The RDF antenna used in Amateur radio is not a true Doppler, but to help understand the Amateur Doppler, we will start by a short study of the hypothetical real Doppler. There are four figures. We suggest that you down load these figures so you can refer to them while reading the text. The real Doppler is illustrated in Figure 1 "Simplified Waveforms of the Doppler System," looking at the top wave form.
The Doppler principle states the apparent receive frequency will differ from the transmit frequency as a function of the relative velocity between the transmitter and receiver antennas. If the two are moving towards each other, the received frequency will be higher that the transmitter frequency. If moving apart, the received frequency will be lower than the transmitter. This is the same effect as the famous whistle of a passing train.
Download Figure 1 as a
GIF image. Its name is DopFunc1.GIF.
Display the figure as a separate
page. You might want to open it as a second window.
In Figure 1 "Simplified Waveforms of the Doppler System," we see a representation of a single antenna spinning around a horizontal circle (green). At the point on the circle marked t0, the antenna is moving directly towards the transmitter. Looking at the waveform (a green sine wave) we see that at time t0 there is a maximum increase in the received frequency.
If this wobbling signal is received by an FM receiver, the receiver output will be a sine wave of voltage. Let us assume that it is + polarity for the + excursion of the frequency.
The electronics system of our hypothetical real Doppler keeps track of where the antenna is on its circular path. Another part of the electronics is able to detect the exact time that the FM radio output passes from positive polarity to negative polarity. We call this the negative going zero crossing detector (NGZCD).
Our hypothetical real Doppler will display the antenna position on its circle path when the NGZCD detects the crossing. (Remember, the electronics "knows" where the antenna is at all times.) This position is the point on the circle that is closest to the transmitter. This display is held until the next crossing is detected.
Unfortunately, a real Doppler is not practical to build. In order to develop enough frequency shift to overcome the noise and / or modulation on the signal, the antenna must rotate at a very high speed. It would be very difficult and expensive to build one that would spin fast enough.
In 1976 there was an article published in QST about the DoppleSCant RDF. That started a long line of Doppler designs and products. I built one of my own design soon after. It had 8 whips in the antenna. The display had 32 LEDs in a ring (the most I have seen). The display also had a row of 4 more LEDs across the center of the ring of 32. These were verniers to divide the angle between adjacent ring LEDs into 4 smaller angles. The display therefore had the same resolution of a ring with 128 LEDs! (32 actually in the ring multiplied by 4 from the vernier LEDs). Its resolution was therefore 360 / 128 = 2.8125 degrees! The only time I could make use of that much resolution was if the signal was in the clear on top of a mountain.
I did a lot of experimenting with this antenna, looking at the waveforms with a two channel oscilloscope. I made diagrams and did thought experiments.
The following discussion is somewhat simplified. For instance, the method of calibration is not discussed. Also, there can be variations on the number of whips, the filtering system, the display, and some other bells and whistle that are not discussed.
The 8 antennas of my Doppler are selected one at a time in sequence going around the circle. The output of the receiver looks something like the second waveform of Figure 1. The pulses shown are delayed lightly (about 250 microseconds) after the antenna actually switches. This is due to the time it take the signal to move from the radio's input to its output. The antenna switches to the next element at each of the times marked t0, t1, t2,.....
This signal goes from the receiver speaker output into the Doppler control electronics. There is it passed through a special filter. This filter is really eight R-C low pass filters. They are selected by an electronic switch in synchronism with the antenna element selection. Thus each of these 8 filters charges its capacitor to a DC value corresponding to its average input value present while THAT antenna is selected. All eight of these filters feed a common point. Looking at that point with a `scope you would see a set of steps of voltage similar to the outline of the steps shown in the lower part of Figure 1.
These steps are then passed though another Low Pass Filter. Its cut off frequency is just high enough to pass the spin frequency of the antenna. the output of that filter will be a near sine wave, as shown in blue. This sine wave will lag behind the step-like waveform slightly, as is shown.
This cleaned up near sine wave then goes into the NGZCD circuit. The output of that detector is a pulse that is used to tell the display circuit to capture, hold, and display the position of the antenna at that instant. There is a delay circuit which is used to calibrate the system. I'll get to calibration later, but first I need to discuss the digital part of the electronics.
Download Figure 2 as a
GIF image. Its name is DopSys.GIF.
Display the figure as a separate page.
You might want to open it as a second window.
There has to be some sort of clocking signal and clock counter to keep everything coordinated. For the sake of discussion, lets assume the antenna sequences all the way around at 100 revolutions per second. Therefore we need to make 800 switches per second (8 antenna elements times 100 Hz). The switch is control by a device ("3 to 8 decoder" IC) that converts 3 counter output lines to active each of eight output line one at a time in sequence. That is, it counts 0, 1, 2, ...7, 0, 1, 2, .... Once though the count is also once around the antenna circle. It is also once though all 8 R-C Low Pass Filters to produce the step-like sine wave.
If you look at the cleaned up sine wave (blue) with an oscilloscope while slowly moving around the antenna with a transmitter, you will see the (blue) sine wave slowly and smoothly move across the screen. (The scope is synched to the above 3 to 8 decoder, say the 7-count event.) Not that the sine wave does not move in jumps of 45 degrees (the spacing of the antenna elements. It moves smoothly! As you are walking the transmitter around the Doppler antenna, the individual steps are rising and falling SMOOTHLY. These, after going thought the next low pass filter (cut off of about 125 Hz, produce the smoothly shifting blue sine wave of Figure 1.
Download Figure 3 as a
GIF image. Its name is DopFunc2.GIF.
Display the figure as a separate
page. You might want to open it as a second window.
Examine Figure 3 carefully. These two sets of waveforms are produced by signals arriving from directions that differ by 22.5 degrees, which is 1/2 the angular spacing between antenna elements! Look at and compare the difference in the pulses coming from the receiver, and comparing the difference in the step heights of the two step-like sine waves. You may be able to see from this how the smoothed sine wave can shift smoothly.
But how do we get control of 16, 32, or even 128 LEDs when we have only 8 antenna elements? This is sometimes a difficult thing for people to understand. I will try to explain.
To simplify this discussion and diagrams, we will assume only 16 LEDs. (Twice as many as there are antenna elements.) Look at Figure 3, at the bottom. There we see a series of Clock Pulses. Note that there are 16 for one revolution of the antenna around its circle. Note the numbers along the bottom, 16, 1, 2, 3, ....16, 1, 2... These are the LED numbers in a ring of 16. Every other one of these pulses (total of 8) control the spinning of the antenna, while every one is used to control the LED display.
For my Doppler the clock would be 128 LEDs x 100 Hz = 12800 Hz! Still only 3 are needed to control the antenna.
Figure 3 shows two waveform drawings for signals coming from two different directions. Look at the top one, at the first zero-crossing (circled in red). Note that this happens when we are in the clock interval corresponding to LED # 5. The crossing detector fires. (We will ignore the calibration delay for now.) The electronics triggers a memory circuit to remember that number (5). Then memory feeds this number to the display electronics, and LED # 5 is turned on and left on (until the next crossing detector firing).
Now look at the bottom waveform. Here the blue sine wave is displaced a little to the right from the upper one. This reflects the fact that the RF is arriving from a different direction (22.5 degrees different). Now look at the first zero crossing of this lower waveform, and note that it occurs during the clock interval corresponding to LED # 6. Thus LED # 6 will light.
In my Doppler, the clock signal at the bottom of Figure 2 has 128 intervals per spin instead of 16, but otherwise the operation is the same. It just takes more electronics to do it.
There must be an adjustable means of calibrating the system. This is to take into account the delay of the signal passing through the receiver. This delay can be different with different receivers. Also there is some delay of processing inside the Doppler electronics, especially the 125 Hz low pass filter.
Download Figure 4 as a
GIF image. Its name is DopCal.GIF.
Display the figure as a separate
page. You might want to open it as a second window.
The top drawing of Figure 4 shows an ideal system that does not require calibration. The next part of the figure shows the error with a real Amateur Radio Doppler without any compensation to calibrate. Due to the system delays the display is "Late" (error is towards the right, or Clock Wise (CW) direction). This is because the electronics has moved on to a later LED by the time the trigger event occurs. Calibration is done by first placing the Doppler antenna on the car with a Counter Clock Wise (CCW) "placement error," as is shown in the third part of the figure. This CCW placement error must be big enough to make the display error now too Early. After this placement error is created, the electronics delay circuit is adjusted until the display reads correctly, as in the bottom part of Figure 4. The Doppler is now ready to use.
Note that this calibration example assumes the antenna spins CW. If yours spins CCW, the CW and CCW references in the text and in the figures must be reversed.
Russ, K6BMG
Jump to...Doppler Weakness....The "Doppler" is not a Doppler.
George R. Andrews (Russ, K6BMG) BMG Engineering, Inc. 9935 Garibaldi Temple City, CA 91780, USA Voice: 1(626)285-6963 FAX: 1(626)285-1684 (24 Hour, automatic) America OnLine: Grandrews Web: http://members.aol.com/bmgenginc (18 Mar 1996)
Send E-mail to grandrews@aol.com.
(A message window will open.)
Return to: Top Page.
Return to: Tutorials
Contents.
