accuracy and sensitivity

[Pells]

I have problems understanding the difference between range accuracy and range sensitivity and if it will be possible for me to measure a very small signal without filtering it with a highpass to get rid of a dc-level.

The signal I want to sample is a sinus signal with an amplitude of 3mV but on the signal is a dc level wich moves the signal up so it varies for example between 3-6mV or even up to 12-15mV,  I amplify this signal with an amplification of 600 so that it instead varies between 1.8V-3.6V or up to  7.2V-9V.

My question is: how small variations in this signal will I be able to detect with a NI USB-6210 card? Sometimes I'm interested in variations lower than 10mV, will there be any problems with that since I have to use the maximum voltage range of the 6210 card to fit the signals that are up to 9V?

[PelleS]

Hi,

I believe that whether or not you can measure this signal depends on how well you can eliminate the noise. If I understood you correctly, you amplify the signal in hardware since you use the maximum range. If your conditioning takes place near the source and thus yields a high signal-to-noise ratio, there will be no problem to measure your signal. However, if the conditioning occurs after the noise is added, then you will have problem extracting the signal from the noise without any filtering. 

Sensitivity is a measure of the smallest signal the instrument can measure. According to page 3 in the Specs the sensitivity on the maximum range is 0.0916 mV (determined by a function of noise). That is well enough for a 10 mV variation.

This link explains the different terms. Please let me know if you have any further questions.

Best regards,