Forum: WHALER
  ContinuousWave
  Whaler
  Moderated Discussion Areas
  ContinuousWave: Small Boat Electrical
  A Primer on Modern Recreational SONAR

Post New Topic  Post Reply
search | FAQ | profile | register | author help

Author Topic:   A Primer on Modern Recreational SONAR
jimh posted 02-24-2014 10:55 AM ET (US)   Profile for jimh   Send Email to jimh  
A Primer on Modern Recreational SONAR for Non-Mathematicians and non-Electrical Engineers

by Jim Hebert,
a non-Mathematician and non-Electrical Engineer

Modern recreational SONAR devices have, in the past few years, evolved considerably in the sophistication of their technology. Aided by the relatively very low cost of very high-performance digital signal processing chips, recreational SONAR devices have been able to incorporate extremely sophisticated techniques of echo-ranging into devices that are affordable for use on even small recreational boats. This article will attempt to explain in a very simple manner how the new technology differs from existing technology.

First, a review of the conventional SONAR technology will be beneficial. Most recreational SONAR devices up to now have transmitted a very short, high peak amplitude (but low average power) pulse at a single frequency, created by exciting a narrow-band and very sharply-resonant transducer. The short pulse was needed in order to obtain reasonable resolution of targets. This is easily understood without much mathematics. Imagine you have a loud air horn. You make a one-second blast of sound with the air horn. When you listen for echoes, it will be hard to hear a distinction between echoes if the separation time of echoes returning from two different reflecting surfaces is less than one second. The first portion of the returning echo from a slightly more distance reflector may overlap the last portion of the returning echo from a closer target. It is difficult to resolve or separate the two targets when their echoes overlap in time. This general problem led all echo-ranging systems to tend to use very short pulses of transmitted signal to provide target resolution.

Another characteristic of conventional SONAR is the narrow bandwidth of the signal sent. Because a highly resonant transducer was used, the emitted signal tended to be confined to only one frequency. When listening for return echoes, the echoes will be at this single frequency. This is known as a monotone or monotonic SONAR.

A further constraint in conventional SONAR is the notion of creating energy in the transmitted signal. Because the duration of the transmitted signal was limited to a very brief pulse, the amplitude of the pulse had to be made very high if enough power was to be introduced into the pulse to be able to carry across a long distance to a reflector and return. The high pulse power also creates problems for listening for return echoes, as the same transducer is typically used for both creating the transmit signal and listening for the return echo. When the transducer is excited on transmit by a very strong pulse, it tends to continue to produce some signal after the exciting pulse is removed. This is called ringing, and is much like a bell that continues to create sound after the bell has been struck by the initial blow of the clanger. The transducer ringing can obscure returning echoes from close-in reflectors.

In conventional SONAR, the detection of return echoes was done by listening for variation in amplitude at the transmitted frequency over time. Success at detecting a return signal required that the amplitude of the return signal be higher than the noise floor of the detection system and other noise picked up by the receiver. Echoes from large targets return with a higher amplitude than echoes from smaller targets. Echoes from close targets return with higher amplitude than echoes from distant targets. To hear echoes from small distant targets in the presence of noise becomes a limiting factor.

The modern SONAR differs from the conventional in three distinct ways:

--the emitted signal is no longer a very short pulse but a longer burst of a continuous-wave signal;

--the emitted signal is no longer a simple unmodulated signal but a frequency modulated or swept signal;

--the detection of the signal is no longer done simply by looking for amplitude change over time but correlates the frequency of the echo by using sophisticated signal processing techniques.

The results of this method are to create a system in which the effective pulse length of the transmitted signal is actually shorter than its actual length. This outcome is referred to as pulse compression. The effectively shorter pulse improves the resolution of the echo-ranging system. An additional benefit is a significant improvement in the sensitivity of the detector for weak signals. Since the pulse is effectively compressed on receive by the signal processing, the total power of the pulse is effectively applied over a short time. This has the effect of increasing the effective power of the pulse at the receiver.

These benefits both occur as a result of the sophisticated manipulation of the received signals in the detector or processor of the echo sounder. In the simplest sense, one can understand this by considering that the signals are processed in three domains at once: time, amplitude, and frequency. This is in contrast to conventional systems which tend to only process the signal in the time and amplitude domains. By looking for correlation or matching in the frequency domain in the reflected signals, better performance is obtained both in resolution and in signal-to-noise. It is now possible to detect signals which are so weak and obscured by noise that they would be undistinguished in a conventional system, and to separate signals from closely spaced reflecting targets.

To understand how the frequency modulation of the transmitted signal helps improve the detection of a return echo from the signal, consider that the detector can now be designed to look for returning signals that exhibit a pattern of frequency change that corresponds exactly with the pattern of frequency change of the transmitted signal. Because the noise in the signal tends to be random, when a pattern matching detection system is used the noise is effectively filtered out.

If you want to look into the mathematics of the pulse compression method of echo location, refer to the Wikipedia article at

http://en.wikipedia.org/wiki/Pulse_compression

The mathematical description of the method being used is sophisticated and a bit hard to grasp, particularly for boaters who otherwise are not necessarily very acquainted with complex mathematics and electronic theory. An easier to grasp understanding can be gained in a visual representation of the signals, where color is used to indicate frequency. For a very good explanation of pulse-compression with excellent illustrations using this method of representation of frequency by color, please refer to this short presentation on the subject:

http://www.tritech.co.uk/uploaded_files/What%20are%20CHIRP%20Sonars.pdf

There you will find some illustrations which will help you further understand the method. (Please note that I have not presented these illustrations in my article as if they were my own work. However, you may recognize some of the illustrations as they have been widely distributed by others without any citation of their source.)

The simple name for the whole process of echo sounding with this technique is sometimes referred to as "chirp." The term "chirp" refers to the acoustic signature of a signal that is frequency modulated with a sweep modulation. If an audible signal is heard with these characteristics is tends to sound like a bird's chirp. The term "chirp" has been used to describe continuous-wave frequency-modulated short signal bursts since the 1920's, when it was used to describe radio telegraphy signals where the stability of the transmitter frequency was not very good, and the signal carrier frequency tended to drift up or down during the duration of each dit or dah being sent.

This method of improvement in echo-ranging with frequency modulated signals was invented by an American, Sidney Darlington, when working at Bell Laboratories. The technique was developed in 1949 and used analog circuitry, which I find quite interesting due to the contrast with today's methods, which tend to employ digital signal processing to implement the same techniques. Darlington's patent application describes the method. Compare at

http://www.google.com/patents/US2678997

Manufacturers of marine electronics have often presented very technical features employed in their devices by giving them simplistic names. It is not surprising that pulse-compression echo ranging has been given a simpler term. Many manufacturers have adopted the term chirp to describe the entire process, even though the chirp is actually only a part of the transmitted signal. You could send frequency-modulated signals all day long and get no benefit unless you detected them with a the appropriate pulse compression signal processing. Nevertheless, the term chirp has come into common use.

The descriptor "chirp" has been recently presented as being an acronym, often as CHIRP. A number of interesting explanations have been given to be the meaning of the acronym. For example, one or more manufacturers have taken to using CHIRP and telling boaters it means "Compressed High Impact Radar Pulse" and others suggest "Compressed High Intensity Radar Pulse."

These explanation may not provide much help in understanding the method, and may, in fact, be misleading. First, the actual pulse sent is not compressed compared to the conventional echo ranging systems, but is typically longer. It is only compressed in a virtual manner by the effect of the signal processing applied to it in the detector. The words "High Impact" convey no real meaning. They appear to have been chosen because they start with "H" and "I", and for little other reason. The notion that the pulse is "High Intensity" is also somewhat paradoxical. In a chirp system the peak power or intensity of the transmitted signal is lower than in conventional systems. The chirp signal is a lower intensity but longer signal. Finally, in the case of SONAR systems, it also seems paradoxical that one would refer to the method as being a RADAR pulse, as RADAR is generally considered to be a technology that operates through air, not water, and uses signals whose frequencies are 10,000-times higher. In short, CHIRP is a simplistic term used by makers of pulse-compression SONAR systems in order to have a shorthand descriptor. But you should not expect to be enlightened about CHIRP from its inventive acronym. Chirp does not "actually stand for Compressed High Impact Radar Pulse," as is often claimed. Chirp refers to the sweep frequency modulation of the frequency modulated signal, because such modulation sounds like a bird chirp. The acronym CHIRP was probably created long after chirp was used initially to describe the frequency modulated transmitted signal.

In some casual research I have found that citation of CHIRP as an acronym with various terms given in explanation occurs most often on boating websites or in boating electronic sales literature. I have not found any really authoritative technical literature that references the acronym. I believe that chirp SONAR is a reasonable term, but I don't see any value in or importance to the rather odd claims for the acronym or the notion that the acronym is fundamental to understanding the process.

A fundamental principle of a chirp echo-ranging system is the notion that the range of the frequency modulation sweep is an important factor in the amount of improvement provided by the method. The wider the range of frequency sweep (or bandwidth) of the transmitted signal, the more improvement (in theory) is possible. Because the transmitted signal must be sent over a wider range of frequencies than in a monotonic SONAR, the transducer must be capable of operating over a much wider range of frequencies. Typically the usual monotonic SONAR transducer is sharply resonant at a single frequency of operation. For chirp SONAR to work, the transducer must be able to operate over a much wider bandwidth. This tends to result in a transducer that is physically larger and more expensive. Some chirp transducers cost as much as the electronic units they are paired with. This is in contrast to conventional SONAR devices where the transducer cost is only a small fraction of the electronic device cost, perhaps as little as one-twentieth the cost of the electronics.

That a particular brand and model of SONAR device employs some form of pulse-compression or chirp technology does not automatically mean it is equivalent to all other chirp SONAR devices and will produce improved or superior results. As in any method, the implementation of the method and the success of the results are subject to variation.

Another on-line resource that offers a readable description of pulse compression techniques can be found at

http://www.radartutorial.eu/08.transmitters/Intrapulse%20Modulation.en. html

The above article has some good illustrations which help to explain the method.

In doing more reading about pulse compression echo ranging systems, I notice that the term chirp, as a word, not as the invented acronym, is generally used only in regard to linear-sweep frequency-modulated transmitted signals. There are other methods of frequency modulation of the transmitted signal which are not a linear sweep or chirp. And, there is never any mention of CHIRP as an acronym.

The general term for the method seems to be pulse compression. On that basis I am going to prefer the designation pulse-compression SONAR to chirp SONAR. I also find the acronym CHIRP to have little meaning or value, and will avoid its use as a way to describe pulse-compression SONAR. So, good-bye CHIRP and hello pulse-compression.

Post New Topic  Post Reply
Hop to:


Contact Us | RETURN to ContinuousWave Top Page

Powered by: Ultimate Bulletin Board, Freeware Version 2000
Purchase our Licensed Version- which adds many more features!
© Infopop Corporation (formerly Madrona Park, Inc.), 1998 - 2000.