Power Supply for NI 9203 module

[chriswilliams]

I am using one of your power supplies (PS-2 p/n:777584-01) to power a 2-wire Rosemount 3051S transducer that is being read by your NI 9203 module (in a cDAQ-9178 chassis).  The signal I am getting is extremely noisy.  + from PSU goes to + of transducer, - of transducer goes to AI0, and PSU common goes to com of 9203 module.  I can connect up to 8 transducers.  I have done this for many years with your legacy Fieldpoint AI-110 modules with no problems whatsoever, although I did have to add a 100-ohm resistor in-line for each channel.  I was under the impression that external resistors were not needed for the 9203 because they were built into the module itself.  The transducers power on just fine.  They wouldn't even power on with the old AI-110 module without a 100-ohm resistor in-line.  Just looking for reasons why my signals are bouncing all over the place.  Is this power supply OK to use with the 9203 module or do I need one of better quality?  I tried a switching power supply but that made it even worse.  What do you recommend?  Any other suggestions on what would be causing the signal to bounce?  Is there a way to filter the noise?  I think the AI-110 module had built-in filtering you could apply, but I cannot find anything using MAX to apply any filtering for this 9203 module.  I just need 1 reading per channel every second from the module.  I have been fighting this for a week now and my bosses are getting impatient with me.  I'm not sure what else to do at this point.  Thanks in advance for any help...

[NIthan]

Hey Chris,

According to the 9203 manual (http://www.ni.com/pdf/manuals/374070c.pdf), you should be fine using that power supply. What kind of noise are you getting, meaning what is the range of your signal, and what is the range of your noise? You said it worked fine when using Fieldpoint, can you test it with the fieldpoint again and see if you get a smooth signal? (it could be that your signal has become more noisy for a reason outside of the measurement system). How are you inputting the signals? Are you programming using LabVIEW and DAQmx, SignalExpress, or a 3rd party software? How do you have everything wired up? Make sure you have the connections wired like figure 3 in the manual.

-Nathan H 

[chriswilliams]

Hi Nathan,

Thanks for your response.  The transducer outputs a 4-20 ma signal.  With the transducer sitting on zero pressure, the output should be 4.00 mA.  The signal will bounce from about 3.5 mA all the way up to over 5 mA, and everywhere in between.  I can put pressure on the transucer to output a higher mA signal but it fluctuates wildy regardless.  I did as you suggested and set it up with my old Fieldpoint system and it worked perfectly.  It did not fluctuate at all, even up to the third digit.  I am using Visual Basic 6 with DAQmx for programming, but I also use MAX to confirm all my readings.

The old Fieldpoint AI-110 module wires up differently because the power is supplied by the module itself, it does not require an external power supply.  You simply jumper the V and C terminals from the FP-1000 brain to the V and C terminals of the AI-110 module.  You then connect the transucer, for instance if I use channel 1:  I place a 100-ohm resistor in terminal 17 and then connect the transucer + to the other end of the resistor.  The - of the transducer then connects to terminal 2 of the AI-110 module.

I have the 9203 module wired exactly as figure 3 in the manual using the external PS-2 power supply I specified in the first post.  The only difference is there is no resistor in-line anywhere because the 9203 has this built-in to the module itself.  I tried adding a resistor but it didn't seem to help at all.  This seems to be a problem for the 9203 because there is no built-in filtering/rejection in the module, whereas the old AI-110 module did.  This is just a theory though but it makes since to me.

Thanks for your help!

Chris

[NIthan]

Hey Chris,

Noise issues are definitely tricky, since there could be a number of things that are causing noise. If you've already tried some basic troubleshooting steps (trying different channels, trying a better shielded cable for the signal, different power supply, etc), the next thing I would try is to accurately measure your input signal's frequency. This requires you to sample at much faster rate than 1 sample per second (probably more like 100KS/s), then measure the signal and try to determine if the noise has an inherent frequency to it. For example it could be that there is a noise signal that oscillates +/-1ma at 60Hz, in which case your signal may look like a 60Hz sine wave oscillating +/-1ma around the true pressure value; however, if you only measure at 1 sample per second, it will just look like the signal is randomly bouncing between 3 and 5 ma.

If you are able to determine a that your noise has an inherent frequency to it, then you can setup a filter to remove that noise (you can either do this in software or hardware). For example, if you determine you have a 60Hz noise, you can setup a lowpass filter with a cutoff frequency below 60Hz. Since the 9203 has no built in filtering, this could be why you are getting noise, so adding a filter may fix your issue.

Another thing to try is a different wiring setup. Depending on how you are currently grounding your system, you may want to make sure you are not creating ground loops by having your power supply, sensor, and cDAQ on the same ground loop. Refer to this document for how to remove noise by wiring signals properly: http://zone.ni.com/devzone/cda/tut/p/id/3344 (table 1 is very useful).

I hope this helps!

-Nathan H

[chriswilliams]

Thought I would reply just to finish up this thread.  The problem was definitely caused by the 9203 module itself, which has no filtering/rejection.  I ordered a NI 9208 module, which has built-in filtering/rejection and it worked perfectly.  Steady signal up the 4th digit.

[NIthan]

Hey Chris,

Glad to hear you fixed this issue. Thanks for the update!

-Nathan H