Modifying an ultrasonic sensor to work underwater?

Other than control. (Navigation, Sonar, Ect.)
derelicte
Posts: 292
Joined: Aug 1st, 2011, 3:08 pm

Re: Modifying an ultrasonic sensor to work underwater?

Post by derelicte »

well I can confirm that my sonar sensor will not work in a housing. as soon as I covered the sensor, it reported the minimum distance of 6". I tried a few different types of plastic (plastic bags, thin plastic sheet, pvc). all reported 6".

I may pick up one of those harbor freight fish finders if I find it on sale.
Seawolf
Posts: 33
Joined: Oct 15th, 2011, 10:22 pm

Re: Modifying an ultrasonic sensor to work underwater?

Post by Seawolf »

Since there was some interest, I'll give a quick update. I ended up getting one of these: http://www.ebay.com/itm/Waterproof-Ultr ... 0721354139 and putting it in a waterproof housing. Preliminary results in the swimming pool show that it can measure from 4 to 12 feet pretty accurately. In theory it should work out to about 35 feet, but I won't get around to testing that until the end of this week/early next week.

During tests in air, I noticed that the sensor wasn't getting good returns off of a person at 6 or more feet away. It did work well with large-ish flat objects, however.

This opens up possibilities for having cheap AUVs be able to navigate around a pool (my project), or having rovs that can detect (or possibly hold) their distance from the bottom (with a minimum range reading of 4 feet). It could also help you avoid running into large (flat?) objects in murky water.

Hope someone is able to use this information to make something cool.

-Seawolf
derelicte
Posts: 292
Joined: Aug 1st, 2011, 3:08 pm

Re: Modifying an ultrasonic sensor to work underwater?

Post by derelicte »

care to detail the housing you built?

thanks!
Seawolf
Posts: 33
Joined: Oct 15th, 2011, 10:22 pm

Re: Modifying an ultrasonic sensor to work underwater?

Post by Seawolf »

I just put a hole in a tupperware container and sealed the hole/lid with silicone. Not terribly showy, but it gets the job done well enough for testing. I'll try to remember to get a picture later this week.

In anyone is interested, the datasheet for the sensor is here: http://www.sure-electronics.net/downloa ... anual.pdf/
Remember that sound travels ~4.3 times as fast in water, so the ranges will need to be scaled appropriately. I believe the sensing angle will also change in water, but I haven't confirmed that yet.

If anyone is using an arduino, the demo code for using the Ping sensor can be easily modified to work with this sensor.

Cheers,
-Seawolf
derelicte
Posts: 292
Joined: Aug 1st, 2011, 3:08 pm

Re: Modifying an ultrasonic sensor to work underwater?

Post by derelicte »

I see. the transducer is watertight so you made a housing for the electronics. my transducer is open which is why I can't seal it up.

thanks!
Seawolf
Posts: 33
Joined: Oct 15th, 2011, 10:22 pm

Re: Modifying an ultrasonic sensor to work underwater?

Post by Seawolf »

I'm a little slow, but here's the picture of the waterproof ultrasonic sensor I made.
Image

Like I said previously, it measures out to about 12 feet accurately. I'm going to try adding some kind of cone around the transducer to try to filter out unwanted reflections from the surface, etc. I'll post an update if there's any improvement.

-Seawolf
perfo
Posts: 151
Joined: Jan 27th, 2015, 12:05 am

Re: Modifying an ultrasonic sensor to work underwater?

Post by perfo »

I'm interested in trying to do this as well for a least as possible (as per usual).

I'm also a little confused about some of the comments. Are we talking about a TX RX pair ?
I'm no expert but I would think you would get a reflection from each interface ie air water ...water plastic etc etc. The sensors would have to be mounted in a fluid similar in density to the medium you're in but no conductive so maybe distilled water for a fresh water set up. If you put both TX and RX in the same container you will get a reflection from the housing and this will give you the minimum distance as the gent said above. However if they are in separate housings then i'm guessing you may not get any reading due to the energy of the pulse being absorbed by the container. If the pulse makes it out of the container then it should be detectable coming back in and doing it this way should eliminate the multiple reflection from the various interfaces..

So in short my view (and very open to be totally corrected) minimise the interfaces, Minimise the energy absorbed by the container, then its just a matter of power and calibration. To calibrate it (and if they can be made cheap enough) have a RX TX either side of you ROv at the greatest distance possible and use this known length to calibrate the time of flight for the main range finder. The time of flight (I guess) will change with pollutants (Salt , silt etc) and even temperature /depth (thermoclines) so a real time calibration method may end up the only reliable way.

Has anyone manged to do this yet ? I'd like to be able to measure 10 meters or more so may need more power than the standard arduino things but I may give them a try when i get a chance. Maybe we may even need to change the code to select other reflections rather than the primary..
Post Reply