Can I use Several NI cDAQs to Replace a NI cRIO-9040?

[MichaelBoyte]

I would like to use three NI9401s and a single NI9215 to measure and control a prototype system. Originally, I had planned on using the NI cRIO-9040, but now I am wondering if I can use 4 NI cDAQ-9171s instead? My understanding is that each of the modules should be able to go inside a cDAQ-9171 and I can monitor and control the modules from a PC with LabView. Is this understanding correct?

[santo_13]

Technically you can but you will not be able to synchronize all 4 modules as they are in different chassis.

In a cRIO, you can deploy an FPGA code to run independent of the host now you need the host always.

[MichaelBoyte]

Thank you for the information. I'm not certain I completely understand what it means that I won't be able to synchronize all the modules. Since they will all be plugged into the same PC would that not synchronize them?

I had also found another possible alternative to the three NI9401s and the one NI9215 all in the cRIO-9040. It would be using a PCIe-6346 in place of the NI9215 and USB-8451 in place of the three NI9401. What need to be able to do is measure a couple analog voltage signals, communicate with 4 SPI devices, and send PWM signals to two motor controllers. With that in mind, would either of the methods I mentioned work for that? The whole design was initially built around using a complete cRIO system, but we won't be able to get a cRIO-9040 for a long time.

[James_McN · Accepted solution]

There are a few reasons that this might not be a good idea. Especially as you mention control:

• As Santhosh said synchronisation between modules isn't possible in the 9171 chassis.
• This forces you to use DAQmx and normally for control with DAQmx you want to minimise latency by using hardware timed single point mode but this is not supported on cDAQ modules.
• Similar latency concerns exist with USB2. Especially if you have multiple devices connected into the same hub.
• USB and Windows systems are not designed for higher reliability and you are more likely to hit issues like the system putting the USB bus to sleep or deciding it needs to perform an update.

So the short answer is it depends but it probably isn't a good idea.

The synchronisation problems could be solved by using a 9174 4-slot chassis instead. But if you need good control performance or reliability then the cRIO is a far superior option.

[MichaelBoyte]

Thank you, I believe that answers my question. The 9174 seems like it's going to be my bets option at this point, because unfortunately we can't wait for the cRIO to become available. Thanks!

[James_McN · Accepted solution]

Sorry, you posted while I was replying!

The synchronisation means at the signal level. i.e. the samples from one mode line up with another.

The analog voltages and PWM would probably work OK on the cDAQ. The SPI wont and would need a dedicated device. That is because to talk to SPI over a general purpose digital line requires fairly tight timing. As you have found, NI make devices dedicated to this.

It still depends a bit on reliability concerns but using the USB SPI device and either the C Series or PCIe DAQ may work depending on the performance and reliability required.

The performance concern would be if you are doing closed loop control on the motors which require frequent updates. Under Windows it is difficult to guarantee rates much faster than 100Hz as other processes may interfere.

The reliability is more about how long you want it to run for