I've seen some comments/reviews that the ADMP401 is obsolete but my understanding is that this product line was sold by Analog Devices to InvenSense (hence the new code INMP401) and is not obsolete. https://www.invensense.com/wp-content/uploads/2015/02/INMP401.pdf

The amplifier design on this board is terrible at common mode rejection from the power supply. I have a 200mv 1Hz ripple on my battery powered application and it swamps out the amplifier and distorts the signal. The RC time constant of the resistor divider should be much larger than the input RC filter to reject common mode noise, and this design has it the other way round. See this guide.

I was pretty frustrated with this at first. I had the audio output tied directly to the ADC input and used an PWM-controlled LED to visualize the amplitude of the audio. It seemed that I had to tap on the microphone directly before I could see any change in the LED brightness, and even then sometimes it seemed to go the wrong way.

Then I hooked up my scope and it was a head-slapping moment. The signal is 1.5V +/- 1.5V. A quick (though maybe flawed) change to my code involved more or less rectifying the signal. I saved the midpoint value of the signal before monitoring the ADC value, then subtracted the current value from the midpoint value. The magnitude of the difference was then written to the PWM register. Now I can talk several feet away from the mic and it picks it up. :)

seems like a lot of comments from people having issues, nearly didn't purchase... finally did, and happy to report no problems testing it out in a few minutes time.

soldered leads, hooked the VCC to the regulated 3V from my Arduino Nano, tied in GND, hooked the AUD output to an analog pin on the Arduino. then hooked 3 LED's to 3 digital outputs on the Arduino in order to create a little VU/SPL meter simulation:

const int pinMic = A5; //analog input connected to mic AUD

const int pinLED0 = 9; //digital output connected to LED 0
const int pinLED1 = 10; //digital output connected to LED 1
const int pinLED2 = 11; //digital output connected to LED 2

//baseline level for my mic in a fairly quiet room
//determined by watching serial monitor on first run
const int valBaseline = 350;

void setup() {
    Serial.begin(115200);

    pinMode(pinLED0, OUTPUT);
    pinMode(pinLED1, OUTPUT);
    pinMode(pinLED2, OUTPUT);
}

void loop() {
    int valMic = analogRead(pinMic);

    Serial.println(valMic);

    if (valMic > valBaseline + 10) digitalWrite(pinLED0, HIGH);
    else digitalWrite(pinLED0, LOW);

    if (valMic > valBaseline + 15) digitalWrite(pinLED1, HIGH);
    else digitalWrite(pinLED1, LOW);

    if (valMic > valBaseline + 25) digitalWrite(pinLED2, HIGH);
    else digitalWrite(pinLED2, LOW);
}

with this code i can speak in a normal volume at 12"+ from the mic and light up the first led pretty solid and the second follows the volume peaks in my speech. if i move to within 6" of the mic or speak a little louder, the first & second LED's light consistently and the third peaks.

on these 5v Arduinos, i believe you could also use analogReference(EXTERNAL) with 3.3v to AREF in order to get more resolution: http://arduino.cc/en/Reference/AnalogReference - is that correct?

Anyone had any luck connecting this to an Arduino? I'm using the 3.3V rail, and A0 as the input, and best case, yelling into it, I get maybe a 13-point offset from the nominal read of 300.
Hooking the audio output up to a scope, I see almost nothing unless I tap it on the table top.

This mic is very heat sensitive! You can easily kill it with a heat gun!!

I was making some voice-reactive jewelry with this little mic. My jewelry uses U-mold plastic; I use a heat gun to mold the plastic and add layers to the jewelry. Heating the plastic to make it playable resulted in the mems mic board not working. I did this twice before realizing what was happening. Please be aware!!!

If I put in backorder over 60 pcs , when can it be shipped?

This mic works surprisingly well, but one word of warning: Do not connect the audio output pin to a low impedance load. My son was playing with the mic and tried to connect a small 8 ohm speaker directly to audio out and ground (with a +3.3V supply). The little op-amp on the board was killed by this. I don't know what the minimum recommended load impedance is, but probably to be safe send the signal to at least a 1K load or more.... maybe at least a 10K load.

Hi, I've connected this to a 3,5" jack then a headphones but I'm just earing some "scratch". I'm using a 3.2V 1000mAh battery, linked like this : Battery [-] : ** Gnd breakout and Gnd 3,5" jack / **Battery [+] : Vcc of the breakout 3,5" Jack [left] pin : AUD of the breakout. Maybe I can't use this like this ? thanks for your advice, G.

Is there a good way to make this somewhat less sensitive to moisture? I would need the microphone to be used in outdoor conditions. I plan to use heat-shrink cover, but as the moisture could seep in, I would prefer to varnish the circuit board as well. I tried spraying a silicon based varnish, but this makes destroys the microphone, especially when the vanish enters the small hole at the bottom of the breadboard. Would covering this small hole while spraying be a solutions? Or would you completely discourage any vanish all together?

Is there any chance of your doing a single or dual breakout board for the ADMP441? It is in the same family as the ADMP401 except it has I2S output which is kind of novel.

I'm using these successfully in a project, but now it seems the part is obsolete??

I was wondering what your process was for reflowing the Mems?

Somehow the PDF layouts have the correct pinning but the SparkFun Eagle library has the part backward. This would have been nice to know a few k$ in boards ago.

Is it possible to connect this directly to an preamplifier and record the audio output?

I get a very clear voice sound, but there is a constant high frequency noise. Then I connected the probe directly to the output of the ADMP401 MEMS device, bypassing the OPA334 OpAmp, and all the noise vanished. Any suggestions without bypassing it?

Got this from the ding-n-dent (other).... I ordered 15x of that ding-n-dent.

I now have more mems mics than I will need in my life :O

An interesting project to do with the "broken" ie blow only sensors is to use a piezo speaker and make an altitude sensor. the principle here is to drive the piezo with a constant 3V p-p at around 2kHz then space the module a distance of a few cm. this should reliably detect altitude if also encased in a plastic tube with dessicant to keep out moisture.

Can someone send me some broken ones for analysis please? i think i know what went wrong. my opinion is that during soldering the fumes got into the tiny hole and gummed up the works. if so then this explains a lot...

i am on 4hv under "conundrum" :-)

well i went ahead and bought one. being speedy and trying to rush to finish my project, i applied 5v... and today reading this i scared myself, its a 3.3v device. but it works at 5v (not saying its safe) im going to go home and fix this asap. its no real change, its swaping the vcc and analog ref.

the day i got it i was already with some code from http://neuroelec.com/2011/03/fft-library-for-arduino/ and i wired it up and used http://code.google.com/p/serialchart/ to graph it, and it worked as expected.

than i modifyed the code to interface into the "Analog" RGB strips. and now i have a nice spectrum analizer. workes beautiful for music, going to play BlackOps and see how it looks. wrapped behind my tv.

I bought 5 pieces of this. Only 3 are working when I received them. Very high failure rate.

Want to buy this and feed it DIRECTLY to the ADC input of my PIC and capture audio (@10 KHz) onto the SD card on board. So what is the verdict?
All I see is people having difficulty in using this board.
Should I buy or not?
Thanks for your input.

Those aren't Eagle files.

Hey guys,
I have a howto question...first time working with sound sensors. So kindly indulge my ignorance.
I have the ADMP401 hooked up to a machine tool (dont ask why..) via an Xbee module. I am powering the sensor and Xbee with a li-poly 3.7 V 1000 mAhr supply. The output is being read into Matlab. Did the same thing with the ADXL335 accelerometer before worked like a charm.
Now with the sound sensor, how am I to take the raw ADC'ed signal and convert to sound on matlab. In all, the reading that comes out is in mV, how do you go from mV to dB for this device so I can hear the machine 'talk' to me.
Thanks guys...

This part did not work as expected.
Used 3V power supply.
Measured AUD on oscilloscope.
The output was 1.5 V DC. No audio.
Tried a different power supply. Tried a different set of probes. Tried 3.3V Vcc. Never worked.

Tis circuit seems to be inappropriate-it uses a relatively low impedance (1.5K) inverting amplifier, instead of the much higher 10K non-inverting configuration in the data sheet.

I's be interested to try it as a independant microphone (I mean not connected to Arduino or another board). How could I adapt it to a regular microphone +48V power ?

I looked at the specs for this mic at: http://www.analog.com/en/mems-sensors/microphones/admp401/products/product.html#similar_product_table....
It indicated a possible issue. This mic system has a VERY low current out 250 uA, into a VERY low 200 ohm impedance. That is NOT much juice. Use an Op-amp step up, or a FET amplifier? It's not working because there is nothing THERE!

I have tried to hook up this sensor to an Arduino Mega and I see almost no change in serial out put unless I blow in the microphone (still this only cause very little change in the output: a couple of quanta).
I noticed that other people have seen this problem. Is there any solutions? Problem with sensor? Problem with arduino? For me this seems to be a problem with the pre-amp...
Thanks in advance

toddkrein: Anyone had any luck connecting this to an Arduino? I'm using the 3.3V rail, and A0 as the input, and best case, yelling into it, I get *maybe* a 13-point offset from the nominal read of 300.
Hooking the audio output up to a scope, I see almost nothing unless I tap it on the table top.
i'm getting the same problem. Tried powering with 2.5v (2s rechargable nimh) and a 3v ldo and getting almost nothing on the scope. If i blow into the hole under the mems or hold my mouth right next to it, i can get a 100-200mv p/p signal, else nada.
SFE, have you had any other reports of this problem? Thx!

The schematic link links to the datasheet.