Rick Dickinson wrote:
>
> I've seen quite a large number of postings by people looking for sonar devices
>  for robotics use.  The commonly suggested solution (the Polaroid sonar
 module)
>  has several drawbacks, however: 1) it's expensive ($50 each or so), 2) it
 needs
>  modifications to measure short distances, 3) the transducer used is fairly
>  large (1.25" or so diameter), and 4) your processor gets tied up calulating
 the
>  range by timing the reflected chirp.
>
> Owing to the success of the various serial LCD "backpack"-type devices, I
>  thought it would be useful to develop a low-cost kit for an ultrasonic
 ranging
>  unit, that could be accessed using some simple serial protocol.  I envision
>  this "Serial Sonar Unit" (SSU) being a small PIC-based PC board, with
>  connections to a pair of Murata 40 KHz transducers like the ones sold by All
>  Electronics for $2.50 a pair.
>
> The SSU would have two modes: continuous reading, and on-command.  If set to
>  continuous mode, the SSU would continually send out, and time the echos of,
>  tone bursts, storing the most recent result internally.  When polled, it
 would
>  return the most recent reading.  In on-command mode, the SSU would wait for a
>  "go" command of some sort, then send a pulse and time its return.  If polled
>  before the measurement was complete, it would return some sort of "busy"
>  indication.  If polled after the measurement was complete, it would return
 the
>  measurement.
>
> My intent is to come up with a design that will save my fellow hobbyists some
>  time and money, and will make it easy to incorporate sonar ranging into PIC
 and
>  other small microcontroller-based projects.  Offloading the timing,
 transducer
>  drive, and calculation tasks to a dedicated SSU will allow the programmer to
>  concentrate on the tasks to be done, not the ranging implementation.
 Allowing
>  single-shot or continuous ranging will allow simpler implementations to use
 the
>  continuous mode, and multiple-unit configurations to use single-shot mode to
>  avoid interference between transducer pairs.
>
> Details to be ironed out include:
>   -  Protocol for commands, including mode changes
>   -  Data format for measurement data returned
>        Inches?
>        Milli/microseconds?
>        Binary or BCD?
>   -  I2C, SPI, or none of the above
>   -  Baud rate/clocking (async 9600/4800/2400/1200? Autobaud? Sync?)
>   -  Distribution mechanism
>        PIC code/Schematics/PCB layout on web for download?
>        Preprogrammed PICs and etched PCBs for purchase?
>        Both?
>
> I am just at the "fleshing out the idea" stage, and I would appreciate any
>  feedback as to whether or not anyone would be interested in such a device.
>  Also, I want the SSU to be as useful as possible, so I would love to hear
>  everyone's input on the implementation details I outlined above.  I figure
 that
>  the discussion would be most valuable if everyone could participate, so
 please
>  send replies to the PICLIST instead of directly to me.
>
> Thanks in advance,
>
>  - Rick Dickinson

Such a product already exist, and the cherry on top of the cake is
that it's pic based! (a basic stamp chip) It's called RadStamp, has a
range of 0.5 to 12m with a resolution of +/-10mm. The US sensor is
mounted on a stepper motor to make measures on 360 degrees. The device
has a RS232 output. It costs around 60$ but you need to add a Basic
Stamp2 to make it work. I guess it should be possible to replace the
BS2 with a "Pure" PIC chip.

It's distributed in UK by Milford Instruments.

Anyway, it will be fun to develop such a device by ourselves!

Cheers

Gael