For nearly the same size, you'll get FAR better runtime by using 3 AA batteries (4.5V, ~2,000 mAH) instead of a 9V Battery (~325 mAH), especially when you are burning off half of the 9V as heat.

I cannot think of a switch that sensitive that also wouldn't be damaged/set off by wind or any other environmental issues (rain, hail, snow, very small rocks...).

 

Hi That one guy,

I was planning to use a voltage regulator to reduce the input voltage regardless of the battery size just as an extra measure for extending the battery life. But you are right, I'll probably use your suggestion.

That being said, I don't think it's possible to reach the desired battery life w/o pulsing the IR diode, and then using comparator.... That's why I think the chip I mentioned above can work, I'm just not sure if its possible to change its core logic. What do you think?

In order to reach 3-4mo of operating time for the trigger (the goal of the IR pair is to trigger an IC btw).

 

What size insects are you trying to detect? Daylight or 24 hours? What environment will this be in?

Daylight only, a lot can be detected with a photo-transistor (how do you stop spelling photototo?), night gets more difficult.

I don't have any experience with the linked IC, I just looked at the datasheet and it looks like it would fit your application somewhat, it's only "looking" for 0.1mS every 10 seconds, however.

An internal oscillator strobes power to the smoke
detection circuitry for 100us every 10 seconds to keep
standby current to a minimum.

An IR LED running during darkness would kill the battery more than any other component. PIC Microcontrollers draw nano-amps in sleep mode, and only a few mA when running. In contrast, the IR LED will draw 15mA or more continually. Using the sun for illumination may be better, but that limits the time you can detect... bugs.

 

The bug will go a predetermined narrow path in a black box setting (I.e. very minimal light).

That's why I think I have to pulse the Diode for only micro secs and once the event has happened I'll use the I/O pin to trigger a transistor that will in turn switch on another voltage regulator for the IC.

But idk if this chip duty cycles can be modified significantly from what's in the data sheet.

 

Ohh and it's abt an ant size, and the path is designed for it.

 

If the ant is slow, and crosses 5mm of photo-transistor area in over 10 seconds, it could fit your application.

You may wish to use more than one, spaced a bit away from where you have the giant magnifying glass to burn them, so the controller can tilt the magnifying glass toward the sun before the ant gets to it.

Substitute whatever other detector/processor stuff you plan to use if this isn't an automated ant cooker.

 

lol, no it's not an automated cooker, though it can be a good idea .

That's where I got stuck in the data sheet "every 10 sec" and then the "battery check every 40 sec." I'd like to change these two values, but it doesn't look like it's possible.... Right?

I wish I could find a chip that can be changed for these values.

 

It comes down to a tradeoff between sampling rate and battery life.

Larger capacity batteries could be used, such as CR123 lithium 3V, the decision is either low power draw with 10 second tests that lasts for months on a set of batteries, or a check every second, which would last for about 3 weeks on a set, assuming you only used the IR LED when it was dark out.

 

For pulsing it every second I'll have to build an array of 10, so it's not very practical; unless I'm missing something.
The other way, a short pulse every 10 sec won't work either, since the time spacing is too wide.

what a pickle....
There has too be a better option

 

I'm about out of ideas as well for off the shelf solutions.

You could make your own version with a PIC and wake on timer event, then set the timer prescaler and count to generate an interrupt as often or infrequent as you'd like, at which point it would enable the LED, check the transistor output, process, reset the counter, and go back into sleep mode. The phototransistor may need an amplifier if measuring very small changes, which should also only be powered by a PIC pin rather than idling on continually.

 

Thanks, I'll explore this option in the coming week.