Cheap, innocent looking piezo elements do okay as greetings cards sounders. They seem to cause grief when used as contact mics. They seem to promise a lot, but easily sound rough as commonly applied. The problem with piezo guitar pickups and contact mics is that they are not well matched to typical audio inputs. By their nature they can generate a lot of signal, but they cannot drive a 50 kilohm typical line input. The pickup needs to work into a much higher impedance, typically 1 megohm or so.
Two solutions are proposed – the cheap and cheerful FET buffer which probably meets all the requirements of 90% of users who are using the devices on outside structures, and the high performance version for those who want higher performance and lower noise. First of all though, why do these things sound rotten the way they are most commonly used?
The sound to voltage conversion isn’t noted for its high quality – most piezo contact mics are tuned speaker elements used in reverse. The brass disc on which the element is glued is designed to resonate at the design frequency of 2-4kHz so that a large audio output is achieved with a small power input. This will tend to lead to a peakiness at mid-frequencies, which is why I glue these to a magnet to try and spread this peak and make the coupling to the source better.
However, the main reason these have gotten a bad rap is that many people couple them into a standard audio load, which loses low frequencies.
The reason why these devices often sound tinny is because the piezo sensor presents its signal through a series capacitance which is small, typically 15nF or less. When wired to a normal 50 kilohm line input this forms a 200Hz high-pass filter, which eliminates the bass.
If wired to a consumer plug-in-power microphone input of about 7 kilohms impedance, the result is a 1kHz high-pass filter. Hence the reputation for poor bass performance…..”
A piezo element can be thought of as a sound-dependent voltage source in series with a large capacitance of about 15nF. Here the trouble starts. You need to put this into a load which is higher than the impedance of the series capacitor at the lowest frequency of interest. If that is 20Hz, since the impedance of the capacitor is 1/2pifreq*C then you want that to be > 530k. I’ve terminated mine in 3.9M because that’s what I had to hand.
So what to people do? They go and stick this into the line input of their recorder, typical impedance 50k, or the plug-in-power mic input of their recorder, typical impedance about 7k, and they start to grouch that this damn thing sounds tinny. Which is does – from the frequency response you can see that the mic input starts losing low frequencies below 1kHz (red line) and the line starts to go below 200Hz (green line).
We don’t really like that very much. This tends to be compounded with the problem that these things aren’t really designed as mics and resonate at 2-4kHz anyway unless you stick them down so you get a harsh ugly result. What you need is a high impedance input, which gets you the blue line in the plot above.
You could try and equalise in post production. The trouble with that approach is you need the specs of the mic and amplifier input to know where to start your 6dB/octave LF boost, so you will probably have to tune that by ear if you don’t know these. Plus you will boost all the noise and hum at low frequencies. With the mic input you are looking at putting in a 25dB boost by the time you get to 50Hz which is just asking for hum problems, and even the 8dB boost on the line input version is more than doubling any hum you may have.
These things are very much like the crystal mics people used to use with valve open reel recorders in the 1950s/60s. Valves do high impedance really well, and inputs were often in the area of 1 megohm input impedance. The move to transistor circuitry meant consumer audio inputs tended to end up around 50k which is reasonably easy to do. Crystal mics also used the piezoelectric principle, and were valued for their low cost and high output voltage which overcame the preamplifier noise better. Transistorising audio gear killed off the crystal mic as they sounded tinny and harsh with the lower input impedances for the same reason.
Hello, this article (blog) is very descriptive and easy to understand for almost everybody, thank you very much for providing all these details. My question is; is there a way of understanding with shortcuts electric current to grasp “piezo” “contact” or “pick-up” mics? Actually to get a deeper understanding of your presentation and conclusions?
I plan to record wild life tiny micro sounds and piezo mics are the conclusion to my filtering of most of all the microphones. This is the why my interest in electricity emanates suddenly. (Bachelor in Audio Engineering)
Cheers
> “piezo” “contact” or “pick-up” mics?
Piezo elements are most often used in contact mics because they match well to solids, I guess because they are solids so the mechanical impedance match is good. Instrument pick-ups can be piezo at the lower quality end, electric guitar pickups tend to be magnetic at the better quality end, these won’t work for your application.
Take a look at the work of David Dunn, I believe he uses piezo elements glued to a metal nail but it’s a guess, he’s cagey about how he does it.
I don’t know of a source of ready made high impedance preamps – the easiest one to construct would be the FET version. I’d imagine someone must make one. Alternatively get in touch with the electronics engineering department at a university or a maker community.
Hello,
thank you for your article, I did learn a few thing in here.
Would it be possible to couple various piezo elements together (in parallel) before entering the buffer, or would it add complications? it’s the 15nF intern capacitance that makes me think that it may cause some issues.
thanks again,
Paul
I’ve never personally tried it, but using these in parallel should work. You will lose some signal from any one device, because, assuming all elements are similar, you are effectively shunting the 15nF capacitance of the active device with a similar capacitance. This makes a capacitive divider, so you will get 6dB (voltage) loss. If both devices receive the same stimulus in phase then the extra output from the second device will cancel the 6dB loss. The LF roll-off is ever so slightly improved with the extra shunt capacitance
If the devices are different then the device with lower source capacitance will suffer more loss than the other.
Hello, I’ve built your NE5532 preamp and I love it. Works beautifully!
I would like to experiment using multiple (identical) piezos on percussion instruments. Example: piano, tuned metal drums or ‘hang’, cajon, xylophone (this is the ambitious one). I’d want to place piezoelectric elements in multiple spots to capture as much sound as possible, In the case of xylophone I would need to place a piezo on every wooden block since they are mechanically isolated from each other. I would like your input before I start so I know what to expect.
My question is this: Building individual preamps for every piezo element would be tedious, and in the case of my xylophone idea, would need over 30-40 preamps as well as a recording interface with that many inputs – which are not cheap to say the least. How safe would it be to connect multiple piezo discs in parallel to a single preamplifier? Can I safely add the signals together by putting a resistor in series with each piezo disc? Will the resistance value change depending if I have many or few piezos?
I should qualify this by noting I am not a musician, more a field recordist, so I don’t have personal experience to this approach. However, you can wire multiple piezos to one input, in theory you lose 6dB of signal each time you double the number of piezo elements if they are all the same. In practice you will lose a bit more due to the extra shunt capacitance, you have to use shielded cable otherwise the hum pickup will drive you bonkers. And you will tend to pick up more hum as you extend the system. Don’t add resistors in series with the inputs. It’s a high impedance input, all you will do is add noise at higher frequencies.
Now if you want to use multiple preamps you could sum the outputs by putting something like 4k7 in series with each output, joining the other end all together and taking your output from the common point. It’s not your canonical virtual earth summing point but it will do in this application. You get some control of relative signal levels by varying the output resistor. You could do a mixture of the two – you usually have a lot more signal at lower frequencies so you could common elements on the input and use more amps at higher frequencies.
You can’t harm the elements by paralleling them. The main way I’ve destroyed piezos is trying to solder to the silvering, and they are a little bit brittle, but other than that they seem to take all the abuse you can give them.
Thanks for the post!
Right up front you mention why so many guitar piezo mics sound bad… As a guitarist, that’s what brought me here. I’m curious, would it be economical to make guitar pickups/preamps using the methods you describe? Is there any chance you can get a friend with a nice guitar to show the difference in tone qualities between a standard guitar contact mic one of yours ? Or would that require a custom set of components that you’re not about to put the money/time into?
I am a field recordist, and came to this from the piezo contact mics angle. I am looking to try and get the sound of the things I am attaching it to, and often these will be resonant structures with low frequencies because they are large, or long i nthe case of wire.
By no means all guitar pickups are piezo elements. Many electric guitar pickups are magnetic. You will not have this problem with those. Having said that I believe they are sensitive to excessively low impedance but for different reasons.
A custom guitar amplifier (or DI box) will be designed to have a higher input impedance – tube amps for instance have naturally higher impedances, so if you are using a Marshall amp this will be sorted for you.
Field recordists are often using piezo contact mics with small recorders designed for plug-in-power inputs, and there is a fundamental electrical mismatch. If you’re a musician in a studio you have much more control over your system, and the DI box is the classic tool for this job. Unless it gets the input impedance over 1M a DI box is a teeny bit low for your average piezo contact mic, but presumably OK for guitars, your lowest E string is in the region of 80Hz.
These guys talk about impedance from a guitar perspective as does Sound on sound
So is it possible to obtain piezo elements which aren’t meant for greeting cards? (ie optimized for producing sound of a few kHz?). I want to make a contact listening device for mechanical troubleshooting and would prefer a more distributed frequency response.
What you’re probably looking for is a piezo accelerometer – this white paper shows you can get better frequency response, and indeed vibration sensing and checking bearings is one use case for these devices. They often have internal signal conditioning.
That paper also talks about frequency response, if you can operate the system below 1/3 of the resonant peak you seem to get satisfactory results. You do get ultrasonic piezo devices – a common application is water fog machines and u/s cleaners.
Other things you could look at is piezo film – note that for contact mic apps piezo film is a lot less sensitive than the ceramic sort. But it has a flatter frequency response because it is more internally damped.
Strain gauges take you above 100kHz though aren’t piezo, and don’t have the fabulous sensitivity
Awesome, thanks. The cost of the film sensors is actually very reasonable.
Great reading , thanks
Am I right to assume that the inductance value is negligible for that type of mic ?
electrically, yes, you will get minimal inductance in the wiring, but the Q of the series circuit will be hammered by the several megs resistance.
I guess there’s an equivalent circuit for the motional impedance of the piezo device will be analogous to that of a quartz crystal and have inductive and capacitive elements that will show when driven. These also set the limitations on frequency response, your piezo driver still peaks in the several kHz range
Hi Richard,
Great write-up, I am enjoying learning about how to use piezos! I am hoping to modify this tutorial by reading the signal with a fast ADC (~50 kHz for audio-esque signals). For this reason, I was looking closely at the output impedance of the piezo and the input impedance of my ADC. Thanks again for taking the time to explain all of this.
To estimate a piezo’s impedance, you suggest imagining it as a voltage source in series with a ~15 nF capacitor. For 20 Hz, that puts the impedance (1/(2pifC)) at >580kOhms. I believe there may have been a small typo…if you calculate for 25 nF you get >320 kOhms. This may have been straight-forward to others, it made me stop so I thought I would mention it. Regardless, it’s a very small change and most sources I have recommend at least 1 MOhm anyways.
A quick follow-up question: How does this relate to the impedance and capacitance reported for different piezo buzzers? For example, a 10nF @ 1kHz with a 1kOhm impedance. My guess would be that we just add them up as impedance sources in series. I believe that reported impedance on piezos is going to negligible compared to the capacitance (commonly around 15nF as you reported!)
> For 20 Hz, that puts the impedance (1/(2pifC)) at >580kOhms. I believe there may have been a small typo…
I can’t argue with that, my bad! Thanks for picking it up!
Piezos in air have a reactance that is primarily capacitive up to the resonance frequency. At resonance the impedance goes to a peak and then drops – they are modelled as a motional inductance and series capacitance, with a lowish capacitance at the output. For operation below resonance (which is where you want to run a piezo mic) the motional series capacitance in series with a voltage source is good enough.
Buzzers are specified usually at resonance, because you get most output from them for a given drive level. You can’t necessarily translate that to use as sensors, because you won’t want to be using them up that high in frequency.
Hi Richard!
I’m planning to buy a Meinl SEMPU magnetic pickup. I dont really understand if this is a piezo with a magnet (hence it stays on the instrument), or a magnetic pickup. I will use it with my Morfbeats Marvin (Basicly a large cowbell instrument with springs). I would like to get the best bass response possible because its a key area of this instrument.
https://meinlsonicenergy.com/en/products/sepu-m17857.html
Do you think I’ll have the same problem with this device that you mention in your post?
On the interfaces now you can have Hi-Z (basically instrument level, for guitars) inputs, would that solve this issue? Another idea of mine was to use a buffer guitar pedal.
What do you recommend to get the best possible bass reponse from this pickup?
Use the instrument input or a DI box. That’s what it’s there for. You’ll be fine. This is for people using field recorders.
Hi Richard,
A very brief question on a passive piezo-disk ukulele pickup…….
All my guitars are equipped with internal condenser and dynamic microphones and amplified through the 2-channel 12kOhm low-impedance input preamplifier (mic preamp), 1st channel does also powers the ECM mic by 4.5DCV through the cable…..
I do also have ukulele equipped with ECM mic and passive piezo-disks, ukulele does not work with my basic preamp, it has a battery box built-in….. I do plug ukulele directly in to the combo.
Well, I just want to have a single universal preamp system for guitars and ukulele, so I would like to get the piezo element’s impedance down passively and then plugging it in to the mics inputs.
So the question: have you ever tried converting piezo’s high-impedance into low-impedance passively with a single 1+ mOhm resistor in-series ?? in other words, wiring some 1-10 mOhm resistor in-series within the signal-chain ??
Thank you!
Yep. That experiment is run here with audio samples.
Long story short – it can work. You have to suck it and see. You take more loss the higher the resistance is, which ends up as hiss. So there’s a tradeoff to be made. something like 220k to 470k, depending on how far down you want to preserve, less resistance = more signal, less hiss, but less LF response. OTOH you don’t need to go down to 20Hz for a ukelele. To a first approximation you need ~ 600k to get down to 20Hz, so 300k will start rolling off at 40Hz, 150k at 80Hz, so you could probably get away with a 68k resistor which would roll off around 160Hz which is probably OK for a ukelele. Not sure what the frequency of lowest string is. YMMV because your piezos may have a lower capacitance etc – suck it and see. If it’s tinny with one 68k resistor, use two in series. You won’t get the optimal noise performance, but piezos chuck out a lot of signal, so it may work for you, and is probably fine in a larger mix if you gate it.
Thanks, Richard!
Sounds good! Seems that I do not need any mOhm resistors, some 56-68kOhm should work OK……..
My preamp has a decent specs of THD<0.002% and SNR 28.384µV, it offers x100 of Gain. My condenser/dynamic mics have 2/0.2mV sensitivity, so I would wish the piezo-discs to go down to that 0.2mV output with hum of <O.
Do you think this might be possible – cutting piezo output to 0.2-0.5mV ??
Amazing article.
I just want to ask, what is the typical output level of a piezo contact mic?
Into a high impedance – lots. I measured 60V tapping the surface with the end of a screwdriver.
All depends on what you have them on and what you are loading the with. You will get less on one stuck to a guitar, probably closer to 1V into a high impedance, and a lot less if you load it with a typical 7k mic input.
Hey Richard!
Thanks for this article, despite being quite old it’s still one of the most condensed and concise sources of information I’ve found.
My question is how buzzer elements calibrated for different frequencies fit into what you’ve talked about here? For example Digikey supplies buzzer elements ranging from 18kHz to 520Hz, my intuition is that elements tuned for lower frequencies may be better suited for picking up lower frequencies.
That is unless only the brass plate is different between calibrations. I’m a bit new to eletronics so forgive me if I’m incorrect, but isn’t this problem avoided by using an instrument input rather than a microphone input when available? My cheap scarlet solo’s instrument input sports an impedance of 1.5 Mohms, which if I understand the article correctly would eliminate any perceivable low-passing.
The project I just finished is a Kalimba that makes use of two buzzers wired directly to a 6.35mm port in parallel, the buzzers are secured using epoxy with hotglue encasing the outside (which I believe replicates the magnet in your build?).
This has worked great for me, but now I’m a bit concerned that I’ve been designing my project for my unique use case, and that it may be non-functional in a normal users audio setup. I don’t have any experience with a larger audio setups and I’m not sure if a user is more likely to want to wire an electro acoustic instrument into a microphone input or an instrument input.
> my intuition is that elements tuned for lower frequencies may be better suited for picking up lower frequencies.
From an engineering POV the high frequency response falls off above the free air resonant frequency. Whether that matters depends on your instrument and the sound you are after – you may lose harmonics if you use a piezo that is too large, they are cheap enough that experimentation is the order of the day. The resonant frequency cited by the manufacturer is usually for the element in air – in application it’s usually held in a plastic casing at three narrow plastic points, sometimes with the plastic casing resonating at the same general frequency.
Gluing the element, or pressure coupling it to a larger solid object will lower the frequency you can pick up, the self-resonance in air will limit the highest frequency. Test experimentally. I don’t have the materials design chops to simulate that, but I have experienced it enough. Glue the magnet to the back and stick it to a farm gate and you will pick up much lower than the cited resonant frequency 😉
You’re right, musicians often have a guitar input, or failing that, a DI box. A guitar amp will offer an impedance of 1M or more, any studio expecting that sort of instrument will have some sort of solution like that. The cheaper end of instrument pickups use piezo elements, and though I’m not a musician I think even passive mag pickups want a high impedance in. Use what you already have.
> I’m not sure if a user is more likely to want to wire an electro acoustic instrument into a microphone input or an instrument input.
Most music facilities should offer an instrument input. If not, you can use a DI box, which are designed to patch instruments into the stage snake going to the mixer.
Any instrument featuring a bridge may be better off using an instrument piezo designed to go under the bridge, you get the mass loading by the physical pressure of the bridge against the body of the instrument, applied by the strings. I don’t see why you can’t use piezo disks there, though the unsupported part may rattle, but something designed for the job may be better.
I wrote this as a field recordist wanting to use contact mics on big resonant objects or wires in wind, so guitar amps and DI boxes are out for me 😉