|Hawes Mechanical Television Archive||
A Col-R-Tel® Camera
Cliff Benham took up the gauntlet! His latest project is a Col-R-Tel camera. This camera proves that mooncams used a color standard from a small company in Indiana: Col-R-Tel!
Many historical accounts report that the Apollo color mooncams operated on the "CBS System." As I've said on the other pages in this series, this statement is wrong. Although NASA documents don't refer to Col-R-Tel, the camera descriptions support my point. Although the Apollo cameras had color wheels, these cameras produced standard NTSC monochrome pictures. Yet the original CBS System was incapable of forming NTSC monochrome pictures. Fact: The Apollo mooncams were Col-R-Tel cameras.
Color wheel from Cliff's camera. The sensor notch is on the lower left.
As my Web pages prove, the CBS System is incompatible with the Apollo mooncams. In fact, the CBS System is incompatible with any other world TV standard.
Col-R-Tel is another matter. Col-R-Tel, which descended from the CBS system, is partly compatible with NTSC. The same is true for mooncam pictures. Mooncam pictures operate at NTSC scanning frequencies. Analog conversion electronics of the 1960s and 1970s could adapt the field-sequential moon pictures. The adapted pictures could play on everyday TV sets across NTSC countries.
Further conversion allowed PAL and SECAM viewers to view the pictures on normal sets. (Note that the British 405-line system never supported color. Of course, this system could only receive the moon pictures in black-and-white.
The mooncams are lunar Col-R-Tel
To prove that moon pictures are in fact lunar Col-R-Tel, you must build a mooncam. Then connect this mooncam simulator directly to a Col-R-Tel-equipped TV monitor. Between camera and monitor, remove the Col-R-Tel scan-conversion electronics. Only the wheel sync electronics should remain. Now Cliff has built the lunar Col-R-Tel camera. He then linked it to a Col-R-Tel-equipped TV without the scan converter. Result: Successful color pictures! What follows is an interview with Cliff, including his photos of the equipment he's using...
Cliff: This camera is a project that I started about eight years ago. I'd lost interest until Stan Lebar spoke at ETF (Early Television Foundation convention)this year.
Tonight, I got the first reasonable color pictures from the camera. The board camera has an AGC circuit that makes good images with very little light. The only illumination that I need is normal room lighting. To view the results, I used my Philco TV and Col-R-Tel color wheel. And yes, it definitely flickers!
For the camera mount, I plan to use a standard, quarter-twenty tripod mount.
The disc motor is free-running. Eventually, I'll use an optical sensor to take a pulse from the notch in the wheel. To keep the camera wheel in phase with the TV wheel, I apply a little "finger braking." The motor is a synchronous clock type.
Later, I'll sync the wheel to the camera and the wheel-notch pulse. To achieve sync, I'll feed 59.94 Hz to a motor-drive amp. This amplifier will power the motor. Then the wheel-notch signal will sync the wheel monitor.
Jim: Maybe you could allow the Col-R-Tel disc sensor to sync the camera disc. A preamplified sample of the sensor output could drive a power MOSFET or SCR to power the camera motor. Since the camera provides sync to the monitor, the monitor's disc sync would be accurate. The reverse feedback loop from monitor to camera would only drive the camera disc. This method would reduce the hardware requirements. You could use the method temporarily, for tests, or maybe even permanently.
How about building the little camera into a simulation of the color mooncam case?
Cliff: That's a nice idea, but the shapes of the board cam and wheel limit my choices.
This view show the camera (left) in position behind the color wheel
Jim: By the way, you might try substituting a couple of two-color discs. Then you'd have a way to test ideas about flicker reduction. My prediction: The flicker would nearly go away.
As I mention on my pages, two-color discs would have reduced the downlink hardware. Without the flicker, the slo-mo disc machine would be unnecessary. The Doppler correction and color picture signal addition would still be necessary.
My friend Adam Ross built a hand-synced, two-color cam. The same disc scans the camera and monitor. I think that I suggested this shortcut to Adam.
Jim: I just received this from Adam: "My two-color system had virtually no flicker!"
Cliff: I took a two-color disk set to the ETF (Early Television Foundation convention). Everybody agreed that the disc had less flicker than Col-R-Tel. Yet the two-color disc still flickered. The amount of flicker varies with color saturation and contrast. Suppose that you watch the set as you increase the color saturation or the contrast. As you pass beyond a medium setting, you notice more flicker.
Color wheel details
Jim: Out of what did you make your color disc? What's the disc diameter?
Cliff: The color wheel is a manufactured item. It's plastic, and has three dichroic glass filters for a digital reality helmet. The wheel rotates in the order R—B—G.
The outer diameter is 3-3/8 inches. About 12 years ago, I got the wheel for five bucks at an MIT hamfest in Cambridge.
The wheel filters color balance with the chip camera at about 3200 K. I'm referring to tungsten lighting. I can add a standard 85B filter and shoot good color images in daylight.
Closeup of solid-state, pinhole camera PCB
Cliff: The wheel rim is black plastic. The glass dichroic filters fit into three very thin compartments. These compartments allow the filters to move a little. When the wheel spins, centrifugal force holds the filters against the inside of the outer rim.
Jim: How did you attach the motor to your color wheel? Did you use a press-on gear?
Cliff: The color wheel has a plastic flange. With great accuracy, this flange presses onto the motor shaft.
Field-sequential color screen shot from Cliff's camera, as reproduced by Col-R-Tel
Cliff: I viewed the images tonight. I fed the field-sequential camera video directly to the video input on a Col-R-Tel setup. The TV is a 1950 Philco set.
Jim: Wow. This screen shot off proves that a mooncam is really lunar Col-R-Tel! Cliff, I know that you restored the vintage TV and the Col-R-Tel unit. Obviously, both work like new. For readers who don't know, I'll briefly describe Col-R-Tel converters. With the converter in place, the host TV becomes a field-sequential TV. The conversion is both mechanical and electronic. Reproducing NTSC video requires extra Col-R-Tel electronics. These electronics modify the incoming signal to produce field-sequential video for the converted set. This field-sequential video enables Col-R-Tel's wheel to accurately paint color details across the monochrome tube. Cliff's camera introduces a slight change. The camera itself outputs field-sequential video. That is, with this camera, no NTSC-to-field-sequential conversion is necessary. For this reason, Cliff can reproduce color pictures "barefoot." That is, without using the Col-R-Tel NTSC electronics. Only the color wheel and Col-R-Tel synchronous drive are necessary. That fact alone proves that a mooncam is a native Col-R-Tel device.
Cliff: Yes. Here are more details on the camera. The camera consists of an NTSC monochrome, one-inch square board camera with a pinhole lens. This is a solid-state camera with a CCD imager. The camera has an IR filter over the front of the lens. A 1200-rpm, five-watt clock motor spins a three-section, dichroic color filter wheel.
This sync motor direct-drives Cliff's color wheel.
Jim: I see that there is no gearing. The motor drives the wheel directly. I also notice that the motor manufacturer is Hurst. Like Col-R-Tel, Hurst is an Indiana company.
Cliff: Yes. Back to the camera board. I use this particular board for a very good reason: It must mount between the motor and color wheel. This board is the only type thin enough to mount that way. The camera board is about three-eighths of an inch thick.
The wheel required only a minor change. Over the red dichroic filter, I glue-taped an ND filter with a transmission characteristic of 30 percent. With this filter, the red disc wedge achieves color balance in 3200K tungsten light.
Now I've built the camera board into a Radio Shack project box. See the photo. What looks like a lens is a 52mm coated, UV clear-glass filter. This filter keeps ultraviolet light out of the board camera.
I screwed a closeup attachment lens onto the front of the UV filter. This lens compensates for the very wide view of the board cam's pinhole lens. On the front is an adjustable polarizing filter that reduces reflections.
Cliff's camera inside its case. What looks like a lens is actually a UV filter.
Cliff: For this camera, I plan to build a dedicated wheel monitor set. The new set will have better video characteristics than the Col-R-Tel has. No doubt an old VHS image enhancer will improve the detail of the camera video, too.
This camera is roughly equivalent to the Apollo color mooncam. I'd like to find two field-store boards. Then I can build a field-sequential to simultaneous NTSC color video converter. The output would be standard NTSC. The process would be the way that NASA made NTSC for the TV networks.
Jim: Of course the Benham Aeronautics & Space Administration would use more electronics and less mechanics.
Go to Page: 1 2 3 Back
Editing and comments by James T. Hawes, copyright © 2009 by James T. Hawes. All rights reserved. Photos and comments by Clifford Benham, copyright © 2009 by Clifford Benham. All rights reserved. Comments by the late Adam Ross, copyright © 2009 by Adam Ross. All rights reserved.