The Doppler Principle wasn't discovered by a ham. Its originator died two decades before Marconi was born. But when most hams think of radio direction finding (RDF), Doppler sets come to mind. This first installment of a classic Homing In series traces the development of technologies that began with a scholar whose 200th birthday was being celebrated when the column was published in 2003.
A great deal has been written about Christian Doppler and the effect named after him, but much of it is inaccurate. Many sources even have his given name wrong, stating it as Johann Christian, or the reverse. Dr. Alec Eden, himself a pioneer in the use of the Doppler principle for medical applications, found Doppler's baptismal records at a small church in Salzburg.[1] Dr. Eden discovered that the Austrian physicist/inventor's full name was Christian Andreas Doppler.
Mozart wasn't the only famous person born in the "Sound of Music" city. This plaque on the house at Makartplatz 1 in Salzburg notes that Christian Doppler was entered the world here on November 29, 1803, just across the river from Mozart's birthplace 47 years earlier.
Christian Doppler was born into a family that had built a successful stonemasonry business for 125 years, but his health was always too bad for that arduous occupation. After studying mathematics and astronomy at two institutes in Vienna, he sought a career as a professor. Despite being quite ambitious and having already published four papers in mathematics, he was turned down for several positions. To make ends meet, Doppler spent almost two years keeping the books for a cotton cloth factory, becoming more despondent as the months went by. At one point, he sold his possessions and went to Munich to arrange a permanent move to America. He ended up not going, because shortly thereafter he was offered a position at a technical school in Prague, then part of Bohemia.
Some have claimed that Doppler ascribed his principle of motion affecting perceived frequency only to sound waves, then others extended it to electromagnetic waves. Not so. His 1842 work was titled "On the Coloured Light of the Double Stars and Certain Other Stars in the Heavens." He gave an oral presentation to the Royal Bohemian Society of Sciences at the Patriotic Hall of Carolinum on the same topic. He theorized that light from double stars (stars mutually orbiting each other) should display differing colors. The star rotating toward earth at a given time should appear more blue to an earthbound observer and the one rotating away should be redder. There was no way to measure tiny shifts in apparent color from stars in those days, so Doppler couldn't directly prove his hypothesis by experiment. Most astronomers pooh-poohed it. Six years later, Doppler set astronomy aside and took on a new scientific interest, studying the earth's magnetic field as a teacher at the Academy of Mines in Slovakia. His publication topics included changes in earth's magnetic declination, another subject of interest to hams who are mobile T-hunters.
Political unrest made him decide to return to Austria in 1850, and he soon was fortunate to be selected as the first Director of the new Institute of Physics at the University of Vienna. But by November 1852, his health had deteriorated to the point that he sought relief by moving to sunny Venice, Italy, where he died four months thereafter.
This public domain Daguerreotype of Christian Doppler is thought to have been from 1845, about six years after that silver-on-copper photographic process was invented.
If you read somewhere that Doppler went to great lengths to validate his principle, take it with a grain of salt. He was a visionary with a love of science and an inventor of many novel optical instruments, but his grasp of mathematics wasn't first-rate. He insisted that light waves were longitudinal, like sound, instead of transverse as French physicist Augustin Jean Fresnel had correctly theorized in his landmark paper of 1821. But Doppler did successfully predict that his principle would someday provide an important method for determining the movements and distances of stars.
It was the Director of the Dutch Royal Meteorological Institute who staged the most vivid verification of Doppler's hypothesis on June 3, 1845.[2] Christoph Buys-Ballot was actually a disbeliever in the principle and thought he could disprove it using sound waves. He enlisted help from a half-dozen trumpeters with perfect pitch. Half of them rode in an open Rhine Railroad train car at high speed through Maarsen station playing a G-note, where the other half of the trumpeters stood on the platform and observed the pitch.[3] The experiment was repeated on the return train trip, but this time the platform musicians played and the ones on the train listened. Buys-Ballot had calculated that if Doppler was right and if the train was moving at 46 MPH, the perceived trumpet pitch would change by a half-tone (6 per cent). Over the noise of the roaring train, the musicians agreed that it did. Despite this dramatic evidence, Buys-Ballot remained a skeptic.
Doppler published another paper further describing his principle in 1846, including movement of the observer as well as the source this time. He likened the stars' color shift to a boat heading out of port into incoming waves, making the wave impact rate appear faster than to an observer on shore. That part was right, but he didn't realize that the relative motion of double stars is too small to create a noticeable color shift. They appear different from other stars simply because their lightwave spectrum is different.
French physicist Armand Hippolyte Louis Fizeau performed the first non-astronomical measurement of the speed of light with a rotating toothed wheel in 1849. He suggested that some amount of Doppler shift could be expected from all stars that are moving relative to the earth. It was almost two decades later that English astronomer Sir William Huggins, the pioneer of spectroscopy, made instruments sensitive enough to show this to be true. As instruments improved, scientists measured the relative movements of stars and galaxies to develop theories of how our solar system was formed and what its future will be.
The average non-ham citizen connects Doppler with just two things, speeding tickets and severe weather forecasts. A patrolman's radar set puts out a microwave signal and measures the frequency difference of the return signal that results from the vehicle moving toward or away from the radar. The greater the difference, the greater the speed.
Joe Citizen might not realize that ground-based radars tracking aircraft and boats make use of the principle too. They display only the pips from Doppler-shifted moving targets and ignore all the non-moving reflections (clutter) from land objects such as buildings and hills. This Doppler-processing trick works especially well in airborne radars, which eliminate ground clutter on their displays by suppressing the echoes that appear to have exactly the same velocity as the aircraft.
Does this mean that a Doppler weather radar measures the speed of the clouds to tell which ones carry severe storms? Not exactly. Clouds usually move too slowly to create enough Doppler frequency shift, so he best way to determine cloud density and precipitation intensity is to simply measure the magnitude of the signals bounced back to the radar. That produces the multi-color "reflectivity" displays that you see on TV. But Doppler weather radar has the additional ability to look at localized wind disturbances such as tornadoes and microbursts, by measuring the velocity of the raindrops, and even the blown-about insects! In the "radial velocity" mode, which is seldom shown on TV, a tornado produces a distinct Doppler "signature" of inbound winds next to outbound winds. Sometimes there is a pocket of no wind sandwiched between --- that's the eye of the storm.[4]
The train-like situation of object moving at a constant rate toward the observer, passing, then moving away at the same rate is rarely found in VHF radio direction finding. The only example that comes to mind is the Search and Rescue Satellite Aided Tracking (SARSAT) system. Each of the USA's four SARSAT birds, in near-polar orbit at 528 miles altitude, passes over every point on earth several times a day. Each one carries a Search and Rescue Repeater that receives and retransmits 121.5 MHz and 243 MHz signals to ground stations. As a satellite passes over a squawking beacon transmitter, its frequency appears to shift lower at the point of closest approach (PCA). The exact amount of the frequency shift tells the perpendicular direction from the satellite's track at the PCA. The exact time of the shift gives the PCA along the track. This is sufficient data to compute position of the beacon on land with accuracy of 2 to 3 miles, or better. Multiple satellite passes refine this accuracy further. Rescue Coordination Centers communicate SARSAT fixes to local organizations such as Civil Air Patrol.
No matter how fast we hidden transmitter hunters drive, our vehicles don't produce measurable frequency changes in the hidden T's signal. So ground-based Doppler RDF sets produce such shifts by putting a vertical receiving antenna into a very rapid circular motion. For a signal incoming from any azimuth, the antenna moves toward it, away from it, toward it, and so on. When moving toward, the apparent received frequency increases and when going away, it decreases.
The Doppler equation below gives the peak frequency shift (S in Hz) as a function of signal frequency (Fc in MHz), circular rotation rate (Fr in inches), and radius of the circular track (R in Hz). Positive and negative peak frequency shifts occur at the points on the circular track where the tangential antenna velocity, relative to the incoming signal, is maximum and minimum respectively (points b and d). From the instantaneous phase of the induced sinusoidal Doppler frequency shift, and from knowing where on the track the antenna is at that instant, we can determine the azimuth of the incoming signal.
Movement of a single vertical antenna in a circular track (top) makes received signals appear to vary sinusoidally in frequency (bottom), in accordance with the formula. This is simulated in ham Dopplers by sequentially switching three or more antennas spaced around the track, then filtering all harmonics from the stepped waveform induced in the receiver audio.
For a practical Doppler RDF system using typical FM receivers, the circular rotation rate must be such that the recovered sinusoidal frequency shift information (the "Doppler tone") is within the audio output passband (about 150 to 2000 Hz). That corresponds to rotation rates of 9000 to 120,000 RPM. We can't physically spin a two-meter vertical whip at that rate, but we can simulate the moving whip by putting three or more whips at equal intervals around along the track and connecting them in sequence to the receiver at an audio rate with some sort of electronic switch.
A moving antenna for Doppler-based direction finding was first published in a 1947 technical journal by British engineers C. W. Earp and R. W. Godfry, who used a mechanically spun antenna mount. In his May 1978 QST article, Terrence Rogers WA4BVY described the first practical VHF Doppler RDF set for hams, employing eight whip antennas in an octagonal pattern and a PIN diode RF sequencing switch to achieve an effective rotation rate of 38,000 RPM. Many improved versions followed, including the Roanoke Doppler, which was the most popular home-built Doppler project for a decade.[5]
SARSAT takes measurements over thousands of wavelengths of distance, which is why it's called a wide-aperture RDF system. On the other hand, a typical ham Doppler is narrow-aperture because its antenna system smaller than a wavelength. Bearing accuracy suffers in a narrow-aperture system because signal reflections cause localized disruptions in the incoming signal wavefront. The solution is to take many bearings (hundreds per second) and average them as the Doppler set moves down the road. That effectively lengthens the baseline. The averaging is done by a narrow audio filter with a passband of as little as 2 Hz. To keep the filter passband centered on the Doppler tone, the same oscillator drives both the antenna sequencing (which produces the tone) and the tone filter.
The basic Doppler RDF block diagram has changed little since 1978, but many of the functions are moving from hardware to software and firmware, reducing the parts count from a dozen ICs down to about five. Signal processing, analog or digital, detects potentially erroneous bearings and suppresses them. Improvements are also being made to the display of bearing information. Today's Dopplers can be part of integrated mobile RDF systems that include compasses, GPS, and computer displays, as you'll learn in later parts of this series.
A very simplified block diagram of a Doppler RDF set. Functions of the blocks have remained virtually unchanged over the years, but the components within them, using hardware or software, are adapting to new technologies.
If you go on a transmitter hunt with a bunch of Doppler fans, you'll find that most of them use four-whip antenna sets. As few as three would work, and some have as many as sixteen. No matter how many whips there are, it's important that they be placed so that they are all exactly the same distance from the center of the array circle, and they are equidistant along circumference of that circle. Using four makes it easy to put magnetic-mount whips on a car roof, because they can be placed in a square pattern.
Is four whips the best number for a mobile setup? How is it possible to get accuracy of five degrees or better with only four whips? What does the Doppler-induced waveform from a typical 4-whip set look like, and why does it lead some hams to say that it's not a "real Doppler?" Those topics lead off the next article of this series, which delves into practical Doppler antenna sets and ways to make them work better.
---------------------
[Note 1] Children were usually baptized within a few hours of their birth in the early 1800s because so many died in infancy. Based on the date and time of the entry in the records, there is no doubt that he found the right baby Doppler.
[Note 2] He had attempted it four months earlier, but had to abandon the effort because snow was blowing into the faces of the trumpeters and the cold air was detuning their instruments.
[Note 3] The pitch of the train whistle was not pure and steady enough for the experiment.
[Note 4] For more about Doppler weather radar, read this NOAA article.
[Note 5] Complete plans in "Transmitter Hunting---Radio Direction Finding Simplified" by Moell and Curlee. This book also has a comprehensive explanation of how a Doppler RDF set obtains bearings using the phase of the sinusoidal frequency modulation imposed on the incoming signal by the pseudo-rotating antenna.
Text and photos © 2003 and 2025 Joseph D. Moell. All rights reserved.
Go to
Doppler Series, Part 2
Back to the Homing In home page
This page updated 24 April 2025