MSP430FR5869: Glass-breakage detector for MSP430FR5869

Dear Son,
many thanks for your inquiry.
I am not sure about the reasons to select the MSP430FR5869 for the glass detector application, as the only functionality, which is evidently used in your design proposal is the ADC integrated in this derivative.

But let's focus first on your design review request. First I have to make a generic statement related to design reviews. You have to keep in mind, that our review cannot cover all required aspects of your design, and you need to maintain responsibility for your design, and irrespective of our comments on your design, have to apply common good practices on verification and testing of your design.
To comment specifically on the suggested schematic, I can make the following comments:

- I am missing the power supply section for the application. According to the labeling, it seems you intend to power the MSP430 by 1.8V, while the microphone is connected to 3.3V. The 1.8V for MSP430 are an issue for multiple reasons.
a) The integrated SVS is active per default on power up. The specified threshold limits in the datasheet are 1.77V-1.99V. This means in worst case the device needs a supply voltage of 1.99V to be released to operation.
b) According to your schematics, you will be operating the ADC with the internal reference. With 1.8V supply, only the reference of 1.2V will be usable, which will limit the ADC range to 0V-1.2V max. This means a very low voltage swing.
c) It is not recommended to operate the device at the limit of the specified supply voltage range, as it is 1.8V min, and usually e.g. with an LDO with typ. +/-5% tolerance, you will be easily out of specification, and have no head room for any supply voltage drops.

- The first OPA circuit is set up incorrectly. It is supplied from GND (negative supply voltage) and 3.3V (positive supply voltage). There is no biasing for the positive input of the OPA, and the connection to the microphone is AC coupled only. This means, the OPA will clamp the positive signals of the input to GND, and only the negative wave portion will be visible at the output. As in our application example with MSP430F2274, you need to set the positive input to half of the supply range of the OPA, by e.g. a resistor based biasing network.

In general, I would rather recommend you to use an MSP430 with the integrated LEA module. This is very powerful arithmetic function support module, which can perform a lot of math processing extremely effectively. It supports also FFT. Thus very useful for applications like glass breakage and other sound processing applications. The derivative having LEA integrated are e.g. MSP430FR6037, MSP430FR6047, MSP430FR5994, MSP430FR59941 and a few other memory and package variations. There are also some application examples on audio processing in combination with the MSP-EXP430FR5994 and BOOSTXL-AUDIO.

To address your 2nd question. The difficulty in general will be, to identify some characteristic attributes of the sounds, which you want to detect. Here I can make some top level suggestions, to look at the time and frequency profile of such signals, trying to find commonalities comparing e.g. the crying sound of different babies.

Link to glass breakage page www.ti.com/.../glass_break_detector

Best regards
Peter

Hi Peter,

Thank for your review and comments.

I will consider and fix issues from your comments.

In current our design, MSP430FR5869 will work with 1.95V ( 1.8V only label of rails, it is not correct). 

Currently, my product is using MSP430FR5869. We want to add Glass-Breakage detector function for Rev 2. When reusing MSP430FR5869, Software will not lose more time for development. Besides, another issue is the price. The price of MSP430FR5869 is cheaper than the price of MSP430FR5994 and others.