If you're lucky enough to have
a kino tube, it might not work. If it does, you probably
don't want to wear it out. Crater tubes are even rarer than flat plate tubes.
The LED situation is a different matter. Super LEDs are plentiful, and you
can choose from a lot of colors.
Some hobbyists also use cold cathode fluorescent tubes. Hobbyists who build
projection sets have tried lasers. Please don't use a laser in a direct-view
set. It will blind the viewer! Still want to try a neon tube? You'll probably
have to drive it through a step-up transformer. A reversed filament
transformer should do the job. I once built a direct driver for a neon tube.
I used a deflection transistor and ran it on rectified 120-volt line power.
The gadget worked, but wasn't as bright as some LED displays.
The problem is programming. Here's the answer:
The Narrow Bandwidth
Television Association sells three or four CDs with programming. If your
radio has a CD player, then you can play the CD through the radio and my
circuit. Through some automatic or manual means, bring your disc up to sync
speed. Now scan the LED with your disc. The LED must fill the entire frame
evenly, so I recommend a diffuser. Operate the LED a few inches behind the
disc frame.
By the way, NBTVA member Gary Millard offers a
free program to
simulate NBTV on your PC. You can watch the disc on your computer, and then
compare your scanner's performance.
A better (and cheaper) way to go is a
program disc from NBTVA. This will
run on a 32-line disc system. Also, I recommend using LEDs instead of a
neon tube. Flat-plate and crater neon tubes are antiques. Nobody that I
know of makes new ones. Anyway, LEDs operate better with solid-state
circuits.
1--Add a second amplifier channel. Use the same disc, but operate the
electronics as if you have two, separate TV monitors. Now, you have a
two-color system. One color can be red and the other can be green. Most
scanners already have the red. Just add the green. If that works, you can
add a blue channel. I have a version of the LED driver board with interlaced
rows of red and green LEDs. I haven't experimented yet with full color. My
derived color experiments worked. The green was too dim. I responded by
writing
color LED planning programs. One is on my web site. It helps you
to work out how many LEDs to use in each channel. The three-channel,
disc-based system is what
Ives invented for AT&T in 1929. Unlike Baird's
color system, the Ives system was practical, and could reproduce motion.
Ives' system was the basis for the RCA color system.
2--Divide your Nipkow disc into three sections. Put a different color gel
over each section. Switch to a white LED for the light source. Maybe add
another white LED to make up for the gels. Now you have something like
Baird's color system. This was the basis for the CBS color system.
If you have a color source for the monitor, then you just need to
sync it up. The source could be a CD or a scan converter such as the NBTVA
one. If you use a camera, then convert that to match the monitor. If you use
the AT&T system, you need two or three pickup channels. Each channel must
have its own color gel. If you use the Baird system, then your camera must
have a Nipkow disc with color filters. The filter positions on the camera and
monitor must match. To make the match, you'll need a commutator. Green should be
the brightest color, followed by red and then blue.
Run each color signal to its own amplifier. The circuit doesn't
need to I and Q (or U and V) encode or decode the signals. You have plenty
of bandwidth,
because everything is closed circuit. The colors occur simultaneously.
For that reason, you also don't need a field-sequential color wheel.
Of course, the input signal must be a color TV signal. Converting a black and
white signal would be an experimental effort. The late
Nam June Paik was a member of our Experimental Television Society.
Nam pioneered artistic color
conversions of monochrome, NTSC TV pictures. I believe that the results were
surreal, rather than naturalistic. That is, an artificial process derived
arbitrary colors. Nam's circuits tricked the NTSC color decoders into
producing fake hues. But that's a horse of another color.
In color-wheel terms, we call such opposite colors "complements."
Mixing any two complements results in white. In television, some complements
are better than
others. For example, red and cyan are a better pair than yellow and blue. You
can determine the best pairs from the proportions of each color in white.
For example...
Cyan makes up 70% of white, and red makes up 30% of white.
Green makes up 59% of white, and magenta makes up 41% of white.
Yellow makes up 89% of white, and blue makes up 11% of white.
With red and cyan, varying either color makes a lot of difference in the
signal. These
two colors are a fine match for color work. I suppose a 50-50
split would be
better yet, but you probably won't find one.
I've never experimented with
green and magenta. I don't know if I can find the right shade of magenta
LEDs to make such experiments practical. Green and magenta should work. A
green and magenta vector is part of the NTSC television signal. Our
color TV sets phase shift, detect and decode this "Q" vector.
Now let's look at yellow and blue. Most of the picture is in the yellow signal.
Yellow, of course, is a combination of red and blue, the two main primary colors.
Since you can't individually vary the red and blue, you lose control of a lot
of the picture. Variations in blue only account for 10% of the picture, which
isn't very much. Obviously, red and cyan are a better choice than yellow and
blue.
I wrote a script
that performs all these calculations for you. After you
have your numbers, you should apply common sense and examine the results. See
if everything rounds out the way it should.
You should know that the program
doesn't compensate for viewing angle or off-axis transmission. Your best bet
is to buy LEDs with nearly matching viewing angles. After mounting them,
adjust them for even illumination. You might need to rotate some of them until
you get the transmission axis right. Rotation can make a huge difference.
