Problem with ultrasonic transducer

[Mad ID]

Hi,
I'm developing a system for ultrasound distance measurement in the range od 30-200cm and have a very peculiar problem with the sensor. The sensor is 400EP250 from ProWave (around $9 in quantity >100pcs)
http://www.prowave.com.tw/english/products/ut/ep/40ep250.htm

I have a test fixture outside which is about 120cm high and is almost fully enclosed. Temperature inside goes up to 60C with air not being able to circulate (as in my application).

It happens sometimes that some of the sensors report wrong distances, like 30cm instead of 120cm. I always take 32 measurements in a row, sort them from lowest to highest, cutoff first and last 8 and do the average of the rest. So it's not a glitch.

One particular sensor was fine measuring around 120cm and then simply went crazy and started measuring 32cm all of the time. I brought it back to the lab and took some scope measurement of the analog signal.

This is how a good sensor looks like when measuring something far away, on the right you can see the return signal



And a bad sensor measuring the same:



It resonates all the time, like it's mechanically damaged. But few moments later for no reason it's OK again?!
Here are the photos of the sensor:
https://carousel.dropbox.com/photos/cc/5MbDeAr8buwDFWU

When taking measuremtn I was holding the sensor by hand on the connecting wires.

I'm using RTV Silicone to attach the sensor to the enclosure as you can see on this photos:
https://carousel.dropbox.com/photos/cc/xCR5dx6AgymgWm9

Is it possible that the acid in the silicone damages the sensor in combination with high temperature?
I used silicone because it's soft.

This happened on couple of sensors.

Looks like ProWave is a shitty company because you can barely get support from them. Any recommendations for a different brand for IP67 sensor @ 10€ pcs max which would provide support?

Thanks!

[ion]

I would avoid using silicone acetate near electronics.  I don't know about transducers, but I've had the acid eat through copper wires in just a few months at room temperature.

Neutral cure RTV silicone is a good alternative - still soft and doesn't release acetic acid when it cures.

[Mad ID]

I would avoid using silicone acetate near electronics.  I don't know about transducers, but I've had the acid eat through copper wires in just a few months at room temperature.

Neutral cure RTV silicone is a good alternative - still soft and doesn't release acetic acid when it cures.

Hi, I'm aware of that and use Sikasil-C which should be neutral. Here is the datasheet:
http://www.tulppa.com/files/8613/9771/6199/Sikasil_C.pdf

[Niklas]

What kind of driver circuit do you use? Is there any passive component that will dampen the ringing?

When the piezo element is fixed to a mechanical structure, its properties will be altered to match those of the complete structure. The dimensions and stiffness of the system can also change with temperature. Not only will the speed of sound in the medium change, the resonant frequency, amplitude and ringdown can also change.

From the pictures you posted it is also visible that the amount of silicone is uneven over the piezo's backside. You have two transition points (piezo-silicone, silicone-air) that can cause reflections and false echoes.

[Mad ID]

What kind of driver circuit do you use? Is there any passive component that will dampen the ringing?

When the piezo element is fixed to a mechanical structure, its properties will be altered to match those of the complete structure. The dimensions and stiffness of the system can also change with temperature. Not only will the speed of sound in the medium change, the resonant frequency, amplitude and ringdown can also change.

From the pictures you posted it is also visible that the amount of silicone is uneven over the piezo's backside. You have two transition points (piezo-silicone, silicone-air) that can cause reflections and false echoes.

We use an IC from the same company (ProWave, PW 0268) which drives the transducer with MOSFET over 1:10 transformer. I suspect the problem is related with the way we attach the sensor to the enclosure. There is 20k resistor in parallel with the sensor.

I'm aware that it is uneven but didn't think until know that it could be a problem. What would you propose? What is important is that we have IP66 at least, so it must be tight. That's way we used silicone.

Silicone is present on the sides and also at the back. I'm thinking of using some kind of soft silicone rubber around the sensor?

I would appreciate any help. Thanks.

[Niklas]

As the sensor module is equipped with a connector, why not swap a "good" and a "bad" sensor and check the result? I also suspect that the solution is somewhere in the assembly method. The sensors I work with are encapsulated and compressed together with backing material. In order to get repeatable results, a special grease is applied to the piezo element before it is assembled into the housing.

- Repeatable amount of silicone or any other type of glue
- Symmetry in the assembly
- Think about coefficient for thermal expansion