Honcho's Blog

HOOORRRRAAAAYYYY, So after modeling the heck out of envelope detectors I finally was able to build the darn thing. So I clapped the first image is of the audio input, and the envelope output together. Yellow is the audio, blue is the envelope.


The second image is just the envelope output.


The last is of the audio input

What I was going for here was to 'normalize' the audio input so that it would be a smoother, easier to detect signal, which would trip the interrupt. Now that I have built this, I need to make a more sensitive mic. (the above signal was from a clap). So now I am going to dabble in auto gain control, and a log amplifier. Hopefully Friday will be a fruitful blog post. until then.

I am currently designing an envelope detector circuit. What it will do is kind of average out the audio signal and make it easier to analysis the signal.

Above is the circuit, the diode makes it so that only a positive signal comes out, the capacitor takes out some of the AC signal while the resistor controls how the signal decays. Below is the magnitude response.
So hopefully soon Ill have some circuit board to show you. :)

Well after a rough start I feel as if I have the right questions to find a direction in which will lead to victory. On Tuesday I came in and built a very simple comparator circuit, using the below circuit, which I found onhttp://ashishrd.blogspot.com/ So thanks Ashish.

The first version looked like:	This is what my final product looks like:
(The big blurry thing is the mic)

The circuit makes a LED illuminate when a clap or snap occurs. Problems that I have encountered where that when the potentiometer (RV1) let a voltage through of 2.3V-2.6V through a feedback loop would be created and the LED would constantly blink due to the noise of the microphone. With some tweaking, I got that problem fixed. But other than that and removing the short on the potentiometer (VR1) the circuit works fine. I have larger issues than this at the moment.

The team wants a scream or gunshot to activate or trip some rules in the system. My current problems are, this circuit can barely hear a scream from the other side of the room. When we probed the mic directly the amplitude of Stephen talking normal, close to the microphone and me screaming were basically the same. Our initial solution was to build a high pass filter to get rid of “normal” talking frequencies. This sounded like a good idea initially, then realized that just because when I think of screaming I think of a female screaming (Higher pitched) doesn't mean a guy screaming will be a higher pitch, it will just be really loud. So now I need to figure out a way to compare decibel levels in order to trip the system.