Need advice on DAQ

[fredjonze]

Hi,

I'm new to LabView and DAQ and am in need of advice for a small research project.

I want to develop a data collection point using a DAQ. I'm anticipating using a small mini-ITX board and then gathering the data with transducers connected to a USB DAQ. Since this is a low cost effort, I'm wondering whether the USB-6008 is an OK choice. Some of the particulars for this project:

1. Running on a mini-ITX...probably on Win2k.

2. Will use LabView to create a small executable for collection

3. Might develop a web interface for remote monitoring

4. Transducers will be analog and include ambient temperature, device temperature, water flow, water level and possibly circuit amperage.

5. High accuracy is not critical as this is just a training experiment

6. May use 700Mhz radio modem for interconnection.

Thoughts on a good, low cost USB DAQ? Also, any suggestions on sources for inexpensive transducers is greatly appreciated.

Thanks,

Fred

[ryan_d]

Hi Fred,

The USB-6008 is a good low cost option, but might be limiting. Take a look at the KnowledgeBase Article 3K793P88: Thermocouple Measurements with the USB-6008 or USB-6009. This article makes a good point about the resolution limitations of the 6008, which is relevant to transducers in general. Also, you will be limited to software timed analog output in both the 6008 and 6009. If this isn’t a problem for your application, then the USB-6008 should be a great fit! I recommend browsing through the faceted navigation on the left side of the screen for our USB DAQ.

Please post back if you have any questions. Have a great day!

[fredjonze]

Hi Ryan,

The kb article seemed to imply that the 6008 had a hard time identifying thermocouple temperature changes of less the 9 degrees celsius. This is probably to coarse for my experimentation. Does this mean the next step up is the M-series? My general inference is that the M-series has a powered bus...does that account for the increased accuracy?

Also, do you have any general pointers for finding probes/transducers for the types of measurements I mentioned? I've been searching around the net, but there are so many manufacturers and alternatives it's a little daunting.

Thanks for your reply,

Fred

[ryan_d]

Hi Fred,

The M Series and cDAQ USB devices generally have a higher precision because they tend to have a much higher resolution than the USB-6008/6009. For example the 6008 has 12-bit resolution giving it 2^12 = 4096 divisions over a range. Its lowest voltage range is +/-1V. This means that it has a resolution of 2/4096 = 488 microvolts. The USB-9211A, in comparison, has a 24-bit resolution giving it 2^24 = 16777216 divisions over a range. Its lowest voltage range is +/- 80 mV. The 9211A has a resolution of 0.16/1677216 ≈ 10 nanovolts.

If you’re not interested in the cDAQ or other M Series devices, you could amplify your signal externally and scale accordingly in software. This will reduce the total range of your transducers, but could be appropriate for your application.

As for finding transducers, I’m not sure that I would look any differently. I generally trust my favorite search engine! NI does offer thermocouples and RTDs.

[fredjonze]

Hi Ryan,

I now have a 6009 and a type T thermocouple I've been experimenting with. However, I don't know quite what to expect from the thermocouple.

1. The connection diagram helper color coding may not match the actual thermocouple coloring. It shows copper (Cu) and copper-nickel (Cu-Ni) for the inputs. I'm assuming the Cu lead (Ch+) is copper color (goes to the AI0+) and the Cu-Ni (Ch-) is silver (goes to AI0-). However, when I connect the leads, the signal bounces around quite a bit. I did some reading in the forums and found that a lowpass filter can sometimes help this. However, after applying that at 50 hz, it stills bounces considerably. I tried lowering the measurement cycles with no success. I verified that the thermocouple is responding by throwing some heat on the ends. I looked at the other ports and found similar bouce...so it may be induced noise. Tried moving the DAQ and probe away from electronics with no change.

Any ideas?

[ryan_d]

Hi fredjonze,

You are correct with your assumptions regarding the leads. The Taking Thermocouple Temperature Measurements tutorial is great place to start for more information on thermocouple measurements. In the section called Eliminating Noise there are several suggestions, including the lowpass filter, for reducing noise.

Try changing some of your input parameters. Setting your Input Configuration to Differential will reduce noise and can be used with thermocouples. Changing your input mode to something other than On Demand (i.e. Finite or Continuous) might also improve your results. Finally, double check your y-axis scale to see that it makes sense with the range of your expected data.

As far as reducing your sampling rate, you need to make sure that you’re observing Nyquist’s Theorem. NI recommends sampling your signal at a rate ten times faster than your expected signal’s rate. In the case of a simple thermocouple measurement, you should be able to sample fairly slowly – on the order of 100s of Hz.