Hardware choice questions

[bob_mcbob]

I was hoping I could get some advice on how to proceed with upgrading my NI hardware.

The current system is an SXCI-1520 with 1314 terminal block, housed in an SCXI-1000 chassis, connected to a PCI-6034E for data collection.  Two channels are used for conditioned strain measurement with small button load cells, and two are connected as voltage inputs from a sync signal and a separate charge amplifier for piezo devices.

The 6034E does not have the necessary speed (200ks/s) to read the voltage inputs with enough resolution, especially when sampling multiple channels.  In addition, the synchronization of all channels is very important.  I have been going through the available DAQ options, and the S-Series PCI-6122 appears to be the most appropriate card for my needs.

I am concerned about the small number of available inputs (4 analog), since the 1520 takes up all the inputs on the DAQ card.  The only options I can see for getting around this involve either having a DAQ card with more than 8 channels, or hooking the 1520 up with a multi-chassis cable that routes all channels from one module to a single channel on the DAQ card.  The latter option would seem to defeat the purpose of buying a fast simultaneous DAQ card in the first place, and the former is not applicable.

A few relevant question:

1) Is it a problem that the 1520 has 8 outputs, but the 6122 has only 4 inputs, or will the other 4 channels just be ignored in parallel mode?

2) Is a straight voltage input to the 1520 limited to 100ks/s as with a condititioned channel?  I am able to read from the 1520 at the maximum rate of the 6034E (200ks/s) without any issue, and the voltage inputs do not appear to be limited in this manner.  Could I read all 4 channels from the 1520 at 500ks/s, even though the conditioned channels will not have as much real resolution?

3) I understand it is not possible to synchronize the counters and analog inputs of the 6122 because they use separate clocks.  I have some optical encoders that I don't use right now, but may need to be added later.  In the past, they were synchronized with the 6034E analog input using an RTSI cable to a PCI-6602 counter-timer board.  Is there any reason why I couldn't do the same thing with the 6122 and 6602 in the future?

4) Does the on-board memory of the 6122 and other such fast-sampling DAQ cards require any special programming, or is it read as normal?

It is likely the strain module and load cells will eventually be completely eliminated from the system, in favour of piezo devices connected to external amplifiers.  Because of this, I would like to minimize spending on any specialized hardware to read from the 1520.  I would like to know whether it's possible to work around the issues described above, and whether the 6122 is appropriate for the use described.

Any advice would be appreciated.

[Doug_L._Bear]

1. The 1520 should be fine in parallel mode and it will just ignore the other 4 channels.

2.You should be able to read as fast as the DAQ card can from the 1520as long as you don't care about the settling time to get a "good" measurement from the card.

3. The 6122 should be able to talk to the 6602 just like the 6034 through RTSI.

4. The 6122 is supported by DAQmx driver as are most of the other S series cards so should "normal"

http://ftp.ni.com/support/softlib/multifunction_daq/nidaqmx/9.0/readme.html

http://digital.ni.com/public.nsf/allkb/3D45A66A691E4BD886256A6C004A4313?OpenDocument

http://www.ni.com/pdf/manuals/372583e.pdf

[bob_mcbob]

Thanks, Doug.  I decided to go with the 6123 for the extra channels.

If I understand your reply to question 2, basic voltage inputs to the 1520 are simply passed through without any conditioning, and the sampling rate is a function of whatever the DAQ card can handle?  Settling time isn't a big issue with a simultaneously sampling DAQ card, is it?  Does the 1520 introduce settling time issues?

I was able to get an excellent deal on some charge amplifiers, so with any luck I will be able to completely eliminate the strain module from the system, and just hook up a basic terminal block for voltage signals.

[Doug_L._Bear]

Yes, as long as you set the filter to just be a bypass should remove any conditioning on the signal.  For how to do this see http://www.ni.com/pdf/manuals/372583e.pdf#36.