how to reduce the noise in a daq USB 6251

[Daniela16]

Hi everyone!!!

I have some questions about a system i'm developing using a NI usb 6251 data acquisition board. The system controls pressure.

First, the signals are acquired using an Omegadyne pressure transducer. (Tipically, the signals are between 0.1 mV and 38 mV , the pressure is between 0 to 50 PSI). The transducer is connected to the daq usb 6251 board (I'm using pinouts 1 and 2 for acquiring the signal, also, i have a 100 K resistor between pinouts 2 and 3) and the board is connected directly to the computer. Aditionally, i have an omega meter, i'm using it only to see the pressure in the display. But in the long term it will be useless.

One of the problems i found is the noise, the values of the  measured voltage change a lot. Even though the measurements in the meter display don't change so much...(or don't change at all)
The system have to regulate the preassure so, if every second it changes a lot, the system was not working properly. How can i make the mesurements more similar to the meter's one? In the labview program i'm using, i'm using a "basic DC" block, it reduced the noisy signals a lot, but is not enough for the experiment, i require values more stables.

Thanks

[Henrik_Volkers · Accepted solution]

I did not look into the pinout /manuals and don't open .doc files of unknown sources (a simple .png graphic is fine 😉 )

However, I assume you already measure in differential mode and the 100k is a path for the bias currents to AGND 

The reason why your meter is stable might be the fact, that it has a build in low pass filter (average) with a quite high time constant.

First things to do: 

Did you make sure that there is a aliasing filter appropriate to your sample rate?  

Read the signal at the highest possible samplerate for at least 100ms   (120ms on 60Hz line )and take a look at the signal (and post a graph as a .png or .jpg picture) 

This will give you an idea how the signal is looks like . Then you can choose a filter according to your system and the signal or do further investigations to not even catch the noise (groundloops,shielding...)

You can use one of the example vis or the testpanel in MAX to do that measurement.  (I personally prefer an analog scope 🙂 )