Transducer design.

[Sture Hultqvist]

I have disassembled some commercial transducer


from well known companys.


The crystals was glued with silver epoxy


on a steel bar. On the sides and back it


was acoustic isolated with a hard foam


probably of polyurethane. The rest was


made of a poluurethane, probably


Conathane EN-9. The advantage of using


EN-9 is that have almost the same proprties


as water.


Another transducer from another well known


company that sells cheaper sidescans did have


a plastic U-shaped "chassi". In this they put


cork sheat on side, bottom and between the


crystal. This was then covered with epoxy.


Both transducer have the same principal:


Isolate bottom and side with something


ultrasound absorbing and then let the side


against water be transparent for sound.


Both unfilled epoxy and hard polyurethane


is transparent and cork and corkoprene is


absorbing.


Sture

[ForumEmail]

Re: Transducer design.

Sture,


This is very helpful! I had a few follow on questions..


Did you notice the dimensions of the steel bar? Do you think this is important? I was thinking that the waves might resonate in the steel. Perhaps, by using the right dimensions, the transducer could become more efficient.


Also, how important is the alignment of the elements. I would imagine with a beamwidth of less than 1 degree, this would be important. I could see why a steel bar might be used to keep things lined up. How do you deal with element alignment in your current design?


Thanks!


-Bob


By the way, if anyone is interested in Conathane EM-9, go here:


http://www.conap.com/cona_156.htm


There is also a list of distributors on their website.

[Sture Hultqvist]

Re: Transducer design.

The steel bar:


By some rough calculation I found the steel bar


thicknees to be 1/4 wawelenght, probably to awoid


"standing wawes " in it. The steel bar was also


sliced between each crystal. This is probably


to avoid longitudinel resonances.


The gaps was filled with cork.


The aligment:


Since directivity becomes from inteferences between


different part of the transducer I belive that


the tolerances should be calculated in degrees or


rather COS(degrees). 10 degrees will then give


a angle error of around 10% and for 500KHz this


is 0.1 mm. A normal workshop easily do better than


that.


Sture

[bob]

Re: Transducer design.

Sure,


1/4 wavelength at 500 KHz would be .75mm. That seems like a pretty thin bar! What size were the cuts?


Perhaps this was a lower frequency transducer..


Also, if the bar needs to be rigid enough to stay within .1mm of perfectly straight, the steel would have to be very strong.. Perhaps the absorber behind it also helps to support the array. Do you know of an absorbent material that is also mechanically strong?


-Bob

[Sture Hultqvis]

Re: Transducer design.

Bob


In steel the speed of sound is 6000 m/s.


For 500K you have a 1/4 wawelenght of 3 mm.


Sture