Re: TI-H: Scanning Sonar

[Prev][Next][Index][Thread]

Re: TI-H: Scanning Sonar

Cool! I was doing someting like this last year but I never actually 
built the hardware because I didn't have a TI at that time. I was going 
to use my PC, but I couldn't program the graphics in C or Pascal so I 
quit... anyway, my idea was kinda similar to yours. here it is:

---
1. calibrate period (especially necessary with TI's) by putting an 
object in front of the sonar and allowing it to turn, detecting the 
object twice and timing that.

2. calibrate speed of sound (changes w/temp)

3. Now on the calcs' screen, there is a rotating arm. From step 1, its' 
position is set to always match that of the sonar. The sonar normally 
will get no response until there is something in the way (and in range). 
It will send a signal to the calc whenever it sends a pulse, resetting 
the calcs' timer each time. If it receives an echo pulse it will send a 
different signal to the calc. The calc will, at the current location of 
the rotating arm on the screen, place a dot in correspondence with the 
time elapsed between sending the pulsse and receiving an echo.

Maybe something like a PIC will be needed to to the timings?
---

Or maybe I am just overcomplicating a simpler thing? Is it meant to be 
like radar or is it meant for to tell the distance between an object and 
the itself?


Mj


>
>> if youre interested, my friend and i designed an eight bit parrallel 
i/o
>> system a while ago.  we never actually drew up the schem, but we got 
all
>> the design in our heads (chips numbers included).
>>
>> we never built one because we don't have a use for it yet.  the 
initial
>> idea was for motion control but other things have taken up time 
since.
>> also, we knew how the algorithms would work but never commited them 
to
>> asm.  those are still in suedo-code waiting to be written into a 
driver.
>>
>> where does the data end up when the sonar is done with it?
>>
>> a sonar sounds like an awsome idea; how is it done?
>>
>> Matthew
>
>I'm interested if you could put it down.  The program would just 
translate
>the time it tookfor the sound to bounce back and map it on the calc.
>Depending on the speed and distance before the sound won't return. I'd 
do
>something like 1 pixel = 2 meters or so.
>IF you wanted it to scan (Turn) the RPM of the actual motor comes into 
the
>program.
>
>Here's what I think the output might look like.
>
>
>
>
>--XXXXXXXXXXXXXXXXX
>----XXXXXXXXXXXXXXXX
>---------------XXXXXXXXXX
>------------------------XXX----
>---------------------------------
>---------------------------------
>---------------------------------
>
>


______________________________________________________
Get Your Private, Free Email at http://www.hotmail.com
Follow-Ups: