Smallest components for motion-sensing LED project

Projects General Guidance
1

(First post, and first real Arduino project!)

I want to put some LEDs onto an object, and have them start to flash within a few seconds after the object starts moving, and to stop flashing shortly after the object stops moving. But the lights should flash only if it's dark. The object will be idle for long periods of time, and the project will be battery powered, so the ability to save power while the object is idle is imperative.

I've looked at a lot of different sensors for this, and have ruled out piezo vibration sensors because I can't find one that can "interrupt" an Arduino to wake it up from sleep mode. If someone knows of something like this that CAN trigger a wake-up, I'd like to know about it, as it seems like they require no power at all when idle.

I think the following components may be good choices for my project, but I would love to hear opinions of those with experience doing this kind of thing:

Here are some specifics that may help you evaluate the suitability of these components, or to suggest others:

  1. The power and controller parts of the project needs to be small. (Don't worry about the physical size of the LEDs.) No more than about 3 cubic inches - the shape isn't all that important - it can be long and narrow, or rectangular and thin. I just don't want it to take up much space.

  2. It may end up being produced in lots of 1,000 or so, and at that volume, the cost needs to be no more than about US$5.00. I know the prototypes will cost more - I'm saying this so no one suggests a component that might cost several dollars each, even in volume.

  3. I want to be able to distinguish between the object being "bumped" (moved for just a few seconds) and actually being "in motion". Ideally, the Trinket won't even wake up when "bumped". But aside from that distinction, I don't care how much the object is being moved. In my programming, I will base all of my important decisions on "Is it moving or not?", rather than "How much is it moving?".

  4. The object will be idle far more often than it will be moving, so low power consumption is critical. Rough guess - the lights will be flashing only about 1/2 of 1% of any given day. I'd love to power it with an AA battery or two, but a 9V would be OK. Hopefully nothing bigger than that will be required.

  5. The whole point of this project is to visibly draw attention to the object when it's moving and it's dark. (Not pitch black, just dark, like a normal night, outside.) I don't know which LEDs to use, nor what color is better, nor how to flash them for maximum visibility, nor how bright to flash them, etc. I figure that will be a matter of trial and error, after I get the basic project working. (But if anyone has thoughts on this, I'd love to hear them.)

  6. The object that will be moving or not - the object to which the LEDs will be attached - is about the size of a briefcase, and only one of the "big" sides of the object will ever be visible. (That may help someone who wants to comment on the LEDs.)

Thanks so much for any input!

Brian

2

How long do you want the device to run, weeks, months or years?
Weedpharma

3

weedpharma:
How long do you want the device to run, weeks, months or years?
Weedpharma

A battery change once per year would be fine. Let me restate that I think the LEDs will be flashing only about 1/2 of 1% of the time, and probably only a couple LEDs will be flashing at a time even then.

4

You may wish to revise your run time estimate.

365 X 24 = 8760 hrs. 1% is 87.6 hrs.

Even ignoring the LED power, running the Arduino for a year is still a big ask if you want it to be sensing continuously.

Gammon Forum : Electronics : Microprocessors : Power saving techniques for microprocessors. May help with power saving ideas.

Weedpharma

5

Hi,

Forget using arduino, waste of a controller.

You should be able to make a small PCB with LED, mercury switch, LDR and a couple of transistors.
Use a "flashing LED" for a start, there is half your problem solved.

Tom.... :slight_smile:

6

weedpharma:
You may wish to revise your run time estimate.

365 X 24 = 8760 hrs. 1% is 87.6 hrs.

Even ignoring the LED power, running the Arduino for a year is still a big ask if you want it to be sensing continuously.

Gammon Forum : Electronics : Microprocessors : Power saving techniques for microprocessors. May help with power saving ideas.

Weedpharma

I said 1/2 of 1% of the time (so only 44 hours), and more important (I think), I said I want to use a movement sensor source with an interrupt, so that the Arduino isn't really doing anything except waiting for the wake-up call the other 99.5% of the time. But I'm so new to this, maybe I don't understand - even in its deepest sleep mode, is an Arduino still consuming more power than a few AA's could support for a year?

I will look at Gammon. Thanks.

7

TomGeorge:
Hi,

Forget using arduino, waste of a controller.

You should be able to make a small PCB with LED, mercury switch, LDR and a couple of transistors.
Use a "flashing LED" for a start, there is half your problem solved.

Tom.... :slight_smile:

I don't think a mercury switch will work, as the object could come to rest in any orientation. But I'll look into flashing LEDs - I didn't know they existed. That will certainly be simpler than programming the Arduino. Thanks.