Microphone to Receive Sound of Car Horn

Not liking the horn solution, (though I note your comment about the neighbour's distance), I'm wondering whether a sensor to detect your headlights while you're reversing would do the job. While headlights detected, illuminate driveway.

Sylvia.

Nice idea, but unfortunately the driveway has a bend in it, which is the main reason I would like the extra light to make sure I don't find myself against the power pole that the bend is placed to avoid.

It's possible that my ignorance of microphone capabilities is the reason I think the horn idea will work. Do you have any reason to suggest it wouldn't?

Actually, a light sensor and a time delay might work. I dismissed a manually triggered time delay because I'm not very consistent in the time it takes for me to go from leaving the house to begining down the drive. But the headlights would enevitably be pointed at the sensor until the bend, and the time taken to negotiate the bend is much easier to determine. There will have to be an ambient light sensor as well, to disable it during the day.

I might stay with the horn idea for the moment unless someone educates on why it won't work. My worry with the light sensor is that I'd have to fiddle with it to make sure light from the house and reflected light from the driveway light itself didn't trigger it. It would probably also mean that I couldn't leave the carport light on for when I got back, if desired.

In any case, many thanks for the suggestion. I might end up going with it.

--
__          __ 
#_ < |\| |< _#

One of the problems is legality, unless on private property, there are laws on use of horns which do not include turning on lights. :)

Not really, but you'd need an amplifier and a filter to separate the horn from other noises (unless spurious triggering doesn't bother you). If you want to use the horn to turn it off, you'd also need to implement a toggle.

You might find that it triggers in heavy rain (the noise of which has a wide spectrum). If you're using the horn to turn off as well, then heavy rain might cause spurious turn-offs. No doubt some effort could be expended to detect the wide spectrum, and ignore it, but you'd then want to make sure it still responds to the horn even in heavy rain - it all starts to get rather complicated.

Putting the sensor inside a tube (painted matt black inside) pointed down the driveway should eliminate spurious operation - headlights are pretty bright, so the sensor need not be so sensitive. You may even find that you don't need daytime suppression.

Sylvia.

Oh it's very much on private property. What I'm calling the drive is really just the drive to the carport. I reverse out of that, past the house, then begin the drive proper (facing the right way 'round).

The road (lane technically) at the end gets one car an hour on a busy day, and probably hasn't seen a policeman in many a year.

Still, I'd be happy to petition for an ammendment to the road rules if my idea catches on. :)

--
__          __ 
#_ < |\| |< _#

The circuit from ETI should take the place of a filter, and be more adjustable. It detects peaks in the sound wave and measures the period between them to determine the frequency. If the frequency is within a certain range, it produces a constant HIGH output.

Yes, perhaps. I hadn't thought of rain, but one other issue will be wind. There aren't many wind breaks around and, looking at the local forcast, it would have to work in winds at (very) least up to 50Km/h. You'd have both the wind noise and the effect of the wind blowing away the sound of the horn (although it would be rare for a strong wind to be blowing from the bearing directly behind the microphone). I wouldn't be surprised if I had trouble myself hearing a car horn

15m away while standing in a 100+Km/h wind gust, and there are a lot of other noises introduced by surrounding structures in those conditions as well.

The wind, if it were to blow near parallel with the line between the car and the microphone would also distort the frequency due to the doppler effect. Well I think that's right anyway...

The problem is that the drive isn't very level (it's gravel), so the headlight beam will move vertically as pot-holes and pot-hole repairs are encountered. There should be a middle ground where a tube or box will allow enough light in, I just fear it might take a fair bit of tinkering in the dark of night.

Once set up perfectly, it sounds like it might be more reliable than the horn though. Yes, I'll go gather up what light sensors I've got.

Comments on microphones still appreciated though, both is always an option!

--
__          __ 
#_ < |\| |< _#

No, the Doppler effect only depends on the difference in speed between the sender and the receiver - the effects of the wind speed cancel out. The speed of the vehicle has an effect, but I gather you'll not be going very fast at that point.

From what I gathered, this solution requires a delay timer anyway, to deal with the bend in the driveway. Just be sure that it gets reset every time the sensor detects the headlights. Then as long as the sensor is not outside the beam for a very extended period, it will work.

Sylvia.

Ah, I understand. Thanks for that.

Yes, that's true. I'll probably set a decent time of three minutes or so as there's no great rush to turn the light off, so it should be fine even if it's only triggered when the vehicle first takes off.

OK, I'll go with this solution unless someone convinces me that most of your and my worries about the microphone are false.

Thanks for the advise.

--
__          __ 
#_ < |\| |< _#

** Most cars have two horns with slightly different frequencies - so the combined sound is a dual tone chord with rapid amplitude modulation.

I have no info on that old ETI circuit so cannot say if it is compatible with a car horn sound.

Electret mics have more output than dynamic capsules so that would be my pic for the task - with some pre-amplification and filtering of low frequencies by careful choice of coupling caps.

Good luck.

.... Phil

Wait till the lyre birds start copying your car horn...

As long as there's a pair of peaks in the waveform with a fixed period, the circuit should be able to detect them. It effectively ANDs the peak detector with a signal from a monostable which goes high for a period (determined by how wide the desired range of frequency acceptance is), after waiting for a time after the first peak is detected (the period of the desired response frequency).

At the microphone, would a cone or something attached to the Electret mic be an advantage to picking up the sound?

Thanks.

--
__          __ 
#_ < |\| |< _#

** Car horns are loud, picking up the sound at 20m should not be a problem.

... Phil

Anyting suited to recording voice should give acceptable results. if you've got any song birds around be aware that they learn to mimic repeated sounds, dinner calls, crossing signals, ringtones, car alarms , probably car horns too.

Good to know.

--
This email has not been checked by half-arsed antivirus software

only when the wind changes speed will the frequency vary from the source frequency (as this changes the length of the delay), but the wind speed will modulate amplitude.

--
This email has not been checked by half-arsed antivirus software

But it may still be difficult to avoid it triggering on other nighttime movements. They've always seemed very unreliable to me. Heck, even the automatic doors on shops tend to ignore me unless I have the foresight to wildly wave a limb at them (alright they're probably just poorly set up, but I stubbornly dislike PIR sensors).

I think the light sensor would be better, of the options that have me using a time delay, and I have no problem building it. An LDR in the end of some poly pipe, connected in a voltage divider on pin 2 of an NE555 IC in monostable configuration, which drives a relay which turns on the light. Easy.

--
__          __ 
#_ < |\| |< _#

Great!

--
__          __ 
#_ < |\| |< _#

I haven't heard any such behaviour from the local birdlife so far. If it were a problem, I could hopefully set the sound amplification so that only the loud car horn could trigger the circuit.

--
__          __ 
#_ < |\| |< _#

Although I don't really need any more reasons not to go with a PIR sensor, it has occoured to me that the cold rear of a car that's just been started wouldn't likely be significantly visible in infrared compared to the cold ground. The exhaust fumes would be warm, however I imagine they would have quite a weak infrared signature, and they would be easily affected by wind.

Turning up the sensitivity would much increase the (already high) likelihood of false triggering.

I may want to spend some time in the car before setting off, so triggering on my presence wouldn't be convenient.

--
__          __ 
#_ < |\| |< _#

** Your tail and brake lights would be picked up though - long as they are not LEDs !

.... Phil