Ultra-low-power Beam Break

[dapug]

We're all familiar with the classic IR emitter/receiver method of detecting a break in the light.  But does anyone know of a lower power way of doing this?

The problem:
Assuming each LED uses 20ma, if I were to power 20 of them, that would draw 400ma, and would kill a (for example) 1200mAH battery in a matter of 3 days or so!  I'd like to find a way to get the total power way below this nearly half amp draw.

What?
Right, so what I am doing is counting the bees exiting and returning to my beehive.  A beehive has an entrance that is typically several inches wide, and I am building a micro turnstile or baffle (like in a subway station) of 10 gates that allow only 1 bee through each.  I want to know the direction the bee is traveling, so that means 2 beams (forward and aft) so that I can calculate which beam broke first and assume an in or out flow. 

But dang... that's a lot of LED's.   Any other options out there?  I thought of a pressure switch like in Minecraft, but um... never mind.

BTW, if it's not obvious, the reason for battery power is that a beehive is not typically near a power source.  I could rig up solar I suppose, but I still really want to keep the wattage to a tiny amount.  Also FYI, I am transmitting data via xBee, but that is the least of my power concerns.

[mij59]

You could try a remote control receiver and transmitter.
The receiver is low power e.g. 200 uA, and is very sensitive, so the transmitter can also be low power.

[SeanB]

Run them in series, that way the 20mA is used to power about 6 IR LED's which will help with a 12V supply. Then you will only have 80mA draw on the 20 ways.

With this as well you can pulse the LED's at say 40-80kHz, using a duty cycle of 10% and get even more power saving. As well the pulse train will make it easier to detect passage, as there will be no slow change with ambient light.

[tggzzz]

Use natural sunlight as the source? Possibly "chopped" at a known frequency by a rotating spoked wheel to make detection easier?

Maybe connect a transparent hollow tube to the entrance to make it easier to "see" the bees on their way in and out. Demonstration hives in zoos/museums often have such a tube which is several feet long, so people can see the hive interior and the bees can still go outside.

[Codemonkey]

Use only one LED feeding into a load of fiber optic strands.

[electr_peter]

You can decrease power consumption 50% by using [1 LED + 2 photo detectors] instead of [2 LEDs + 2 photo detectors]. 1 LED shoud be enough for 2 sensors.

[mikeselectricstuff]

Use a PIR sensor - may need to fiddle with optics to get the right sensitivity field of view, but you can get PIR sensors down to a few uA

Another option may be some sort of capacitive sensing - plates either side of the exit, along the long axes. Capacitance increases when a (presumably conductive) bee passes.
 Surface insulation to avoid effects of damp. By splitting the plates into seperate strips you may be able to detect direction.

[Zero999]

Use a PIR sensor - may need to fiddle with optics to get the right sensitivity field of view, but you can get PIR sensors down to a few uA

The problem with that is, unlike humans, bees don't emit much IR

Another option may be some sort of capacitive sensing - plates either side of the exit, along the long axes. Capacitance increases when a (presumably conductive) bee passes.
 Surface insulation to avoid effects of damp. By splitting the plates into seperate strips you may be able to detect direction.

That could work but I believe bees can charge up with static electricity, perhaps the change in field could be detected?
http://phenomena.nationalgeographic.com/2013/02/21/bees-can-sense-the-electric-fields-of-flowers/

[mikeselectricstuff]

Use a PIR sensor - may need to fiddle with optics to get the right sensitivity field of view, but you can get PIR sensors down to a few uA

The problem with that is, unlike humans, bees don't emit much IR

Another option may be some sort of capacitive sensing - plates either side of the exit, along the long axes. Capacitance increases when a (presumably conductive) bee passes.
 Surface insulation to avoid effects of damp. By splitting the plates into seperate strips you may be able to detect direction.

That could work but I believe bees can charge up with static electricity, perhaps the change in field could be detected?
http://phenomena.nationalgeographic.com/2013/02/21/bees-can-sense-the-electric-fields-of-flowers/

Plessey make some field sensors that may be worth looking at - results may vary with humidity, wetness of bee etc.

[tggzzz]

Consider that inductance sensing might be more reliable than capacitance sensing. See
http://www.ti.com/ww/en/analog/dataconverters/inductance-to-digital-converter/index.html
for an IC that makes it easy,

[mikeselectricstuff]

Consider that inductance sensing might be more reliable than capacitance sensing. See
http://www.ti.com/ww/en/analog/dataconverters/inductance-to-digital-converter/index.html
for an IC that makes it easy,

Only if the bees are made of metal.

[electr_peter]

Consider that inductance sensing might be more reliable than capacitance sensing. See
http://www.ti.com/ww/en/analog/dataconverters/inductance-to-digital-converter/index.html
for an IC that makes it easy,

Only if the bees are made of metal.

Bees forge their stings out of metal, don't they?

[mikeselectricstuff]

You could try a remote control receiver and transmitter.
The receiver is low power e.g. 200 uA, and is very sensitive, so the transmitter can also be low power.

IR remote receivers can be good for detection, but most aren't designed for low power as they're designed for mains powered gear.
However you can get similar devices designed for 3d-glasses, which are more power-optimised, and also designed to wake quickly from powerdown
Vishay do some
e.g.
http://www.vishay.com/company/press/releases/2010/100802optoreceivers/

[bktemp]

Those sensors should work well:
http://www.silabs.com/products/sensors/infraredsensors/pages/si114x.aspx

Because of the low duty cycle the avarage current is in the uA range.

I use them in a battery powered application to count persons walking by at about 1m distance. This is the upper limit of the sensor, but for a short distance a much lower led current can be used.

[Artlav]

AFAIK, bees don't fly at night.
So, why not replace the LEDs with sunlight/daylight?
I.e. pass lightguide fibers from a big white light collector on top to the holes for LEDs above the turnstiles.
Now, all you need is to read off the photodiodes, which takes less power than a bee's brain.

[Paul Price]

(near) Zero Power Solution

Use wind power (windmill) to run/generate HV by means of a Van De Graff.  Store HV in large Lyden Jar so power is supplied when wind has stopped. Lyden Jar is large glass jar lined with Aluminum foil inside/outside
A small battery powered piezo HV generator (from gas stove lighter) could also be used to occasionally charge up Lyden Jar if no wind.

Use the DC HV for sensor tube/bee freeway bee tagging.
 
Bees travel though a hard plastic tube/square channel etc. detection pathway with a HV needle discharge at one opening of pathway.

Bees pick up a significant HV charge upon entry into pathway.

Electrolet sensors (really cheap and sensitive: metal plates connect to input of a CMOS gate like 4004UB) detect bee passing.
A multi-meg resistor quickly discharges plate after it has registered its pulse output, then ready for next bee to pass.

Store the collected data storage using micro power PICC etc chip, very large data storage: could be SRAM using micro power with tiny battery backup.

Maybe needed:  a conductive fiber matrix for bee to pass through to discharge bee upon entry/exit so it can be sensed again.
This solution requires only micro-watts.

[mikeselectricstuff]

(near) Zero Power Solution
This solution requires only micro-watts.

Until it rains.
Or gets humid
And assumes the bees aren't sensitive to electrostatic fields.

[Paul Price]

Lightning works well in wet weather.

Basic Ebonics Explanation:

A soaking wet bee can carry the same charge as a dry one when flying or traversing a plastic tubeway.
Bees might/might not bee bothered, even if they bee sensitive to electrostatic fields, but they wanna go where they wanna go..out.

De-ionized water can be a good insulator.

Closely spaced metal(AL foil) or conductive plastic sensor areas separated(insulated) by closely spaced insulating raised thin ridges could detect the drop in resistance due to wet bee feet contacting adjacent metal sensors on rainy days. Metal plates/humidity detector could detect when to switch over to wet bee feet detector.

[mikeselectricstuff]

Lightning works well in wet weather.

The creepage distance is somewhat larger....
Anyone who has ever used a van De Graff will tell you that humidity is a major issue.
Anything relying on very high impedances in a semi-outdoor environment is doomed to failure.

[SeanB]

You are not doomed, it is absolutely sure to fail.

[TerraHertz]

I expect a bee would make quite a racket, walking across a piezo disk transducer wired to a high gain amp.

How you go from sensing a bee walking across a piezo plate, to reliably sensing the direction they walked, I don't know. Two sensors just means two bees are going to walk on them at the same time.

Maybe train the bees to press a door bell?  No seriously, can bees be trained? "It presses the button, or it doesn't come inside" sort of thing? Then you need very very tiny buttons. That bees can see. Even tinier UV LEDs?

'xbee' is a terrible joke, in these days of bee colony collapse disorder.

[bookaboo]

It's a shame bees don't emit IR as these have an analogue o/p with which you could (with some clever algorithms) detect direction.
http://www.conrad-electronic.co.uk/ce/en/product/172526/B-B-Thermotechnik-PIR-LP-PIR-SMD-MODUL-3-5-V80-UA-SMD-PIR-Motion-Detector-Operating-voltage-3-5-V40-A-Angular-ape?ref=searchDetail
Just rip off the fresnel and play around with various size holes, perhaps you could introduce a small about of IR that the bees would reflect/block.

The other idea is to cut down your emitter current consumption by using a single light source and an array of light pipes (though I guess those are opaque at IR. Then further cut consumption by using some form of PWM.

Or.... remember the really old tv sets with capacitive buttons? Can insects trigger those?

[voltsandjolts]

You might be able to use one or more infra-red proximity sensors, such as the Vishay VCNL4000 series. One small device with built in infra red transmitter and receiver (you can add extra emitters if required) which detects objects (10 to 200 mm away) by the amount of infra red bounced back. Also has a daylight sensor built in so you could shutdown at night. All controlled over I2C interface. The infra red is pulsed (you control sample rate, i.e. power consumption) and functions in daylight (modulated IR). I've used a VCNL4020 in a non-contact gap measurement application (10 to 50mm) with 100uA (average) consumption. They are intended for smart phones to detect your head against the phone to switch screen off. Some folks are using them in robot object detection.

www.vishay.com/docs/83476/vcnl4020.pdf

eBay keywords: Adafruit VCNL4000

I have a Python Bus Pirate script here somewhere for the VCNL4020 I2C interface, let me know if you want it and I'll dig it out,

With the right arrangement at the entrance to the hive these might work for you - probably need something to keep the rain off of it though. I like your idea, the more data we can gather on bee numbers the better our understanding of their (sad) decline. Good luck.

[rob77]

what about a camera + motion detection software (RPi or other small computer).  it definitely won't be low power, but it will be one of the most accurate solutions to count the bees.

[Zero999]

(near) Zero Power Solution

Use wind power (windmill) to run/generate HV by means of a Van De Graff.  Store HV in large Lyden Jar so power is supplied when wind has stopped. Lyden Jar is large glass jar lined with Aluminum foil inside/outside
A small battery powered piezo HV generator (from gas stove lighter) could also be used to occasionally charge up Lyden Jar if no wind.

Use the DC HV for sensor tube/bee freeway bee tagging.
 
Bees travel though a hard plastic tube/square channel etc. detection pathway with a HV needle discharge at one opening of pathway.

Bees pick up a significant HV charge upon entry into pathway.

Electrolet sensors (really cheap and sensitive: metal plates connect to input of a CMOS gate like 4004UB) detect bee passing.
A multi-meg resistor quickly discharges plate after it has registered its pulse output, then ready for next bee to pass.

Store the collected data storage using micro power PICC etc chip, very large data storage: could be SRAM using micro power with tiny battery backup.

Maybe needed:  a conductive fiber matrix for bee to pass through to discharge bee upon entry/exit so it can be sensed again.
This solution requires only micro-watts.

The idea of using locally generated renewable energy is a good one, even if the high voltage/static idea is flawed.

You could use solar or wind power to charge a battery and power a not so low power solution.