DAQ + DAC... NI PCI-6143 + PCI-6731 vs. only PCI-6229?

[clas004]

Hi, I am trying to control a XYZ stage using a feedback program in LabVIEW, and am wondering about my computer card options for (1) data acquisition of analog inputs and (2) digital to analog outputs.

A suggestion I've gotten is the NI PCI-6143 for data acquisition of analog inputs plus the NI PCI-6731 for digital to analog output.

However, I'm wondering if the NI PCI-6229 can do the same thing that those 2 cards together can do, and for a fraction of their price. The only issue I can foresee is a slower data transfer rate with the PCI-6229. Any other problems with this idea?

Please advise.

Thanks! 

[Seth_B.]

Hello clas004,

How fast are you wanting to sample the incoming data and how fast do you want to update your output values?  Also, what voltage ranges are you looking for?  Based on these, we can help you pick a card.  For basic operations, though, any of the M-series cards can do both analog input and analog output.

Regards,

[clas004]

Hi Seth, thanks for that.

To answer your questions... first, voltages: our XYZ stage amplifier / servo-controller takes -2 V to +12 V. So the typical NI range of -10 V to +10 V should be OK.

Next, how fast? In short, we're probably not going to feel 1-order-of-magnitude differences.

A rough estimate...

Analog inputs, we expect to use 4 channels, so for the PCI-6229 that's (250 kS/s)/4 channels = 62.5 kS/s/channel, versus 250 kS/s/channel for the PCI-6143.

Analog outputs, we expect to use 3 channels, so for the PCI-6229 that's (833 kS/s)/3 channels = 277 kS/s/channel, versus 1024 kS/s/channel for the PCI-6731. 

If I've got that right, then the one-card PCI-6229 solution would be about 4x slower on input and output than the two-card PCI-6143 and PCI-6731 solution. But it would also be about 4-5 times cheaper, and that's a key consideration for our shoestring budget.

Worst case scenario, my experiments would take 4 hrs to run instead of 1 hr. I can deal with that. But more likely the system will be limited by experimental factors totally unrelated to data transfer speeds, so we won't even be able to tell the difference between  62.5 kS/s/channel versus 250 kS/s/channel speeds.

So, for our requirements of at least 4 analog inputs and 3 analog outputs, would the PCI-6229 be our most cost-effective solution?

To go along with that, was looking at the BNC-2110 connector block - all of our analog inputs / outputs going through BNC wiring - and a SHC68-68-EPM cable. Again, looking for the most cost-effective solution.

Related to that, is this a mistake on the NI.com page... "Each NI PCI-6229 requires: 2 Cables, 2 Connector Blocks":
http://sine.ni.com/nips/cds/view/p/lang/en/nid/14136 
I only see one 68-pin input on the card, so why would two 68-pin cables (and connector blocks) be needed?

Looking forward to placing the purchase order and getting started with it 

[clas004]

The other question I had was about compatibility.

The XYZ stage and amplifier /servo-controller manufacturer is PI (Physik Instrumente). In the data sheet they write:

"digital control via an external D/A converter is possible. For several D/A boards from National Instruments, PI offers a corresponding LabVIEW driver set which is compatible with the PI General Command Set (GCS)"

[from http://www.physikinstrumente.com/en/products/prdetail.php?sortnr=601800 ]

Unfortunately they don't list what those boards are (or I haven't been able to find this list). Have sent them an e-mail recently, but have yet to receive a reply.

For what it's worth, I know that a research group has gotten the PCI-6143 and PCI-6731 to work with that PI hardware. My hope is that the PCI-6229 is similar enough to work as well. But I'll wait for PI's answer on that.

[Seth_B. · Accepted solution]

Hello clas004,

The PCI-6229 looks like it would be ideal for your situation.  As far as the two connectors are concerned, the 6229 does have two front connectors.  If you check out the NI 622x Specifications and look at the pinout toward the bottom, you can see what signals are routed to each connector.  Since you need 3 Analog Outputs, you will need a cable and terminal block for the second connector as two of the AO signals are on each connector. For cost effectiveness, if noise is less of a factor for AO signals, you could consider the CB-68LP unshielded terminal block for the second connection.

As far as the compatibility of the PI functions are concerned, it's likely that they will work, but PI would be able to answer that the best.

I hope this is helpful,

[clas004]

Perfect, thanks for all the help Seth!

[GabrielaK]

Hello class004,

I just read this thread by chance and I am wondering if you got this answer from PI that you were waiting for. Is everything solved now for your problem? If not, please add a short post here. I haven't found any possibility to contact you directly, so I am asking here.

Thanks and best regards,

Gabriela

[clas004]

Hi Gabriela, yes, PI answered us that another customer of theirs uses the USB version of the 6229 without any issues. So at this point it all looks good.

By the way, are you also working with PI products? 

[GabrielaK]

Hi class004,

yes, I do 🙂

So if you have any further question regarding PI products, let me know.

Regards,

Gabriela