Spikes in Data stream on CRIO with Realtime Scan-Mode (IO Variable)

[mastada]

Hello

I have an analog NI-9205 input card configured in Real Time Scan (I0 Variables) for a labview program which writes data to file.
When running the Labview program, consisting of a fairly involved acquisition but essentially a While Loop reading off the channels, I experience a Noise of spikes in the 4 acquiring channels on the 9205. (note. 4 channels populate therefore 6 open channels).
The noise is more of a random spike and it will hit each of the channels. Although one of the channels, which is a relay signal to a channel thus using a pull resistor when open, infact does not experience the frequent but random spikes.

My hookup config on the 9205 is:

Chan

AI0      Diff connected

AI1      Diff Connected w pull down resistor to Com for open chan.

AI2-4   open chans (transducers not currently hooked up but Program set for them),
AI5-6   Diff connected
AI7-9   Open chans (Not needed channels)

Now. When I create an NI-DAQMX TASK in NI-MAX and use the 4 channels that I have transducers on, I get a very clean signal. No spikes with a voltage change of only +-.005  V per chan oscilitating.
As per seen in the image of NI-MAX Task.
But when recording data to disk in Realtime (IO Vars) the voltage spikes can be the magnitude of .02 V as seen in the time history plotted to excel from the collected data.

This can be in the magnitude of .02 volt on the same input as seen on a static Clip displacement transducer of the Orange and Green line. (those lines are already scaled to mm)

Is there something happening in the hardware input scan that varies between Real-Time(NI-DAQmx) and Real Time (IO Var) when looking at a Task vs a Labview VI scanning the Channels?

 

[ZYOng]

NI-9205 has an aggregate data rate hence it is a multiplexed sampling module. Most likely what you are seeing is the ghosting effect from the unconnected channels.

See Incorrect Readings on Unconnected or Open Channels of DAQ Device and How Do I Eliminate Ghosting from My Measurements?

[mastada]

Thanks for the reply.
So I shorted out my open inputs to keep them from floating all over. I did a simple short on the +- of the Diff channels and they do not experience the spike.
But no Change in regards to the spikes on the channels that are populated, noting that Channel A01 is the only channel in normal operation that uses a pull down resistor to Com for its open status from a relay and it has never experienced the spike.

So. I should note that I am running a VFD about 6 feet away from the NI CRIO and this 9205 input card.
When the VFD is not running, the signal are sound.  But when the VFD is put operational, that is the moment that the signal spikes drop. 

So if it is a EMI spike, I still sit confounded as to why I don't see it in the NI-DAQmx task mode but do see it in the NI-Realtime mode for a labview program.