Can I connect NI USB 6008 to thermocuples, pressure sensors and control valve

[jackgrove]

 I'm endevoring to build a bench scale Biomass gasification plant for testing process control schemes. Will NI USB 6008/6009 be suitable for use as an DAQ. I will be having RTD's, thermocouples and pressure sensors. I don't want to use industrial automation controllers.  Also is it possible to use the analog output channel for sending position signals to a control valve (using adequate current/voltage drivers in between).

[johnsold]

The answer is: It depends.

1. The USB-600x devices use software timed outputs. If that is fast enough, it might work.  Since nothing happens very fast in gasification processes, it probably will be fast enough.

2. The RTDs and thermocouples will require external signal conditioning circuitry.  You cannot get useful results by connecting them directly to a USB-600x.

3. Do not rely on a device like the USB-600x and a desktop operating system computer to assure safe operation of a process which can produce dangerous outputs. What if the computer crashes or your software hits a bug (your code never has errors, right?) right after you leave for the night and the valves are wide open? In the morning your lab is full of H2 or CO!

Lynn

[jackgrove]

Thankyou for the suggestions.I require a few more clarifications.

1) As I mentioned I building a benchscale unit purely for testing purposes (and completely on my own). So unsupervised operation is not a requirement.

2) What sort of signal conditioning is required for RTD's and thermocouples. I was expecting that their mV output can be directly measured by the analog channels in the USB 6008.

3) As you rightly guessed my control software will be in a PC and written in MATLAB or Octave. One of the reasons I want to use NI DAQ is because I've read these software platforms can accept data from NI DAQ's. Please correct me if I'm wrong about this.

4) Can you suggest any other DAQ's from NI in the same price range as 6008 which would be more suitable for my application?

Siby

[johnsold · Accepted solution]

1)  OK.  Just wanted to be sure you were aware of the potential hazards.

2) An RTD is a resistor which has small changes in resistance per degree of temperature change. To measure this you need a current source and enough resolution to be able to detect small changes.  At 25 degrees C a typical RTD is 109.73 ohms and the resistance changes 0.38 ohms per degree.  If you had 1 mA flowing through this RTD the voltage across it would be 109.7 mV and the voltage would change by 0.38 mV per degree.

The resolution of the 6008 on the most sensitive range is 0.49 mV or > 1 degree.  The accuracy of the 6008 is 1.5 mV typical.

For a Type K thermocouple the voltage at 25 degrees is 1.407 mV and the change per degree is 39 microvolts.  The half millivolt resolution of the 6008 translates to about 12 degrees.

So you need an excitation source for the RTD and some kind of amplification for both RTD and thermocouples before it would make any sens to try to use the USB-6008.

3) I have not used anything except LabVIEW with DAQ devices and drivers.  I think DAQmx can be used with MATLAB and other languages.

4) The 6008 is the low end device made by NI. You will need to go to a more expensive device or add the signal conditioning circuits.  Talk to your NI representative for help in selecting a suitable device.

Lynn

[AnalogKid2DigitalMan]

I just assisted another member with chip-based temperature sensors, you may want to look at this thread for some links:

http://forums.ni.com/t5/Multifunction-DAQ/Add-CJC-channel-for-thermocouples/td-p/2351018

-AK2DM

[jackgrove]

1) Thanks for information about signal conditioning for RTD's and thermcouples. I think I can manage that.

2) I have found enough documentation on interfacing NI DAQ (including USB 6008) with MATLAB's data acquisition toolbox once the DAQmx driver is installed. So I think it's straightforward.

3) I will be getting in toch with an NI representative. But I'm probably going with the 6008 or 6009 since I don't have the budget for a higher model.

Siby

[jackgrove]

Thanks but chip based sensors tend to be on the lower temperature range whereas my application is mostly above 500 deg C.

Siby