Sanity check on CAN + DAQ instrumentation

[aeroAggie]

I'm trying to figure out the right instrumentation solution for a project in our lab. I would appreciate a sanity check on what I have so far.

Requirements:

1. Measure 12 analog inputs (0-5V) at 1 kHz
2. Communicate with 2-4 motor controllers on a CAN bus at 10 Hz.
   I need to send CAN messages to control the devices (such as setting a target speed).
   I also need to receive CAN telemetry messages from the devices (such as device temperature, actual speed, current draw, etc.).
3. Write data to a file.

Tentative solution:

CDAQ 9174 compact DAQ (USB connection) with 2 9201 analog input modules (2x8 channel = 16 channels) and a 9862 CAN interface module.

Program a while loop to run at 10 Hz that reads the analog inputs with a DAQmx buffer size of 100 samples (1 kHz sampling rate).

Questions/confusions/hesitations:

1. Can the DAQ and LabVIEW keep up if I add CAN code to the main while loop to write 2-4 messages and read up to 50 messages at 10 Hz?
2. Can a USB connection to a PC keep up with this data? When should I consider an ethernet connection?

Thanks!

[santo_13]

---

Questions/confusions/hesitations:

1. Can the DAQ and LabVIEW keep up if I add CAN code to the main while loop to write 2-4 messages and read up to 50 messages at 10 Hz?
2. Can a USB connection to a PC keep up with this data? When should I consider an ethernet connection?

---

1. It depends, you can always adopt a better architecture to handle the DAQ side and CAN side separately and eliminate any impact
2. 1kHz sampling rate and 10Hz CAN rate don't sound like a large data, USB should be able to keep up with that

here is an article to calculate throughput on a cDAQ - https://knowledge.ni.com/KnowledgeArticleDetails?id=kA00Z0000019QYVSA2&l=en-US

[aeroAggie]

Thanks for the link! Looks like I should be well below a USB port's throughput.

What architecture would you recommend to separate the DAQ and CAN?

I thought about CDAQ + CRIO. The DAQ would measure the analog inputs (forces), and the CRIO would control the devices (motor controllers). However, I also want to synchronously measure (record) telemetry from the motor controllers (RPM, temperature, etc.). I've always thought of RIOs as a control unit, not a data-acquisition unit. Can I record data from a RIO?