hummingbirds

I have a great view of the hummingbirds I feed. what I would like to do is to get a way to sense their presence.

I can easily hear their hum so the pressure wave is there to use.

The feeder is outside, under the awning so it is out of the rain. I could easily mount it above the feeder and have it 'look' down.

I am hoping to get a count of how often there is a hummer in the area. not looking for rocket surgery accuracy, just an ides if there is one there every 5 minutes or maybe every 10....

any suggestions on sensors ?

ultrasonic?

I wonder if you could use an audio microphone then somehow pick the right frequency? I contemplated trying to detect when my boiler is operating by the low-frequency sound you can hear but I have yet to try it out.

rw950431: I wonder if you could use an audio microphone then somehow pick the right frequency? I contemplated trying to detect when my boiler is operating by the low-frequency sound you can hear but I have yet to try it out.

In theory, yes, but it's one of those devil is in the details things.

I built a smoke detector detector, which uses an electret mic input converted to a square wave (simple opamp circuit). What I do is use the square wave input to an interrupt, which just counts square waves. When I start a count at 0, I record the start time (I set up a 0.5 ms. timer to do this, because delay() only gives resolution to 1 ms., and I don't trust it even then for reliable timing). After I've counted 100 square waves (pick your own number), I check how long it took. If my smoke detector is in fact the source of the signal, I get a total time consistent with 100 cycles of a 3160 Hz. sound. If the total time is off from that by, say, 10%, I assume it's not caused by the smoke detector.

This turns out to work well, and can distinguish frequency differences to a fairly high accuracy. I've left it on for days, and never had a false alarm. I don't know what freqs your hummingbirds generate, but this principle might work, and clearly could be used even for several different frequencies.

I don't know about sensors, but you might find these pictures interesting (I shot them through the kitchen window a couple of days ago) if you are interested in the little buggers https://www.facebook.com/mike.fields.121/posts/10209867527274771

aarg: In theory, yes, but it's one of those devil is in the details things.

Indeed.. that's one of the reasons it hasn't progressed beyond the ideas notebook.

I've a suspicion that the arduino isn't going to have sufficient memory and CPU to do the fancy maths while sampling at a high enough rate.

What about using a PIR sensor like this pointing down onto the feeder.

As others have hinted at, maybe a suitable microphone passed through a band pass filter and fed to a FFT sketch will allow you to isolate the hum.