as you say, 20IRE to 100IRE) before trying to get eyeball-perfect color bars.

As I said, never had hands-on TV repair experience or did I ever work in broadcast video, so much of this detail stuff is new to me. Add to that the fact I’m an old guy….

Appreciate all the comments and suggestions received along the way. In particular, Old_TV_Nut with all his envious TV-engineering experience who kindly nudges me in the right direction now and then.

Have attached a shot taken this morning of simultaneous ATSC-NTSC broadcasts. The two ends of the engineering spectrum match fairly well in this shot, but I’ll admit to toning back the HD set in brightness and contrast a bit.

Pete,

A great picture! Keep at it.

Getting a precise gray scale setting is difficult without a color meter, but IF the bias and drive controls do not interact too badly you can usually get a pretty good result by eyeball. It really helps if the controls can be rocked back and forth through the correct point - you can go for decreasingly obvious misadjustment as you rock the control back and forth, and zero in on a good point. The ideal is to be able to set the lowlights and the highlights separately, doing the blacks/dark bars first - but not all sets will let you do this. Looking at the CT-100 schematic, if the DC clamping is working OK, you should be able to set the lowlights close and then set the drives for proper white. Then touch up the lowlights for precise tracking, and go back and forth from highlights to lowlights again if necessary.

Because of differences in phosphor spectra, a filter colorimeter may be slightly off in its readings. The Zenith optics lab had to use a master spectroradiometer to calibrate filter-type colorimeters to the current phosphor set - the colorimeters would then have a multi-positon switch for different calibrations. The spectroradiometer was calibrated against a NBS traceable standard lamp, and then the color was calculated from the measured spectrum using the CIE standard obsrerver curves. (These curves have been shown to have some errors, but that's another story). I believe that nowadays there are better ways to calibrate than a standard lamp, using power measurements of laser sources, but I haven't kept track of what optical instrument companies are actually doing.

The best white point setting (for one observer) can be made using a split-screen D65 lamp. This device has a window that is half open to the CRT and half mirrored to reflect the D65 lamp. The view is of the screen gray bar immediately adjacent to the lamp gray, with no intervening frame line. Any one person can get a near perfect match (for him/her self) of monitors having any phosphors by using this. However, other individuals likely will see the match as slightly off due to normal color vision variations. I and one of the older engineers always disagreed on how much blue drive produced a match. (He needed more blue.) Now that I am older and slowly developing cataracts, I probably need more blue than I used to.

The colorimeter substitutes its filtered spectral responses for the human eye, which is reasonable but not as accurate as a spectroradiometer, hence the need for multiple calibrations depending on the phosphor characteristics. I would guess that the colorimeter will be fairly correct for the different greens, both of which have smooth spectra, but may show errors due to the difference between the CT-100 red spectrum and the spikey rare-earth red spectrum of a modern tube.

So - if you get a colorimeter, and the two sets don't end up looking like the same gray scale, you may have to trust your eyes.

Just do that old psychology trick I read in one of my old tv repair books - hire somebody to set the greyscale for you and tell them to ask YOU when they have it white : )
They said repairman should do that to prevent the customer from a callback complaining "this thing is green", because they are the one that decided it was white to begin with, heh heh.

Weekend Udate...
 

I have been experimenting with the relative setting of the G2 grids. The attached photo is with a low-brightness grey G2 setup (didn't upload for some reason; musta been too big. Substituted an ad c.u.). It is fairly decent and still viewable in the morning light. When a brighter grey is obtained with a relatively greater G2 setup, the resulting picture will lack adequate red phosphor output, an effect that is particularly noticeable where green is an element of the color being reproduced (as in that odd hue in the background of the GMA-set).

A low G2 grey setup provides greater G1 (control grid) response; a high G2 (brighter grey) setup delivers a brighter picture. With the greater G2 setup, I was able to watch the set during a bright sunny day with just the blinds closed.

So far, I haven’t been able to have it both ways. :thumbsdn:

Next step is to backup farther and readjust the red purity, which has deteriorated since set up weeks ago, as discovered last weekend during G2 adjustments. This will be at least part of the solution. :thmbsp:

Hey Pete,

I have been following your restoration efforts with great interest. Seems there's no end to the technicial issues in trying to achieve the perfect image on a CT-100. As you, and the rest of us CT-100 owners have learned, this set is very high maintenance, and a bit temperamental. And, while it's fun, for a while, to tweak this control and that, you may just want to walk away for a time. Sit back and watch the beautiful results you have achieved. Enjoy the fruits of your labor. I love the photo of your CT-100 sitting in the corner purring away and displaying it's magnificent living color picture.

-Steve D.

Does the CT-100 have any kind of set-up switch that sets the video bias to a fixed value while playing with G2's? If not, you may be walking yourself away from midrange on the G2's and black level (brightness + any other video black level settings if there are any). If you can check the G2's against a service-info value, and/or look at the red video drive with a scope to see if it's clipping, that might help.

The piece of schematic that comes to hand shows R,G,B, drive capacitively coupled to the G1's, each with its own diode DC restorer to set the black level (birghtness) - plus the blue and geen have individual background controls.

I think it goes like this: When you raise the G2 voltage, the tube is harder to cut off, but that means you need more negative swing on the grids, which would not cause the red amp to clip on highlights - maybe the opposite. So, I think it's possible that when you turn up the G2's, you are reaching the max current that the red gun will deliver on the red highlights - just a guess, but if it's true, there's nothing to do but run it at lower settings.

It's amazing how you can tweak the tint on a black and white picture and it apparently becomes slightly red-tinted... but keep watching it for a minute or two and it eventually just looks like black and white again. Then put it back the way it was before and suddenly it will look green/bluish! Try this one - stare at a test pattern for about 30 seconds, then put it on a program. Wow look at all the pretty colors! : )
All this is probably why I don't spend too much time trying to perfect all the color settings etc., as long as it looks pretty good and people faces are flesh colored, that's 95% of it.

frenchy and Steve D...
 

Matter of fact, last Saturday the set and I had a marathon date lasting 5.5 hours in the morning through early afternoon, then another 3.5 hours of quality time that evening. Tomorrow however the back comes off and the top gets popped in prep for that red purity setup I've been threatening to redo. Before that I'll scope the red gun to check for possible distortion under varying G2 settings. It's all good fun with the only bad thing a hold on the next project, restoring the '56 Philco.

Enjoying retirement,
Pete :D

Pete is correct, retirement IS fun!

I just wanted to say:

A few years ago the conversations taking place here could have only happened by phone or maybe mail. (assuming that, without this site, folks like Pete & Old TV Nut would even have heard of each other) Half of the phone conversations would have been forgotten. Now, here they are, in writing so you can go back & reread. We can all benefit from the knowledge here, today & hopefully for all time.

Hear, hear.

Or should that be: read, read (pronounced red red).

The thing that stands out to me in this thread is what a perfectionist will to to make a TV set be as accurate as possible ...just a pity that most programs today are pretty sloppy in terms of color and covered in dreadful station logos.

But then when you watach a film where the colour is part of the whole desing of the movie you realise why getting the best you can out of a TV is worth it.

But then being colour blind means for me getting the colours accurate is rather a subjective thing!

and of course ....here in PAL land the colour looks different anyway .... it is interesting how when changing from a PAL source to an NTSC one requires changes to the settings on the set... mainly because the black level is different.

On my Sony the "preset" color, brightness and black level auto-settings are clearly designed for NTSC sources as they are way off for a PAL source ...too dark and contrasty a dn way too much chroma but for NTSC are pretty good.

I'll be lost when the analog transmitters shut down in a few years...
 

About a week ago I set off to investigate characteristics of the red gun signal. Data from the task has been added to a new Troubleshooting section on the CT-100 site. There are four entries so far; check the one from 11-19-05.
http://home.att.net/~pldexnis/input/...eshooting.html

I'll be lost when the analog transmitters shut down in a few years...
 

Three weeks ago I readjusted the CT-100 and achieved a better hue match between the 'reference' digital set and the Merrill. It ameliorated my earlier supposition that intense green was a result of wide green gamut. Here's a link to the update page where the 12-9-2005 update appears:

http://home.att.net/~pldexnis/ct-100...tion_log8.html

Very interesting, Pete! How does the comparison look on a gray scale now?

It seems the gray scale is a bit contaminated because the color control will not kill it completely, and so I don't have a recent shot. I will have to confirm this w/a scope, however, and try for one Monday morning. Last weekend, channel 13 ran pledge programming during their usual color bars. Hopefully they'll have raised enough $ to go back to color bars this weekend.

Automatic balck level and white balance
 

Now it would be real neat to design a circuit that would adjust black level and white balance for a 15GP22 in a similar fashion found in today's modern TV sets. During the blacker than black and peak white pulses within the blanking periods the gun current for red, blue, and green would be individually compared to a reference and then G1 and G2 adjustments made automatically!

It would make an interesting project.

At this point, if you adjust the G2 for a lower gray and view the set in subdued light the CT-100 picture compares favorably with a digital set tuned to the same analog broadcast. There are differences of course: at those times when the digital set shows virtually no phosphor glow, the same area on the CT-100 screen will sometimes show a very dim glow.

On a similar note, I’m borrowing a virtual multimeter. Hope to be able to develop a data-logger-type monitor that will keep a txt file record of circuit voltage. I’d kinda like to learn something about how critical the 115-Vac line-voltage spec is to optimum CT-100 operation. If it works I’d particularly like to gather data on how the high voltage varies with line voltage.

The value of remanufacturing vintage parts...
 

The last of my three vertical convergence transformers failed. One is used in a CTC2, the CT-100 chassis. Sunday morning June 18, 2006, the fuse protecting the H-V cage blew on turn-on. Unlike my two other failed transformers, however, which still display correct dc resistance, this failure wiped out the secondary. The tap is completely open. There are about 12 Meg between the top and bottom of the secondary where about 12 k-ohms should be. This part has an extremely high failure rate. Probably second only to the 'white' peaking coils that rot open.

For the repair, I tacked in one of John Folsom’s newly manufactured replacements I've had for about a year now. It was static-tested at about 4 kV across primary-to-secondary for a few hours when it arrived last June, and so I expected a no-issue installation.

On the bench, the H-V ran a steady 21 kV with the new transformer in place and the H-V control set wide open with a 115-Vac line. The focus voltage, which passes through the secondary and is the reason for all the stress on the transformer, is steady at 3.5 to 4 kV, depending on the focus control position.

Picture is of the new transformer on the rear of the H-V cage.

In the future, I’ll remove the failed core from the transformer housing and use the original case around the new transformer.

Returned the chassis to the cabinet today, cranked the H-V control back to the specified 19.5 kV and while I haven’t reconverged it yet, it's perfectly watchable.