project idea realisation?

stepfresh

Hey guys,

I have an idea for a project but im not sure of it's plausibility, and i was wondering if anyone could provide some enlightenment.

I would like to have several seperate units, which can be oriented in any way (i.e. not necessarily facing eachother), recieve an send signals to eachother inaudibly. I would like one unit to begin by sending a signal to the world around, when another unit recieves that signal, it derives the distance it is from the sending object, waits for a time relative to the distance, then sends it's own signal. Whenever a unit recieves a signal and after the delay, it will make a sound.

How possible is this and what is the best medium to use. Infra-red is out of the question to to the dis-orientation, and audible sound due to the requirement of surrounding silence. Would it be possible to send a timestamp with the sonar which the recipient can calculate (using the speed of sound) the distance? Could it also send an identifier to recognise which unit sent the sonar?

Thanks very much for any assistance, it will be very appreciated.
- Dan

Istha Powron

Dan,

Unless I am mistaken, there's no way that a unit will be able to determine the time taken for the signal to reach it unless it knows when it was sent or both sender and receiver are working from the same time base.

One way you could do it would be to send a visual pulse at the same time as you send the audio pulse. Lightspeed is almost instantaneous, therefore the receiver could start counting as soon as it sees the light.

Issy

stepfresh

Thanks for your reply, I was thinking i would include a timer on each unit and would globally reset them all before starting.

edit: And i was hoping to be able to communicate the time at which the signal was sent using the sonar signal. So i would know the time sent, time arrived, and speed of sound.

Post Edited (stepfresh) : 10/2/2006 1:48:27 AM GMT

allanlane5

This kind of very high-resolution, high-accuracy timing of things is pretty hard to do with the BS2. The BS2 takes about 300 uSec per instruction, and I'm sure it will take several instructions to implement this.

Now, the BS2 CAN measure pulse-widths down to 2 uSec resolution, but for that to work some 'local' to the BS2 device (like the PING sensor) must generate both the START of the pulse and the END of the pulse.

Tracy Allen

We've used ultrasonic tags for the purpose of detecting the presence of items in a room. There is a pulse code, but the bit rate has to be very slow (~0.1 second per bit). That is due to the random echos off the walls and objects in the room, which at the speed of sound can take quite a long time to die out, and you do not want to mistake them for a new bit.

There is quiite a bit of interest in animal behavior circle to have systems that can locate the direction of a sound and with two more more receivers triangulate. They are also using that in cities to pin point gun shots.

Maybe you could do something with an RF back channel that links all of your devices, so that all of them would know the instant that any of the others receives a sound pulse, so each could calculate its relative distance to the source.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com