Hello again ,

After some more research regarding PIR/photodiodes/piezoelectric sensors/etc. I've refined the requirements for the system.

Quick overview:
The device I'm building have 2 systems, one is an embedded circuit, and the second is, hopefully, will be a sensor that will trigger that embedded system to switch on.

The device itself will run on batteries (again hopefully for 3-6mo). In efforts to save power I'd like to turn on the embedded circuit only when an event happens and for that I need a sensor that will switch it on (maybe through a transistor).

The problem:
Again, in order to have the device run as long as possible (and w/o using 12 9V batteries , I've looked into piezoelectric sensors that can generate a 0-5V analog output for the switching. But, the bugs I'm interested in are about 4mg in weight, and I've yet to find a piezoelectric sensor that can detect such low force (corresponds to abt 4microNewtons I believe).
Perhaps a piezofilm? Or a resistor after the piezoelectric to boost the voltage?

Any comments/suggestions about making a sensitive & completely power source free switch will be very appreciated.

Furthermore, I've ran the numbers of running an average LED (IR~1.2V) on an average 9V battery and got anywhere from 20-100hrs depending on the resistor, additional time can be obtained using a voltage regulator (that will prevent the heat....). Though this is not enough, since I need it to run at least 3-6mo.

 

You could use something like a piezo speaker element and then amplify the output voltage. Also you should realize that piezo´s don´t measure force directly, the output voltage is proportional to change in shape, so a positve pulse when you push on it and deform it, and a negative pulse when you let go and it takes back its original shape.

 

Hi Kubeek, Thank you for your reply.
I'm aware of the electromechanics of a piezo material, the issue is that typical piezoelectric sensor has a ~1.5V/g sensitivity, so the output will be 1.5mV. And I don't think it's enough to make a transistor switch on let alone if mV range is a valid input rather than just noise.... Perhaps if it'll be rectified, but then it won't be power source free switch.

Piezo speaker element might pick up random noise from the room....

 

Is this "device" an outdoor unit?

Is this "bug" going to land on this "sensor" or fly-by? If you intend to get any useful answers you need to be REALLY specific about this bug/how/when/where/etc... sensing a 1-5 mg insect doing something is far from a simple task....its very small obviously. It sounds like you are searching for a specific insect too..is that right?
Please tell me its not a "bed bug" sensor.

Whats the LED for?

3-6 months is really nothing for "sensor" type device. Heck my wireless outdoor weather station (and its red LED that flashes when its sending) has been running for 7+ years reporting info to the indoor display every 15 seconds on a couple AA batteries that I've NEVER changed. Low power micros/wakeup,etc.. IMO is the least of your worries.

 

Hi mcgyvr,
Point taken.
The device is indoor, and the bug crawl along a predetermined path (the bug doesnt fly or jump).
No not another bed bug "miracle" sensor.
Well, at first I was debating if I should use a photodiode and an IR LED pair to detect the bug passing by to trigger the embedded system, but this consumes too much power.

Idk how you managed to make it run for 7+ yrs, pretty nice. I suspect that the hardware specs of my embedded system are more intense, idle/sleep sleep modes save about 1/3 of the power....
Please let me know if there is anything else you'd like to know.
And thank you.

 

Idk how you managed to make it run for 7+ yrs, pretty nice. I suspect that the hardware specs of my embedded system are more intense, idle/sleep sleep modes save about 1/3 of the power....
Please let me know if there is anything else you'd like to know.
And thank you.

I didn't build it..just got it from home depot. And there are plenty of micros with ultra low power consumptions.

So whats gonna happen when the sensor is triggered?

 

Piezo speaker element might pick up random noise from the room....

Correct, the will be trouble with that. You would need a sensor that is not microphonic - high mass would help. The sensitivity and amplifying the output is not that big problem, the problem is in the sensing method. I guess you want to be sure it doesn´t get tripped accidentally. Are you do some testing, something like a box with bugs and a sensor?

 

Ideally, the sensor will switch on a transistor (or other simple circuit) that will in turn switch on the power to the embedded system. It'll be great if the former can operate completely free of power source.

 

Yeah, I can do testings, got plenty of them bugs...lol.
I know photodetector and IR diode will work for sure, but it requires a power source.
In my perspective, the sensing method is the crucial part, since it should not depend on a power source at all.

 

A properly weighted/balanced microswitch/scale type setup could "wake" a micro up and require very little power.

 

Interesting, I've never used a microswitch... I'll look it up. Although idk how the scale setup will be.

 

And apparently I can't find it on google, unless it's something like this:
http://www.newark.com/jsp/search/pr...58&CMP=KNC-GPLA&mckv=|pcrid|19038772341|plid|

which I doubt.

 

Is the environment quiet enough that you could sense the bug crawling over a microphone?

Or is it lighted well enough that you could use a passive optical sensor with no need for illumination?

 

Hi John,

No, the room/lab isn't quiet enough to detect only the bug.
As far as lighting, I'd like this to stay as restriction free as possible, but if that's the only solution I'll use it.

 

What do you think about this one:
http://www.imagesco.com/catalog/sensors/film.html

 

It could be a start, do you have access to an oscilloscope?

 

Thanks kubeek, I'm writing them an email at this moment actually.
Yeah, I have a few scopes around here. What do you have in mind?

 

You could simply start with putting the sensor with the bugs and see what output you get. It also might not be a bad idea to put (stick?) some soft foam under the sensor to dampen the ambient noise.

 

You are right, I'm going to wait until Monday afternoon to hear back from them and then order; hopefully they'll answer my mail before.
Thats a great idea about the stick foam, I'll get some this weekend.

I'll keep you updated....

 

Hello again,

As it turns out, that company emailed me back and said that their sensors will not work with it.

So, I've done a little thinking and remembered that it was mentioned here smoke detectors/temperature detectors, so I've started looking into those applications.

This might be obvious, but I didn't know that the power on, standby, and power of every X number of seconds. At any rate, I was thinking of using this chip:
http://www.mouser.com/ProductDetail/Microchip-Technology/RE46C144E16F/?qs=sGAEpiMZZMtaH4%2fiokLhnxfkWNF8lVM%252bUCAtBSkMfEc%3d

But I'm not sure it's pulses are programmable (i.e. duty cycles) from the data sheet, I think it comes already programmed idk. Ideally, I'd want it to pulse every 500msec for 100usec and going to idle the rest of the time. Can you please check the data sheet here? You think it might work with an IR photo/diode pair?