On 23/10/2010 19:14, PICdude wrote:
> I should've mentioned that directional is a problem.  6" would be
> nice, but 1-ft should would fine also.  Here's the scenario --
> quad-copter in a (large) room with some "obstacles" (partitions, etc
> to form a maze).  We'd like to determine the position of the
> quad-copter and guide it around the maze (which the "base" will know
> about).  We can use two such sensors with an altimeter or three sensors.
>
> Well have GPS on board, but that won't work in a room.  We'll have
> bluetooth (or some other RF transciever) on boad, and it would've been
> nice to use that, but that does not seem feasible, unless there's some
> analog way to determine the slight shift in the returned signal.
>
> Cheers,
> -Neil.
Ultrasound pulses is really best for 6"

Night and I'm tired, so excuse any mistakes...

you are not limited to 40KHz or large transducers..
A miniature electret can work to 70kHz
Some small ceramic "beepers"/Speakers can be resonated at various=20
ultrasound frequencies above 50KHz.

There are miniature 40kHz transducers too

You do need to short the input when you send the pulse or an analogue=20
filter on the microphone  can take too long to recover. !

70kHz is about 4mm wavelength so you can make small light weight horns.=20
The RF equivalent is 7GHz. In practice 10Ghz band is closest with cheap=20
devices and licence free SRD spot frequencies. But the Ultrasound is=20
simpler, cheaper and lower power and needs no "RF SRD" approval. The=20
range isn't great (< 20m).

You need pulses as then the receiver gain can be high. Any continuous=20
wave system looking at phase shift tends to be deaf unless it's > 400GHz=20
with very directional transmitter.

A bat actually has a flap that momentarily closes its ear when it emits=20
the "click".

You measure time between sending pulse and hearing it.

for 50mm minimum distance  using sound round trip is 100mm. Speed is=20
300m/s so time is 50/300 ms, approx 500/3 us =3D 166us, pulse wants to be=20
about 80us
but... 70kHz that's only about 6 cycles... Not much for a filter. 12=20
cycles for 160us Pulse, limit is 100mm (about 4"). Or 18 cycles approx=20
at 150mm / 6" about 240us pulse

RF 10GHz?
Speed is million times more. So you need 240 pico second pulse for=20
minimum 6" and about 20 pico second measurement accuracy is about 0.5" I=20
think
BUT 1 cycle is 1 nano second at 10GHz... see the problem? At 18 cycles=20
you are at 18 nanoseconds., about 500"  Even 10GHz RF is too low a=20
frequency for 6" minimum distance I think.

Car parking sensors are all Ultrasound I think.





--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.