DSA signal follows current input

[pratty]

Hi,

I am having problems with vibration measurements taken using 4496 DSA cards. I have a PXI system running phar lap that contains two 4496 DSA cards, a DeviceNet card and a 6238 current input card. My program scales the incomming current inputs (4-20mA outputs from RTD signal conditioning units), takes rms vibration measurements, calculates the speed of the machines from tacho inputs to the dsa and stores all the data in a tdms file.

I am getting strange readings from the vibration inputs. The rms values are not correct. When i remove the accelerometer there should be no vibration however the rms signal drifts with the temperature signal. I have verified the vibration signals using portable data collectors all of which give correct readings however the pxi does not appear to reading correctly.

All dsa signals are from buffered outputs from signal conditioning modules. The accelerometers are iepe type, i have set the supply power option to none as the signal conditioners take care of all sensor power requirements.

I suspect that the current inputs and the dsa inputs are interacting somehow? I have the current inputs set to differential with AIGND connected to the negative side of one of the current inputs. Is this correct?

The 4-20mA current signals are coming from an Entek XM361 temperature module.

Any help would be appreciated

[dsb@NI]

I recommend that you verify that coupling for the vibration measurement channels is set to AC or use a digital filter to remove low frequency bias (due to temperature changes). The rms level includes both AC and DC components. You can band-limit the input using a hardware filter (coupling) or a digital filter in software. Without changing the coupling or adding a filter, you can measure the AC level by using the AC & DC Estimator VI that ships with LabVIEW Full or the SVL AC & DC Level VI that ships with the Sound and Vibration Toolkit.

[Jason_D]

Hi pratty,

When you say "remove the accelerometer" what exactly do you mean?  Do you remove it from the system with the system unconnected?  If so there is no expected value as it will be "floating."  In order to see zero you would need to tie this to ground.

Message Edited by Jason_D on 01-08-2009 01:17 PM