Exact match. Not showing close matches.
'[EE] Separating signals'
On Wed, 2 Dec 2009 17:51:34 -0200, Bruno L. Albrecht wrote:
> I'm trying to separate 3 frequency components from an analog signal. The
> original signal is composed by the sum of 3 frequencies (1500, 4000 and
> 10200 Hz) and some noise. The objective is to separate only those 3
> frequencies and evaluate their magnitudes. I thought about lots of ways of
> doing it, but couldn't get an effective one.
> Any ideas?
If there were two tones, perhaps a PLL. But I think this would be a nice
application for the Goertzel algorithm if you're willing to use DSP. The
dsPICs would easily be able to do this - in fact, I think wikipedia has a
C implementation as well as a description (for DTMF decoding). I don't know
how the majority of the maths work, but the implementation is straightforward
enough to follow. I suppose your solution will also depend on how much noise
you have/anticipate (and the frequency if it's from a known source; aliasing
may be a problem), but sounds like a good application for Goertzel anyway.
Would certainly be an interesting project.
Bruno L. Albrecht
I'm thinking about using a MC1496 (Balanced Modulator / Demodulator). It
seems to fit my needs, once I can sum the carrier to an on/off switch, and
then just demodulate the signal with the same carrier. And all of it with
only two cheap IC per frequency!
What are your opinions?
On Thu, Dec 3, 2009 at 4:39 PM, Peter Restall <restall.net> wrote: pete
I was amazed that MC1496s were still being sold at DigiKey ( ON
I had a lot of fun with them in the past. Using the MC1496 seems
harder than using
a switched capacitor filter. Do you have the 3 frequencies available ?
Bruno L. Albrecht [GMAIL]
yes, I actually have complete control over everything. My idea is to
drive 3 LEDs in diferent frequencies and then measure them with a
photodiode. The photodiode response would be a DC level for ambient
light plus the 3 frequencies of the LEDs. So I would cut off the DC
voltage and measure the amplitude of each frequency. These amplitudes
would be how much the LED is affecting the photodiode.
Bruno L. Albrecht
Eng. da Computação/06
Falker Automação Agrícola Ltda.
YES NOPE9 escreveu:
More... (looser matching)
- Last day of these posts
- In 2009
, 2010 only
- New search...