
#1
Jan3012, 10:20 AM

P: 44

Hi all
I have measured the resonance frequency of a resonator as a function of time by mean of a LaserDoppler vibrometer and a frequency counter. In this measurement there are some fluctuations in the measured frequency as time goes by. My question is: How do I determine the resolution of this frequency measurement? Or in another way: How precise can I resolve the frequency? Do I take the standard deviation for a period and then define this as my frequency resolution? I hope someone can help me. Best regards Excom 



#2
Jan3012, 11:27 AM

Sci Advisor
PF Gold
P: 11,363

That would be the RMS error. A pretty universal way of stating errors  I say go for it.




#3
Feb112, 07:13 AM

P: 44

Thanks:)




#4
May1913, 11:24 PM

P: 44

Resolution of F vs t measurement
A more correct way of doing it is to calculate the Allan deviation as a function of integration time and then take the minimum value.




#5
May2013, 04:33 AM

Sci Advisor
PF Gold
P: 2,197

Indeed. the standard deviation is not a good measure in these situations. The reason is that it only converges for white noise, for all other noise types the STD will give you the "wrong" result which is why the Allen deviation should always be used for this type of analysis.
Have a look at the NIST handbook of frequency analysis (free download). Also, there are free Matlab routines for calculating the ADEV available on Mathworks file exchange. 


Register to reply 
Related Discussions  
Actual earth measurement contradicts measurement predicted by special relativity  Special & General Relativity  93  
Resolution measurement for sem  Atomic, Solid State, Comp. Physics  0  
about time measurement vs measurement of movement  Special & General Relativity  21  
resolution of eye  General Physics  5  
eye resolution  General Physics  1 