Searching \ for 'Signals?' in subject line. ()
Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=signals
Search entire site for: 'Signals?'.

Truncated match.
'Signals?'
1995\03\28@085300 by

Why don't you calculate the integral:
T
/  2         1/2
( 1/(2T) * / x  (t) dt )
/
-T
You can approximate this integral with a sum. The T is a multiple of one
period.There may be some problems with the root but you could solve this with a
lookup table. If you want, you can store a couple of squared x:s and "recycle"
them to create a moving rms process (You might need a lot of RAM to store
one period of data.)
---------------------------

>Hello picer's
>
>Does anyone have any bright ideas for signal processing on a pic?
>I would like to find an rms value for for every 2 kHz to 20 kHz
>on a 8 bit digital signal.  I know how to do this using fft's and
>all that, but I don't think an fft would be easy to write for a
>pic.
>
>I also know of several references on signal processing; however, they
>always assume that your postprocessing on something bigger than a pic.
>
>Derrick Early

There is an FFT algorithm in Microchip's 'Embedded Controller Handbook'.
The application note may therefore be available on Microchip's Web Page.

Andy

>
> Why don't you calculate the integral:
>                    T
>                  /  2         1/2
>      ( 1/(2T) * / x  (t) dt )
>                /
>             -T
> You can approximate this integral with a sum. The T is a multiple of one
>  period.There may be some problems with the root but you could solve this with
a
> lookup table. If you want, you can store a couple of squared x:s and "recycle"
> them to create a moving rms process (You might need a lot of RAM to store
> one period of data.)
> ---------------------------
>
Hi Conny,

OK, this will give me an rms for all frequencies, but I would like to
know the rms for discrete frequency bands.

Maybe I should do this with an analog circuit. :(  I'm not sure if a 20 MHz
pic has the muscle to tackle this problem.

Yours,
--
Derrick Early
earlyfinite.nrl.navy.mil

>
> There is an FFT algorithm in Microchip's 'Embedded Controller Handbook'.
>  The application note may therefore be available on Microchip's Web Page.
>
> Andy
>
Andy,

Thank you for redirecting me.  I thought that I read the book from cover
to cover.  So much for my long term memory.  Where can I buy more ram for
my brain?

Yours,
--
Derrick Early
earlyfinite.nrl.navy.mil

1995\03\28@130647 by
>From:         Derrick Early <earlyFINITE.NRL.NAVY.MIL>
>Maybe I should do this with an analog circuit. :(  I'm not sure if a 20 MHz

I also thought of implementing a DTMF decoder with the '71 (with the built
in A/D).  I soon realized that the limited ram on these chips severely
restricts the resolution of any FFT program.

You may want to use a PIC controlled bandpass filter and sweep
the pass band around.  Have the PIC output a frequency that is multiplied
by the incoming signal using a 4 quadrant multiplier.  Low pass filter the
output of the multiplier and peak detect the result.  This can serve
as the amplitude of your signal.

+---------------------------------------+-----------------------------------+
|          Edward Cheung, Ph.D.         |  The opinions expressed herein    |
|Satellite Servicing Robotics Laboratory|    do not necessarily reflect     |
|    NASA Goddard Space Flight Center   |       those of my employers'      |
|         Code 714.1, Bldg T11B         |                                   |
|          Greenbelt, MD 20771          |                                   |
|   301-286-1269(office) 286-1717(fax)  |  My next book:                    |
| Internet: oadebcrobots.gsfc.nasa.gov | Statistics, Demos and Other Lies  |
+---------------------------------------+-----------------------------------+

Edward Cheung wrote
...
>
>You may want to use a PIC controlled bandpass filter and sweep
>the pass band around.  Have the PIC output a frequency that is multiplied
>by the incoming signal using a 4 quadrant multiplier.  Low pass filter the
>output of the multiplier and peak detect the result.  This can serve
>as the amplitude of your signal.
>
I'm not sure about the implementation, but what about a switched capacitor
filter for the bandpass (just a suggestion, since I#ve never worked with
either analog multipliers or switched cap filters) ? Of course this will
(like the multiplier setup) only be feasible if your input signal is
constant during the time needed to sweep the filter.

Just my .02 \$

Bye
Markus

> I also thought of implementing a DTMF decoder with the '71 (with the built
> in A/D).  I soon realized that the limited ram on these chips severely
> restricts the resolution of any FFT program.
>
> You may want to use a PIC controlled bandpass filter and sweep
> the pass band around.  Have the PIC output a frequency that is multiplied
> by the incoming signal using a 4 quadrant multiplier.  Low pass filter the
> output of the multiplier and peak detect the result.  This can serve
> as the amplitude of your signal.
>
Edward,

Thank you for the suggestion.  I did a little more reading about the
digital filter routines written for the 17c, and I began to realize that
I don't have a hope in digitally processing above 5 kHz.  As a result,
I'll have to do some off pic work.

I like your idea of controlling a bandpass filter with the pic.  This
sounds like a really good method to do what I want.

Yours,
--
Derrick Early
earlyfinite.nrl.navy.mil

>  I'm not sure about the implementation, but what about a switched capacitor
> filter for the bandpass (just a suggestion, since I#ve never worked with
> either analog multipliers or switched cap filters) ? Of course this will
> (like the multiplier setup) only be feasible if your input signal is
> constant during the time needed to sweep the filter.

Good point about the signal being constant.  Since I'll be looking at the
signal from an accelerometer on an engine, I'm sure the signal will not
be constant.

Yours,
--
Derrick Early
earlyfinite.nrl.navy.mil

> I like your idea of controlling a bandpass filter with the pic.  This
> sounds like a really good method to do what I want.
>
> Yours,
> --
> Derrick Early

A similar system was developed at my former university for undergrad
signal processing labs.  They used a MAXIM clock tunable switched cap
filter to sweep the band to display the spectrum of the signals in the
labs.  I never got a chance to use it but from what I hear I am glad.
Apparently (for reasons unknown to me) the performance was very slow and
it would take seconds for the unit (plugged into a PC port not sure which)
to perform a sweep.

Kalle
--
Kalle Pihlajasaari          kalledata.co.za

> >  I'm not sure about the implementation, but what about a switched capacitor
> > filter for the bandpass (just a suggestion, since I#ve never worked with
> > either analog multipliers or switched cap filters) ? Of course this will
> > (like the multiplier setup) only be feasible if your input signal is
> > constant during the time needed to sweep the filter.
>
> Good point about the signal being constant.  Since I'll be looking at the
> signal from an accelerometer on an engine, I'm sure the signal will not
> be constant.
>
> Yours,
> --
> Derrick Early
> earlyfinite.nrl.navy.mil
>
I tried to reply to you earlier but I think I've got an email problem.
Anyway, iff you decide you need to do some external processing it sounds
like you could do worse than have a look at Harris Semiconductor's
HIP9010 which is described as an "engine knock signal processor". It's
designed to be connected to a microprocessor as part of an engine
management system. If you don't know any Harris reps you can always get
in touch with them via their web page at:

<URL:http://rel.semi.harris.com:80/>

They've got some data sheets online as well, although I can't guarantee
that will be one of them! Cheers.
-Alasdair

[ The opinions expressed are my own and are therefore undeniably correct ]
[ beyond a shadow of a doubt.                                            ]

Alasdair MacLean, University of Edinburgh, Dept. of Artificial
Intelligence, 5 Forrest Hill, Edinburgh, EH1 2QL
tel. (+44) 0131 650 4497        fax. 0131-650-6899
email alasdairaifh.ed.ac.uk
<URL:http://www.dai.ed.ac.uk/staff/personal_pages/alasdair/index.html>

More... (looser matching)
- Last day of these posts
- In 1995 , 1996 only
- Today
- New search...