I toyed with the idea of using some sort of conical-scan back in 1993. It was pretty crude at first, but that's how you have to start out, sometimes. Would you believe a car-window motor (sans gears) spinning a subreflector with a Casegrain 1296-MHz backfire-helical feed? The whole structure vibrated, and the idea of that 8" metal subreflector becoming an 1800 RPM Frisbee was a little scary. I'm just not mechanically adept enough to dynamically balance something like that. The dish was the one below, but the feed, of course, was different:
I did manage to buy a surplus X-band conical-scan passive tracker from an older Navy anti-radiation missile, just to study it. It's an extremely well-balanced, precision piece of equipment. It has an eccentric parabolic dish that rotates around a fixed center feed, perfectly balanced. After spinning the rotor up to 1200 RPM with a Dremel tool (I don't have a source for the 400 Hz motor), the thing continues to spin for more than three minutes! That's precision, and I don't think something like that can be produced cheaply.
An interesting discussion about conical scan radars is the Lee
Davenport interview. Lee was the project manger of the SCR-584
conical scan radar in WWII. Here's a poignant quote:
" We had no compunction whatsoever about bending the rules a little bit if necessary to get something done. We would buy a washing machine if it would bring a quicker result....we were expected to hurry and we were not to be put aside for anything. That permeated the building of XT-1. The first experimental truck was a project that we moved heaven and earth to get done quickly. We had no reluctance whatsoever about going to an outfit that could spin a 6-foot aluminum parabola..." Wow.
"Hold a pencil with two hands. Hold the eraser end still, and move the pencil around so that the point describes a circle. This is nutation."
The Navy text gave the final hint I needed:: it said that the feed does not appear to the unaided eye to be rotating, only vibrating slightly. A-HA! As these things often do, a four-dimensional, working vision of a nutating feed appeared in my mind: an orbital sander with a feedhorn bolted where the sandpaper goes!
After making a 12:00 AM trip to the local discount store, and buying a cheap $15 Chinese palm sander, I started tearing it apart. I saw that the "nutating" motion was produced by an offset-drilled rotor fitted with a bearing. The motor turns the eccentric (but balanced) piece, and the outside of the bearing drives the sanding pad. I figure by mounting a feedhorn to the sanding pad, and replacing the AC motor with a stepping motor (with a "home" sensor), I can generate a nutating motion.
I'm going to try to bolt all this onto my RCA DSS dish that I have from earlier DARTS experiments. I'm going to see if I can't produce a sinusoidal amplitude modulation when the dish is pointed at a stationary source, and the "nutator" (sander) is on.
Conical scan is quite simple for a system that is receive-only, which
explains my interest in the next subject (one-way operation).
Suppose the rocket has a phase-locked oscillator (1), PN generator
and timing reference (2), BPSK modulator (3), and power amplifier (4).
There is also an antenna and power supply, but these are not shown in the
diagram.
The phase-locked oscillator (1) generates a carrier frequency accurate to within a few parts per million. The PN generator (2) generates a long PN sequence, which is then modulated onto the carrier by the BPSK modulator (3). The timing reference in the PN generator is also used as a reference for the carrier frequency PLL. The power amplifier provides power to the antenna.
I even thought of a way to use a 900 MHz transceiver design of mine for this purpose (at 5.7 GHz downlink, 900 MHz upllink). The uplink would be used to sync the PN generator. A sextupler (made up of diode doubler and MMIC tripler stages with microstrip filters, like the TX5B design) would take the 250 mW 950 MHz signal up to around 13 dBm at 5.7 GHz.
The ground station has a carrier-tracking BPSK receiver (1), PN
generator and timing reference (2), sliding digital correlator (3), and
range/doppler processor (4). There are also antennas and power supplies
that are not shown in the diagram.
Just before firing, the ground station synchronizes its internal PN generator to the received PN sequence.
When the rocket is fired, the increasing distance between the rocket and ground causes an increasing “slip” between the ground PN sequencer and the received PN sequence. If the rocket is 51 range units away, for instance, the ground PN sequencer will be generating the 51st bit, but the ground station will be receiving the 1st bit in the sequence. Therefore, the number of shifts between the two sequences is the distance between the ground station and the rocket.
Just as in the Apollo ranging system, the system resolution can be improved to far less than one range unit by Doppler tracking. The latter, of course, requires either multilateration or an extremely accurate frequency source in the rocket. Moreover, the higher the carrier frequency, the easier it would be to track the Doppler. At 5.7 GHz, the Doppler is about 8.4 Hz/mph; at rocket speeds (e.g., 300 mph), the shift will be 2.5 kHz or so.
The accuracy of this system depends on several variables: