how to reduce noice in garph without changing the sample rate

[sridhar..reddy]

HELLO...

I am using NI-CDAQ9184 module with NI- 9203 Analog input module.

i am using this module to caluculate currents and pressure.

 samples to read is 1 and rate 25hz.

with this sample rate i am getting more accurate values but the thing is ,in the graphs lot of noise is appearing.

i want to remove that noise without changing the sample rate,i tried by changing the samples to read as 1 and rate 10 its is working fine but

when i used this sample rate i am losing the data .

samples to read 1 rate 25 ---- with this configuration i am able to get 10 samples in my excel sheet.

samples to read 1 rate 10---- with this configuration  i am able get 5 samples in my excel sheet.

my requirement is to get 10 samples in my excel sheet and i must not change the sample rate or else i can increae increase ,but i should not decrease the sample rate.

Kindly suggest me that how to reduce the noise and how to get the graphs clearly noise less.

[James.M]

There are many diffferent filters you can use to eliminate noise. You could simply run an averaging function on your data before the graph. You could also use a lowpass filter. It really depends what aspect of your data is most important.

[Bob_Schor]

What do you know about signals, noise, and statistics?  There are many forms of noise, one of which is the variability inherent in making any measurement (try measuring the size of your room with a ruler -- do it 5 times and you'll get 5 slightly different answers).  If the measurement is "static" (not time-varying), a simple arithmetic mean is a good technique to reduce the variability of the signal.  [If you've taken any statistics, you'll know that the variance of a mean is the variance of the sample divided by the number of samples).

If you are analyzing a time-varying signal, it gets a little more complicated as there are two sources of variation -- the inherent variation of measurement and the time variation of the signal itself.  Functions that manipulate time-varying signals are often called "filters" -- do you know about them?  Have you heard of "low-pass filters" and do you understand what they are and do?

It helps to have a good understanding of what you want to do (to improve the precision of your measurements) before worrying about how to implement it.  Actually, with the understanding, the implementation will "suggest itself".

Bob Schor

[sridhar..reddy]

hi...

thanks for your reply..

what you said is right , i am going to put mean function befor the graph ....i think it may works.

Thank you......