Remember that the pictures are clickable to bring up larger, bandwidth-eating versions.
To view the AVI clips on this page, you'll need to install one of the Intel CODECs below. Sorry they're not Mac compatible.
My early efforts concentrated on the tracking system employed for the
Trailblazer
sounding rocket series in the early 1960's. There were two
influential papers:
It was my idea to update this method with modern technology, using DSPs to reduce the data "real-time" at a "master" station, which received data via ground radio links from the other stations.
I had no idea what frequencies to use, but Mr. Bjelland and Mr. Henning had used VHF to good success. I went hunting for a VHF amateur radio band to use. At that time, I was licensed as a Novice operator, so to send data on VHF I had to use 220 MHz. I decided that was the band to use. I also flirted with several other approaches, including the audio subcarrier of a low-power UHF TV transmitter.
At the time, I was very impressed with Bjork's approach, because it
required only minimal hardware in the rocket. And, the error margins didn't
seem too bad, but they were expressed as percentages of maximum altitude
(1 million+ feet), so they were much worse than they seemed...
The DSP processor was an Analog Devices ADSP-2105, running at 10 MHz (large square chip, lower right). There was 32Kx32 of RAM on the board (gold-topped chips). The board had high-speed ADCs and RS232 communications.
However, at the time, I didn't have the engineering skills nor the money to develop three full stations. I determined that measuring altitude was the most important thing, and I could do that with just one station.
One thing I didn't realize at the time is that errors in measuring the Doppler shift, when integrated, contribute doubly to the error in position, since two mathematical integration operations are required. Bjork's data showed lots of variation, but I though it was due to his data reduction methods. Lesson: do the math!
About this same time I had the (dumb) idea of making this my senior
project in electrical engineering school. I didn't realize that doing that
was tantamount to handing over control of the whole project to a group
of impractical academics and students (no offsense intended to academia
in general).
What the good professor proposed that we do was this: bounce a low-microwave signal (we used the 1296 MHz ham band) off the rocket, and measure/process the Doppler shift with the DSP unit. The fact is, no amateur rocket has a large enough radar cross section for a low-powered Doppler radar, but that was of no concern of my professor and the student zealots. They were concerned about getting an 'A' in the class, not solving the problem.
Needless to say, of course... it didn't work. Even though it took me more than a semester to get the beast working (and by then the other students were long gone, back to their respective countries-of-origin), it was an utter failure. With 50 mW of power at 1296 MHz, the "radar" wouldn't even track a TRUCK very far. We would have at least had some success had we used 10 GHz, a fact I pointed out many times.
Unfortunately, I cannot find any pictures that I took of this unit. It looked like a plumbing project, with all the semi-rigid coax cable that I used to interconnect the components.
However, I took some pictures of the 1296 MHz LO I built. This is an H. Paul Shuch (N6TX) design from 1979. I scanned the board layout from his magazine article, cleaned it up in a bitmap editor, and printed it onto transfer film to make the PC board.
Ah, well. Live and learn. This phase of DARTS taught me a lot about human nature, and gave me a chance to develop my microwave construction skills.
History Top
Generation II: FM/FSK (1993 - 1996)
[DARTS Main Page]
This document copyright Steve Bragg, KA9MVA. Updated: 06/19/2001