Our full technical support staff does not monitor this forum. If you need assistance from a member of our staff, please submit your question from the Ask a Question page.


Log in or register to post/reply in the forum.

Measure 10KHz sinewave timehistory


AlexChao Jun 14, 2011 11:29 AM

Dear Sir

We measure 10KHz sinewave timehistory data by CR1000,how do it ? The voltage of sinewave is about 1000mV.We hope to get 10K data to analysis.Thanks a lot.


Sam Jun 20, 2011 02:57 AM

I think the best a CR1000 could do is to take bursts of 1.3 KHz measurements - not continuous, but bursts. About 750 usec per measurement.

VoltDiff (Dest,100,mV5000,-1,False,500,250,1.0,0)


aps Jun 20, 2011 10:59 AM

Sam's suggestion would allow you to capture the raw waveform which could then be processed by the datalogger or post collection to give measures of frequency and peak, however you would need to be very careful doing this because you will be undersampling.

If you just want to measure variations in frequency then I suggest you look at using the period average measurement to sample the frequency (it can deal with the voltage and frequency OK). You can then just store the frequency or period at given intervals or do more advanced statistics.


AlexChao Jun 20, 2011 01:01 PM

Thanks a lot.
We don't get right frequency to use period average to measure frequency that varied as time,and we think that samples are not stable.So we want to get samples and FFT to get peak frequency to replace period average.Is it right?

Best Regards


aps Jun 20, 2011 01:30 PM

The period average will measure frequency to an accuracy of 0.01% or better, that is providing you have selected the right range. What you are probably seeing is either noise on the signal or real variation in the frequency. Chose the highest range possible to help avoid issues with noise and a reasonable number of cycles to count too which will act to average out short term fluctuations.

Using an FFT analysis will allow you to pickout a frequency where these is some noise but you wont be able to do this well with the CR1000 as ideally you should sample the waveform at least 2X the maximum frequency of the signal, i.e. 20 kHz in this case. The CR1000 can only run at about 1.3 kHz so it is a non-starter.

Log in or register to post/reply in the forum.