signal

[madd]

Hi

I have signals generated in matlab "wave file" and i am reading these signals in Labview through "read file" there is a diffrence in the frequencies of the signal generated in Matlab and Signal detected(read) in Labview eg.

signal frequency generated in matlab                  signal frequency detected in labview

                1100Hz                                                                        1096.79Hz

                1200Hz                                                                        1195.99Hz

                1500Hz                                                                        1510.09Hz

                1700Hz                                                                        1750.31Hz

                2000Hz                                                                        2000Hz

                2200Hz                                                                        2173.23Hz

                2500Hz                                                                        2387.23Hz

               2700Hz                                                                        1345.99Hz

why are these frequencies different at many points and why are they same at some(2000Hz) point. i have used Tone measurement and Transition measurement VI's to determine the frequencies of signal.

Regards

Madd  

[Brian_T]

madd,

Are the signals generated in Matlab exactly 1100 Hz, 1200 Hz, etc?  Or is that a different displayed number that might have the significant digits set to 2?  If you can explain how a little more about how the signals are generated in Matlab it may help.  Thanks

Brian T
Applications Engineer

[madd]

Tnx

For ur response. I have signals which i have modulated in Matlab using "MPSK" simulink block providing it with the frequencies as mention above.The modulated waveform ,when i claculate the frequrncy by claculating time period of waveform and taking its inverse(f=1/T) just by watching the plot of modulated waveform it comes exaxtly same as i have mentioned in "MPSK" block.But when i read these frequencies in Labview they have a pattren as i have mentioned befor.Is there any thing wrong with my version i am using . I am using 8.5 evaluation version. which i dont think will be. Looking for ur further guidence.

Regards

Madd

[Brian_T]

How exactly are you calculating the period of one waveform?  And what sampling rate are you using in order to derive this?  You may need to increase the sampling rate when you are reading the signal in LabVIEW. 

[madd]

HI

LOOKING AT  THE SCOPE SIGNAL I HAVE GENERATED I CAN FIND THE TIME PERIOD OF THE SIGNAL AND CAN CALCULATE THE FREQUENCY . ALSO HERE I HAVE ATTACHED THE SIGNALS I AM GENERATING UISNG "mpsk" I AM GENERATING ALL THE FREQUENCIES USING THIS CARRIER FREQUENCY.

REGARDS

MADD

[Brian_T]

Could you post your code, both from Matlab and LabVIEW, so I can get a better idea of whats going on?  Thanks

Brian T

[COMSIS]

Hi

i am really very sorry everyday i was watching ur response but i now i saw that i have no posted the attachements i have mentioned in my previous reply.here are the matlb attachments and the labview attachments i have made .i am setting the carrier frequecncy by using matlab "Carrier Frrequency"(2000) block option.the scope plot i am using is to physically calculate the time period of wave and taking inverse to find frequency. In labview two Numeric indicator tells different story regarding frequecy of the generated signal(in Matlab).one shows a shlight diffrence and other has more.

Regards

Madd(Now COMSIS)

[COMSIS]

ooops

[madd]

HI

can u also explain me about timing recovery in PSK/QPSK(i have gone through white paper on NI platform) .Or can u guide me about some forum where i can find it by implimentation perspective not by long theories and equations. how would i implimnet it in Labview. any good clue or book or paper which can be helpful in labview implimentation of timing recovery. I know this is not a proper platform to ask this question but Generalllly if u could help

[Eli_S]

COMSIS,

The reason why you are observing different measurements using the Tone Measurement VI and the Timing and Transition Measurements VI is because the way each of those VI's actually work. With the Tone measurement VI, which measures the period of a signal and takes the inverse, it is possible that the funtion misses the actual period by a few samples, or includes a few additional samples in the same period. This creates what is called spectral leakage, which when taking the FFT for a final frequency determinatio, makes it seem as if the FTT is shifted slightly and that is why you seem to be reading a slighly different frequency. Conversly, the Timing and Transition VI, actually counts the rising edges and measures the dt. This is a more accurate measurement because the is no room for spectral leakage to take place.

For more information about spectral leakage please refer to the following links;

http://zone.ni.com/devzone/cda/tut/p/id/4892
http://digital.ni.com/public.nsf/allkb/2A8C234B3F3F4B81862567E60073D997
http://digital.ni.com/public.nsf/allkb/4985224270D324A086256936005879BB