Videokarma.org TV - Video - Vintage Television & Radio Forums

Videokarma.org TV - Video - Vintage Television & Radio Forums (http://www.videokarma.org/index.php)
-   Early Color Television (http://www.videokarma.org/forumdisplay.php?f=36)
-   -   Prototype set is here (http://www.videokarma.org/showthread.php?t=251369)

Penthode 06-23-2011 08:35 PM

This prototype I find incredibly interesting in light of the early development of the NTSC standard. Certainly, it should be possible to generate a 1952 CPA signal to run the set. I would certainly hope that the set not be modified from the CPA design if it indeed is still a CPA receiver.

Is the subcarrier crystal 3.89MHz?

I came across another interesting paper published in the March '53 Proceedings of the IRE. It is entitled "Generation of NTSC Color Signals" by Joseph Fisher and pertains to the earlier CPA specification.

It appears that for the NTSC '52 tests, field rate R-Y switching was employed. Vestigial sideband equal bandwidth R-Y and B-Y chroma channels were used. Color subcarrier was 3.89MHz and it appears only standard 1/2 line offset interleaving was used.

CPA at field rate as opposed to line rate probably was a better choice in 1952. The reason for CPA was to cancel phase errors in transmission and to facilitate vestigial sideband chroma channels without quadrature crosstalk. Because there was no way to electronically delay and combine alternate field or line information economically in 1952, the errors would be cancelled visually. This is very much akin to simple PAL in Europe. The field rate equivalent to "Hannover Bars" would result from transmission phase errors.

The real failure of the 1952 CPA tests therefore must have been due to the flickering akin to Hannover Bars and the 610 kHz chroma subcarrier to audio carrier beat (this was before the line and field rate was shifted to 15,734.26Hz and 59.94Hz respectively to ensure that the sound carrier was a multiple of the scan rates). In addition, the alternating R-Y axis would prevent proper 1/2 interleaving of the chroma within the luma.

Note that PAL later adopted the 1/4 line offset to help mitigate the interleaving problem. PAL also introduced a picture rate subcarrier frequency offset (25Hz) to further conceal the interference. (In hindsight, a 29.97Hz offset could have been added to help conceal the NTSC dot crawl. But the standard was already established and the simpler luma to modulated chroma arrangement in NTSC probably makes this unnecessary).

old_tv_nut 06-23-2011 08:59 PM

Quote:

Originally Posted by Penthode (Post 3007026)
I
It appears of the two options, the 1952 field test used field alternation rather than line alternation. Is this correct? Slight phase errors would introduce flicker on image edges or chroma transitions due to the crosstalk not being properly integrated or cancelled.

Yes, field alternation was used. Wentworth's book on color television engineering states that flicker was bad due to difficulty of maintaining the exact relationship between the alternate phases in transmitter and receiver. The air-core coil that Nick found looks suspiciously like it was built to reduce variations with temperature, spacing to the shield can, etc., etc., but we won't know until we discover whether there is a phase switching circuit.

Also, the crystal appears to be missing, so we don't have that confirmation.

old_tv_nut 06-23-2011 09:05 PM

An aside regarding frequency interleaving: during the analog to digital transition period in the US, digital channels that were lower adjacent to analog channels were required to have a precise frequency offset to reduce the possibility of visible beats between the digital pilot carrier and the chroma of the upper adjacent analog signal - a re-application of the principle developed by NTSC long ago.

JBL_1 06-23-2011 10:30 PM

Quote:

Originally Posted by miniman82 (Post 3006885)
That's exactly what I'm wondering. The other end of the resistor string has a wire that goes under the chassis, with connections to many different parts. So perhaps it does have dynamic, but it's done in a different way?:

The transformer is only needed for the vertical component. The horizontal component can be AC coupled. We looked at the same thing for dynamic focus back in the late 1980's. The horizontal component was easy to add the vertical component more difficult. Is that really a high voltage vertical transformer?

miniman82 06-23-2011 10:44 PM

Quote:

Originally Posted by John Folsom (Post 3007022)
Nick, I am comforted by your discovery of convergence transformers. :-) BTW, that schematic is from a RCA CRT symposium, not a set I have.

D'oh! Sorry about that, I've been pretty hasty about reading these last couple days. There was some bad weather that came through recently, and it knocked out my internet connection. Because of that, I've had to use spare time at work to post. Now that I have my internet back though, I'm free to post at will. Muhahahaha! :banana:

miniman82 06-23-2011 11:05 PM

Quote:

Originally Posted by Penthode (Post 3007055)
This prototype I find incredibly interesting in light of the early development of the NTSC standard. Certainly, it should be possible to generate a 1952 CPA signal to run the set. I would certainly hope that the set not be modified from the CPA design if it indeed is still a CPA receiver.

Is the subcarrier crystal 3.89MHz?


*sigh*

If only there had been a crystal installed in the chassis, it would have made the mystery of what standard this chassis uses a lot easier to figure out that's for sure! Since it didn't have one though, the best I can do is trace the circuit and try to determin if it's still CPA or if it was converted to use the later 3.58MHz subcarrier. I'm still leaning towards CPA though, considering the 'CPA transformer' is the one that's still connected.

Quote:

CPA at field rate as opposed to line rate probably was a better choice in 1952. The reason for CPA was to cancel phase errors in transmission and to facilitate vestigial sideband chroma channels without quadrature crosstalk. Because there was no way to electronically delay and combine alternate field or line information economically in 1952, the errors would be cancelled visually. This is very much akin to simple PAL in Europe. The field rate equivalent to "Hannover Bars" would result from transmission phase errors.

One of the things I'm itching to figure out with this thing is if the edge effects and bars with CPA really are as bad as people make them out to be, it's one of the motivating factors to get it working as it last did. I'm hoping my brother will stop by this weekend, so we can do some digging in the color section and determine for sure what it's supposed to do. He should be able to sketch the circuit in only a couple hours, and usually takes payment in beers. :thmbsp:

miniman82 06-23-2011 11:09 PM

Quote:

Originally Posted by JBL_1 (Post 3007076)
The transformer is only needed for the vertical component. The horizontal component can be AC coupled. We looked at the same thing for dynamic focus back in the late 1980's. The horizontal component was easy to add the vertical component more difficult. Is that really a high voltage vertical transformer?

I don't think so, it looks much too small. Everything is on the other side of a dooknob cap, so I'm assuming right now that everything having to do with convergence is also low voltage.

Penthode 06-23-2011 11:47 PM

I came across another article entitled "General Color-Receiver Design Considerations" presented at the fall meeting of the IRE and RTMA in October 1952 by staff at Hazeltine Labs. It is a description and details on CPA receiver design.

On the final page of the article is a complete diagram of a CPA demodulator. The color oscillator has no crystal! A 12AT7 is used: one half the reactance control and the other half a Colpitts oscillator. Only LC components and no crystal.

I will try and post the drawing and article tomorrow. But the decoder line up in the drawing includes the following:

12AT7 Reactance/3.89MHz Oscillator
1/2 6AL5 Clipper
6CB6 Bandpass amp
6AS6 B-Y demodulator
6AS6 R-Y demodulator
1/2 12BH7 B-Y matrix amplifier
1/2 12BH7 R-Y matrix amplifier
1/2 12AT7 G-Y matrix amplifier
6CB6 Isolation Amplifier (or Color oscillator buffer output)
1/2 12AT7 Field Recognizer Driver
1/2 12AT7 Field Recognizer Phase Detector
12AT7 CPA Multivibrator
1/2 12AT7 Field Recognizer Keyer

miniman82 06-24-2011 12:22 AM

Quote:

Originally Posted by Penthode (Post 3007085)
I will try and post the drawing and article tomorrow. But the decoder line up in the drawing includes the following:

12AT7 Reactance/3.89MHz Oscillator
1/2 6AL5 Clipper
6CB6 Bandpass amp
6AS6 B-Y demodulator
6AS6 R-Y demodulator
1/2 12BH7 B-Y matrix amplifier
1/2 12BH7 R-Y matrix amplifier
1/2 12AT7 G-Y matrix amplifier
6CB6 Isolation Amplifier (or Color oscillator buffer output)
1/2 12AT7 Field Recognizer Driver
1/2 12AT7 Field Recognizer Phase Detector
12AT7 CPA Multivibrator
1/2 12AT7 Field Recognizer Keyer


Oh, baby! This might be a huge help in locating what circuits do what in this set, I can't wait to see what it looks like!

ChrisW6ATV 06-24-2011 12:22 AM

Quote:

Originally Posted by miniman82 (Post 3007082)
*sigh*

If only there had been a crystal installed in the chassis, it would have made the mystery of what standard this chassis uses a lot easier to figure out that's for sure!

Since the crystal is missing, I wonder if it would be feasible to connect either a sweep generator plus scope, or a spectrum analyzer with tracking generator, to various circuits in the chassis to figure out their resonant frequencies and bandwidths as currently tuned. Or, get a 3.89 MHz crystal along with a 3.57 MHz one, isolate and apply power to the oscillator circuit with each crystal, and see whether either or both work in the circuit and measure relative RF outputs to determine which frequency the circuit might have used when it last operated.

Here's another idea: Try using a dip meter near the various circuits, if it is possible.

JBL_1 06-24-2011 07:18 AM

Quote:

Originally Posted by ChrisW6ATV (Post 3007091)
I wonder if it would be feasible to connect either a sweep generator plus scope, or a spectrum analyzer with tracking generator, to various circuits in the chassis to figure out their resonant frequencies and bandwidths as currently tuned.

Sweeping the bandpass amplifier is a great idea! The response may not be peaked at the subcarrier. The peak could be staggered to make up for any roll off in the IF or bandpass/lowpass filtering.

But the information still could be usefull if you looked at the overall video response of the chassis.

Penthode 06-24-2011 11:35 AM

1 Attachment(s)
Please find attached two papers, both from Hazeltine. The first one on general color receiver design pertains to CPA only. The last page has a nearly full CPA demodulator. Note the demodulator does not use a crystal! The design is a straight reactance controlled Colpitts oscillator. This is interesting in as much as it is expected to keep sync over the entire line from onlyeight cycles of gated burst reference! Note the color hold control- if there is no crystal on this prototype I would trust there is a color hold control on the front!

The second Hazeltine paper is by Bernard Loughlin. His name conjures up a lot of early NTSC development. His 1966 paper is timed with the launch of PAL in Europe. It is interesting to read his reflections since the NTSC did much of the earliest PAL and PAF work.

I find his interesting note in passing on the second page (numbered 154) under the heading Stationary Axis. Mr Loughlin refers to his October 1951 IRE paper whereby the R-Y axis was stationary and the B-Y axis switched or inverted 180 degrees. He goes on to state the 1952 PAF NTSC testing made the B-Y axis stationary and the R-Y axis switched. In PAL, the R-Y switches at line rate.

I would be interested to see how closely the prototype receiver follows the Hazeltine decoder.

Pete Deksnis 06-24-2011 12:28 PM

re discussions of the oscillator:
 
1 Attachment(s)
It's a minor thing, but it made me recall the crystal socket that was discovered last Saturday. Went back to my photos and confirmed that the subcarrier oscillator was, in fact, built on a removable chassis. The tube is a 6U8, one RCA used in this circuit in pre CTC2 chassis before introducing the 6AN8. Unfortunately, the name of the builder of this chassis is no closer to being known. The crystal in this photo was from a CTC2.

Pete

miniman82 06-24-2011 01:49 PM

Quote:

Originally Posted by Penthode (Post 3007118)
if there is no crystal on this prototype I would trust there is a color hold control on the front!

Perhaps I should have been more clear:

There was no crystal INSTALLED, though there is a SOCKET for one.

JBL_1 06-24-2011 01:56 PM

Was thinking you will need to get some sort of power supply to power up sections of the chassis. I think the power supply from one of the many 1948-1950 B&W RCA projection sets would be a good fit. Has three 5U4G, it is an RCA chassis, just about the right period. And there has to be a couple of guys with spare chassis for these.


All times are GMT -5. The time now is 02:50 PM.

Powered by vBulletin® Version 3.8.4
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
©Copyright 2012 VideoKarma.org, All rights reserved.