power measurement using field point

[SimonH]

sumi,

Forgive me if I am interpreting this thread incorrectly but it looks like you are using VIs such as the RMS.vi which calculates the RMS value of an array of values.  Typically examples built around DAQmx or simulated signals will read/generate many samples on every iteration - returning an array of values which can be passed to RMS.vi. 

A FieldPoint read on the other hand reads in one point at a time.  Consequently taking the output of a FieldPoint read and passing it directly to the RMS.vi will generate the RMS of a single value.  This would explain why your RMS and peak values are just the instantaneous values (the average of x is x and the peak of x is x).

As Ching mentioned earlier in the thread take a look at the point-by-point VIs such as RMS PtByPt.vi.  These VIs store previous values and add the new value to the array of historical values before computing the new output value.

This being said I would still echo centerbolt's concerns that you aren't sampling the input signal quickly enough to gain an accurate representation of the input waveform.

Regards,

Simon H
Applications Engineer
National Instruments
http://www.ni.com/support/

[sumi]

i here attach three sample programs to measure the rms values. In the first two programs rms.vi and rms_new.vi the rms value showed instantaneous variations as mentioned. We also tried many vi's related to the rms value measurement and the result was the same. the third program, basic scalar measurement.vi running smoothly as such with the waveform generation vi's. But when i modified the program with field point vi's i got the usual result. How can i solve my problem?

[sumi]

i got the rms value constant by using the rmsPtByPt.vi

Only when large sampling length of order 10000 are employed the rms value is converging. I am measuring signals at 50 Hz frequency. But when large sampling lengths are employed the vi take large time to keep track the changes.

Is there any method to specify the sampling length for a complete cycles of an 50 Hz frequency?

[Henrik_Volkers]

back to my first reply ...: if you want to measure a 50HZ rms signal you need a (much) higher sampling rate than 100ms!!!

If you sample a 50Hz sinewave every 100ms you get a pur DC signal because you make one measurement at the same point every fives wave.One 50Hz wave take 20ms (1s/50)

However you can try a sampling rate a little of the main frequenz 11ms (if possible ) or 99ms ....

If you sample with 99ms you need 20 values to get an idea of your 50Hz wave. Since your system isn't ideal you might need 10 or 100 times more samples to get a reasonable result. 

Make youself familay with BASICS of nyquist theorem and undersampling and how to determ the minimum sampling rate for your signal. Sorry for sounding a little picky. It's learning by *DBOING*