Asking for the best option

[Francois Ouellette]

Hello,

I need to chose the best board for my application. I have 4 signals to digitize simultaneously, at 5 kHz each, with at least 20 samples per modulation period. 12-bit resolution is enough. I also need to perform a simple operation on the data (difference between two samples in one cycle divided by their sum), and average the results once per second (ideally all 5000 cycles), so I get 4 averaged measurements per second (1 per channel). 

So either I need throughput of 400kS/s to the computer (and perform the calculation on the data), or, if it is at all possible, perform the calculation on the board itself (by programming the FPGA)? I normally use Labview for the computations and the user interface. 

I have used a RedPitaya board so far, which is overkill in terms of speed and resolution, but also the throughput to the computer using LabView is very small, so I can only average about 4 modulation cycles per second, and there are only 2 input channels. So I am looking for a better option to build a working prototype. 

It seems to me that the USB-1208HS could do the job, but I am not sure. 

Can you help? Thanks in advance!

[JRys]

Hello,

The USB-1208HS is a good device and will meet your requirements.

Something to think about is that it uses multiplexing to switch the channels one at a time to the  A/D converter. This mean there is a small time skew between each of the channels. Sampling four channels at 100k Hz (400k samples per second) would result in approximately 2.5uS between each channel. If you were to bump up the sample rate to 250k Hz per channel (1M samples per second), the channel-to-channel time would drop to 1uS. There is a scan option called BURST mode that forces the channel-to-channel time to the minimum, but it's not available in our ULx for LabVIEW support. 

If you wish to have virtually no time skew between channels, consider the USB-1608FS-Plus or the USB-1808X. These devices have a per channel A/D converter to read the inputs at the same time. 

Regards,

John