topic Re: ( active noise control using loudspeaker ) in LabVIEW

( active noise control using loudspeaker )

ALI2020 — Wed, 25 Dec 2013 16:51:17 GMT

Actually I have project to cancel noise ( active noise control using loudspeaker ) and my signal processing is labeled and am using labeled 2011 I want to squire signal from speaker that is connected to signal generator and the parameters of this signal want to send it out to loudspeaker connected to the laptop so it can cancel eachother and no sound detected >>> i have used DAQ to eliminate the delay as I worked before without DAQ and there was delay >>> now when the signal came out to the canceling speaker isn't match with the signal that is generated >>>>>I want to make the signal out of phase 180 so that it can cancel >>>>> I hope if u can help me i will appreciate this very very much thanks again

Re: ( active noise control using loudspeaker )

johnsold — Wed, 25 Dec 2013 17:45:47 GMT

Please do not post the same message to multiple threads. The other threads were for similar but not identical questions and are several years old, so please keep this discussion here.

It is not clear from your post what particular problems you have. Please be as specific as possible.

What is the frequency of the signal you want to cancel?

What is the DAQ sampling rate? How many samples do you acquire with each read?

How are you measuring the delay you mention? What is the timing reference?

How are you generating the output signal? How are you trying to create the phase inversion?

How are you determining the effectiveness of the cancellation?

What is the physical configuration of the speakers and microphones?

Lynn

Re: ( active noise control using loudspeaker )

ALI2020 — Wed, 25 Dec 2013 18:54:55 GMT

Thanks a lot and I will keep the discussion in here
my aim is to cancel frequency less than KHz and my experiment I use 300 hz
my daq sample rate is Thousand samples per second
the samples I a quire is 10000
I tried to measure the time delay between signal input and output by autocorrelation ( timing reference I don't used I hope if I know)
I am generating output signal through DAQ output that is connected to speaker

i am trying to make the phase inversion but i donot know how this is one of the problems that i want to solve and asking about it I just tried by getting the same parameters from the input signal and send it to the speaker and play with phase by slider

effectiveness of the cancellation I know by the delay in output and no sound reduced between input and output

The speaker and microphone configurations( the microphone is between the two speakers ( one speaker as input connected to signal generator(getting signal input ) and another speaker connected to the PC as the output device )

I have attached the code that I am trying and I hope u can help me because I amount that professional in labeled am in learning time
thanks million in advance

Re: ( active noise control using loudspeaker )

ALI2020 — Wed, 25 Dec 2013 18:56:59 GMT

this is correction lyn i apologize
my frequency is less than 1 khz
my daq sample rate is 120000 samples per second

Re: ( active noise control using loudspeaker )

johnsold — Wed, 25 Dec 2013 20:19:21 GMT

You have several problems.

The Acquire Sound VI collect 10 seconds of sound before it returns to the main VI. Thus, you have at least a 10 second delay between the sound produced by the external signal generator and the beginning of the processing in your VI. That will absolutely prevent any realistic cancellation.

The sound acquisition rate is 48 kHz. The Simulate Signal sample rate is 10 kHz. The DAQ Assistant for the output appears to have a 10 kHz sample rate although you posted a correction saying it is 120 kHz. While not absolutely necessary, it makes things easier if all the sampling rates are the same.

Array to Cluster -> Unbundle -> Bundle -> Unbundle ??? A simple Index Array expanded to two elements would give the same result. The first two elements of the Power Spectrum array coming out of the Power Spectrum VI represent the power at DC and the power at frequency df, where df = fs/N. If fs = 48000 Hz and N = 10 seconds*48 kHz, then df = 0.1 Hz. (Read the Detailed Help for the Power Spectrum.vi). Most sound cards are AC coupled so it is likely that the values of the first two elements are very close to zero. The frequency should be set to Index of first PEAK (not first element)*df and the amplitude to the amplitude of the first peak. EXCEPT that the spectral peaks may not be exactly at the frequency of the signal due to spectral leakage and other effects. Try the Extract Single Tone Information.vi. It finds the frequency and amplitude of the dominant tone.

To get effective cancellation the time delay between the input signal and the output signal must be minimized. To get effective frequency estimation, the number of cycles of the input signal should be large. Any program running under Windows (since you are using the DAQ Assistant, you must be using Windows) will experience unpredictable delays on the order of milliseconds to tens of milliseconds. Such delays are much larger than 180 degrees of phase shift at frequences of a few hundred hertz. So you may be able to demonstrate proof of concept but you will not get really good, consistent cancellation.

Lynn

Re: ( active noise control using loudspeaker )

ALI2020 — Thu, 26 Dec 2013 07:35:58 GMT

thank you very very much lynn
firstly i got all ur useful information and i will take advantage of it >>>

second thing i am afraid that my concept is not that clear so can u tell me by edittting my programm attatched so i can know more and if u can add or remove any fucntion that u can see it is useful i will apperciate that .... lynn i amnot in engineeing sound feild rather i am in mechatronics so i donot have background for things like this thats why i m stcuk in my project .... i have searched and i couldnot find certain guide that can guide me to zoom in the idea in my mind ....

Re: ( active noise control using loudspeaker )

ALI2020 — Fri, 27 Dec 2013 14:32:36 GMT

lynn firstly i would like to thank u and i want to say that i need ur help hope u can reply me
""""To get effective cancellation the time delay between the input signal and the output signal must be minimized. To get effective frequency estimation, the number of cycles of the input signal should be large."" how i could do that

Re: ( active noise control using loudspeaker )

johnsold — Fri, 27 Dec 2013 19:22:40 GMT

That is what engineering is all about - choosing the compromise between conflicting requirements which best matches the needs of the customer!

The VI attached simulates cancellation in two ways. I hope this will help you understand some of the issues involved.

It generates a sine wave of frequency 10.1 Hz with amplitude 1 and phase 0. This is displayed on the graph Sine phase 0. It uses the Tones and Noise.vi. Although I used only one tone and no noise, this VI was chosen so that a more complicated simulation could be added if desired. Note that reset phase is True on all the Tones and Noise VIs.

A second copy of the Tone and Noise.vi generates another sine with the same frequency and amplitude but with the phase controlled by the phase control on the front panel. The phase is in degrees. It is displayed on Sine plus phase. Odd multiples of 180 degrees should give cancellation. The two sine waves are added and displayed on Sum of sines.

The second method is a model of what you are trying to do with your microphone, speakers, and DAQ system. It takes a subset of the original sine wave (displayed on Subset) and uses the Extract Single Tone Information.vi to get the frequency, phase, and amplitude of the subset. The use of the subset is to show the effects of trying to measure the frequency quickly. The outputs of the Extract Single Tone Information VI are used as inputs to another Tones and Noise VI (with 180 degrees added to the phase). It is displayed on Sim sine. This represents the cancellation signal you are trying to generate. It is added to the original sine and the sum is displayed on sim sum.

Run the VI with various values of phase and length to see the effects. Notice that the cancellation is not perfect and not identical at 180, 540, and 900 degrees. This is due to round off effects related to the finite size of the representation of numbers in the computer. Notice that the sim sum "cancellation" is orders of magnitude worse although still quite good until the length gets below 200 which results in 2 full cycles of the sine wave going to the Extract Single Tone Information.vi.

Now consider that noise, sampling effects, and the phase shift due to the time the sound takes to travel from speaker to microphone will all add more uncertainty to the inputs to the cancellation sine generator.

Lynn

Re: ( active noise control using loudspeaker )

ALI2020 — Fri, 27 Dec 2013 20:00:08 GMT

thanks alot and i really i am thankful for u
by the way how can i merge this code with my old code as i am confusing and i don't get the whole idea about the code so i can try and see
it is required that i aquire the signal and send the output to the daq that is connected to the speaker >>>> my question is how i match my old code with the new file u send me ???? hope u got my question
thanks in advance and i have to no way to express my grateful to appreciate ur help

Re: ( active noise control using loudspeaker )

ALI2020 — Tue, 31 Dec 2013 17:37:49 GMT

hello lynn
hope i donot disturbe u but i really need ur help i got ur useful porgam that u sent to me but i couldnot use it as i donot understand it well
can i use this ( as i explained to u my project ) >>>> how please ?
thanks alot
regards

Re: ( active noise control using loudspeaker )

johnsold — Tue, 31 Dec 2013 18:23:09 GMT

Please try to use complete words such as "you" rather than "u". For many of the people who use these Forums English is not their primary language and the use of shortened forms makes it hard to understand.

Consider the steps which must be performed.

-The program must acquire a signal from the microphone.

-It must analyze that signal to determine the frequency, phase, and amplitude of the largest component.

-It must generate a new signal at that frequency and determine the appropriate phase and amplitude to produce cancellation.

-It must send that new signal to the DAQ AO line.

Now determine which parts of my program perform those tasks or simulations of them. Replace those parts with the code which uses the real signals. Modify the program control structure to work with real signals.

For example you have a VI which obtains a signal from the Sound In port of your computer. That corresponds to the Tones and Noise.vi connected to Sine phase 0 in my program. You may also need to change the timing as I discussed earlier. I suggest starting with segments of 0.5 to 1 second and then make chages as needed.

Try this process and see how far you can get.

Lynn

Re: ( active noise control using loudspeaker )

ALI2020 — Tue, 31 Dec 2013 18:35:28 GMT

Hello Lynn
thank you very much and I apologize for the short form I have used as I am new and next time I willnot do that for the benefit to others ..
Actually Lynn I found program from internet which can do my project as now I have no time and I have tried but I am quite sure all my theory that I have implemented and no point achieved so hope if you can analyze the file attached and tell me why input and output data are missing file or have been moved ?
Please Lynn this will help me a lot
thanks
regards

Re: ( active noise control using loudspeaker )

NIquist — Tue, 31 Dec 2013 19:26:15 GMT

The file you're looking for will be different for every microphone/speaker combination. It represents the audio path between them so it is required for the code to work properly. However, even if you could generate that file you will still need a fast, real-time PC to actually do ANC (as Lynn has already mentioned).

Sorry to say it but I think you may be attempting a project that is beyond your skills and resources. Perhaps there's a way to try something a little simpler?

Also being handled here: http://forums.ni.com/t5/LabVIEW/Enquiry-about-missing-file-in-program-attached/td-p/2680287/highlight/false/page/2

Re: ( active noise control using loudspeaker )

ALI2020 — Tue, 31 Dec 2013 19:39:30 GMT

hello NIquist
thanks
"I think you may be attempting a project that is beyond your skills and resources" this is totaly true and i know it basically labview and its functions needs time and beside that sound properties and the interfacing and detecting beside the elemination of delay is beyond my knowledge which i cannot do this withing three months and i have faced many problems regarding labview labraries and i spent time learning labview but i still consider as beginer ... my plan was to get the code to help me analyze it , understand and then be able to achieve my project as labview coding isnot in my major and i wish if i have time to do anything >>>> thanks all

Re: ( active noise control using loudspeaker )

NIquist — Tue, 31 Dec 2013 19:47:36 GMT

Well again, perhaps there is a simpler application that you could try like showing why the real-time cancellation of one sound signal by another is so difficult when you must deal with the physical properties of acoustics and the limitations of computers. There is a reason that all the noise cancellation products on the market are based on headphones. Headphones minimize the variables like the audio path and reflections and allow a small, fast processor inside to do the math in real-time.

Your idea was very sound. Don't put it aside completely. Just look for a project that is within your resources and push ahead. Learning LabVIEW takes time but there's no better way than creating a project that interests you to make the learning fun. Good luck and post back if you need further help...

Re: ( active noise control using loudspeaker )

ALI2020 — Tue, 31 Dec 2013 20:08:19 GMT

hello NIquist
firstly thank you very much based on your advices i should look back for resources help me understand it this is for future analysis and study for the time being as student and don't have that much knowledge I was trying how to implement based on any code whoever make it as I tried days but no point ...
I hope if can do it and then if I have any enquiry I can post it back
really I appreciate all your help your guides and your valuable advices as expert .....
Regards

Re: ( active noise control using loudspeaker )

ALI2020 — Tue, 31 Dec 2013 20:15:12 GMT

hello NIquist
i have just basic questions and i need to learn such things hope you can answer me

how can i know the exact delay between input and output signal ? is there such function in labview ? i have used co and crossrealtion but alawys give me the same value even if i changed and also repeated thing in 100 ms ...
is there such function in labview can do phase shifting ?
what is the meaning of output array basically ?

Re: ( active noise control using loudspeaker )

NIquist — Fri, 03 Jan 2014 15:28:40 GMT

Stick with the time domain if you want to measure time. There's no reason to mess with phase. Just record a sharp noise like a hand clap or a digital pulse and measure the delay by subtracting the timestamps.

Re: ( active noise control using loudspeaker )

ALI2020 — Fri, 03 Jan 2014 15:55:14 GMT

Hello NIquist
thanks a lot for your answer am trying to do small experiment first to understand more and read resources then if i face any problem i can highlight here thank you very much

regards

Re: ( active noise control using loudspeaker )

ALI2020 — Wed, 08 Jan 2014 15:27:28 GMT

Hello everyone
hope you can help me by seeing my file and tell me what is the things that is not needed I am trying to get signal and take this signal parameters and send it to the daq (that is connected to amplified speaker ) with phase shift 180 so it can cancel out the signal acquired but when I implement this in lab I couldn't get the result expected can you please help me figure it out >>> sorry to all of you I,m not expert in this field ..... Thank you very much