30 years ago, on May 2nd, 1982, a new satellite channel debut made history. TWC went online thanks to the work of dozens of pioneering meteorologists and technicians, including my friends John Coleman and Joe D’Aleo.
This is my personal story, never before told here.
I had a small part in supporting TWC in the early days, though some technology I developed, which I’ll talk more about later. But first, the launch of TWC. Here’s the pre-launch ceremony from Las Vegas and Atlanta (30 minutes) followed by the first 30 minutes of the Weather Channel’s broadcast from Sunday, May 2, 1982:
I was invited to the 30th reunion, which is being held this weekend, due to the fact that I myself was an early pioneer in weather technology for television, and TWC was one of my early customers using what was then some “revolutionary” technology I developed.
Few people who read WUWT know this, but I developed the first ever interface to allow computer weather graphics done on the IBM-XT @4.77Mhz, and later the IBM-AT @ 6.0 Mhz, to be broadcast on television. This was no small feat, because back then, such devices were usually rack mounted dedicated boxes. Using a dual slot frame buffer card from an early CAD/CAM terminal company called Vectrix, I designed the first ever PC based broadcast quality RGB to NTSC encoder card for the IBM-PC platform.
I’ve kept a sample of each piece of technology, which I photographed this morning on my desk. Here’s the complete solution, the dual slot frame buffer card, plus my NTSC encoder card:
The other half of the frame buffer card (not shown, underneath) is nothing more than rows and rows of 64K DIP memory chips. This ISA buss full length dual card was driven by an Intel 80188 CPU with a command set programmed to take ASCII commands (over a parallel or serial port) like [draw pixel, x,y] It was crude by today’s standards, but revolutionary then. Today, any cheap PC video card for $49 will run circles around what you see above.
Notice the long white ceramic chip in the upper left of the frame buffer card. That’s the heart of the graphics engine, the NEC 7220 graphics display controller chip, one of the first graphic chips ever invented. It allowed us to do things never before done outside of mainframes and was designed to be the heart of this beast, the NEC APC Advanced Personal Computer.
The trick to making the NEC7220 produce broadcastable RS170A (NTSC) video came with a mandatory need for something called “genlock“, which allowed all devices in a TV studio to be synchronized into a common switcher, so that video effects like green screen chroma key (essential to TWC) could be done.
Making the NEC7220 do genlock, was no easy task, since it had never been designed to do that job, and had no sync input of any kind. The task was something I took up, because I wanted to open up the IBM-PC to the world of broadcasting. When TWC started, they were using Z80 CPU/S-100 buss based Cromemco Z2 rack mount beasts with a 16 color frame buffer card done with an external rack-mounted NTSC encoder. The price tag on these things with software, broadcast encoder, and training was easily $30,000.
Cromemco (named for CROthers MEMorial Hall – the Stanford residence where the founders lived), came into existence in the mid 1970s, and grew to become a major player in the S-100 business systems market. Check out the dual 360K floppy drives whoo! More here.
The Z2 was adopted by Terry Kelly of Weather Central in Wisconsin to broadcast some of the very first weather graphics supplied by WSI corporation. The did custom programming (in BASIC no less) to enable some of the very first weather graphics to be displayed on TV, prior to that, we had Alden Fax images and magnetic symbols on metal boards. I used such a setup in 1978 when I first started in TV and I salivated over computer graphics. One time I tried to adapt an early Apple IIc computer for use on TV and in 1979 I called up Apple and asked to speak to the “chief engineer” about the video output quality. It wasn’t until a few years later that I realized I had given Steve Wozniak himself an earful about video signal engineering. I still remember the sound of the little floppy drive after typing in PR#6 from the console to get it to read the disk.
So when the IBM-PC came out, with a standardized and smaller buss, plus open sourced technical documentation (unlike Apple who with the Mac in 1984 created a tech prison) it made sense to try to make a broadcast system out of it. The broadcast video encoder was the big hurdle, and I solved it with this card below:
Note all the analog circuitry. There were delay lines (the big copper coiled tube is a 400 nanosecond delay line to match the 3.58 Mhz chroma subcarrier to the luminance signal) filters, scads of bypass caps to keep the noise down, plus subassembly chips and boards that were NTSC composite and RGB signal distribution amplifiers respectively. That plus a phase locked loop on the NE564 chip design that kept the clock of the VX384 frame buffer card in sync with studio gen-lock signals. It was analog black magic, all hand-made and hand-soldered.
Tuning this card was not unlike trying to tune the SU carburetors on a British Leyland Jaguar V-12. I had 12 trimpots plus a trimcap that had to be adjusted “just right”. Setup was accomplished using a TV monitor, an oscilloscope, and several test points on the card and usually took about two hours to get right. In those days, component drift could be a problem, and if you didn’t get the card up to temperature in the chassis first, you could miss the sweet spot and you’d lose genlock…which is a disaster on air in chroma-key when the satellite picture behind the talent in front of the green screen would go wonky.
Tech savvy readers might have noticed the “breadboard” area of the NTSC encoder board I designed. There was a reason for that, thanks to the Grass Valley Group corporation, whose GVG broadcast production switchers had nuances that required me to adjust the blanking signal in the RS170A output in order to get properly horizontally phased gen-lock at some TV studios.
[Trivia Sidebar: most people don’t know this, but the scene in the first Star Wars movie in 1977 where they fire the “death ray” from the Death Star, show a scene with a hand pulling a lever…it is actually a T-handle from a GVG model 1600 video switcher as seen above, and I think it was filmed at the studios of KGO TV by the ILM/Lucas crew]
So, I’d often have to add a switched delay line, and that breadboard section allowed me to do that on-site if need be. Yes, I’d tweak these systems onsite with a portable oscilloscope, wire wrap, a soldering pencil, and my wits.
Both WSI corporation and Accu-Weather used weather display systems I designed for them in the 80’s and 90’s. That little NTSC broadcast encoder board enabled hundreds of TV stations to put weather graphics on the air.
So, enough about the technology. The point is that John Coleman and Joe D’Aleo, who made TWC happen, thought my contribution to early TV weather and TWC was significant enough that they invited me to attend, even though I was never on the air at TWC, though back then it was a dream I had. I thank them for the gracious invitation.
I’ll be attending the TWC 30 year reunion this weekend, and reporting on this once and only event here at WUWT becuase I feel it is important to document this unique piece of American history. I’ll be traveling to Atlanta tomorrow, for the meeting Saturday. Blogging will be light the next couple of days.
I generally don’t like to beg, but all of the travel and lodging is out of pocket, and my “big oil” check still isn’t in the mail, so if anyone feels like hitting the tip jar (orange “donate” button) on the right sidebar, I will be most grateful.
If anyone has any questions about how TWC got started and operated they’d like me to ask while I’m there, feel free to leave a comment.