PCI-6023E DAQ card maximum sampling frequency

[Jasiu]

Hello
I am using PCI-6023E DAQ card in pc-based ETS solution (and writing appilication in LabView 7.1 with RT module). The card has 200kS/s maximum sampling frequency, but it can be set for much higher sampling frequencies and the waveform acquired appears to be correct (i.e. i've tried setting it fo 1MS/s and sampling 400kHz sine, which is obviously above Nyquist frequency for 200kS/s card, but on spectral graph, main peak is at 400kHz). Is the card driver doing some kind of free/coherent sampling?
Moreover, when sampling frequency is set to 200kS/s, the card seems to be doing same thing - i.e. for 200kS/s and sample block size of 200kS, graph should be updated once in a second, but it's updating slighty slower.
I'd really appreciate if someone could explain me (or gave me a link to materials) what exactly is happening here? Is driver doing some background work, or maybe it is problem with network latency/unstability ? What is the impact of this effect on real-time aquisition?


Thanks in advance

Jan Kienig

[parker]

Since the fundemental is 4 times the nyquist, then what you are measuring is an alias of the fundemental. This works well as long as the fundemental is a repetitive signal. Sampling every other peak and every other node looks the same as sampling every peak and node. Tektronix exploited this on their 7S series sampling heads. Another use of this phenomena is the effective demodulation of high frequency signals as long as the bandwidth meets nyquist. As with your card, if the input amplifier supported it, I could extract modulation information from a 500 MHz signal so long as the the bandwith of that modulation did not exceed 100 kHz.


Parker