MEMS microphones Forum Discussions
We made new mems microphone package based on Infineon MEMS and ASIC.
we tried to calibration this package. however, to start calibration,when the package is applied bias voltage, the initial sensitivity of Mems microphone is slightly increasing from -47.2dB to -46.9dB, it took about 10sec and it settled a stable senstivity value. About 0.3dB variation occured, a certain package has a 0.6dB variation. What kinds of causes can make this problem? SNR is OK. THD is OK. Freq. response is OK.AOP is OK. But even after calibration, this phenomenon still happen.
Could you recommend any solution and the theoretical explanations for this issue?
I really appreciate if I get the answer from here.
I'm building a design using a IM73A135.
The datasheet lists a SNR of 73 dBA and an output impedance of 250 ohms in Normal mode.
Can I use 2 of these and tie the positive outputs together (and the negative outputs together) to gain 3dB SNR while having no increase in output level?
It works in a simulation, but I might be missing something.
I know the normal way to do this is through an active mixer using an op-amp.
However, I want to avoid the space/power needed for the op-amp, as well as the level increase.
I was planning on DC coupling the outputs due to space constraints.
How consistent is the DC output voltage from these mics?
I want to use a IM72D128 device in a product that will have IP56 rating. The datasheet for IM72D128 says that it has a rating of IP57 at a device level, but how does that work when the component is mounted? Is there any requirements on the solder used, or the substrate that the component is mounted on?
Markus GrawéShow Less
We are evaluating the IM72D128V01 microphone for a new commercial product.
We would like to use the AudioHub_Nano kit to record the sound, but the 25.0 mm X 4.5 mm evaluation flex board ( https://www.infineon.com/cms/en/product/evaluation-boards/kit_im72d128v01_flex/) which houses the MEMS microphone is unfortunately too short, and the 15.0 mm optional Flex adapter board is also too wide for our application.
Can you let us know if Infineon can provide the design layout and Gerber files for the flex pcb so that we can modify the design by extending the overall length of the flex board and order from a pcb manufacturer a much longer (1-2 ft) flex board that would allow us to interface the microphone with your AudioHub_Nano kit?
Many thanks for your help!Show Less
I would like to use the Audiohub nano with digital MEMS and control the recording (init, start, stop, gain) with a script. Is there a way to do that using the FT9xx programming interface?
What documentation is available to support re-programming of the Battery Operated Glass Break Sensor kit, EVAL_SEN_SMART_ALARM @ https://www.infineon.com/cms/en/product/evaluation-boards/eval_sen_smart_alarm/ ?
Since it's PSoC 6 based, I'm guessing a Miniprog 4, CY8CKIT-005, can be used. Confirmation that the Miniprog can connect directly to the 10-pin X3, or if it should be connected to the 5-pin X2 along with any power, jumper or configuration settings required to program would be appreciated.
Are there different versions of firmware available to test different types of Glass Break or intrusion detection?
Are there ways to connect the Audiohub Nano or a similar product to ethernet? Which alternatives are available?
I am testing an IM69D130 microphone on a Shield2Go evaluation board.
I am getting something that I would call quite a large DC offset and that is heavily fluctuating over time.
It is the same using the I2S output of the ADAU7002 ic on the eval board or the direct PDM output of the microphone itself.
The datasheet doesn't mention anything like that at all. Is that really to be expected? (When testing with a similar product from a different vendor it does not happen).
Sine wave recored to a wave file and opened in Audacity:
You clearly see the movement up and down.
Graph of the min (blue), max (red) and avg/offset (green) for every 3000 samples for a total of 30 seconds. When recording a constant sine wave.
When you subtract the green line the others would look clean. So yes one can probably do some form of post processing/filtering. Not sure how good that would work because it fluctuates so much.
Any ideas? Show Less