Waterproof Microphones?

[CaptnYellowShirt]

I'm involved with a project that needs to make some fairly low range pressure measurements in a water tank.

The area if interest is very small (8mm ?). I've tried using a pressure tap to bring the pressure out though a small tube and to a large (proper) pressure sensor, but there are a lot of issues with getting the tap to work right -- dynamic response, clear of air bubbles... etc.

As a replacement sensor, I've thought a small microphone like this would work:



(4.5mm dia).

However I am having trouble finding any that are compatible with water. And by compatible, I mean will give me a reliable week of operation -- for $3/ea and in the situation I am working in,  I can just replace them every week if need be.

I've tested most of the 4mm and 4.5mm dia microphones that are sold on Digi-Key. Attached is a copy of the excel chart of their nominal impedances measured with time after they were submerged in distilled water. You can see where they failed when the impeadances suddenly dropped out. I didn't bother to include the microphones that gave less than one day of performance.

I'm hopeful that 2 or 3 of them may meet the 1 week performance mark in a larger statistical sampling.  However, I am wondering if anyone here has experience with this kind of thing? Any suggestions on hacking a microphone (or whatever) to work underwater for a week or more?


PS: The one 4mm dia microphone that are sold on digi-key as "waterproof" (http://www.digikey.com/product-detail/en/CME-1538-100LB/102-2190-ND/2364631) is anything but. It fails within the first 10mins of being submerged -- its redeaming value is it tends to 'dry out' and function after a few hours of being pulled out.

[Yago]

I think the transducer you want might be a hydrophone.
Had a look and a couple are based on similar condenser you pictured, probably a matter of sealing.
It also depends on what you wish to measure, is it static pressure or dynamic, what range and frequency etc.

What is the tank like?
Might be worth a try bonding a little condenser mic to the outside of the tank.

[CaptnYellowShirt]

I think the transducer you want might be a hydrophone.
Had a look and a couple are based on similar condenser you pictured, probably a matter of sealing.
It also depends on what you wish to measure, is it static pressure or dynamic, what range and frequency etc.

What is the tank like?
Might be worth a try bonding a little condenser mic to the outside of the tank.

Can you link in a ref to those hydrophones? I wasn't able to find any on the order of 4mm dia.

I need both static head and dynamic. 0-100Pa / DC-100Hz.    But higher would always be nice.

Modification / calibration is not an issue. Just a roughly linear response is all that is required so I can calibrate them in situ.

Outside won't work. It needs to be in contact with the water.

[Fraser]

I just posted but it appaears to have got lost in Cyberspace.

Try sealing the microphone in a condom. No joke. Give it a go.

Note that if you are using electret Mic elements, they use electrostatic charge across plates and so are not compatible with direct contact with water ! A dynamic (speech coil around magnet) based mic may work in distilled water without sealing. You just need to be sure that both sides of the diaphragm can flood to equalise pressure across it. Failure to equalise the pressure will just crush the diaphragm making its useless.

Another possibility is the use of a piezo electric element as a microphone (crystal microphone) A crystal earpiece may work but I am uncertain of the frequency response.

Unsealed microphones would need to be operating in an insulator. Have you considered using very thin oil instead of water as the liquid medium ?

[CaptnYellowShirt]

just a random idea

the mini electrets you have, have you tried to encase them in a enclosure, weighted so it sinks? for a fast experiment, why not try wrapping it up (maybe like microwave wrap?) and throwing it into the water, if through the wraps it works ... well you get the idea.

Like place a membrane across the opening of the microphone? I've thought about it, and its what some manufactures do to make a waterproof mic (like Knowles -- good mics, just too expensive). I haven't tested anything like that yet. I was hoping to find ones that didn't require that extra step. But it may be where this project is headed. Maybe a small bit of Kapton tape over the inlet? (suggested welcomed here!!!)

[Fraser]

The mic would need complete sealing, not just the acoustic aperture as the rear is not watertight. Also consider pressure differencial across the diaphragm and the effect on performance. I use Knowles mics and at GBP55 each they are expensive if killed during experiments !

[CaptnYellowShirt]

Note that if you are using electret Mic elements, they use electrostatic charge across plates and so are not compatable with direct contact water ! A dymamic (speech coil around magnet) based mic may work in distilled water without sealing.

It seems that ones I have tried have some ability to reject water from the electrostatic area. Perhaps its "sealed"?

It would seem from my testing that the seal is only good for ones to several days.

The voice coil sensing method is hard to find in such a small package... and it wouldn't give any static pressure indication.

[CaptnYellowShirt]

The mic would need complete sealing, not just the acoustic aperture as the rear is not watertight. Also consider pressure differencial across teh diaphragm and the effect on performance. I use Knowles mics and at GBP55 each they are expensive if killed during experiments !

Heh... yeah, plus you got my shaky hands soldering them.

[CaptnYellowShirt]

Here's another idea following up on the Kapton tape... I could back fill the port using a viscous silicone gel.

[Yago]

I didn't think the mics mentioned went down to DC.
No Volts generated without the coil or plate moving?

How about something like this?
http://www.omega.co.uk/pptst/PXCPC.html

[Fraser]

Is it possible to provide a little more detail of the test that is being carried out? Members of this forum may have different approaches to offer ?

[oldway]

Look for "pressure transducers" on Farnell....there are some for low pressure H²O.  (Honeywell and Amphenol)

[Fraser]

Another thought...... have you considered a probe tube with passive diaphragm driving it to trasfer pressure changes to an external sensing element ?

[CaptnYellowShirt]

The experiment deals with erosion control. My group is studying hydrodynamic forces incident on small features of models.

You got your lasers and your doppler systems that can give flow velocity. But the incident pressure is hard to determine. And that's where I'd like to instrument these models with pressure sensors.

Drilling pressure taps and attaching them to a bank of large (a few inches wide) pressure sensors is what is in place now. Basically the same way you'd think of looking at pressure on an airfoil in a wind tunnel. But here we have this nasty "water" stuff to deal with. And it is a real pain in the butt to flush these pressure taps between experiments. It can take hours... and one stray bubble can ruin a set up. Not to mention the un-calibratable change in frequency response the fairly long tube (5-10 inches) of water adds. And have you ever tried to drill around a curve? I could go on and on...

So enter the local sensor. These microphones are just about the only thing I've found that 1) measures pressure and 2) is small enough.

I've looked at these DIP-8 sensors that honeywell, et al make (barometers, HVAC sensors, etc). But they are just too big. I've even dissected a few and tried to rip out the chip that's housed inside... fun times were had by none.

[CaptnYellowShirt]

Another thought...... have you considered a probe tube with passive diaphragm driving it to trasfer pressure changes to an external sensing element ?

If its a water / water diaphragm we're basically in the same boat we were before. If its a water / air diaphragm the calibration of the change in material impedance is an issue.

I can calibrate the sensors in another machine, but I cant calibrate a long loop. 

[Fraser]

Thank you for the full and interesting description. It helped me understand your situation. Sorry I have not got experience in this area. I only operate in air environments and not fluid.

[oldway]

http://www.ge-mcs.com/download/pressure-mems/920-303B_LR.pdf

[CaptnYellowShirt]

http://www.ge-mcs.com/download/pressure-mems/920-303B_LR.pdf

Totally.... but its too big. The microphones I'm using could almost fit completely inside of just the pressure on that bad boy.

[oldway]

Are you sure ?
Diameter of the tube of this sensor is 0.127 inch or 3.23 mm.
You have show pictures of microphone of 4.5 mm of diameter.

How deep will the pressure sensor stay?

[CaptnYellowShirt]

Let me back up for a moment.  The over-all size is important too. Not just the frontal area. The sensors have to fit 'inside' features that are on the order of 8mm. So this 15mm wide chip wont cut it, sadly. Thanks for the suggestion though.

[CaptnYellowShirt]

I didn't think the mics mentioned went down to DC.
No Volts generated without the coil or plate moving?

How about something like this?
http://www.omega.co.uk/pptst/PXCPC.html

Sorry, I missed this reply from before....

The sensor is just too big.... 9.9mm to fit into an 8mm space...

Electret microphones change their nominal impedance with respect to DC offset. So there's a way you can back-out the DC component if you interrogate it correctly.

[oldway]

http://sensing.honeywell.com/index.php/ci_id/142179/la_id/1/document/1/re_id/0

[NiHaoMike]

Piezo discs with the ceramic side painted after soldering might work.

[CaptnYellowShirt]

Piezo discs with the ceramic side painted after soldering might work.

I looked at PVD film. You can get close to DC with some clever guarding and DSP... but not DC. Unless I'm mistaken?

[qno]

You can try a MPXV5004DP.
If the environment temperature is reasonably constant you can see 100 Pa.
This is a $10 sensor.

There are better ones but far more expensive, $100 buy's you a Sensirion.


SDP1000-L025  -62 to 62 Pa  -/+ 0.25" H2O 0.5% FS / 1.5% m.v. linear
SDP1000-L05 0 to 125 Pa  0.5" H2O 0.2% FS / 1.5% m.v. linear

SDP1000-L  0 to 500 Pa + 2" H2O 0.1% FS / 1.0% m.v. linear

SDP1000-R 0 to 500 Pa + 2" H2O 0.2% FS / 1.0% m.v. root square

SDP2000-L 0 to 3500 Pa  +14" H2O 0.1% FS / 1.0% m.v. linear