Did RCA use WB IQ after the CTC2/Bs.

[stromberg67]

Trying again to send readable pics. Kevin

[zenithfan1]

Thanks Kevin. They look good on my end.

Here are some shots I took out of the manual for the CTC-121. I'll post one of the board from the manual too If I can make it turn out legible. I buried the set last time I rearranged. I'll try to dig it out today and take some pics since my OT got cancelled due to no snow like we thought.


These pages explain how the chroma is processed. Let me know if they suck, my scanner is broken (darn kids)

[Tomcomm]

Well we’re starting to get some schematics of the CTC133B from Kevin and some description pages of the CTC121 from Zenithfan1. The photos from the CTC121 are 1023 X 769 and are perfectly readable. However the CTC133 schematics at 600 X 424 are still unreadable. The chip block diagram at 800 X 567 is ok but fuzzy. Sorry Kevin. It seems VA will let you do 900, even 1023! I and those really interested in WB IQ would like to see you go for another try at 900 or 1023 this next time, please? I gotta see how they implement the critical LP filtering and the necessary compensation delays. The CTC121 description doesn't mention this consideration at all! Thanks for you guys interest in my silly WB IQ exercise………Tom

[old_tv_nut]

It would be nice to get larger schematics, but I can read them well enough to tell you that this does show wideband I (and narrowband Q). The RCA trick to getting a delay line or filter with one pin is applied to pins 8 and 9, which in the chip block diagram go to stages called "Q-phase filter" and "I-phase filter."

If you look at the schematic of the parts attached to these pins, you see an external buffer transistor on each. This buffer (an emitter follower) drives what would normally be the ground point of a delay line (for I) or a filter (for Q).

Normally, a delay line or filter requires one IC pin to drive its input and another pin to sense the voltage on the output. However, with the clever circuit trick, you need only one pin to drive the delay line or filter input with voltage (via an internal emitter follower). Then, with the ground also driven by the external buffer, the output current of the delay line or filter cannot go anywhere except through the pin and the internal emitter follower. So, the internal emitter follower has the delay line input voltage, but the delay line output current! This current, being in the emitter, is also in the collector (minus a little base current) and is taken off the collector of the internal transistor and goes to the matrix for R-Y, B-Y and G-Y, having been subject to delay or filtering as desired.

[Tomcomm]

Wayne…….I recalled you mentioning this when you described ”how clever” these RCA guys were with the one pin delay-line when you were comparing RCA’s WB IQ with Zenith’s NB R-Y B-Y color systems. I traced out the pins 8 & 9 of the demod and found the tiny DLxxx and buffer xistors on Kevin’s 600 X 424 schematics. I hope he has no problems with re posting at 900 X 638 or larger. Then we can get part numbers and everything! Hitachi has a 176225 listed for $6. I wonder where I can get the delay line? Everything else seems generic, would be fun to build this WB IQ demod up…..Tom

[zenithfan1]

I'll try taking some pics of my schematic and see if they are legible, if so I'll post 'em up. Sorry it takes me so dang long but I'm pretty busy these days....still gotta dig that set out too....

[amptramp]

For what it's worth, Hoffman had a wideband system which used R-Y and B-Y detectors but after a chrominance channel where the high bandwidth chroma signal was delayed while the narrowband was not. This is a bizarre method and was used briefly (around 1955, I think), but I would prefer I, Q demod. The NTSC spent a lot of effort to determine the resolution and criticality of the eye for various colours and I think I, Q demod would be as good as HDTV because the eye does not see large bandwidth in blue or yellow.

[old_tv_nut]

Originally Posted by Tomcomm

I wonder where I can get the delay line? Everything else seems generic, would be fun to build this WB IQ demod up…..Tom

There may be a problem finding out the coil specs for the Q filter also. I believe the values on the schematic are probably resistance readings for troubleshooting, not the inductances. I suspect they may also be wound for specific self-resonant frequencies.

[old_tv_nut]

Originally Posted by amptramp

For what it's worth, Hoffman had a wideband system which used R-Y and B-Y detectors but after a chrominance channel where the high bandwidth chroma signal was delayed while the narrowband was not. This is a bizarre method and was used briefly (around 1955, I think), but I would prefer I, Q demod. The NTSC spent a lot of effort to determine the resolution and criticality of the eye for various colours and I think I, Q demod would be as good as HDTV because the eye does not see large bandwidth in blue or yellow.

It is usually stated that the I axis was chosen to match the axis of maximum color resolution of the eye. But if that is true, the eye can't see if the small details are somewhat off in hue, so the axes chosen by NTSC actually could have been R-Y (wideband) and B-Y (narrowband). Actually, it is probable that using B-Y as the narrow axis would be better in terms of color detail than using Q, that is choosing the axes so the narrowband one falls on the lowest color resolution axis and not worrying if the high res one is slightly off. However, there are other more important reasons to use I and Q. One is that I is very close to flesh tones, so if there is a transient effect on the edge of a face, it doesn't look the wrong hue. This is probably the most important practical result of using the I axis in particular. Another is that the roll-off in color resolution is not abrupt, and you can't force people to sit far enough away to guarantee they won't see edge effects. A further reason is that pictures made of I axis colors often look plausible, while pictures made of only Q axis colors often do not. Most of the automatic tint circuits developed later were based on this. Another reason is that much of the test material was color slides scanned by unmatrixed cameras, both facts which tended to suppress greens, so the test material was notably lacking in strong Q axis colors. So, if you picked the axes to provide the best overall results with a random selection of color transitions, you might have ended up with something different; but the choice of I and Q actually catered to the statistics of the color palette of common source materials and important image elements (faces), which meant it was well suited for general use.

Summary: the choice of I and Q was optimum, but not purely for the reason of resolution that is usually stated.