Hi,
As a hobby, I climb up mountains in Wales and take photographs of military aircraft performing low level training. This requires an extremely high level of concentration as you need to be ready to spot the aircraft appearing. A good example of what is involved can be seen in this video :

http://www.youtube.com/watch?v=1N23miNLAY4

Now, it would be brilliant if we could get a few extra seconds warning of the approach of an aircraft. I would appreciate comments / views on the following.

It would be possible for us to place a device (would need to be battery operated) in a place where the aircraft pass before they appear in our line of view, the position though is close to a road so passing cars and lorries would need to be ignored). If I could detect the approach (or passing) of an aircraft, I could trigger a PMR radio to send a notification bleep to us up on the hill.

The question is - how to differentiate an aircraft from for example a passing lorry, which could be as loud as an aircraft (so a simple sound operated switch would not be enough to prevent large numbers of false triggers).

I had though of measuring the doppler shift of the sound made by anything which passes this device, the shift being proportional to the speed of the object creating the noise. As an aircraft is MUCH faster than a lorry or car, the doppler shift would be much higher and would allow me to differentiate between them (hopefully).

So, firstly, does anyone have a view as to whether this might work, and if so, do you know of any chip which could measure and provide an output which tells me what the doppler shift was so I can process it with a micro processor and operate the radio transmitter at the appropriate time?

All input is very much appreciated.

 

If you got a directional mic (parabolic, etc) then you could just go on loudness.

 

I agree. You could put your MIC on a tripod or stand that places it at 10 or 15ft.

MIC facing straight up.

That will give you a pretty good natural noise reduction from anything under(or behind) the MIC. If the MIC is higher that the automobiles, you would have a pretty good chance of doing this.

A comparator could be used to only transmit the 'BLEEP' when sounds over a certain threshold are heard.

 

Jet turbines are very high rpm. Maybe just an ultrasonic mic would differentiate their sound from that of reciprocating engines.

If they fly a fairly regular course, as they do in the US, perhaps an upward directed PIR detector would pick them up...unless they are F117's. You might try a portable radar speed gun. Oh, that might have a few downsides.

I do like your idea of using the Doppler shift, but I suspect you would need some sort of sound processing to identify which of potentially several concurrent sounds was shifted. Although, once you could identify the sound signature of a jet, you wouldn't need Doppler.

All in all, it seems that what you want to do (i.e., early detection) is something the military would like to know how to do too or at least be able to counteract.

John

 

All in all, it seems that what you want to do (i.e., early detection) is something the military would like to know how to do too or at least be able to counteract.

Im sure they would.

 

The military used to use these:
http://en.wikipedia.org/wiki/Acoustic_mirror

 

Thanks for the input guys. I don't think a PIR would work as the device will need to be 'secreted' in an area with some tree coverage. The mic being too high would make the device conspicuous (which would probably result in it being stolen or vandalised).

The directional mike idea is a possible though.

I don't think any device I could come up with would worry the military inn the slightest though - As it is audio based it would only give warning of sub-sonic aircraft anyway and the detection range would be severley limited. All I am looking for is 3-5 seconds of extra warning to allow us to get the video camera ready.

The fact that the jet engines will probably produce ultrasonics too is an interesting point, and would probably differentiate them from a road vehicle quite easily. Just need to find an ultrasonic detector.

 

If you can install the device in a tree, that will help with the height and disguise to prevent theft. It need only be a little MIC, comparator, transceiver and battery.

A little paint, and you should be good to go.

 

That might not be a lot of warning. I was out checking a fence line when an Air Guard F-4 went overhead. I had only the warning of the shadow before the aircraft was past. Glad I didn't duck into the barbed wire. The sound reached me after the Phantom was past.

Even prop aircraft don't advertise well - http://www.youtube.com/watch?v=ve-MVVQmaRo

 

That might not be a lot of warning. I was out checking a fence line when an Air Guard F-4 went overhead. I had only the warning of the shadow before the aircraft was past. Glad I didn't duck into the barbed wire. The sound reached me after the Phantom was past.

Even prop aircraft don't advertise well - http://www.youtube.com/watch?v=ve-MVVQmaRo

The device will be left about 3-4 miles away from our location, so it will give some additional warning. That f-4 must have been close to supersonic to creep up on you with no sound !!!!!

 

Yeah, he was having fun. When you are out in a rural area, the Air Force gives free shows. I get to see F-16 dog fights (hard deck at 10,000 feet), and occasionally get to be a target (simulated) for A-10's while on the road. We are also in the spot where B-2's out of Whiteman climb out to clear commercial air traffic.

Did you see "28 Weeks"? The last scene has a Tornado making a very low pass over a lake and some hills.

 

My old house was a target, had plenty of fighters over, they used to (imaginary) rocket our house then bank hard away. Had A-10's over too a few times - they are scary things. I suppose it is because it was a white house on top of a hill.

 

The fact that the jet engines will probably produce ultrasonics too is an interesting point, and would probably differentiate them from a road vehicle quite easily. Just need to find an ultrasonic detector.

You might like to consider trying one of these : http://www.knowles.com/search/prods_pdf/SPM0404UD5.PDF

Not too expensive, and with a bit more of a frequency range than the fixed 40KHz type of cheap transducer.

Now how you go about finding something ultrasonic, that is specific to jet engine sounds, I'm not sure!
I know that during my experiments with making devices to detect ultrasonic bat echolocation signals, I discovered all sorts of things that have ultrasonic 'footprints' that you don't hear in the 'normal' hearing frequency range.
The movement of vegetation in the wind, or even the ultrasound emanating from grass when you walk through it are quite significant.
And rattle a bunch of keys? -- deafening!!

Analysing what ultrasonic 'pattern' might be specific to a jet turbine?....hmmm, interesting project!!

 

I wonder if you could get software which could be "trained" to recognize the sounds of jet engines and ignore other sounds. That might be in terms of the spectrum of frequencies, but also the envelope of volume that the microphone would be exposed to over the few seconds that the plane is nearby.

Or, maybe you can set up a laptop computer with a webcam and some kind of software that says "I'm seeing a moving object". Or would there be too many false alarms caused by clouds, birds, sheep on the hillside etc?

 

I know that during my experiments with making devices to detect ultrasonic bat echolocation signals, I discovered all sorts of things that have ultrasonic 'footprints' that you don't hear in the 'normal' hearing frequency range.
The movement of vegetation in the wind, or even the ultrasound emanating from grass when you walk through it are quite significant.
And rattle a bunch of keys? -- deafening!!

Analysing what ultrasonic 'pattern' might be specific to a jet turbine?....hmmm, interesting project!!

Interesting info re the grass / car keys ! I'm now thinking about a combination of ultrasonic detection and Doppler shift detection. The ultrasound from grass and vegetation movement is unlikely to suffer a doppler shift, whereas the jet aircraft will. Also, the aircraft will likely be significantly louder than the other sources of ultrasound. Thanks for the link to the better detector - although a surface mount device might be a bit of a challenge for me.

I'll keep posted on progress.

Out of interest, does anyone know of a relatively cheap recording / playback device which I can use to record the ultrasonic signature of a passing aircraft to use for testing my device.

 

I just tuned in and saw the word, "trained" and it got my attention. You could train a dog to detect ultrasonic frequencies and "point".

This is either a brilliant answer or a stupid one, but dogs do have good hearing, it works at ultrasonic levels, and their ears come complete with a mind that can recognize an airplane engine.

 

Take the dog to airshows. Every time a jet flies by, give it a doggie treat.

Then when you are out in the field waiting for a flyby, just wait for the dog to start drooling.

 

My only worry is that this is, "off topic" because dogs don't run on electricity.

 

dogs don't run on electricity.

I'll bet they actually run quite fast, with a cattle prod stuck up their.......

That would be a dog running on electricity

 

<snip>

Out of interest, does anyone know of a relatively cheap recording / playback device which I can use to record the ultrasonic signature of a passing aircraft to use for testing my device.

you can check out Sensory Inc, they have a line of Voice recognition devices that you program the "sound" that activates a different function on the device (look for the Voice Direct 64), possibly you can record some "real time" sounds where you need to watch out for the jets and then use that to trigger an IO when the same sound is heard....

My .02
B. Morse