Create an Arduino-based full-fledged outdoor sound pressure level station.

My project is to create an Arduino-based full-fledged outdoor sound pressure level station.

Not only one piece just for me, but to create a modular system that everybody could use.

Sound pressure level measurement is not trivial, I have teared-off a couple of SPL meters and they all have a bench of analogue electronic upfront. The challenge is, that a metering range from 30dB to 120 dB requires a dynamic range of 1 to 10 millions.

So basically the project will start with creating an analogue measurement shield consisting of:

  • a low noise amplifier
  • a passive (A) law filter
  • a calibration amplifier
  • a precision logarithmic rectifier based on the chip AD8307

I have got a quite precise idea how to realise it, and I am ready to solder the first prototype on a breadboard. I will also start learning to use EAGLE to create the PCB files for a pre-series order.
I am completely unaware on -once the product type is working- how to find a company that would want to produce and sell it.

When we are looking closer to the microphone, it begins to become really tricky.
Of course the project will include the possibility to use a cheap analog sound level sensor for the lowest requirements. The microphones included are not especially sensitive, have a high intrinsic noise, and do not provide the required long-term stability for a meaningful measurement and you can completely forget to run them outdoor.

So the challenge is to find an affordable microphone that is low noise, sensitive enough, and has a reasonable long-term stability. And just to make it a bit more difficult that microphone will stay outdoor. Depending on the climate and the humidity, we will have to consider some kind of microphone heating as well.

The design of the windshield isn't trivial either, I have also gathered a couple of ideas, but again from the prototype to something that is freely purchasable, there is a long way and I would appreciate some advice.

In a first variant the sound level meter station will be wired, in a second variant I would like to consider a solar powered wireless variant as well, it would be good to have some person who have got some experience in building solar powered wireless solutions as well.

So if somebody gets excited by the project, please feel free to contact me lazlo.lebrun*googlemail.com

regards
Laszlo

...If this is for anything "serious" such as legal or regulatory requirements, it may be impossible for a "homemade" meter to be calibrated & certified by a calibration lab.

The challenge is, that a metering range from 30dB to 120 dB requires a dynamic range of 1 to 10 millions.

My antique analog Radio Shack SPL meter has several ranges selected manually by a switch. My multimeter has auto-ranging, and I assume modern digital SPL meters also have auto-ranging. My sound-activated lighting effects automatically switch the Arduino's ADC between the 1.1V and 5V automatically as necessary.

And, are you sure you need that wide of a range? You probably don't need to go to 120dB unless you are on an airport runway or something like that, and if you are on a runway you probably don't need to go down to 30db.

  • a low noise amplifier
  • a passive (A) law filter
  • a calibration amplifier
  • a precision logarithmic rectifier based on the chip AD8307

You should be able to do that all with one stage with the exception of the A-weighting filter.

So the challenge is to find an affordable microphone that is low noise, sensitive enough, and has a reasonable long-term stability.

It's my understanding that electret microphones deteriorate over time (maybe loose their "permanent" charge?). But, I assume most SPL meters use electrets. You'll have to calibrate the thing initially anyway (with a known-good SPL meter or by a certified calibration lab) and for "serious use" it would have to be calibrated periodically anyway. So maybe you can live with some drift/instability, and maybe you can make the mic element easily replacable if it gets to the point where it can no longer be calibrated.

If this is for anything "serious" such as legal or regulatory requirements, it may be impossible for a "homemade" meter to be calibrated & certified by a calibration lab.

My primary motivation is to create noise gathering stations for the European aircraft noise system.
The system can work with different station types.
The professional ones range about $3000, but you can also use some " near calibrated " stations that sell about $250 + a PC to cope for the Internet transmission.

Those " near calibrated " stations have now a 10 year old design and are still working with V-24 serial communication.

And, are you sure you need that wide of a range? You probably don't need to go to 120dB unless you are on an airport runway or something like that, and if you are on a runway you probably don't need to go down to 30db.

Okay, if necessary we could make some kind of compromises: Anyhow 30 to 110 is already reachable with the existing " near calibrated " stations.

You should be able to do that all with one stage with the exception of the A-weighting filter.

The basic design does include a quad op-amp anyhow, so the number of stages is not critical. The novelty would be to use the far superior chip AD8307 as a logarithmic precision rectifier, which has been primarily conceived as a RF power measurement chip, but according to the data sheet is able to go down to near-DC.

It's my understanding that electret microphones deteriorate over time (maybe loose their "permanent" charge?).

Yes and no. The current " near calibrated " stations is using a noise sensor module with an electret microphone that is theoretically indoor only.
We have got about 300 stations operating outdoor with more or less dilettantic DIY whether protection solutions, and quite all of them are working for years without much trouble.

But, I assume most SPL meters use electrets.
The cheap ones anyhow. The professional ones >$1000 are using true condenser microphones, those which need 200V bias voltage.
The very cheap SPL meters are using whatever electret microphone capsules they get for less money; these will usually deliver low dynamics, and high noise and drift. The better ones will use better capsules that will have a higher dynamic, lower noise and be stabler over time. I would like to find a provider for those ones.

You'll have to calibrate the thing initially anyway (with a known-good SPL meter or by a certified calibration lab) and for "serious use" it would have to be calibrated periodically anyway. So maybe you can live with some drift/instability, and maybe you can make the mic element easily replacable if it gets to the point where it can no longer be calibrated.

Yes, I know. Most of the users will not be willing to pay for periodic calibration, and we have to consider that many devices will be placed on the roof, quite out of reach for a periodic maintenance.
That is the very reason to invest in an improved, but still affordable solution, that will offer the best compromise in weather-roughness and stability.

One basic idea is to build the microphone and its first preamp stage in a standard 1/2inch stainless tube, which is the industry standard in professional SPL microphones.
So we'll be able to use a commercial professional (expensive) windcap.

It seems unimportant, but such a heavy duty windcap is everything but trivial.
It must withstand UV, heat, frost, rain and ice. It must also lead water and humidity away from the microphone, and not act as a sponge, nor as a drum under rain, nor attract birds as a landing target... and last but not least, must not stay in the way between the microphone capsule and the aircraft. :wink:

Finally I would need some experience from guys, who already have built solar powered wireless Arduino solutions,
So, you see it's an interesting project with quite a lot of facets, not only Arduino programming. I hope some people will join the project, maybe a high school, since it offers quite a lot didactic stuff as well.