1
\$\begingroup\$

I'm in the process of designing a battery powered microphone sensor for wildlife monitoring, specifically recording wildlife in forests. I want to design a system that uses a MEMS microphone that can not only detect signals close by but also far away if possible. I want to be able to record signals up to 80 kHz to also research the ultrasonic sounds. However, I'm not very experienced with the analog side of things and I was hoping for some advice.

Currently, I have designed a system based on the AMM-3738-B-R, which is then amplified by a one channel of the OPA2834 with a non-inverting amplifier circuit, as shown here:

Non-Inverting Amplifier

I selected the AMM-3738-B-R based on the ultrasonic response in the datasheet and the low current consumption at 1.8V (+- 90uA). I selected the OPA2834 because of its relatively low Iq (170uA/CH), high bandwidth (20 Mhz) and relatively low noise (12 nV/√Hz). Including the 12-bit ADC (STM32U5A9), resistor currents and 3V voltage reference, I'm looking at an estimated 477 uA.

I have two questions coming from my current setup:

  1. Would it be better to switch to the IM73D122 digital MEMS microphone? It offers a higher SNR (73 vs 62) and higher sensitivity (-26 vs -38) but I lose the ability to amplify the sound with the opamp, it consumes a lot more power (900uA without SAI interface) and the ultrasonic response graph looks worse. It does reduce the circuit complexity since my system has a 3V reference only for this circuit and it eliminates the opamp and its passive components. Would I still be able to pickup on lower volume sounds? What would the implication of converting the PDM signal to a usable signal for FFTs be in terms computational complexity (I would like to sleep the MCU for a minute while a buffers fills up with DMA)?
  2. Would it be better to switch to the IM73A135 analog Differential MEMS microphone? It offers a higher SNR (73 vs 62) and identical sensitivity (-38 vs -38). Additionally, the ultrasonic response graph looks worse. My main issue is that this is a differential microphone and I don't know how to amplify this. Would this be as simple as routing the OUT- and OUT+ pins of the microphone through a channel of the OPA2834 each and using the differential mode of my ADC? Would the difference in SNR be worth it for the power consumption of the additional opamp channel or would I be better of using the second channel to create a two stage amplifier circuit?

On a more general note, I was wondering if I could exchange R2 for a digipot to allow for programmable gain and if so, how important is the bandwidth spec of the digipot. Is it sufficient to be higher than 160 kHz?

\$\endgroup\$
8
  • \$\begingroup\$ The AMM-3738-B-R does not state it can go above 20 kHz. Where did you read otherwise? Also it's not how far away a sound is sourced that limits a microphone's ability to reproduce a signal. Maybe you need to post the responses that you saw. \$\endgroup\$ Commented Mar 6 at 15:53
  • \$\begingroup\$ Hello Andy, thank you for your reply! On page 4 of the datasheet, the ultrasonic frequency response is shown. Are there any other factors that I need to take into account for selecting a ultrasonic capable microphone? \$\endgroup\$ Commented Mar 6 at 16:10
  • \$\begingroup\$ going for 250 gain in one step with a GBW product of only 20M is asking for trouble, use a faster amplifier, or split the gain into two cascaded lower gain sections. \$\endgroup\$ Commented Mar 6 at 17:26
  • \$\begingroup\$ Would I be able to use both channels of the OPA2834 and use a 200 gain first stage and i.e. a 5 gain second stage? \$\endgroup\$ Commented Mar 6 at 19:21
  • \$\begingroup\$ @JorisGravesteijn Your original circuit shows a gain of 250, I would suggest putting sqrt of that in each stage, so two stages of 16x gain. Your comment is suggesting a total of 1000x. \$\endgroup\$ Commented Mar 6 at 21:10

0

Reset to default

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.