LM741-Is there an easier or better way?

[Bose(o)]

Hello all.

I have a quick question here. I have built and tested a microphone circuit which connects to an 8W pre-amp and then is further amplified by a car-amp that I have. Everything is fine here, no worries . Though, my question includes taking an audio signal and biasing it as positive. In other words, taking a sine wave and it's negative counterpart then flipping it such that I will have a cusp-type wave. This will be used with my LM339 comparator such that an LED will turn on from the output of such.

[azira]

Use 4 diodes in a bridge-rectifier configuration. The mechanism you're describing is a called "full wave rectification." You might want a large size capacitor and/or a schmidt trigger configured opamp to help smooth out the signal so that your LED doesn't turn on and off constantly.

Rod Elliot's website has a project for a signal detector (for like subwoofers). It's similar to what you want.
--
Danny

[Bose(o)]

I have many LN5408 diodes, would those do? I'm thinking they may be too slow. I really don't mind about quality or anything of the like, I just have to detect whether there is a change in SPL or not. I looke up full-wave rectification, and it is perfect! I knew there was a proper name for this method. The capacitor you mentioned, would I connect that before the LED? Seeing as it would "hold" the power that the LM339 would send.

[jackinnj]

you're trying to build a VOX -- here are some things to take into consideration:

-- buffer the sound input with an opamp -- then feed into a second opamp configured as a comparator * With Hysterisis * -- thus the opamp will "hang" on while the hysterisis comparator remains charged above the threshold level of the comparator.

i wouldn't be concerned about rectifying since the eye "integrates" anyway.

[Bose(o)]

I hope this isn't too much to ask, but do you have a schematic? (jack)

[azira]

Originally posted by Bose(o)
I have many LN5408 diodes, would those do? I'm thinking they may be too slow. I really don't mind about quality or anything of the like, I just have to detect whether there is a change in SPL or not. I looke up full-wave rectification, and it is perfect! I knew there was a proper name for this method. The capacitor you mentioned, would I connect that before the LED? Seeing as it would "hold" the power that the LM339 would send.

Hmm... if you want a SPL meter (or a VU meter), then there is a chip that would do it for you....

Anyway, so it's kind of like building a power supply. What you do is take the input, run it through the 4 diodes (those would be fine i think, no need for fast ones) and then into the cap to gnd. The cap will "hold" the DC level of the signal and reduce ripple. The DC level should change along with the RMS level of the signal... What you do with this signal is up to you..

One more thing, this will only work for signals that are over 1.4V (2 diode drops). So in an 8-ohm environment, it'll only output a signal if there is more than 175mW RMS of power.
--
Danny

[gmphadte]

Go to the site www.national.com and search for LM3915
This IC will do a amuch better job with less components and less effort.

Enjoy

Gajanan Phadte

[peranders]

Some of the application notes have also good examples of fullwave rectifiers and peak detection.

http://www.national.com/an/AN/AN-20.pdf#page=10

page 10 in the pdf

http://www.national.com/ds.cgi/LF/LF155.pdf

page 17, a classic. I have it in my old VU-meter from the stone age. The VU-meter consists of 80 opamps (= 40 dB, 1 dB resolution), home made voltage dividers, totally unprotected CMOS40?? (was it 4010?) from the mid 70's.