Functional fishfinder sidescan

[Rickard]

Why it works.

There has been no opposition, contrary to what I expected, to my pushing of the idea of optimizing fishfinder arrays by unequal spacing. It works, as I know and have tried to demonstrate. The reasons for this are interesting. One has to perform rather advanced simulations to get the full picture of what happens in these cases.

For that reason I made various simulations in 3D where all influential parameters were studied. One has to consider piezo dimensions, spacings, frequencies, angles in all directions for transmitted and received sound and, this proved to be important, time during a sound wave cycle. Simple simulation only in a plane along the array and perpendicular to to the optical window does not show all phenomena.

The picture shows the beam pattern for an equally spaced and an optimized array with four circular piezos. Diameter is 20 mm and frequency 200 KHz. Array length about 300 mm. A circle represents angles from 0 to 45 degrees from the middle of the sound beam. Thus, a circle covers 90 degrees. Middle of sound beam is in the center and at the edge the angle is 45 degrees. The height represents one-way transmitted energy. The orientation of the array is indicated by a line.

As can be seen side lobes are much lower in the optimized array. The side lobes are not smooth "bulbs". They are divided into many "fingers" so the influence of side lobes depends on in what direction sound is transmitted/received. The finger pattern is caused by a complicated interaction between the circular form of piezos, the array spacings, angle and frequency. I have seen the result of this when using 83 KHz over a flat bottom. Then the fingers can paint faint bands of echo. The main beam itself does not seem to be much affected by this finger phenomenon. Note that vertical width (not only the height) of side lobes decreases the farther away from the main beam they are. This is very good and depends on the shaping effect from the circular forms of piezos.

These figures are not the typical dB charts but merely shows relative energy level. dB is so depressing, especially in this case. Then it's better to think about the one-way/two-way distinction. These figures show the one-way situation. That is, when sound is either transmitted or received. For piezos things multiply as they are transmitting and then receiving the same sound. The beam forms are the same in both cases so when the already formed transmitted sound comes back it is formed once more. This results in further thinning of main beam and further suppression of side lobes. Nice things these piezos! Manufacturers of transducers, particularly fishfinder manufacturers, do not define their specifications very well. It's important for an array builder to know if they mean sound pressure or energy and if it's the one-way or two-way performance. I still don't know this for the transducers in my array but it worked well anyway.

Rickard

[Rickard]

Fishbows and ghosts

Although the arrangement with irregular spacing improves the performance of a fishfinder array the images are far from perfect. In my images one can see several examples of ghost images and a phenomenon that looks like the fishbows you see when the instrument is used for normal fishfinding. In images with three parallell structures two of the structures are almost certainly copies of the real structure in the middle. If you look at the photos in my earlier posts you can find several examples. The "wreck" I found when testing methods for photographing has three parallell lines. It probably should be only one line. The photos of the "mysterious structure" from the mouth of the marina show typical ghost images. These are probably due to strong echoes from hard materials as concrete and steel. On several photos the objects of interest have an overlay of bows. These bows appear when details give echoes so strong that several of the weaker sidelobes still are strong enough to affect the image.

The ways out of these problems have one thing in common, reducing energy so much that only the mainlobe can produce something on the display. One can use lower sensitivity, lower transmit intensity, reduced number of sound wave cycles per ping, reduced ping rate, set a suboptimal frequency and scan at larger distance. These remedies are not particularly effective and there will be a significant loss of information in the images.

The array is not worthless because of these problems since you can detect things very well. But you can not expect true images of the reality on the sea floor. There will be a risk for illusions since the ghost images can form structures. You have to scan and rescan and think and think again not to fool yourself.

In my simulations the side lobe suppression is -4dB when optimizing unequal spaces. The suppression of an array of equal length but evenly spaced transducers is about -1dB. These results are confirmed in simulations with a professional program at the University of Oslo in Norway . Their simulations test the "Thinned array theory" presented by René Pascal. Unfortunately, René's simulations give a much too optimistic view of the potential of unequal spacing. I knew this from the start but I will go on testing the limits of this concept anyway and I'm very grateful for the inspiration René's idea gave me. This has been, and still is, among the funniest projects in my life.

Rickard

[Glennk]

Any new info on your system Rickard?

I would like to come over to Sweden soon, but I have to stay still for a few weeks after my back surgery last monday..

I was thinking of packing my sonar into my car and take a trip to Mälaren and try the equipment on some of your objects. Then we also could try to interface your fish onto my display unit so that we have the record option.
Maybe we also could try my towfish onto your display unit, just to see what differences it would make on the images compared to yours.

Glenn

[Rickard]

The cork worked

Any new info on your system Rickard?

I would like to come over to Sweden soon, but I have to stay still for a few weeks after my back surgery last monday..

I was thinking of packing my sonar into my car and take a trip to Mälaren and try the equipment on some of your objects. Then we also could try to interface your fish onto my display unit so that we have the record option.
Maybe we also could try my towfish onto your display unit, just to see what differences it would make on the images compared to yours.

Glenn

I hope you will recover soon Glenn!

While waiting for a miracle which can improve sidelobe suppression to more than -4dB I have tested the idea to mask the sound window of one of the transducers with cork in order to reduce aperture and thus increasing vertical beamwidth. I glued two paralell strips of 3 mm cork, the kind of cork you can do your own gaskets from, 10 mm apart. This would yield a vertical beamwidth of about 35 to 40 degrees (oneway energy at -3dB). The vertical beamwidth of the unmasked transducer is 12 degrees. The result was better than expected. The beamwidth is much larger so now I can cover a much wider swath. Sound is more evenly distributed. No traces of noise from the cork. But intensity is lower as energy is less concentrated and more than half of the energy is probably absorbed by the cork. However, there seems to be more energy than needed and there are still strong echoes at 100 meters. Now I will put cork on all transducers to see if the effect holds for the whole array. I will look over the optimization of spacings too. Perhaps rectangular sound windows need other spacings? The instrument doesn't lock on the closest range so I have no problems with too high ping rate - yet.

Rickard