At the illutron art ship, we are using a lot of piezo elements to make contact microphones and hydrophones (underwater microphones)

Piezo elements can be a noise nightmare. And if you are using the piezos for hydrophones, to capture underwater sound, the noise can be even more challenging, unless you get everything very well sealed.

We came up with a trick, to get a balanced signal from piezos:

Simply using two piezo elements sandwiched, back to back, you can generate a balanced signal, that will be completely noise free.

Step 1: How to Make It

You need

  • 2 normal piezo elements
  • Microphone cable
  • XLR plug


  • Sandwich the two piezos "belly-to-belly" (crystal and wires in, metal out)
    • We just use a wad of normal silicone sealant in-between the piezos (Think Oreos)
  • Join the two (black) wires from the capsules and connect to Gnd on your xlr / microphone-cable
  • Connect the centre wire (red) from piezo A to + (hot) and the red from piezo B to - (cold) on your xlr / microphone-cable
  • Connect a XLR-plug to the other end of your cable

Step 2: Try Your Noise Free Contact Microphones

Plug the xlr-connector into the balanced input of a mixer or recorder, and enjoy the crisp clear noise free sound from your piezos.

This is a recording we did with two balanced piezos mounted on pieces of burning firewood.

Headphones are recommeneded - Enjoy! :-)

Experience so far

  • The sound from the piezos are very clear and noise free
  • The oscilloscope have verified, that the output truely is balanced
  • The piezo microphone is very sensitive and responsive.

As hydrophones

I tried this piezo-design out as hydrophones. I could just stick them in the water and get sound. I didn't even care to waterproof them, and they worked perfectly and noiselessly under water too.

So, for quick and dirty underwater microphones, just sandwich two piezos like this and dump them in the pond.

I did the soldering, then didn't have silicone so stuck the piezos together with blue tack and it works, finaly a buzz free sound, thanks a million!<br>I was wondering, it is important to make sure that the metal discs don't touch each other, isn't it?
<p>Yes it is: As shown above, the gold/brass discs are electrically connected to each other as &quot;ground&quot;. But when physically mounted, the &quot;white&quot; discs are apparently next to each other. If these touch they will short out the signal.</p>
<p>Nice bit of lateral thinking: The output from a piezo is very high impedance and so very easy to interfere with. The normal way to solve this is to put a preamp right there with the piezo, to give some &quot;robustness&quot; to the signal. This solves the problem in a different way. </p><p>It's like a humbucker pickup technique, but not using magnetics!</p>
This is more than lateral thinking, it's lateraly balanced thinking! :)
<p>Hi there! I just came across this instructable and think this is exactly what I'm looking for. I'm amplifying the sound of a very large and heavy rock wobbling on a stone tile. The sound this produces without any amplification is already quite loud. </p><p>At the moment I'm trying to figure out what kind of preamp I'll need. Can someone tell me how to choose the right preamp for the kind of sound I'm amplifying?</p><p>Thanks in advance!</p><p>Regards,</p><p>Dave</p>
<p>What did you use for the housing (the black ring seal around the edge)?</p>
<p>One great tip I learned from &lt;a href=&quot;http://www.nicolascollins.com/handmade.htm&quot;&gt;Nic Collins&lt;/a&gt; is to dunk the piezo and lead wires in some Plasti-Dip. Two coats is what he recommends. It makes the whole apparatus water proof, protects those delicate wires that are soldered to the piezo, and has surprisingly little effect on the response of the microphone.</p>
This is just the way the piezos we use came from the manufacturer. It's a simple plastic ring. I have used the method described here with the piezos without the plastic ring, and it works equally well.
<p>What would happen with the performance if you were to mount the piezos back to back (metal side in)? I think it would greatly simplify wiring because both metal backs are touching, but I don't know if that would make the piezos not perform well.</p>
Well, we just have to try it and see, I guess :-) <br>But I don't think it will have any influence on the sound. I have placed the two piezos side by side in some setups, and it was just as good as back-to-back.<br><br>- My reason for putting the metal side out is from a mechanical aspect. The crystal exposed on one side are quite fragile and sensitive. <br><br>I just felt that it would give a better mechanical protection having the metal side out. <br><br>I guess if you applied some other form of mechanical protection to the crystal side, having the crystal facing out would be just fine. :-)
<p>Hi. I'm trying to record heartbeat sounds using the piezo element. And I want to record using my computer. Can I use a XLR to 3.5mm converter? Also what kind of preamp would be necessary when I want to connect to a computer.</p><p>Thank you :)</p><p>Nikhil</p>
<p>Hi Nikhil - If you want to use the piezo to record heartbeats, you should get a good preamp before sticking it into the computer. <br>If you want to gain anything from the balanced signal, you should choose a mixer with balanced inputs (like xlr) <br>Personally I use a small Soundcraft mixer for my recordings. Finding a mixer with low noise on the inputs can be a challenge. Some of the small Behringer mixers are really a nigthmare of noise. <br><br>I would try to find a used mixer of good quality on ebay, if your budget is restricted. </p><p>Good luck with your recordings :-)</p>
<p>Heartbeats? I'd get a stethoscope (it's cheap) and a mic capsule (condenser one, probably, so that means add a 3v plug in power) that fits into the tube.<br>https://rockoys.wordpress.com/2013/09/09/electronic-stethoscope-2/</p>
<p>Hello <a href="https://www.instructables.com/member/chrlilje/" rel="nofollow">chrlilje</a>,</p><p>It's really a simple and great idea !</p><p>Don't you think the sealant is a bit too soft and might absorb a significant amount of the vibrations ? Wouldn't a stiffer material like epoxy glue be better ?</p>
Thanks!<br>I can very well imagine, that &quot;something harder&quot; between the two elements would be better. Would love to try it out. :-) <br><br>I just always end up using some sort of silicon based sealant, because that means &quot;Done in 5 minuter&quot; - where epoxy-stuff means &quot;Done in 20 hours&quot; <br><br>So far I think the biggest issue at all with these comes from the piezo elements themselves. The ones we use are from a random China-batch, and we don't have any specs on harmonic response etc.
<p>Hello...! Great idea.... Since you are in to Arduino, I have to ask you for help using your Piezo elements on our ROV ,at 380 meters waterdepth. Do you also have an idea of how to transfer signals from the piezo elements, via rs232/485 ,and to a computer or amplifier? I would like to be able to listen to the subsea equipment ,and hear how noisy our Magnum ROV is underwater :) </p><p> Hilsen Kjetil </p>
Hi Kjetil - Looks like a magnific piece of ROV you got there :-) <br><br>I actually would not be surprised, if the signal from the piezos would be able to travel the full 400 meter of cable, if you mount them in this balanced configuration. <br>I don't know how many leads you have available in your umbilica, but if you have a spare twisted pair, I would just use those and send the audio through those. <br>If you need to boost the signal, then I would suggest you getting a line driver, maybe even an active line driver, powered. <br>But, one of my friends have used piezos at I think 800 meters of cable, so give it a try the simple way first. <br>Good luck with it, and do share your experiences! :-)
<br> Hello..! <br><br> THe problem is, that this system runs the coms on optic fibre. So you need some kind of analog-digital-analog converter. Many years ago, we could have run this on twisted pair, but 800 meter of cable would still be a short cable..! The ROV is rated at 2500 meters, and usually have at least 1200 meters umbilical. ... For our system, we also have about 600 meters of tether... ;)
<p>Amazing Idea!</p><p>I'm really going to try that out...</p>
<p>wow - this looks just like what i need for a project i am working on. It is great when people share stuff like this....</p><p>question - could i use double-sided sticky tape instead, to hold them together?</p><p>does it matter how long the cable is?</p><p>cheers</p><p>RT</p>
<p>Thanks - I hope you find it useful :-)<br>Double-sided sticky tape would work just fine. We just used silicone adhesive, because we wanted them to be waterproof. <br>Cable length : I have used at least 20-30 meter of microphone cable on the piezos without any problems. <br><br>Do share your project when it moves on :-) </p>
<p>This is brilliant! What sort of noise were you getting before - hiss/static, 60Hz hum, or excessive handling/contact noise? I am going to try this on acoustic guitar, where all of those noise sources can be a problem.</p>
<p>We got a variety of noise. Biggest issue was the 50Hz (Europe) hum, that very easily bled into the piezo, if the piezos was not completely electrically isolated. <br>We use the piezos to pick up very faint signals, so it is important for us to be able to crank up the gain of the preamp. <br>With the balanced setup, we can turn the gain right up into the sky, if the preamp is of good quality. </p><p>Basicly the setup improves the signal-noise ration immensely.</p><p>Handling noise will still be there - That is more a mechanical challenge - How and where to mount them. </p><p>Do share your experiences with this setup on a guitar - would love to hear how that will work. <br></p>

About This Instructable




Bio: I am creating interactive installations and performances. Inviting the spectator to participate in various ways I use any media possible to bring "the idea" to ... More »
More by chrlilje:Underwater lights - insulated in old jam jars Dmx from the Electric Imp Retrofitting a Nemo 8C dive light with a LED spot 
Add instructable to: