Using Accelerometer on NI 9220

[arnold05]

Hi all,

I am having trouble connecting my accelerometer to NI 9220. The output does not make sense at all.

Please find attached the photos (file name Figures.pdf of what I use to connect the accelerometer (Figure 1). Please ignore the wires as they belong to another device.

Basically I use the blue and grey connectors to power the accelerometer, and the NI 9220 to obtain the output. As seen in Figure 2, the accelerometer has 4 wires. Figure 3 shows how the wires are supposed to be connected. By following Figure 3, I connected the accelerometer's wires as such:

1. Red and black to the positive and negative power source (i.e. the blue and grey boxes) respectively.
2. Green to one of the AI+ slot on the NI 9220.
3. White to one of the AI- slot on the NI 9220.

This connection gives me some results that do no make sense at all, and with significant amount of noise. May I know what have I done wrong in the connection?

The specification of the accelerometer has also been attached.

Thank you very much. Any help is highly appreciated 😃

[Agamemone]

What kind of signal are you reading in? Is it just noisy? Could you please attach screenshots?

[arnold05]

Thank you for the reply. 😃

Please find attached the output graph from the accelerometer. I haven't done the sacling yet so y-axis is in terms of Voltage. 

I didn't move the accelerometer at all when logging the data. As you can see, the output voltage is stucked at about 10.5V, and can fluctuate down to 8.5V. I don't understand why it is behaving like this. I think there could be something wrong with my connections.

[Agamemone]

Have you tried reading a known signal from a power supply of something like that?

[Henrik_Volkers]

Your source seems to float. ... and there is no bias path ...

From pos and neg exitation connect a 10k ohm resistor to the COM input , so two resistors total ;). 

OR connect the GRN sensor wire to pos input and a 1M Ohm from the pos input to COM

and you can try to connect the shield to COM too.

[Henrik_Volkers]

The max output of the sensor is +-400mV  and your fixed input range is +-10V .

you loose about 4-5  most significant digits. depending of the task you want to solve I recommend a instrmentation preamplifier (INA) with a gain of 10 to 20.

And you should investigate your supply voltage. This type sensor is only specified for a 10V supply and the output is ratiometric to the powersupply.

If you add an amplifier it's woth to monitor the bridge supply.

[arnold05]

Thank you for the reply! After connecting the negative output (AI-) to COM, the noise seems to be lowered. However, two issues remain: 1. The measurement I got (before any scaling) is between 4.6-4.7V if I use an input voltage of 9V (I don't have 10V power supply at the moment). This is very different from the output I expected according to the datasheet, which is 400mV maximum output as you mentioned. 2. I changed the output to real time V-t graph. However, it does not have any response even when I shake the accelerometer. Again, thank you in advance!

[Henrik_Volkers]

Please provide a schematic of the wirering ....  Seems not one of the two I proposed 😉

You have a differential input and COM. IF you have a floating source use the two resistor method. OR   If you look at the sensor datasheet you see two 500Ohm resistors to complete a full bridge, they are connected to the GRN (pos) output, that why I recomment an additional bias path resistor from the GRN (pos signal) to COM with a value of 10k to 1M ... 

If you (ab)use the 10V supply for other sensors/things add them in the schematic including ground impedances.

If you shake the sensor  I expect about 600µV signal amplitude....  easely buried in the noise, that why I recommend a preamplifier ....  look for industrial signal conditioners with 10V bridge supply +-10V output and a gain of ~20.  (DEWETRON and others makes Labversions)