T&L Publications

I am interested in building a small parabolic microphone. I've seen plans in electronic mags before, and possibly in Nuts & Volts.<p>Anyone have any good ideas?

Buy one of those turned stainless steel bowls from a department or home goods store. To find the focal point, hold the bowl toward the sun and probe the space in front of it with a small piece of wood or other material. If the focus is any good, the wood will start to burn at the focal point, but even if it doesn't it should still work. Make a frame to hold the microphone at the focal point, pointing toward the bowl, and wire it to an amplifier.

Hi there,<p>In addition to the other suggestions already
given, you might try buying a cheap 6 volt
lantern battery flashlight, and take out the
reflector and use that. These are usually
already in the shape of a paraboloid and should
have a pretty tight focal point so would work
pretty good for a small directional mic.
You would place the mic where the bulb filament
would normally be.<p>--Al

Even with the much better focus, I wonder if a small 5" diameter reflector would have enough sound-gathering ability to be effective, Al. But it would be compact, wouldn't it? There have been at least three articles that I know of in the past. Two used the stainless steel bowl idea, one I think in an ancient issue of Radio/TV Experimenter and a much newer one in the last two or three years in Nuts & Volts. Popular Electronics had the "Shotgun Sound Snooper" back in the 1960's that used a collection of resonant aluminum tubes in a "shotgun" format. I'm told by a guy that built it that it had great sensitivity and good directional capability, but did suffer from "organ pipe" effects in a good crosswind. I don't have that particular N&V issue and I've just packed away all my other magazines in anticipation of a move, so can't get to those particular issues.<p>Dean

How "small" do you want it? There is quite an easy way to go about this. Build or buy a small audio amplifier. Or better yet get the "super ear" from Radio Shack or other store. These are the size of a personal radio but have a built in speaker to amplify distant sounds to the common personal stereo type headsets. They have sold these on TV as a personal hearing device for those that are hard of hearing. Get one of these and mount it at the focal point of one of the DBS Mini Dishes.

Bought a toy one, for about $5
marked down from something like $25..
Slammed in a 9V and wow, was quite
impressed at the directional gain,
cames with headphones.
Never underestimate the toy department!
Regards,
Scott

Satellite dishes usedto be parabolic, The questions are, Is 18" compact ENOUGH, Are the new style mini dishes still perfectly parabolic(I'm almost sure), And can you find one for $5 somewhere?<p> They are by god heavy though. Maybe you could take a sander or grinder and make it half as thick to reduce weight.

Edmund Scientific sells parabolic reflectors, here is a link to my first search under "parabolic".
http://www.scientificsonline.com/Produc ... =parabolic+<p>Satellite dishes should still be parabolic, or fairly close to it unless they've really cut corners and decided to go spherical like telescope mirrors, which in my thinking, wouldn't make any sense as they are simply stamped out. However, it really doesn't matter whether it is truly parabolic or spherical, even on telescopes, there is only a slight skewing of colors with the final result.<p>Douglas<p>[ August 18, 2002: Message edited by: Isenbergdoug ]</p>

Were dealing with a parabola, a lens. <p>No matter which lens you use, a salad bowl, or a 18 inch satellite lens, you will need Focous it onto any Microphone.. <p>Use a cheap dial type slide tube unit off any cheap telescope unit to mount your Microphone so that where ever the Microphone needs to be, you can slide it in or out to accommodate the focal ratio.<p>Satellite dishes, are pre focused for 25 thousand miles. <p>Making it completely adjustable, no matter what your distance of sound means you can Focous that sound onto you pick up Microphone at the best possibility.<p>Second, use "Fet" technology amp/pick up..... for your sound pick up due to the signal to noise ratio or response. A.K.A., quiet.

Aren't the 18" dishes somewhat oval in shape?<p>And b.t.w. I was thinking that you may get a great deal of gain by attempting to filter out other noises. An adaptive filter is the most effective approach I know of when the noise is measurable, but cannot be characterized.<p>With a parabolic dish, you could mount microphones in an array around the center one, with the purpose of being slightly off-axis. These would pick up unwanted noises in the vicinity of the sound of interest. These signals can be combined, correlated against, and filtered out of the primary signal with a discrete Wiener filter.<p>To implement one requires the use of a FIR filter whose coefficients are based on the Widrow-Hopf Least Mean Square (LMS) algorithm.<p>(1) Initially, set each weight w_k(i),i=0,1,2...N-1, to an arbitrary value, such as 0.<p>For each subsequent sampling instant, k=1,2..., carry out steps (2) and (3) below:<p>(2) compute filter output<p>
<p>(3) update the filter coefficients<p><p>where e_k is the output of the adaptive filter, w_k is the set of filter coefficients, x_i is the input to the filter, and mu is a convergence coefficient that determines the rate of convergence on the noisy signal (ie. how fast the filter adapts to the changes in the characteristics of the noise)<p>[ August 23, 2002: Message edited by: bobsRAC ]</p>

I prefer a 7 or 10 band graphic equalizer for a car stereo system and clean low noise amp as well, so you don't have to build all those filters. <p>They also come in a very neat package and all items are 12 volts. <p>The oval dish is not a problem, its for polarization purposes but doesn't affect the sound gathering that much.

The point of an adaptive filter is that it automatically adjusts the filter parameters to filter out the portion of the measured signal that correlates with the noise signal.<p>This is often desirable over a static filter because it can filter out many types of noise, as well as noise that may exist in overlapping bands with the signal of interest.<p>A graphic EQ approach has a disadvantage if the signal of interest coincides with an unwanted signal in the frequency domain. The EQ can be set up to either filter out part of the signal of interest, or to leave unfiltered components of the unwanted signal, or worse yet, both. <p>Furthermore, the EQ must be retuned for each type of unwanted signal. For instance, if you are listening to a bird on some tree, and another (that you don't want to hear) is nearby, the frequency bands may be overlapping. After you've got your EQ tuned, another bird flys nearby and you have to start over.

But in the real world, not many unwanted noises really bother you that way.<p> My unit was tested on a soccer field at over a thousand yards away, and tuned with a simple slide of one of the ten slides, for each child's voice, and out side noise blocking with the other slides. <p>My only problem was the unexpected scream from one child which blew out my ears because I was using a 100 watt amp into head phones? <p>That should give you an idea of what sort of filtering you can achieve with simpler equipment.<p>The Harley and truck noise of the highway was cut out with the 3 lower band filter slides, while voices were crystal clear, and the over head Jets were also tunable with the upper band slides. <p>Your way is a little complex as much as it is better, while these days 100 bucks will buy all of the above in Fet technology? <p> I tend to concentrate on the pick up end, and customize that for pennies. Circular pattern piezo crystal mikes, times ten, completely focus-able, worked for me, because the area once tuned or focoused to, rarely has more than one noise within the 100 or so square feet area. Think focous, and lens.

What sort of pattern do you have your mics arranged in? I see that you said "circular" but... can you post an image describing the array?<p>I see that you suggested using a telescope focuser, but without having seen the array f mics, I'm having trouble picturing your scheme.<p>b.t.w. -- Why are you using a 100W amp for headphones?