Digital Audio Spectrum Analyzer
Archiver
Posts: 46,084
Hello Stampers
Can someone PLEASE help me with an idea that I have. I am trying to build a
Digital audio spectrum analyzer. I was going to use 31 Op-amp circuits for the
31 bands then use 31 A/D converters to the BS2p then to my display, but that is
allot of parts.
I would like to use an audio A/D converter to the BS2p. The stamp would do the
math for the readings from the A/D. But I have no idea on how the math would
work. I mean how can a digital signal turn into the amplitude for each
frequency?
Any insight would help. Is there any books, sites, ect. that would point me in
the right direction?
Thanks
A. Conti
[noparse][[/noparse]Non-text portions of this message have been removed]
Can someone PLEASE help me with an idea that I have. I am trying to build a
Digital audio spectrum analyzer. I was going to use 31 Op-amp circuits for the
31 bands then use 31 A/D converters to the BS2p then to my display, but that is
allot of parts.
I would like to use an audio A/D converter to the BS2p. The stamp would do the
math for the readings from the A/D. But I have no idea on how the math would
work. I mean how can a digital signal turn into the amplitude for each
frequency?
Any insight would help. Is there any books, sites, ect. that would point me in
the right direction?
Thanks
A. Conti
[noparse][[/noparse]Non-text portions of this message have been removed]
Comments
channel A/D chips. For example the LTC1298 is a 2 channel, 12 bit A/D that
uses three Stamp pins. There are other chips with 4 channels per package
and I am sure a little reasearch would uncover higher density packaging
still.
If I understand your question you can easily build bandpass filters for
the op amps to measure the amplitude at as many frequencies as you like.
There are several good books on the subject of opamps that will tell you how
to design the filters. One book that is quite good is the "OpAmp Cookbook".
Any good electronics store should have these types of books. Also you might
look for a copy of the ARRL handbook at a used book store or a Ham radio
store.
MIke B.
Original Message
From: "Anthony Conti" <aconti@s...>
To: <basicstamps@yahoogroups.com>
Sent: Monday, September 02, 2002 8:31 AM
Subject: [noparse][[/noparse]basicstamps] Digital Audio Spectrum Analyzer
> Hello Stampers
>
> Can someone PLEASE help me with an idea that I have. I am trying to build
a Digital audio spectrum analyzer. I was going to use 31 Op-amp circuits for
the 31 bands then use 31 A/D converters to the BS2p then to my display, but
that is allot of parts.
>
> I would like to use an audio A/D converter to the BS2p. The stamp would do
the math for the readings from the A/D. But I have no idea on how the math
would work. I mean how can a digital signal turn into the amplitude for each
frequency?
>
> Any insight would help. Is there any books, sites, ect. that would point
me in the right direction?
>
> Thanks
> A. Conti
>
>
> [noparse][[/noparse]Non-text portions of this message have been removed]
>
>
> To UNSUBSCRIBE, just send mail to:
> basicstamps-unsubscribe@yahoogroups.com
> from the same email address that you subscribed. Text in the Subject and
Body of the message will be ignored.
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
looking to combine it into something much smaller maybe two A/D converters and
with the right sampling I would find the amplitude for each frequency. But I do
not know where to start.
Thanks
A. Conti
First of all you can dramatically reduce the parts cound by using multi
channel A/D chips. For example the LTC1298 is a 2 channel, 12 bit A/D that
uses three Stamp pins. There are other chips with 4 channels per package
and I am sure a little reasearch would uncover higher density packaging
still.
If I understand your question you can easily build bandpass filters for
the op amps to measure the amplitude at as many frequencies as you like.
There are several good books on the subject of opamps that will tell you how
to design the filters. One book that is quite good is the "OpAmp Cookbook".
Any good electronics store should have these types of books. Also you might
look for a copy of the ARRL handbook at a used book store or a Ham radio
store.
MIke B.
Original Message
From: "Anthony Conti" <aconti@s...>
To: <basicstamps@yahoogroups.com>
Sent: Monday, September 02, 2002 8:31 AM
Subject: [noparse][[/noparse]basicstamps] Digital Audio Spectrum Analyzer
[noparse][[/noparse]Non-text portions of this message have been removed]
A digital signal can be digitized with a single A/D converter, and then
processed to produce the amplitude for each frequency by using a
software Discrete Fourier Transform (DFT) (also loosely called Fast
Fourier Transform (FFT). This is a tall order for a BS2, because it
requires at least double-precision fixed point, or ideally floating
point, calculations. Maybe one of Al Williams's clever PAK
co-processors could provide enough computational power, but the usual
way to approach this is with a signal processing chip. I do recall some
assembly routines for performing FFTs with PICs on the Microchip
website, so it is probably possible. To conserve memory, certain DFT
algorithms perform the transform _in place_, without requiring multiple
arrays. The _spectrum_ can be regarded approximately as the square of
the DFT output, plotted versus frequency.
A good place to start for understanding the theory and mechanics is to
browse through the many digital signal processing books in a bookstore,
and choose one with good examples that appears readable.
Another great method is to find a friend with Matlab, or other
signal-based programs, and go through the signal processing demos.
Dennis
Original Message
From: Anthony Conti [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=fvmG9kdK-T4BLJeFDH4XlCBdI8hFipUQdA0SoNj1ykMkKDQxvw4FTle-VF_uRkgBg3oDD7zowK-ycu3J]aconti@s...[/url
Sent: Monday, September 02, 2002 8:32 AM
To: basicstamps@yahoogroups.com
Subject: [noparse][[/noparse]basicstamps] Digital Audio Spectrum Analyzer
Hello Stampers
Can someone PLEASE help me with an idea that I have. I am trying to
build a Digital audio spectrum analyzer. I was going to use 31 Op-amp
circuits for the 31 bands then use 31 A/D converters to the BS2p then to
my display, but that is allot of parts.
I would like to use an audio A/D converter to the BS2p. The stamp would
do the math for the readings from the A/D. But I have no idea on how the
math would work. I mean how can a digital signal turn into the amplitude
for each frequency?
Any insight would help. Is there any books, sites, ect. that would point
me in the right direction?
Thanks
A. Conti
[noparse][[/noparse]Non-text portions of this message have been removed]
To UNSUBSCRIBE, just send mail to:
basicstamps-unsubscribe@yahoogroups.com
from the same email address that you subscribed. Text in the Subject
and Body of the message will be ignored.
Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
relatively simple R/C circuits. I don't think you could do FFT on a stamp.
On Mon, 2 Sep 2002, Dennis P. O'Leary wrote:
> Anthony,
>
> A digital signal can be digitized with a single A/D converter, and then
> processed to produce the amplitude for each frequency by using a
> software Discrete Fourier Transform (DFT) (also loosely called Fast
> Fourier Transform (FFT). This is a tall order for a BS2, because it
> requires at least double-precision fixed point, or ideally floating
> point, calculations. Maybe one of Al Williams's clever PAK
> co-processors could provide enough computational power, but the usual
> way to approach this is with a signal processing chip. I do recall some
> assembly routines for performing FFTs with PICs on the Microchip
> website, so it is probably possible. To conserve memory, certain DFT
> algorithms perform the transform _in place_, without requiring multiple
> arrays. The _spectrum_ can be regarded approximately as the square of
> the DFT output, plotted versus frequency.
>
> A good place to start for understanding the theory and mechanics is to
> browse through the many digital signal processing books in a bookstore,
> and choose one with good examples that appears readable.
>
> Another great method is to find a friend with Matlab, or other
> signal-based programs, and go through the signal processing demos.
>
> Dennis
>
>
Original Message
> From: Anthony Conti [noparse]/noparse]mailto:[url=http://forums.parallaxinc.com/group/basicstamps/post?postID=NpGC3zk3Ma5z7mcnXl2iFuvEUxrUErzbeLFsibml9WOUZB8s4QGyCLS-By3IhcXFX2hgiiXQNet9S4TOOjg]aconti@s...[/url
> Sent: Monday, September 02, 2002 8:32 AM
> To: basicstamps@yahoogroups.com
> Subject: [noparse][[/noparse]basicstamps] Digital Audio Spectrum Analyzer
>
>
> Hello Stampers
>
> Can someone PLEASE help me with an idea that I have. I am trying to
> build a Digital audio spectrum analyzer. I was going to use 31 Op-amp
> circuits for the 31 bands then use 31 A/D converters to the BS2p then to
> my display, but that is allot of parts.
>
> I would like to use an audio A/D converter to the BS2p. The stamp would
> do the math for the readings from the A/D. But I have no idea on how the
> math would work. I mean how can a digital signal turn into the amplitude
> for each frequency?
>
> Any insight would help. Is there any books, sites, ect. that would point
> me in the right direction?
>
> Thanks
> A. Conti
>
>
> [noparse][[/noparse]Non-text portions of this message have been removed]
>
>
> To UNSUBSCRIBE, just send mail to:
> basicstamps-unsubscribe@yahoogroups.com
> from the same email address that you subscribed. Text in the Subject
> and Body of the message will be ignored.
>
>
> Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
>
>
>
>
>
> To UNSUBSCRIBE, just send mail to:
> basicstamps-unsubscribe@yahoogroups.com
> from the same email address that you subscribed. Text in the Subject and Body
of the message will be ignored.
>
>
> Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
>
>
>
Sean T. Lamont, CTO / Chief NetNerd, Abstract Software, Inc. (ServNet)
Seattle - Bellingham - Vancouver - Portland - Everett - Tacoma - Bremerton
email: lamont@a... WWW: http://www.serv.net
"Do not fear mistakes, There Are None" - Miles Davis
programs that use the standard PC sound card as an audio specrum
analyzer. Up to 16 thousand FFT sample points with a decent sample
rate can be had for the cost of a stamp.
Stamps have there place but FFT analysis is a little out of its
league.
If you want to program in C you could use a TMS series DSP from TI
and use a stamp to control the DSP. A more relastic solution if you
need a custom analyzer. Development tools for DSP's are not cheap
though, can cost in the thousnads of dollars to get setup.
Jason
--- In basicstamps@y..., "Sean T. Lamont .lost." <lamont@a...>
wrote:
>
> It's probably better to utilize tight band-pass filters, which are
> relatively simple R/C circuits. I don't think you could do FFT on a
stamp.
>
> On Mon, 2 Sep 2002, Dennis P. O'Leary wrote:
>
> > Anthony,
> >
> > A digital signal can be digitized with a single A/D converter,
and then
> > processed to produce the amplitude for each frequency by using a
> > software Discrete Fourier Transform (DFT) (also loosely called
Fast
> > Fourier Transform (FFT). This is a tall order for a BS2, because
it
> > requires at least double-precision fixed point, or ideally
floating
> > point, calculations. Maybe one of Al Williams's clever PAK
> > co-processors could provide enough computational power, but the
usual
> > way to approach this is with a signal processing chip. I do
recall some
> > assembly routines for performing FFTs with PICs on the Microchip
> > website, so it is probably possible. To conserve memory, certain
DFT
> > algorithms perform the transform _in place_, without requiring
multiple
> > arrays. The _spectrum_ can be regarded approximately as the
square of
> > the DFT output, plotted versus frequency.
> >
> > A good place to start for understanding the theory and mechanics
is to
> > browse through the many digital signal processing books in a
bookstore,
> > and choose one with good examples that appears readable.
> >
> > Another great method is to find a friend with Matlab, or other
> > signal-based programs, and go through the signal processing demos.
> >
> > Dennis
> >
> >
Original Message
> > From: Anthony Conti [noparse][[/noparse]mailto:aconti@s...]
> > Sent: Monday, September 02, 2002 8:32 AM
> > To: basicstamps@y...
> > Subject: [noparse][[/noparse]basicstamps] Digital Audio Spectrum Analyzer
> >
> >
> > Hello Stampers
> >
> > Can someone PLEASE help me with an idea that I have. I am trying
to
> > build a Digital audio spectrum analyzer. I was going to use 31 Op-
amp
> > circuits for the 31 bands then use 31 A/D converters to the BS2p
then to
> > my display, but that is allot of parts.
> >
> > I would like to use an audio A/D converter to the BS2p. The stamp
would
> > do the math for the readings from the A/D. But I have no idea on
how the
> > math would work. I mean how can a digital signal turn into the
amplitude
> > for each frequency?
> >
> > Any insight would help. Is there any books, sites, ect. that
would point
> > me in the right direction?
> >
> > Thanks
> > A. Conti
> >
> >
> > [noparse][[/noparse]Non-text portions of this message have been removed]
> >
> >
> > To UNSUBSCRIBE, just send mail to:
> > basicstamps-unsubscribe@y...
> > from the same email address that you subscribed. Text in the
Subject
> > and Body of the message will be ignored.
> >
> >
> > Your use of Yahoo! Groups is subject to
> > http://docs.yahoo.com/info/terms/
> >
> >
> >
> >
> >
> > To UNSUBSCRIBE, just send mail to:
> > basicstamps-unsubscribe@y...
> > from the same email address that you subscribed. Text in the
Subject and Body of the message will be ignored.
> >
> >
> > Your use of Yahoo! Groups is subject to
http://docs.yahoo.com/info/terms/
> >
> >
> >
>
> Sean T. Lamont, CTO / Chief NetNerd, Abstract Software, Inc.
(ServNet)
> Seattle - Bellingham - Vancouver - Portland - Everett - Tacoma -
Bremerton
> email: lamont@a... WWW: http://www.serv.net
> "Do not fear mistakes, There Are None" - Miles Davis
Microchip has application notes on FFT. Analog Devices also has "The
Engineers Guide to DSP" on their website -- free to download (but
it's a huge file, even zipped).
A spectrum analyzer isn't that difficult to make for low-end
applications. Zetex has a chip (ZXF36L01) which combines a bandpass
or notch filter with a mixer section. It uses a simple pair of
resistors and capacitors to determine the notch depth and notch
frequency. The stamp can be used to generate the sweep frequencies
at 1/3 octave intervals for the local oscillator ( from a lookup
table -- just use a low pass filter to remove the harmonics). I
wrote a paper on it, but the publisher lost interest. If you are
interested in the paper e-mail me at jack@t...
--- In basicstamps@y..., "jbirnsch" <jbirnsch@v...> wrote:
> Do a search on google for Spectrum and sound card. There are
software
> programs that use the standard PC sound card as an audio specrum
> analyzer. Up to 16 thousand FFT sample points with a decent sample
> rate can be had for the cost of a stamp.
>
> Stamps have there place but FFT analysis is a little out of its
> league.
>
> If you want to program in C you could use a TMS series DSP from TI
> and use a stamp to control the DSP. A more relastic solution if you
> need a custom analyzer. Development tools for DSP's are not cheap
> though, can cost in the thousnads of dollars to get setup.
>
> Jason
>
>
>
> --- In basicstamps@y..., "Sean T. Lamont .lost." <lamont@a...>
> wrote:
> >
> > It's probably better to utilize tight band-pass filters, which are
> > relatively simple R/C circuits. I don't think you could do FFT on
a
> stamp.
> >
> > On Mon, 2 Sep 2002, Dennis P. O'Leary wrote:
> >
> > > Anthony,
> > >
> > > A digital signal can be digitized with a single A/D converter,
> and then
> > > processed to produce the amplitude for each frequency by using a
> > > software Discrete Fourier Transform (DFT) (also loosely called
> Fast
> > > Fourier Transform (FFT). This is a tall order for a BS2,
because
> it
> > > requires at least double-precision fixed point, or ideally
> floating
> > > point, calculations. Maybe one of Al Williams's clever PAK
> > > co-processors could provide enough computational power, but the
> usual
> > > way to approach this is with a signal processing chip. I do
> recall some
> > > assembly routines for performing FFTs with PICs on the Microchip
> > > website, so it is probably possible. To conserve memory,
certain
> DFT
> > > algorithms perform the transform _in place_, without requiring
> multiple
> > > arrays. The _spectrum_ can be regarded approximately as the
> square of
> > > the DFT output, plotted versus frequency.
> > >
> > > A good place to start for understanding the theory and
mechanics
> is to
> > > browse through the many digital signal processing books in a
> bookstore,
> > > and choose one with good examples that appears readable.
> > >
> > > Another great method is to find a friend with Matlab, or other
> > > signal-based programs, and go through the signal processing
demos.
> > >
> > > Dennis
> > >
> > >
Original Message
> > > From: Anthony Conti [noparse][[/noparse]mailto:aconti@s...]
> > > Sent: Monday, September 02, 2002 8:32 AM
> > > To: basicstamps@y...
> > > Subject: [noparse][[/noparse]basicstamps] Digital Audio Spectrum Analyzer
> > >
> > >
> > > Hello Stampers
> > >
> > > Can someone PLEASE help me with an idea that I have. I am
trying
> to
> > > build a Digital audio spectrum analyzer. I was going to use 31
Op-
> amp
> > > circuits for the 31 bands then use 31 A/D converters to the
BS2p
> then to
> > > my display, but that is allot of parts.
> > >
> > > I would like to use an audio A/D converter to the BS2p. The
stamp
> would
> > > do the math for the readings from the A/D. But I have no idea
on
> how the
> > > math would work. I mean how can a digital signal turn into the
> amplitude
> > > for each frequency?
> > >
> > > Any insight would help. Is there any books, sites, ect. that
> would point
> > > me in the right direction?
> > >
> > > Thanks
> > > A. Conti
> > >
> > >
> > > [noparse][[/noparse]Non-text portions of this message have been removed]
> > >
> > >
> > > To UNSUBSCRIBE, just send mail to:
> > > basicstamps-unsubscribe@y...
> > > from the same email address that you subscribed. Text in the
> Subject
> > > and Body of the message will be ignored.
> > >
> > >
> > > Your use of Yahoo! Groups is subject to
> > > http://docs.yahoo.com/info/terms/
> > >
> > >
> > >
> > >
> > >
> > > To UNSUBSCRIBE, just send mail to:
> > > basicstamps-unsubscribe@y...
> > > from the same email address that you subscribed. Text in the
> Subject and Body of the message will be ignored.
> > >
> > >
> > > Your use of Yahoo! Groups is subject to
> http://docs.yahoo.com/info/terms/
> > >
> > >
> > >
> >
> > Sean T. Lamont, CTO / Chief NetNerd, Abstract Software, Inc.
> (ServNet)
> > Seattle - Bellingham - Vancouver - Portland - Everett - Tacoma -
> Bremerton
> > email: lamont@a... WWW: http://www.serv.net
> > "Do not fear mistakes, There Are None" - Miles Davis