A PORTABLE COLOR RECORDER (Mar, 1967)

|<<
<< Previous
1 of 2
|<<
<< Previous
1 of 2

A PORTABLE COLOR RECORDER

Newest type of helical-scan video tape machine has been colorized

By JOE ROIZEN*

RECORDING COLOR TELEVISION SIGNALS on magnetic tape has been practical since 1958 when the first compatible color broadcast recorders went into service. These transverse studio machines use four heads which rotate at right angles to tape travel (see Fig. 1). The machines also contain very complex circuitry and time-base correction devices. The circuits are necessary to achieve studio-quality NTSC playbacks that meet FCC specifications for on-the-air transmission; such VTR’s (video tape recorders) range in price from $40,000 to $100,000.

The development of inexpensive helical videotape recorders (Fig. 2) for monochrome industrial and home applications, coupled with the current color boom, has led to investigations into relatively simple, inexpensive ways to colorize these recorders. The pilot-carrier principle has proved a suitable system. Modifications to a normal monochrome recorder (Ampex VR-7000) and a home color receiver make it possible to record and play back color programs with a fidelity approximately equal to off-air home reception.

The time-base problem The NTSC color signal is composed of interleaved monochrome and chrominance signals amplitude-modulated on an rf carrier. The monochrome portion of the video signal requires only that the horizontal sync coming from tape have less than a 0.15% per sec2 rate change for stable monitor images. This is a fairly large and easy-to-meet requirement for modern videotape recorders with head-drum servos. The chrominance portion, however, has a subcarrier signal of approximately 3.58 MHz. The instantaneous phase of this subcarrier determines hue in the reproduced image. One cycle of the subcarrier (360°) has a 0.279-usec period, and a 10° error in subcarrier phase will produce a noticeable hue shift. 10° represents only about 8 nsec. Allowing for the accumulation of record and playback errors, a time base of better than 4 nsec is needed to reproduce faithful color pictures. Such an extremely fine time base is not easy to attain.

Any rotating mechanism is subject to undesirable movement due to mechanical and electrical eccentricities, dynamic imbalance, walking bearings, etc. The head-drum assembly in a VTR will normally display such variations in angular velocity as a time-base displacement of the reproduced signal. A monochrome picture may exhibit slight jitter, which is usually masked by the flywheel effect of the horizontal sync circuit of the home receiver. But when color is added, the rotating-head displacements show up as constant changes in subcarrier phase and the image looks as though it has lost color synchronization.

The pilot-carrier principle The composite color signal used for recording in the VR-7000-A (the color version of the VR-7000) is also fed to a burst separator which phase-locks a crystal oscillator running at the color subcarrier frequency (see Fig. 3). The output of the crystal oscillator is divided by 7 in a tuned circuit that yields 511 kHz, as shown in Fig. 4. The 511 kHz is then multiplexed at a 5% level onto the FM signal applied to the recording head. The current through the head then has a 5% pilot-carrier content. The level must be high enough to be detectable in the playback circuits yet low enough to minimize interference visibility in the reproduced image.

In playback (Fig. 5) the 511-kHz signal is recovered at the head preamp output, and a bandpass filter isolates it from the FM signal carrying the video information. Two limiters amplify and clip the signal to a uniform level; the pulses now drive a Schmitt trigger whose square-wave output goes to a second bandpass filter centered at 3.58 MHz, the 7th harmonic of the 511-kHz pilot carrier. The 3.58 MHz is amplified and fed out of the recorder to the chrominance demodulation circuits of the modified home receiver. The set’s own quadrature circuits form the 0° and 90° signals to decode the color information.

Since the pilot-carrier signal is subject to the same time-base displacement errors that the composite video signal is experiencing, the time relationship be- tween the pilot carrier and the desired signal remains constant. Hence the color signal can be decoded with reasonable time-base accuracy. The local oscillator in the color receiver is temporarily deactivated during VTR playback.

Recorder operation The signal system of the VR-7000-A (Fig. 6) must be capable of handling a bandwidth of at least 4.2 MHz to not attenuate the color sidebands. To eliminate unwanted noise, spurious high-frequency signals, etc., the input is filtered by a phase-linear 4.5-MHz low-pass filter network.

A fast-switching multivibrator-type modulator converts the video signal to FM. The carrier and deviation frequencies are somewhat elevated from their monochrome counterparts to minimize intermodulation effects between the FM signals and the high-energy color sub-carrier (Fig. 7). The modulator operates between 5.5 MHz at sync tip to 6.6 MHz at peak white. A rising pre-emphasis going up to 14 dB at the color sub-carrier improves signal-to-noise ratio and differential gain and phase. The FM signal goes to a head-driver amplifier which provides a constant-current source to the recording head up to 15 MHz. A rotating transformer with an 8-to-l ratio transfers the amplifier output to the transducer. A 50-microinch head gap is employed.

In playback (Fig. 8) a low-impedance preamp gives a flat frequency response. Aperture correction and equalization are applied to the FM signal before 50 dB of shunt limiters eliminate variations in signal amplitudes.

The output of the limiter is a constant-amplitude FM signal. A pulse-count detector and a 4.2-MHz phase-linear low-pass filter convert the signal back to video and remove residual carrier and deviation components. The output amplifier feeds two 75-ohm outputs, and the monitor (receiver) must be “jeeped” (rf and i.f. stages bypassed) to provide direct access to the video circuits.

Further development of a heterodyne signal-processing system will eliminate the need for modifying the home receiver. At that time it will be possible to modulate the composite video signal on a carrier and feed it into the set through the antenna terminals on an unused channel.

A color-kill circuit in the VR-7000-A detects the presence of bursts on the input signal and activates the pilot carrier in the record mode. If no burst is present, the pilot carrier is shut off so that the recording will not contain the 511-kHz signal. Under certain background conditions, faint vertical lines can be seen in the playback image due to interference from the pilot carrier. The level, however, is not high enough to be objectionable and with normal image conditions, is not noticeable.

The colorized VR-7000-A produces acceptable color pictures for most non-broadcast uses, such as educational, industrial and home applications.

12 comments
  1. Charlene says: December 13, 20084:04 pm

    I seem to remember similar machines at our school. I can’t remember exactly when they were used, but by 1978 they were long gone, replaced either by actual film or by videotapes.

    PS. Your Fig 8 turned into Fig. Smiley. Apparently 8 ) without the space gives you a smiley wearing sunglasses. 8) It’s a cool schematic and everything, but I thought you’d appreciate the heads-up.

  2. Casandro says: December 13, 20085:27 pm

    This sure was a _lot_ cheaper than the usual timebase correctors they had for colour video back then.

    For you non-video geeks out there, essentially the problem is like this:
    NTSC can be seen as simmilar to time-multiplexing of the colour components. So you transmit one component, then the other, then the first again (negative) then the second again (negative).
    Under normal circumstances, the reciever will get a short colour burst at the start of every line. This burst should be enought to keep the “clock” of the reciever in sync with the one at the transmitter. If you have a VTR however imperfections in the mechanics can cause yitter which is easily large enought to cause completely false colours. In professional machines (mentioned in this article) you had to have a special electronic memory which can delay the signal by a varying amount of time. (done with varicap diodes and inductivities) It is set so it corrects the yitter from the mechanics. What they did here is to add a second signal at a low frequency from which you could derive the right yitterey clock to your yittery colour signal thus you can decode it again.

  3. Rick Auricchio says: December 14, 20083:06 pm

    Charlie, this article reminds me of a Popular Electronics article where you could build your own linear-scan VTR. Because of the high frequencies needed for TV, that deck ran at something like 150 ips (consumer audio tape runs at 7.5ips). It flew through a reel of tape in perhaps 10 minutes.

    I used to be an avid “PopTronics” reader in high school, after having read a lot of the 50s-60s PopSci mags. My dad and uncle used to read PopSci, but it was I who went for the electronics end of things.

  4. Casandro says: December 14, 20083:40 pm

    Homebrew linear-scan VTRs sound interresting, but those 150 ips sound a little bit low. It’s just 20 times higher than normal audio tape. So if you estimate a Bandwidth of 20 kHz for the tape (unrealistically high) you’d only get 200 kHz for video. That’s still to low for a propper picture. You’d have to combine 10 channels to get the bandwidth needed for a decent picture. So unless you build special heads a home brew linear scan VTR might be very difficult.

  5. JMyint says: December 14, 20084:09 pm

    Cassandro, I suppose you had never heard of the Fisher-Price PXL 2000.

    http://www.retrothing.c…

    Thirteen minutes of video on a ninety minute audio tape isn’t too bad I guess, thats only 7 times speed.

  6. Casandro says: December 14, 20084:18 pm

    Yes, but that’s cheating, it wasn’t a full resolution VTR.

  7. Rick Auricchio says: December 14, 20088:30 pm

    Maybe it was 450ips?

  8. Casandro says: December 15, 20082:10 am

    Hmm, with 450ips and 2 channels, something which you might potentially be able hack your audio tape machine to do, you’d get 2.4 MHz of bandwidth. That would be enought for decent video. I’ve seen worse on broadcast television. Seriously there are stations out there still using VHS for broadcast, because they never could afford more and cannot afford to switch to computer based editing.

  9. Mike says: December 15, 20088:35 am

    Casandro, I don’t know if I want to laugh at your post. I had a friend/co-worker who worked at a station like that, and it wasn’t professional grade VHS or even consumer SVHS, it was off the shelf VHS. That was maybe 15 years ago, I would hope some of those stations have upgraded or have some sort of DVR set up by now if they are still around.

  10. Casandro says: December 15, 200811:13 am

    Mike, what country was that?

    Well today just doing it via a computerized system is _way_ cheaper. You can get full HD camcorders for less than 200 Euros.

  11. Wendy says: December 17, 200910:25 pm

    The BeBe Sounds Portable Color Video and Sound Baby Monitor is currently on auction.

  12. portable color tv says: December 17, 200911:43 pm

    We had a black and white TV when I was 2 years old. But today, I have a very small portable TV and it is very cool. The signal was very clear. You can watch anything on it. You can also plug in a DVD player and watch video with it. Technology will never stop improving as the year continues.

Submit comment

You must be logged in to post a comment.