DSP Processor for Real Time FFT Calculation

Hi,

I want to build a real time FFT analyzer, my frequency range at the moment is 3Khz to 150KHz. I need to perform 2048-points FFT or even more.

I am looking for a combination of (DSP+Microcontroller) device, actually I am doing modifications in the existing device that uses MCU but I need DSP to process FFT algorithm.

I would like to know if any of you have ever used this kind of processors for FFT application, please help me out in choosing a DSP+MCU unit for my application.

Eagerly waiting for your reply.

BR,
EE

_____________________________________ Start a New Thread

EE-

> I want to build a real time FFT analyzer, my frequency range at
> the moment is 3Khz to 150KHz. I need to perform
> 2048-points FFT or even more.
>
> I am looking for a combination of (DSP+Microcontroller) device,
> actually I am doing modifications in the existing
> device that uses MCU but I need DSP to process FFT algorithm.
>
> I would like to know if any of you have ever used this kind of
> processors for FFT application, please help me out in
> choosing a DSP+MCU unit for my application.
>
> Eagerly waiting for your reply.

Is this a student project? What is the "existing device"?

-Jeff

_____________________________________ Start a New Thread

Hi Jeff,

Yes, it is a student project that can be interfaced with Automatic Meter Reading
equipment. The part I am working on is to measure the quality of
power line communication signals. Some time meter reader doesn't recognize that wheather the coming signal is sent by concentrator (i.e. many meter connected to this unit) or its just noisy signal.

Any help, please ?

EE

________________________________
From: Jeff Brower
To: m...@yahoo.com
Cc: c...
Sent: Mon, April 26, 2010 9:22:10 PM
Subject: Re: [c6x] DSP Processor for Real Time FFT Calculation

EE-

> I want to build a real time FFT analyzer, my frequency range at
> the moment is 3Khz to 150KHz. I need to perform
> 2048-points FFT or even more.
>
> I am looking for a combination of (DSP+Microcontrolle r) device,
> actually I am doing modifications in the existing
> device that uses MCU but I need DSP to process FFT algorithm.
>
> I would like to know if any of you have ever used this kind of
> processors for FFT application, please help me out in
> choosing a DSP+MCU unit for my application.
>
> Eagerly waiting for your reply.

Is this a student project? What is the "existing device"?

-Jeff Start a New Thread

BS EE-

> Yes, it is a student project that can be interfaced with
> Automatic Meter Reading
> equipment. The part I am working on is to measure the
> quality of
> power line communication signals. Some time meter reader
> doesn't recognize that wheather the coming signal is sent by
> concentrator (i.e. many meter connected to this unit) or its
> just noisy signal.
>
> Any help, please ?

The concentrator is sending a signal "upstream"? At what bitrate, a few Hz? Does the concentrator send some type of
periodic signal; i.e. does it modulate a carrier (fundamental frequency)? Or include a carrier to help identify its
signal? If so you might expect that averaging many FFT output frames over a period of time would show the
concentrator's carrier, assuming the noise is random and averages out to be more or less flat (or a predictable
frequency domain profile). What kind of noise do you have?

Regarding hardware, have you looked at using one of the cheap TI DSK boards? Can you interface your signals to a
standard audio type A/D converter (assuming it can sample down to DC)? Or do you need a more specialized A/D front
end?

-Jeff

PS. Please don't cut text from your initial post. I read 100s of post every day and I don't remember your original
question and I'm not going to waste time to go back and look.

> ________________________________
> From: Jeff Brower
> To: m...@yahoo.com
> Cc: c...
> Sent: Mon, April 26, 2010 9:22:10 PM
> Subject: Re: [c6x] DSP Processor for Real Time FFT Calculation
>
> EE-
>
>> I want to build a real time FFT analyzer, my frequency range at
>> the moment is 3Khz to 150KHz. I need to perform
>> 2048-points FFT or even more.
>>
>> I am looking for a combination of (DSP+Microcontrolle r) device,
>> actually I am doing modifications in the existing
>> device that uses MCU but I need DSP to process FFT algorithm.
>>
>> I would like to know if any of you have ever used this kind of
>> processors for FFT application, please help me out in
>> choosing a DSP+MCU unit for my application.
>>
>> Eagerly waiting for your reply.
>
> Is this a student project? What is the "existing device"?
>
> -Jeff

_____________________________________ Start a New Thread

BS EE-

Sorry, you didn't cut text, I just realized... I was looking for it "inline" and didn't skip down far enough. Again,
sorry.

-Jeff
>> Yes, it is a student project that can be interfaced with
>> Automatic Meter Reading
>> equipment. The part I am working on is to measure the
>> quality of
>> power line communication signals. Some time meter reader
>> doesn't recognize that wheather the coming signal is sent by
>> concentrator (i.e. many meter connected to this unit) or its
>> just noisy signal.
>>
>> Any help, please ?
>
> The concentrator is sending a signal "upstream"? At what bitrate, a few Hz? Does the concentrator send some type of
> periodic signal; i.e. does it modulate a carrier (fundamental frequency)? Or include a carrier to help identify its
> signal? If so you might expect that averaging many FFT output frames over a period of time would show the
> concentrator's carrier, assuming the noise is random and averages out to be more or less flat (or a predictable
> frequency domain profile). What kind of noise do you have?
>
> Regarding hardware, have you looked at using one of the cheap TI DSK boards? Can you interface your signals to a
> standard audio type A/D converter (assuming it can sample down to DC)? Or do you need a more specialized A/D front
> end?
>
> -Jeff
>
> PS. Please don't cut text from your initial post. I read 100s of post every day and I don't remember your original
> question and I'm not going to waste time to go back and look.
>
>> ________________________________
>> From: Jeff Brower
>> To: m...@yahoo.com
>> Cc: c...
>> Sent: Mon, April 26, 2010 9:22:10 PM
>> Subject: Re: [c6x] DSP Processor for Real Time FFT Calculation
>>
>> EE-
>>
>>> I want to build a real time FFT analyzer, my frequency range at
>>> the moment is 3Khz to 150KHz. I need to perform
>>> 2048-points FFT or even more.
>>>
>>> I am looking for a combination of (DSP+Microcontrolle r) device,
>>> actually I am doing modifications in the existing
>>> device that uses MCU but I need DSP to process FFT algorithm.
>>>
>>> I would like to know if any of you have ever used this kind of
>>> processors for FFT application, please help me out in
>>> choosing a DSP+MCU unit for my application.
>>>
>>> Eagerly waiting for your reply.
>>
>> Is this a student project? What is the "existing device"?
>>
>> -Jeff
>
>
>
> _____________________________________
>

_____________________________________ Start a New Thread

Hi Jeff,

The concentrator is sending a signal "upstream"? At what bitrate, a few Hz? Does the concentrator send some type of periodic signal; i.e. does it modulate a carrier (fundamental frequency)? Or include a carrier to help identify its signal?
If so you might expect that averaging many FFT output frames
over a period of time would show the concentrator' s carrier, assuming
the noise is random and averages out
to be more or less flat (or a predictable frequency domain profile). What kind of noise do you have?

Regarding hardware, have you looked at using one of the cheap TI DSK
boards? Can you interface your signals to a standard audio type A/D
converter (assuming it can sample down to DC)?
Or do you need a more specialized A/D front end?

-Jeff

The bitrate in the communication between the
concentrator and the energy meter is usually less than 5 kbits/s (FSK and BPSK modulation).

We may have noise from, for example, motor drives (frequency
converters/inverters), swithing power supplies, energy saving lamps and
other communication systems (although there should not be any other than the power utilities system between 3...95 kHz).

Yes, at the moment I am thinking to use TMS320C67x DSP, it will cost
around US $395.

It probably is neccessary to have some analog
amplifier (with
adjustable/programmable gain) and at least an anti-aliasing filter in
front of the ADC.

These are the quantities to be calculated.

-Power Line Communication (PLC) Voltage
-Noise Level at the PLC frequencies
-Signal to Noise Ratio

Average disturbance level at selected frequency
bands
- 50... 95KHz, 70... 95KHz, 95... 150KHz
-max/mean/min

Waiting for your reply.

BR,
BS

________________________________
From: Jeff Brower
To: BS EE
Cc: c...
Sent: Wed, April 28, 2010 12:06:14 AM
Subject: Re: [c6x] DSP Processor for Real Time FFT Calculation

BS EE-

> Yes, it is a student project that can be interfaced with
> Automatic Meter Reading
> equipment. The part I am working on is to measure the
> quality of
> power line communication signals. Some time meter reader
> doesn't recognize that wheather the coming signal is sent by
> concentrator (i.e. many meter connected to this unit) or its
> just noisy signal.
>
> Any help, please ?

The concentrator is sending a signal "upstream"? At what bitrate, a few Hz? Does the concentrator send some type of
periodic signal; i.e. does it modulate a carrier (fundamental frequency)? Or include a carrier to help identify its
signal? If so you might expect that averaging many FFT output frames over a period of time would show the
concentrator' s carrier, assuming the noise is random and averages out to be more or less flat (or a predictable
frequency domain profile). What kind of noise do you have?

Regarding hardware, have you looked at using one of the cheap TI DSK boards? Can you interface your signals to a
standard audio type A/D converter (assuming it can sample down to DC)? Or do you need a more specialized A/D front
end?

-Jeff

PS. Please don't cut text from your initial post. I read 100s of post every day and I don't remember your original
question and I'm not going to waste time to go back and look.

> ____________ _________ _________ __
> From: Jeff Brower
> To: matlab_fft@yahoo. com
> Cc: c6x@yahoogroups. com
> Sent: Mon, April 26, 2010 9:22:10 PM
> Subject: Re: [c6x] DSP Processor for Real Time FFT Calculation
>
> EE-
>
>> I want to build a real time FFT analyzer, my frequency range at
>> the moment is 3Khz to 150KHz. I need to perform
>> 2048-points FFT or even more.
>>
>> I am looking for a combination of (DSP+Microcontrolle r) device,
>> actually I am doing modifications in the existing
>> device that uses MCU but I need DSP to process FFT algorithm.
>>
>> I would like to know if any of you have ever used this kind of
>> processors for FFT application, please help me out in
>> choosing a DSP+MCU unit for my application.
>>
>> Eagerly waiting for your reply.
>
> Is this a student project? What is the "existing device"?
>
> -Jeff Start a New Thread

In case anyone else is following this thread, I'm reposting this because I messed up the cc addr.

---------------- Original Message -------------
Subject: Re: [c6x] DSP Processor for Real Time FFT Calculation
From: "Jeff Brower"
Date: Sat, May 1, 2010 10:37 pm
To: "BS EE"
Cc: c...@yahoogroup.com
-----------

BS EE-

>> The concentrator is sending a signal "upstream"? At what
>> bitrate, a few Hz? Does the concentrator send some type of
>> periodic signal; i.e. does it modulate a carrier (fundamental
>> frequency)? Or include a carrier to help identify its signal?
>> If so you might expect that averaging many FFT output
>> frames over a period of time would show the concentrator' s
>> carrier, assuming the noise is random and averages out
>> to be more or less flat (or a predictable frequency domain
>> profile). What kind of noise do you have?
>
> Regarding hardware, have you looked at using one of the cheap
>> TI DSK boards? Can you interface your signals to a standard
>> audio type A/D converter (assuming it can sample down to DC)?
>> Or do you need a more specialized A/D front end?
>>
>> -Jeff
>
> The bitrate in the communication between the concentrator and
> the energy meter is usually less than 5 kbits/s (FSK and
> BPSK modulation).

Ok... I thought that bit rates in the multi-kHz range would get blocked by transformers and other line equipment...
but maybe not if the distance traversed does not have too many of these elements.

> We may have noise from, for example, motor drives (frequency
> converters/inverters), swithing power supplies, energy saving
> lamps and other communication systems

Ok.

> (although there should
> not be any other than the power utilities system between 3...95
> kHz).

Ya "signal shouldn't be noisy", right. But of course it's noisy as heck -- the story of signal processing.

> Yes, at the moment I am thinking to use TMS320C67x DSP, it
> will cost around US $395.

This is not a bad choice for a starting point. But if you need more horsepower, you might look at C6472 or other high
clock rate fixed-point DSK boards. How important is floating-point in your application?

> It probably is neccessary to have some analog amplifier (with
> adjustable/programmable gain) and at least an anti-aliasing
> filter in front of the ADC.

Amplifier maybe, anti-alias filter probably not. What are your input signal levels? At 5 kbps data rate you might be
sampling anywhere from 10 to 50 kHz depending on how many samples per symbol you need in your BPSK waveform. The
AIC23 codec on the DSK C6713 board is a sigma-delta type with very effective built-in anti-aliasing (16-bit,
2-channel, max sampling rate up to 96 kHz).

> These are the quantities to be calculated.
>
> -Power Line Communication (PLC) Voltage
> -Noise Level at the PLC frequencies
> -Signal to Noise Ratio
>
> Average disturbance level at selected frequency bands
> - 50... 95KHz, 70... 95KHz, 95... 150KHz
> -max/mean/min

If you need to accurately see frequencies about 48 kHz then built-in analog circuitry on the TI DSK boards is not
going to be enough. You might consider a third-party DSK daughtermodule of some type, with higher sampling rates. If
those are not sigma-delta type ADCs, then the issue of anti-aliasing gets back in the picture.

-Jeff

> ________________________________
> From: Jeff Brower
> To: BS EE
> Cc: c...
> Sent: Wed, April 28, 2010 12:06:14 AM
> Subject: Re: [c6x] DSP Processor for Real Time FFT Calculation
> BS EE-
>
>> Yes, it is a student project that can be interfaced with
>> Automatic Meter Reading
>> equipment. The part I am working on is to measure the
>> quality of
>> power line communication signals. Some time meter reader
>> doesn't recognize that wheather the coming signal is sent by
>> concentrator (i.e. many meter connected to this unit) or its
>> just noisy signal.
>>
>> Any help, please ?
>
> The concentrator is sending a signal "upstream"? At what bitrate, a few Hz? Does the concentrator send some type
of periodic signal; i.e. does it modulate a carrier (fundamental frequency)? Or include a carrier to help identify
its signal? If so you might expect that averaging many FFT output frames over a period of time would show the
concentrator' s carrier, assuming the noise is random and averages out to be more or less flat (or a predictable
frequency domain profile). What kind of noise do you have?
>
> Regarding hardware, have you looked at using one of the cheap TI DSK boards? Can you interface your signals to a
standard audio type A/D converter (assuming it can sample down to DC)? Or do you need a more specialized A/D front
end?
>
> -Jeff
>
> PS. Please don't cut text from your initial post. I read 100s of post every day and I don't remember your original
question and I'm not going to waste time to go back and look.
>
>> ____________ _________ _________ __
>> From: Jeff Brower
>> To: matlab_fft@yahoo. com
>> Cc: c6x@yahoogroups. com
>> Sent: Mon, April 26, 2010 9:22:10 PM
>> Subject: Re: [c6x] DSP Processor for Real Time FFT Calculation
>>
>> EE-
>>
>>> I want to build a real time FFT analyzer, my frequency range at the moment is 3Khz to 150KHz. I need to perform
>>> 2048-points FFT or even more.
>>>
>>> I am looking for a combination of (DSP+Microcontrolle r) device, actually I am doing modifications in the existing
>>> device that uses MCU but I need DSP to process FFT algorithm.
>>>
>>> I would like to know if any of you have ever used this kind of processors for FFT application, please help me out in
>>> choosing a DSP+MCU unit for my application.
>>>
>>> Eagerly waiting for your reply.
>>
>> Is this a student project? What is the "existing device"?
>>
>> -Jeff

_____________________________________ Start a New Thread

BS-

> I Got the C6713 DSK but stuck with sampling of analog signal. There
> is one dsk_app.pjt example associated with the
> installation folder but its really a mess. It has two .c files one
> aic23.c and other is dsk_app.c file, both attached
> here. Now I don't know how to use them simply to sample I KHz sine
> wave.
>
> Can you please help me with this ?

Open CCS, select Open Project and then open the dsk_app.pjt file. After opening the project file, you should see
"Source Files", "Libraries", etc (and individual files underneath) in the lefthand pane (window) in CCS. What do you
see?

-Jeff

>> ________________________________
> From: Jeff Brower
> To: c...
> Sent: Fri, May 7, 2010 9:37:04 AM
> Subject: Re: [c6x] DSP Processor for Real Time FFT Calculation
>
> In case anyone else is following this thread, I'm reposting this because I messed up the cc addr.
>
> ------------ ---- Original Message ------------ -
> Subject: Re: [c6x] DSP Processor for Real Time FFT Calculation
> From: "Jeff Brower"
> Date: Sat, May 1, 2010 10:37 pm
> To: "BS EE"
> Cc: c6x@yahoogroup. com
> ------------ --------- --------- --------- --------
>
> BS EE-
>
>>> The concentrator is sending a signal "upstream"? At what
>>> bitrate, a few Hz? Does the concentrator send some type of
>>> periodic signal; i.e. does it modulate a carrier (fundamental
>>> frequency)? Or include a carrier to help identify its signal?
>>> If so you might expect that averaging many FFT output
>>> frames over a period of time would show the concentrator' s
>>> carrier, assuming the noise is random and averages out
>>> to be more or less flat (or a predictable frequency domain
>>> profile). What kind of noise do you have?
>>
>> Regarding hardware, have you looked at using one of the cheap
>>> TI DSK boards? Can you interface your signals to a standard
>>> audio type A/D converter (assuming it can sample down to DC)?
>>> Or do you need a more specialized A/D front end?
>>>
>>> -Jeff
>>
>> The bitrate in the communication between the concentrator and
>> the energy meter is usually less than 5 kbits/s (FSK and
>> BPSK modulation).
>
> Ok... I thought that bit rates in the multi-kHz range would get blocked by transformers and other line equipment...
> but maybe not if the distance traversed does not have too many of these elements.
>
>> We may have noise from, for example, motor drives (frequency
>> converters/inverter s), swithing power supplies, energy saving
>> lamps and other communication systems
>
> Ok.
>
>> (although there should
>> not be any other than the power utilities system between 3...95
>> kHz).
>
> Ya "signal shouldn't be noisy", right. But of course it's noisy as heck -- the story of signal processing.
>
>> Yes, at the moment I am thinking to use TMS320C67x DSP, it
>> will cost around US $395.
>
> This is not a bad choice for a starting point. But if you need more horsepower, you might look at C6472 or other high
> clock rate fixed-point DSK boards. How important is floating-point in your application?
>
>> It probably is neccessary to have some analog amplifier (with
>> adjustable/programm able gain) and at least an anti-aliasing
>> filter in front of the ADC.
>
> Amplifier maybe, anti-alias filter probably not. What are your input signal levels? At 5 kbps data rate you might be
> sampling anywhere from 10 to 50 kHz depending on how many samples per symbol you need in your BPSK waveform. The
> AIC23 codec on the DSK C6713 board is a sigma-delta type with very effective built-in anti-aliasing (16-bit,
> 2-channel, max sampling rate up to 96 kHz).
>
>> These are the quantities to be calculated.
>>
>> -Power Line Communication (PLC) Voltage
>> -Noise Level at the PLC frequencies
>> -Signal to Noise Ratio
>>
>> Average disturbance level at selected frequency bands
>> - 50... 95KHz, 70... 95KHz, 95... 150KHz
>> -max/mean/min
>
> If you need to accurately see frequencies about 48 kHz then built-in analog circuitry on the TI DSK boards is not
> going to be enough. You might consider a third-party DSK daughtermodule of some type, with higher sampling rates. If
> those are not sigma-delta type ADCs, then the issue of anti-aliasing gets back in the picture.
>
> -Jeff
>
>> ____________ _________ _________ __
>> From: Jeff Brower
>> To: BS EE
>> Cc: c6x@yahoogroups. com
>> Sent: Wed, April 28, 2010 12:06:14 AM
>> Subject: Re: [c6x] DSP Processor for Real Time FFT Calculation
>> BS EE-
>>
>>> Yes, it is a student project that can be interfaced with
>>> Automatic Meter Reading
>>> equipment. The part I am working on is to measure the
>>> quality of
>>> power line communication signals. Some time meter reader
>>> doesn't recognize that wheather the coming signal is sent by
>>> concentrator (i.e. many meter connected to this unit) or its
>>> just noisy signal.
>>>
>>> Any help, please ?
>>
>> The concentrator is sending a signal "upstream"? At what bitrate, a few Hz? Does the concentrator send some type
> of periodic signal; i.e. does it modulate a carrier (fundamental frequency)? Or include a carrier to help identify
> its signal? If so you might expect that averaging many FFT output frames over a period of time would show the
> concentrator' s carrier, assuming the noise is random and averages out to be more or less flat (or a predictable
> frequency domain profile). What kind of noise do you have?
>>
>> Regarding hardware, have you looked at using one of the cheap TI DSK boards? Can you interface your signals to a
> standard audio type A/D converter (assuming it can sample down to DC)? Or do you need a more specialized A/D front
> end?
>>
>> -Jeff
>>
>> PS. Please don't cut text from your initial post. I read 100s of post every day and I don't remember your original
> question and I'm not going to waste time to go back and look.
>>
>>> ____________ _________ _________ __
>>> From: Jeff Brower
>>> To: matlab_fft@yahoo. com
>>> Cc: c6x@yahoogroups. com
>>> Sent: Mon, April 26, 2010 9:22:10 PM
>>> Subject: Re: [c6x] DSP Processor for Real Time FFT Calculation
>>>
>>> EE-
>>>
>>>> I want to build a real time FFT analyzer, my frequency range at the moment is 3Khz to 150KHz. I need to perform
>>>> 2048-points FFT or even more.
>>>>
>>>> I am looking for a combination of (DSP+Microcontrolle r) device, actually I am doing modifications in the existing
>>>> device that uses MCU but I need DSP to process FFT algorithm.
>>>>
>>>> I would like to know if any of you have ever used this kind of processors for FFT application, please help me out
>>>> in
>>>> choosing a DSP+MCU unit for my application.
>>>>
>>>> Eagerly waiting for your reply.
>>>
>>> Is this a student project? What is the "existing device"?
>>>
>>> -Jeff

_____________________________________ Start a New Thread

/*
* Copyright 2003 by Spectrum Digital Incorporated.
* All rights reserved. Property of Spectrum Digital Incorporated.
*/

/*
* ======== dsk_app.c ======= *
* Version 1.00
*
* This example digitally processes audio data from the line input on the
* AIC23 codec and plays the result on the line output. It uses the McBSP
* and EDMA to efficiently handle the data transfer without intervention from
* the DSP.
*
* Data transfer
*
* Audio data is transferred back and forth from the codec through McBSP2,
* a bidirectional serial port. The EDMA is configured to take every 16-bit
* signed audio sample arriving on McBSP1 and store it in a buffer in memory
* until it can be processed. Once it has been processed, the EDMA
* controller sends the data back to McBSP1 for transmission.
*
* A second serial port, McBSP0 is used to control/configure the AIC23. The
* codec receives serial commands through McBSP0 that set configuration
* parameters such as volume, sample rate and data format.
*
* In addition to basic EDMA transfers, this example uses 2 special
* techniques to make audio processing more convenient and efficient:
*
* 1) Ping-pong data buffering in memory
* 2) Linked EDMA transfers
*
* Applications with single buffers for receive and transmit data are
* very tricky and timing dependent because new data constantly overwrites
* the data being transmitted. Ping-pong buffering is a technique where two
* buffers (referred to as the PING buffer and the PONG buffer) are used for
* a data transfer instead of only one. The EDMA is configured to fill the
* PING buffer first, then the PONG buffer. While the PONG buffer is being
* filled, the PING buffer can be processed with the knowledge that the
* current EDMA transfer won't overwrite it. This example uses ping-pong
* buffers on both transmit and receive ends for a total of four buffers.
*
* The EDMA controller must be configured slightly differently for each
* buffer. When a buffer is filled, the EDMA controller generates an
* interrupt. The interrupt handler must reload the configuration
* for the next buffer before the next audio sample arrives. An EDMA
* feature called linked transfers is used to make this event less time
* critical. Each configuration is created in advance and the EDMA
* controller automatically links to the next configuration when the
* current configuration is finished. An interrupt is still generated,
* but it serves only to signal the DSP that it can process the data.
* The only time constraint is that all the audio data must be processed
* before the the active buffer fills up, which is much longer than the
* time between audio samples. It is much easier to satisfy real-time
* constraints with this implementation.
*
* Program flow
*
* When the program is run, the individual DSP/BIOS modules are initialized
* as configured in dsk_app.cdb with the DSP/BIOS configuration tool. The
* main() function is then called as the main user thread. In this example
* main() performs application initialization and starts the EDMA data
* transfers. When main exits, control passes back entirely to DSP/BIOS
* which services any interrupts or threads on an as-needed basis.
*
* The edmaHwi() interrupt service routine is called when a buffer has been
* filled. It contains a state variable named pingOrPong that indicates
* whether the buffer is a PING or PONG buffer. dmaHwi switches the buffer
* state to the opposite buffer and calls the SWI thread processBuffer to
* process the audio data.
*
* Other Functions
*
* The example includes a few other functions that are executed in the
* background as examples of the multitasking that DSP/BIOS is capable of:
*
* 1) blinkLED() toggles LED #0 every 500ms if DIP switch #0 is depressed.
* It is a periodic thread with a period of 500 ticks.
*
* 2) load() simulates a 20-25% dummy load if DIP switch #1 is depressed.
* It represents other computation that may need to be done. It is a
* periodic thread with a period of 10ms.
*
* Please see the 6713 DSK help file under Software/Examples for more
* detailed information on this example.
*/

/*
* DSP/BIOS is configured using the DSP/BIOS configuration tool. Settings
* for this example are stored in a configuration file called dsk_app.cdb.
* At compile time, Code Composer will auto-generate DSP/BIOS related files
* based on these settings. A header file called dsk_appcfg.h contains the
* results of the autogeneration and must be included for proper operation.
* The name of the file is taken from dsk_app.cdb and adding cfg.h.
*/
#include "dsk_appcfg.h"

/*
* These are include files that support interfaces to BIOS and CSL modules
* used by the program.
*/
#include
#include
#include
#include
#include
#include
#include
#include

/*
* The 6713 DSK Board Support Library is divided into several modules, each
* of which has its own include file. The file dsk6713.h must be included
* in every program that uses the BSL. This example also uses the
* DIP, LED and AIC23 modules.
*/
#include "dsk6713.h"
#include "dsk6713_led.h"
#include "dsk6713_dip.h"
#include "aic23.h"

/* Function prototypes */
void initIrq(void);
void initMcbsp(void);
void initEdma(void);
void copyData(Int16 *inbuf, Int16 *outbuf, Int16 length);
void processBuffer(void);
void edmaHwi(void);

/* Constants for the buffered ping-pong transfer */
#define BUFFSIZE 1000
#define PING 0
#define PONG 1

/*
* Data buffer declarations - the program uses four logical buffers of size
* BUFFSIZE, one ping and one pong buffer on both receive and transmit sides.
*/
Int16 gBufferXmtPing[BUFFSIZE]; // Transmit PING buffer
Int16 gBufferXmtPong[BUFFSIZE]; // Transmit PONG buffer

Int16 gBufferRcvPing[BUFFSIZE]; // Receive PING buffer
Int16 gBufferRcvPong[BUFFSIZE]; // Receive PONG buffer

EDMA_Handle hEdmaXmt; // EDMA channel handles
EDMA_Handle hEdmaReloadXmtPing;
EDMA_Handle hEdmaReloadXmtPong;
EDMA_Handle hEdmaRcv;
EDMA_Handle hEdmaReloadRcvPing;
EDMA_Handle hEdmaReloadRcvPong;

MCBSP_Handle hMcbsp1; // McBSP1 (codec data) handle

Int16 gXmtChan; // TCC codes (see initEDMA())
Int16 gRcvChan;
/*
* EDMA Config data structure
*/

/* Transmit side EDMA configuration */
EDMA_Config gEdmaConfigXmt = {
EDMA_FMKS(OPT, PRI, HIGH) | // Priority
EDMA_FMKS(OPT, ESIZE, 16BIT) | // Element size
EDMA_FMKS(OPT, 2DS, NO) | // 2 dimensional source?
EDMA_FMKS(OPT, SUM, INC) | // Src update mode
EDMA_FMKS(OPT, 2DD, NO) | // 2 dimensional dest
EDMA_FMKS(OPT, DUM, NONE) | // Dest update mode
EDMA_FMKS(OPT, TCINT, YES) | // Cause EDMA interrupt?
EDMA_FMKS(OPT, TCC, OF(0)) | // Transfer complete code
EDMA_FMKS(OPT, LINK, YES) | // Enable link parameters?
EDMA_FMKS(OPT, FS, NO), // Use frame sync?

(Uint32)&gBufferXmtPing, // Src address

EDMA_FMK (CNT, FRMCNT, NULL) | // Frame count
EDMA_FMK (CNT, ELECNT, BUFFSIZE), // Element count

EDMA_FMKS(DST, DST, OF(0)), // Dest address

EDMA_FMKS(IDX, FRMIDX, DEFAULT) | // Frame index value
EDMA_FMKS(IDX, ELEIDX, DEFAULT), // Element index value

EDMA_FMK (RLD, ELERLD, NULL) | // Reload element
EDMA_FMK (RLD, LINK, NULL) // Reload link
};

/* Receive side EDMA configuration */
EDMA_Config gEdmaConfigRcv = {
EDMA_FMKS(OPT, PRI, HIGH) | // Priority
EDMA_FMKS(OPT, ESIZE, 16BIT) | // Element size
EDMA_FMKS(OPT, 2DS, NO) | // 2 dimensional source?
EDMA_FMKS(OPT, SUM, NONE) | // Src update mode
EDMA_FMKS(OPT, 2DD, NO) | // 2 dimensional dest
EDMA_FMKS(OPT, DUM, INC) | // Dest update mode
EDMA_FMKS(OPT, TCINT, YES) | // Cause EDMA interrupt?
EDMA_FMKS(OPT, TCC, OF(0)) | // Transfer complete code
EDMA_FMKS(OPT, LINK, YES) | // Enable link parameters?
EDMA_FMKS(OPT, FS, NO), // Use frame sync?

EDMA_FMKS(SRC, SRC, OF(0)), // Src address

EDMA_FMK (CNT, FRMCNT, NULL) | // Frame count
EDMA_FMK (CNT, ELECNT, BUFFSIZE), // Element count

(Uint32)&gBufferRcvPing, // Dest address

EDMA_FMKS(IDX, FRMIDX, DEFAULT) | // Frame index value
EDMA_FMKS(IDX, ELEIDX, DEFAULT), // Element index value

EDMA_FMK (RLD, ELERLD, NULL) | // Reload element
EDMA_FMK (RLD, LINK, NULL) // Reload link
};

/* McBSP codec data channel configuration */
static MCBSP_Config mcbspCfg1 = {
MCBSP_FMKS(SPCR, FREE, NO) |
MCBSP_FMKS(SPCR, SOFT, NO) |
MCBSP_FMKS(SPCR, FRST, YES) |
MCBSP_FMKS(SPCR, GRST, YES) |
MCBSP_FMKS(SPCR, XINTM, XRDY) |
MCBSP_FMKS(SPCR, XSYNCERR, NO) |
MCBSP_FMKS(SPCR, XRST, YES) |
MCBSP_FMKS(SPCR, DLB, OFF) |
MCBSP_FMKS(SPCR, RJUST, RZF) |
MCBSP_FMKS(SPCR, CLKSTP, DISABLE) |
MCBSP_FMKS(SPCR, DXENA, OFF) |
MCBSP_FMKS(SPCR, RINTM, RRDY) |
MCBSP_FMKS(SPCR, RSYNCERR, NO) |
MCBSP_FMKS(SPCR, RRST, YES),

MCBSP_FMKS(RCR, RPHASE, SINGLE) |
MCBSP_FMKS(RCR, RFRLEN2, DEFAULT) |
MCBSP_FMKS(RCR, RWDLEN2, DEFAULT) |
MCBSP_FMKS(RCR, RCOMPAND, MSB) |
MCBSP_FMKS(RCR, RFIG, NO) |
MCBSP_FMKS(RCR, RDATDLY, 0BIT) |
MCBSP_FMKS(RCR, RFRLEN1, OF(1)) |
MCBSP_FMKS(RCR, RWDLEN1, 16BIT) |
MCBSP_FMKS(RCR, RWDREVRS, DISABLE),

MCBSP_FMKS(XCR, XPHASE, SINGLE) |
MCBSP_FMKS(XCR, XFRLEN2, DEFAULT) |
MCBSP_FMKS(XCR, XWDLEN2, DEFAULT) |
MCBSP_FMKS(XCR, XCOMPAND, MSB) |
MCBSP_FMKS(XCR, XFIG, NO) |
MCBSP_FMKS(XCR, XDATDLY, 0BIT) |
MCBSP_FMKS(XCR, XFRLEN1, OF(1)) |
MCBSP_FMKS(XCR, XWDLEN1, 16BIT) |
MCBSP_FMKS(XCR, XWDREVRS, DISABLE),

MCBSP_FMKS(SRGR, GSYNC, DEFAULT) |
MCBSP_FMKS(SRGR, CLKSP, DEFAULT) |
MCBSP_FMKS(SRGR, CLKSM, DEFAULT) |
MCBSP_FMKS(SRGR, FSGM, DEFAULT) |
MCBSP_FMKS(SRGR, FPER, DEFAULT) |
MCBSP_FMKS(SRGR, FWID, DEFAULT) |
MCBSP_FMKS(SRGR, CLKGDV, DEFAULT),

MCBSP_MCR_DEFAULT,
MCBSP_RCER_DEFAULT,
MCBSP_XCER_DEFAULT,

MCBSP_FMKS(PCR, XIOEN, SP) |
MCBSP_FMKS(PCR, RIOEN, SP) |
MCBSP_FMKS(PCR, FSXM, EXTERNAL) |
MCBSP_FMKS(PCR, FSRM, EXTERNAL) |
MCBSP_FMKS(PCR, CLKXM, INPUT) |
MCBSP_FMKS(PCR, CLKRM, INPUT) |
MCBSP_FMKS(PCR, CLKSSTAT, DEFAULT) |
MCBSP_FMKS(PCR, DXSTAT, DEFAULT) |
MCBSP_FMKS(PCR, FSXP, ACTIVEHIGH) |
MCBSP_FMKS(PCR, FSRP, ACTIVEHIGH) |
MCBSP_FMKS(PCR, CLKXP, FALLING) |
MCBSP_FMKS(PCR, CLKRP, RISING)
};

/* Codec configuration settings */
AIC23_Params config = {
0x0017, // 0 DSK6713_AIC23_LEFTINVOL Left line input channel volume
0x0017, // 1 DSK6713_AIC23_RIGHTINVOL Right line input channel volume
0x00d8, // 2 DSK6713_AIC23_LEFTHPVOL Left channel headphone volume
0x00d8, // 3 DSK6713_AIC23_RIGHTHPVOL Right channel headphone volume
0x0011, // 4 DSK6713_AIC23_ANAPATH Analog audio path control
0x0000, // 5 DSK6713_AIC23_DIGPATH Digital audio path control
0x0000, // 6 DSK6713_AIC23_POWERDOWN Power down control
0x0043, // 7 DSK6713_AIC23_DIGIF Digital audio interface format
0x0001, // 8 DSK6713_AIC23_SAMPLERATE Sample rate control
0x0001 // 9 DSK6713_AIC23_DIGACT Digital interface activation
};
/* --------------------------- main() function -------------------------- */
/*
* main() - The main user task. Performs application initialization and
* starts the data transfer.
*/
void main()
{
/* Initialize Board Support Library */
DSK6713_init();

/* Initialize LEDs and DIP switches */
DSK6713_LED_init();
DSK6713_DIP_init();

/* Clear buffers */
memset((void *)gBufferXmtPing, 0, BUFFSIZE * 4 * 2);

AIC23_setParams(&config); // Configure the codec

initMcbsp(); // Initialize McBSP1 for audio transfers

IRQ_globalDisable(); // Disable global interrupts during setup

initEdma(); // Initialize the EDMA controller

initIrq(); // Initialize interrupts

IRQ_globalEnable(); // Re-enable global interrupts
}
/* ------------------------Helper Functions ----------------------------- */

/*
* initMcbsp() - Initialize the McBSP for codec data transfers using the
* configuration define at the top of this file.
*/
void initMcbsp()
{
/* Open the codec data McBSP */
hMcbsp1 = MCBSP_open(MCBSP_DEV1, MCBSP_OPEN_RESET);

/* Configure the codec to match the AIC23 data format */
MCBSP_config(hMcbsp1, &mcbspCfg1);

/* Start the McBSP running */
MCBSP_start(hMcbsp1, MCBSP_XMIT_START | MCBSP_RCV_START |
MCBSP_SRGR_START | MCBSP_SRGR_FRAMESYNC, 220);
}
/*
* initIrq() - Initialize and enable the DMA receive interrupt using the CSL.
* The interrupt service routine for this interrupt is edmaHwi.
*/
void initIrq(void)
{
/* Enable EDMA interrupts to the CPU */
IRQ_clear(IRQ_EVT_EDMAINT); // Clear any pending EDMA interrupts
IRQ_enable(IRQ_EVT_EDMAINT); // Enable EDMA interrupt
}
/*
* initEdma() - Initialize the DMA controller. Use linked transfers to
* automatically transition from ping to pong and visa-versa.
*/
void initEdma(void)
{
/* Configure transmit channel */
hEdmaXmt = EDMA_open(EDMA_CHA_XEVT1, EDMA_OPEN_RESET); // get hEdmaXmt handle and reset channel
hEdmaReloadXmtPing = EDMA_allocTable(-1); // get hEdmaReloadXmtPing handle
hEdmaReloadXmtPong = EDMA_allocTable(-1); // get hEdmaReloadXmtPong handle

gEdmaConfigXmt.dst = MCBSP_getXmtAddr(hMcbsp1); // set the desination address to McBSP1 DXR

gXmtChan = EDMA_intAlloc(-1); // get an open TCC
gEdmaConfigXmt.opt |= EDMA_FMK(OPT,TCC,gXmtChan); // set TCC to gXmtChan

EDMA_config(hEdmaXmt, &gEdmaConfigXmt); // then configure the registers
EDMA_config(hEdmaReloadXmtPing, &gEdmaConfigXmt); // and the reload for Ping

gEdmaConfigXmt.src = EDMA_SRC_OF(gBufferXmtPong); // change the structure to have a source of Pong
EDMA_config(hEdmaReloadXmtPong, &gEdmaConfigXmt); // and configure the reload for Pong

EDMA_link(hEdmaXmt,hEdmaReloadXmtPong); // link the regs to Pong
EDMA_link(hEdmaReloadXmtPong,hEdmaReloadXmtPing); // link Pong to Ping
EDMA_link(hEdmaReloadXmtPing,hEdmaReloadXmtPong); // and link Ping to Pong

/* Configure receive channel */
hEdmaRcv = EDMA_open(EDMA_CHA_REVT1, EDMA_OPEN_RESET); // get hEdmaRcv handle and reset channel
hEdmaReloadRcvPing = EDMA_allocTable(-1); // get hEdmaReloadRcvPing handle
hEdmaReloadRcvPong = EDMA_allocTable(-1); // get hEdmaReloadRcvPong handle

gEdmaConfigRcv.src = MCBSP_getRcvAddr(hMcbsp1); // and the desination address to McBSP1 DXR

gRcvChan = EDMA_intAlloc(-1); // get an open TCC
gEdmaConfigRcv.opt |= EDMA_FMK(OPT,TCC,gRcvChan); // set TCC to gRcvChan

EDMA_config(hEdmaRcv, &gEdmaConfigRcv); // then configure the registers
EDMA_config(hEdmaReloadRcvPing, &gEdmaConfigRcv); // and the reload for Ping

gEdmaConfigRcv.dst = EDMA_DST_OF(gBufferRcvPong); // change the structure to have a destination of Pong
EDMA_config(hEdmaReloadRcvPong, &gEdmaConfigRcv); // and configure the reload for Pong

EDMA_link(hEdmaRcv,hEdmaReloadRcvPong); // link the regs to Pong
EDMA_link(hEdmaReloadRcvPong,hEdmaReloadRcvPing); // link Pong to Ping
EDMA_link(hEdmaReloadRcvPing,hEdmaReloadRcvPong); // and link Ping to Pong

/* Enable interrupts in the EDMA controller */
EDMA_intClear(gXmtChan);
EDMA_intClear(gRcvChan); // clear any possible spurious interrupts

EDMA_intEnable(gXmtChan); // enable EDMA interrupts (CIER)
EDMA_intEnable(gRcvChan); // enable EDMA interrupts (CIER)

EDMA_enableChannel(hEdmaXmt); // enable EDMA channel
EDMA_enableChannel(hEdmaRcv); // enable EDMA channel

/* Do a dummy write to generate the first McBSP transmit event */
MCBSP_write(hMcbsp1, 0);
}
/*
* copyData() - Copy one buffer with length elements to another.
*/
void copyData(Int16 *inbuf, Int16 *outbuf, Int16 length)
{
Int16 i = 0;

for (i = 0; i < length; i++) {
outbuf[i] = inbuf[i];
}
}
/* ---------------------- Interrupt Service Routines -------------------- */

/*
* edmaHwi() - Interrupt service routine for the DMA transfer. It is
* triggered when a complete DMA receive frame has been
* transferred. The edmaHwi ISR is inserted into the interrupt
* vector table at compile time through a setting in the DSP/BIOS
* configuration under Scheduling --> HWI --> HWI_INT8. edmaHwi
* uses the DSP/BIOS Dispatcher to save register state and make
* sure the ISR co-exists with other DSP/BIOS functions.
*/
void edmaHwi(void)
{
static Uint32 pingOrPong = PING; // Ping-pong state variable
static Int16 xmtdone = 0, rcvdone = 0;

/* Check CIPR to see which transfer completed */
if (EDMA_intTest(gXmtChan))
{
EDMA_intClear(gXmtChan);
xmtdone = 1;
}
if (EDMA_intTest(gRcvChan))
{
EDMA_intClear(gRcvChan);
rcvdone = 1;
}

/* If both transfers complete, signal processBufferSwi to handle */
if (xmtdone && rcvdone)
{
if (pingOrPong==PING)
{
SWI_or(&processBufferSwi, PING);
pingOrPong = PONG;
} else
{
SWI_or(&processBufferSwi, PONG);
pingOrPong = PING;
}
rcvdone = 0;
xmtdone = 0;
}
}
/* ------------------------------- Threads ------------------------------ */

/*
* processBuffer() - Process audio data once it has been received.
*/
void processBuffer(void)
{
Uint32 pingPong;

/* Get contents of mailbox posted by edmaHwi */
pingPong = SWI_getmbox();

/* Copy data from transmit to receive, could process audio here */
if (pingPong == PING) {
/* Toggle LED #3 as a visual cue */
DSK6713_LED_toggle(3);

/* Copy receive PING buffer to transmit PING buffer */
copyData(gBufferRcvPing, gBufferXmtPing, BUFFSIZE);
} else {
/* Toggle LED #2 as a visual cue */
DSK6713_LED_toggle(2);

/* Copy receive PONG buffer to transmit PONG buffer */
copyData(gBufferRcvPong, gBufferXmtPong, BUFFSIZE);
}
}

/*
* blinkLED() - Periodic thread (PRD) that toggles LED #0 every 500ms if
* DIP switch #0 is depressed. The thread is configured
* in the DSP/BIOS configuration tool under Scheduling -->
* PRD --> PRD_blinkLed. The period is set there at 500
* ticks, with each tick corresponding to 1ms in real
* time.
*/
void blinkLED(void)
{
/* Toggle LED #0 if DIP switch #0 is off (depressed) */
if (!DSK6713_DIP_get(0))
DSK6713_LED_toggle(0);
}
/*
* load() - PRD that simulates a 20-25% dummy load on a 225MHz 6713 if
* DIP switch #1 is depressed. The thread is configured in
* the DSP/BIOS configuration tool under Scheduling --> PRD
* PRD_load. The period is set there at 10 ticks, which each tick
* corresponding to 1ms in real time.
*/
void load(void)
{
volatile Uint32 i;

if (!DSK6713_DIP_get(1))
for (i = 0; i < 30000; i++);
} Start a New Thread

Jeff-

Open CCS, select Open Project and then open the dsk_app.pjt file. After opening the project file, you should see
"Source Files", "Libraries", etc (and individual files underneath) in
the lefthand pane (window) in CCS. What do you
see?

I have alreay done this and ran the example successfully, now I want to modify this code according to my requirement. Is the aic23.c file just configuring the contents of AIC23 and dsk_app.c is the actual file where I need to make modifications according to my requirement ?

-BS

________________________________
From: Jeff Brower
To: BS
Cc: c...
Sent: Fri, May 21, 2010 8:04:30 PM
Subject: Re: [c6x] DSP Processor for Real Time FFT Calculation

BS-

> I Got the C6713 DSK but stuck with sampling of analog signal. There
> is one dsk_app.pjt example associated with the
> installation folder but its really a mess. It has two .c files one
> aic23.c and other is dsk_app.c file, both attached
> here. Now I don't know how to use them simply to sample I KHz sine
> wave.
>
> Can you please help me with this ?

Open CCS, select Open Project and then open the dsk_app.pjt file. After opening the project file, you should see
"Source Files", "Libraries", etc (and individual files underneath) in the lefthand pane (window) in CCS. What do you
see?

-Jeff

>> ________________________________
> From: Jeff Brower
> To: c...
> Sent: Fri, May 7, 2010 9:37:04 AM
> Subject: Re: [c6x] DSP Processor for Real Time FFT Calculation
>
> In case anyone else is following this thread, I'm reposting this because I messed up the cc addr.
>
> ------------ ---- Original Message ------------ -
> Subject: Re: [c6x] DSP Processor for Real Time FFT Calculation
> From: "Jeff Brower"
> Date: Sat, May 1, 2010 10:37 pm
> To: "BS EE"
> Cc: c6x@yahoogroup. com
> ------------ --------- --------- --------- --------
>
> BS EE-
>
>>> The concentrator is sending a signal "upstream"? At what
>>> bitrate, a few Hz? Does the concentrator send some type of
>>> periodic signal; i.e. does it modulate a carrier (fundamental
>>> frequency)? Or include a carrier to help identify its signal?
>>> If so you might expect that averaging many FFT output
>>> frames over a period of time would show the concentrator' s
>>> carrier, assuming the noise is random and averages out
>>> to be more or less flat (or a predictable frequency domain
>>> profile). What kind of noise do you have?
>>
>> Regarding hardware, have you looked at using one of the cheap
>>> TI DSK boards? Can you interface your signals to a standard
>>> audio type A/D converter (assuming it can sample down to DC)?
>>> Or do you need a more specialized A/D front end?
>>>
>>> -Jeff
>>
>> The bitrate in the communication between the concentrator and
>> the energy meter is usually less than 5 kbits/s (FSK and
>> BPSK modulation).
>
> Ok... I thought that bit rates in the multi-kHz range would get blocked by transformers and other line equipment...
> but maybe not if the distance traversed does not have too many of these elements.
>
>> We may have noise from, for example, motor drives (frequency
>> converters/inverter s), swithing power supplies, energy saving
>> lamps and other communication systems
>
> Ok.
>
>> (although there should
>> not be any other than the power utilities system between 3...95
>> kHz).
>
> Ya "signal shouldn't be noisy", right. But of course it's noisy as heck -- the story of signal processing.
>
>> Yes, at the moment I am thinking to use TMS320C67x DSP, it
>> will cost around US $395.
>
> This is not a bad choice for a starting point. But if you need more horsepower, you might look at C6472 or other high
> clock rate fixed-point DSK boards. How important is floating-point in your application?
>
>> It probably is neccessary to have some analog amplifier (with
>> adjustable/programm able gain) and at least an anti-aliasing
>> filter in front of the ADC.
>
> Amplifier maybe, anti-alias filter probably not. What are your input signal levels? At 5 kbps data rate you might be
> sampling anywhere from 10 to 50 kHz depending on how many samples per symbol you need in your BPSK waveform. The
> AIC23 codec on the DSK C6713 board is a sigma-delta type with very effective built-in anti-aliasing (16-bit,
> 2-channel, max sampling rate up to 96 kHz).
>
>> These are the quantities to be calculated.
>>
>> -Power Line Communication (PLC) Voltage
>> -Noise Level at the PLC frequencies
>> -Signal to Noise Ratio
>>
>> Average disturbance level at selected frequency bands
>> - 50... 95KHz, 70... 95KHz, 95... 150KHz
>> -max/mean/min
>
> If you need to accurately see frequencies about 48 kHz then built-in analog circuitry on the TI DSK boards is not
> going to be enough. You might consider a third-party DSK daughtermodule of some type, with higher sampling rates. If
> those are not sigma-delta type ADCs, then the issue of anti-aliasing gets back in the picture.
>
> -Jeff
>
>> ____________ _________ _________ __
>> From: Jeff Brower
>> To: BS EE
>> Cc: c6x@yahoogroups. com
>> Sent: Wed, April 28, 2010 12:06:14 AM
>> Subject: Re: [c6x] DSP Processor for Real Time FFT Calculation
>> BS EE-
>>
>>> Yes, it is a student project that can be interfaced with
>>> Automatic Meter Reading
>>> equipment. The part I am working on is to measure the
>>> quality of
>>> power line communication signals. Some time meter reader
>>> doesn't recognize that wheather the coming signal is sent by
>>> concentrator (i.e. many meter connected to this unit) or its
>>> just noisy signal.
>>>
>>> Any help, please ?
>>
>> The concentrator is sending a signal "upstream"? At what bitrate, a few Hz? Does the concentrator send some type
> of periodic signal; i.e. does it modulate a carrier (fundamental frequency)? Or include a carrier to help identify
> its signal? If so you might expect that averaging many FFT output frames over a period of time would show the
> concentrator' s carrier, assuming the noise is random and averages out to be more or less flat (or a predictable
> frequency domain profile). What kind of noise do you have?
>>
>> Regarding hardware, have you looked at using one of the cheap TI DSK boards? Can you interface your signals to a
> standard audio type A/D converter (assuming it can sample down to DC)? Or do you need a more specialized A/D front
> end?
>>
>> -Jeff
>>
>> PS. Please don't cut text from your initial post. I read 100s of post every day and I don't remember your original
> question and I'm not going to waste time to go back and look.
>>
>>> ____________ _________ _________ __
>>> From: Jeff Brower
>>> To: matlab_fft@yahoo. com
>>> Cc: c6x@yahoogroups. com
>>> Sent: Mon, April 26, 2010 9:22:10 PM
>>> Subject: Re: [c6x] DSP Processor for Real Time FFT Calculation
>>>
>>> EE-
>>>
>>>> I want to build a real time FFT analyzer, my frequency range at the moment is 3Khz to 150KHz. I need to perform
>>>> 2048-points FFT or even more.
>>>>
>>>> I am looking for a combination of (DSP+Microcontrolle r) device, actually I am doing modifications in the existing
>>>> device that uses MCU but I need DSP to process FFT algorithm.
>>>>
>>>> I would like to know if any of you have ever used this kind of processors for FFT application, please help me out
>>>> in
>>>> choosing a DSP+MCU unit for my application.
>>>>
>>>> Eagerly waiting for your reply.
>>>
>>> Is this a student project? What is the "existing device"?
>>>
>>> -Jeff Start a New Thread