🚀 go-pugleaf

Chad Edward Albrecht (albrecht@csd4.csd.uwm.edu) wrote:
: 1)    ...I need to measure distances within .001 of an inch ... It needs to
:       be a fairly long wavelength, so as to not give me cancer! 
: I will ... have each transmitter send a slightly different frequency
: and the same receiver distinghush between them.

I'm not sure exactly what you're after, but I think your expectations are
rather higher than you will be able to achieve easily. Why have you decided on
such a high level of accuracy? You can't hold your hand still to anything like
that precision, so why measure it?  Assuming you are working in a coordinate
system, 20 feet on a side, you'll be working in a space 240000 pixels on a
side, or a volume of 1.38x10^16 points! I hope you've got some nifty hardware!
That's a LOT of points to recalculate every frame!

Also I think you're wanting to use RF(Radio Frequency) signals for the
measurements. This is tricky stuff to set up, and extracting several
close-packed frequencies will be a nightmare!

I think a much more realisable scheme would be based on ultrasound. You then
require one transmitter, and four recievers to fix the transmitter's position
in an x,y,z coordinate system. Generate an ultrasonic chirp, and time the delay
before each reciever gets it. The propagation speed of sound in air is *MUCH*
slower than RF, so you should be able to do this with a micro (but watch out
for return signals from two detectors arriving very close together!!)

The delay varies with distance, and sound travels at 330m/s in air. That means
33 cm/ms or 3.3 mm every 10 microseconds. I really believe that an accuracy of
5 mm will be PLENTY! This assumes that you can check for the return signals
every 10 microseconds, which is not difficult on a decent PC. 

I don't know how you plan to measure the orientation of the transmitter point -
I really don't have very good suggestions here. A joystick-type affair would
give you pitch and roll information, but it isn't much help with rotation about
an axis perpendicular to the floor! Sorry I can't be more helpful!

I would only take a position reading about twice a second, because you'll need
the CPU time to triangulate the position, and redraw the screen. THIS WILL TAKE
TIME! 

In summary, I would advise you to think VERY carefully about what you NEED to
get out of this system. If you have limited time for this project, make sure
you have achievable goals, or at least realistic ones! Your assessors are more
concerned with the way you tackle the problem than its eventual solution! Don't
try to build a system that does everything. Spend your time on getting a simple
system that works well, and approach it sensibly. You will reap the rewards in
a much better assessment than if you pack in all the features you can think of,
don't quite get it working, and the whole system is a mass of patches and bugs
waiting to get out.

Hope this is of some help. Sorry I can't E-mail! All the best for a very
interesting and challenging project!
--
*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*
+   Michael Thomson    + The opinions expressed here, while definitely NOT  +
+                      + those of my employer, may OR MAY NOT be mine!      +
+  BNR Europe Limited  + --------------------------==========-------------- +
+ Maidenhead,  England +        This space intentionally left blank         +
*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*=*