[RFI] Ultrasonic Dish

[N1BUG]

On 01/02/2013 12:32 PM, Lee K7TJR wrote:
>> Virtually anything that is roughly the desired shape will work
>
>     I have to take issue with this statement as I spent several weeks a couple
>   of years ago building an ultrasonic detector. I too thought that a snow dish
> would work, however when compared to a real parabolic curve it was
> nearly useless. YMMV

I agree 100% with Lee. A parabola is a very specific curve derived 
from a mathematical formula. "Close" shapes may work to some extent 
but will almost certainly be very inefficient compared to a proper 
parabolic surface of the same size. Even the best of ultrasonic 
detectors is none too sensitive for this application, so anyone 
serious about using them for power line noise source verification 
would be well advised to optimize in every way possible. Before 
using any object as a dish I would suggest taking at least 10 
*accurate* measurements of the surface along a radial from edge to 
center and verifying that it is a parabola using the proper 
formula**. Also, as I said earlier, to live up to its full potential 
it needs to be the proper depth or flatness of parabolic surface for 
the characteristics of the detector being used. Guessing games will 
likely produce perfectly horrible performance in *most* cases.

** a possibly easier alternative would be to measure the dish depth, 
then plot/cut a parabola out of stiff paper, cardboard, or whatever 
and then place that in the dish to compare the curve.

73,
Paul N1BUG