LabVIEW application targeting DSP

[wired]

I am developing an application that requires algorithms to be developed for the detection of voice activity.  I plan to target these algorithms on an embedded DSP, and am currently looking at the Blackfin, the Sharc, and the TI 6000 family of DSPs.  I need to have the ability to customize the settings for these algorithms using a PC application that can send configuration info to the device over USB.

To minimize development time, I would like to develop the PC application in LabVIEW, and would also like to perform some rapid prototyping of the algorithms using LabVIEW.  I see that the LabVIEW DSP Module would allow me to do rapid-prototyping on certain evaluation boards, such as TI's 6713 board.  I also see that LabVIEW Embedded has a Blackfin interface for essentially doing the same thing, although the Embedded app can be applied to custom targets.  My questions are:

1) Can a LabVIEW application written with LabVIEW DSP Module be built into a standalone application that would allow me to load the application and configuration parameters to the board, or do I require the entire LabVIEW development environment?  Can I set up the A/Ds and DACs on the eval board from LabVIEW, and control the digital I/O as well, i.e. can the entire application be written in LabVIEW, or will I need to do some work with Code Composer/VisualDSP++?

2) What are the limitations of DSP Module?  Can I target different memory types, program on-board peripherals, etc. ?  Is there any way to target a custom board using the generated code from DSP Module?

3) Would LabVIEW Embedded allow me to do what I have described above? Am I limited to the Blackfin, or are the Sharc and TI chips potential targets?

[Justin D]

Hi wired,

1) The LabVIEW DSP module essentially builds the exe that runs on the DSP module and that's it.  There is no further communication between the embedded application and windows.  As such you can't build an exe that runs on the windows environment.  You can build your embedded exe and put it on however many DSP targets you want though (as long as they are all of the same type and one of the targets supported in LabVIEW DSP).

2) The DSP module was built to be as easy and user friendly as possible.  As such you sacrifice some of the flexability of your application.  You cannot target different memory types etc.  Also, there is no way to target a custom DSP board.  You can only use one of the 4 targets supported by the DSP module.

3) LabVIEW Embedded would allow you to do all that you have described above.  LabVIEW Embedded allows you to define how your code gets compiled down to your custom DSP module.  You are not limited to any specific embedded target.  If you use the Blackfin edition you will be able to use LabVIEW Embedded to generate code for the Blackfin board right out of the box.

-Justin D

[Caputo]

Greetings to all.

I develop applications in contrlole acoustic using technology DSP6713 and LabVIEW 8,2 DSP module. I am having difficulties in the programming of acquisition of sound data using the I/O of my Started KIT DSK 6713. Somebody can help me?

E-mail

rodrigo_caputo@yahoo.com.br

[Riconquistiamola]

Hi Caputo,

You're going to have to expand on your problem thus far.

What have you tried?
What hasn't worked?
What is your programming algorithm?

With more information I am confident we can solve your problem.

Cheers.

[Caputo]

Hi Michael,

I study would engenaharia in Brazil and participate of a project of research in acustica.

We are trying to implement the Algorítimo of control LMS in technology DSP - in particular DSK 6713.

using Lab VIEW for DSP.

My objet is reduction noise in closed place.

I configured the I/O of line of my DSK6713 to use the canals left and right with being been the refers of entrance of (LMS-Algorítimo of control). But I am finding problems… the exit signal is not the waited one…

You know this algorítimo of control? LMS Algorítimo? Already it worked with DSK 6713?

In  LabVIEW you can see the example LMS Filter...But My signals of entrance are real and not controlled as of example of the LabVIEW…

Best Regards.

Caputo

[Riconquistiamola]

Hi Caputo,

I am having language problems, but do not worry. Your English is better than my Portugese . Here is what I think you are asking (and my answers).

You are running a DSP project using the DSK 6713. You want to use the LMS algorithm (least-mean-squares) in this project. You are using the LMS Adaptive Filter (C:\Program Files\National Instruments\LabVIEW 8.2\Targets\NI\DSP\vi.lib\comm\LMS Adaptive Filter.vi), but it is not working as expected.

"the exit signal is not the waited one. My signals of entrance are real and not controlled as of example of the LabVIEW"

I admit I do not understand what this means. I found this document on the LMS Adaptive Filter: How Does the LMS Adaptive Filter for DSP Work? [link removed; article no longer available].

Does this help? If not, can we find a person to translate your problem more specifically? I apologize again if this sounds insulting, as I am having difficulty solving your problem because of the language difference.

Cheers.

[wired]

Michael,

I think Caputo is saying that he is not getting the expected output from the filter when the input signals are real-world signals, e.g. from a microphone, rather than an artificially generated signal, which is probably what is being done in the LabVIEW example.  The first question would be 'how do you know what the expected output is?'.  What is being used as a reference?  Perhaps Caputo should collect a data sample and save it to a file, then use LabVIEW's regular signal processing VI to examine the output.  Maybe there is an an instrumentation problem.  Were it me, I would use a DAQ card to collect the real-world data, perform all of my signal processing and analysis on the PC, and verify that my algorithm is correct.  Only when I was confident that the LabVIEW code was working as expected would I port it to the DSP environment.

[Riconquistiamola]

Hi wired,

Thanks for the input. I think you have adequately repeated and answered the question.

I'm willing to bet there is a general example with results for expected output somewhere online.

Caputo, does this suggestion work for you?

Cheers.

[Caputo]

Hi Wired and Michael!!!

Thanks a lot for the attention.

My English is not ones of the best ones : but I go to more clearly.

My aplicattion is :Adaptive Noise Cancellation is used to remove background noise from useful signals. This is an extremely useful technique where a signal is submerged in a very noisy environment. A typical example is inside a jet aircraft. The jet engine can produce a noise over 140dB. Since normal human speech is at a level between 30 and 40 dB, the pilot's communication is impossible in such a environment if there no noise cancellation equipments inside the cockpit.

I go to follow the advise of the Wired. Testing the trustworthiness of my equipment of first acquisition and later leaving for the tests with algorítimo of control - LMS. I´m using no professional equipaments.

Michael, What are you suggest  to use to eliminate the errors of acquisition of available acoustic data for the National Instruments?

Greetings!!!

rodrigo_caputo@yahoo.com.br

[Riconquistiamola]

Hi Caputo,

I think the best answer here is to ask if you've implemented wired's suggestions? If you've got file information, can you test it with LabVIEW's regular signal processing VI to examine the output?

Cheers.