What I am trying to do is record 20 different AC voltage channels ranging from 0 to 7Vac RMS @ 3000 Hz each (sine waves). The recording software I want to use is NI SignalExpress. The voltage range should not be an issue due to the max voltage being right below 10V.
I need to record for several hours at a time so I want to avoid just using a very high sample rate due to the size of the data. Normally I would just convert from AC to DC with external circuitry before running this signal into my DAQ card, but would also like to avoid that method due to having 20 different channels I would need to do it to.
Is there a way with any of the NI DAQ cards to convert this input signal to an RMS value (DC voltage) so that I can record at a slow sample rate (200 Hz or less)? I checked out all the voltage DAQ cards but I could not find an answer. I have used many different cards for DC voltage measurements but needing to record AC signals is new to me.
I am not familiar with all of NI's products but standard DAQ cards typically do not have rms converters. The DMM modules may have that but are probably overkill (in both performance and price) for your application.
Usually people measuring AC signals sample at a high rate and do the RMS conversion in software. The data size would not get large. Suppose you sample each channel at 25 kHz (500 kHz aggregate). Read every 100 ms. That gives you 2500 samples per channel. Calculate RMS for the 2500 samples and save 1 value.Throw away the raw data after the RMS conversion. 200 data points per second. 720000 point per hour. Calculating the RMS value of 2500 samples takes a few milliseconds. Doing it for all 20 channels in 100 ms should be feasible.
If the AC amplitude is your information, why do you use RMS? It will accumulate the noise and offset too !
I would do it with tone detection and grap the amplitude ...
(OK, depends on how the source is controlling the AC amplitude, if harmonics are involved and the feedback also uses RMS...)
If you use multiplexed input cards and your source output impedance isn't very low , have a look at the settle time error!
(With 500k total Samplerate with 5k Ohms source impedance I found >1% error)
At least I would give it a check, one channel a 'stable' source signal, the rest to the generator with 3000Hz and stepped amplitudes, look at influences at the stable channel.
If you don't really need 'simultanious' values (slowly changing signals) but more accuracy, sample each channel for some ms individually , calc the amplitudes and save them.