accelerometer data acquisition from NI 9218 : getting non-zero amplitude

[Nitin.K.S]

Dear Members,

I am using a NI 9218 DAQ module for acquiring vibration data from PCB accelerometer M353B15. When I acquire data, for no physical excitation of accelerometer, the graph should show zero amplitude. But the signal graph shows amplitude in the range of 2.6-3.3g when the accelerometer is at rest(no physical excitation). Also, I have entered the correct sensitivity value and the excitation current value in the settings. 

The accelerometer has a bias level of 8.8 VDC. 

I have not used an external signal conditioning module. I have used the internal IEPE excitation provided by the NI 9218 module.

I am not sure where I am going wrong. When there is no physical excitation, I want to get zero amplitude in graph. I request you to please provide your suggestions.

Thanking you,

Nitin

[Norbert_B]

Looking on the graph, you obviously have a HUGE issue with EMC. You have a really bad noise on it and it seems to have a frequency of 50Hz.

Regarding the offset, there could be multiple reasons:

- Missing offset compensation (esp. if the sensor doesn't provide a full bridge)

- Physical offset (e.g. sensor has already a strain due to mounting)

- External noise

- High impedance of sensor

Tasks for you:

1. Eliminate the external noise (turning of source or shielding cables and connectors appropriately)

2. Verify offset compensation and impedance

[Henrik_Volkers]

Norbert is rigth with the EMC issue, however it's a IEPE equiped quartz shear accelerometer (link to datasheet) , not a bridge sensor 😉 😄

But most probably the EMC issue is due to a wrong connection of your sensor and/or a configuration problem, since in IEPE mode you have a AC coupling (0.5Hz HP), so without shaking the sensor, the readout should settle at zero!

Please provide your connection schematic.

Since you measure 8.8V (with an external DMM?) and have an exitation current of about 2mA your sensor impedance is 4.4k.

A quick check: Instead of your sensor connect a 4.7k Ohm resistor at your DAQ. Should result in 9.2V bias and no noise and should settle at zero!

(OK, every resistor at room temperatur has noise 😉  ... but it's less than your DAQ noise :D)