Regarding Inconsistent Glass break Detector Algo implemented on MSP430F2274

Brajesh Sharma

Hi All,

I have got an OLIMEX kit designed for Glass break detection using TI's Algo. The problem what i am facing is that the glass break detection is very inconsistent. I have modified the code (downloaded from TI's website for Robust design) to change some LED settings so that i can track the state of the device properly. By that i found out that sometimes device remains in thud valid mode without any significant activity on the mic. I have modified the HIGH_LVL, LOW_LVL and THUD_LVL parameters as defined in application Note of GBD. But the major issue is the inconsistency on detecting the glass break and the range is just around 1 feet (although application note says around 8 feet).

Can anyone please help me out on this issue, is this. Any inputs would be appreciated.

Thanks

Brajesh Sharma

over 10 years ago

0 YiKai Chen over 10 years ago

Guru 735685 points

From the application report, MSP430 only wakes up 20us every 2.5ms and I think it is OK for power consumption. I think the detection depends heavily on wake-up period so you suppose not to change that.

0 Brajesh Sharma over 10 years ago in reply to YiKai Chen

Expert 1785 points

Yes, i agree but our device is coin cell battery device which needs certain battery back up and this much power consumption is a but high for to achieve that. Can you suggest something to reduce power consumption. or do you think it's not feasible with this TI Algo ?

0 YiKai Chen over 10 years ago in reply to Brajesh Sharma

Guru 735685 points

If You need very long battery life, I would say it's not feasible using coin cell battery for this application.

0 Jens-Michael Gross over 10 years ago in reply to YiKai Chen

Guru 227245 points

Indeed, one cannot eat the cake and keep it.
Low power is mainly a result of not doing much. Requiring constant monitoring of the signal requires power.

You might be able to implement the 'detection algorithm' in hardware (filters etc), so at the end, only a 'detected' signal is applied to an MSP digital pin. It might consume less energy than constantly monitoring the microphone and analyzing the analog signal.
Implementing thing sin software is convenient and reduces hardware (and therefore production) cost. But it is not the most efficient way regarding power consumption, and not even always the most efficient way to use development power.

0 Brajesh Sharma over 10 years ago in reply to Jens-Michael Gross

Expert 1785 points

I guess so. Thanks Michael ..... :)

**Attention** This is a public forum

MSP low-power microcontrollers

MSP low-power microcontroller forum

Regarding Inconsistent Glass break Detector Algo implemented on MSP430F2274