Grounding the DAQ (NI USB 6363 and NI BNC 2110)

[Testeraj]

I am working on measurement systems in an engine test facility with multiple test cells. We mainly use NI BNC 2110 and USB 6363. The USB 6363 has a separate grounding terminal which is currently connected to the common electrical ground, we are working under the assumption that the 2110 is internally grounded through the PC. This means all the DAQs are connected to the common electrical ground which is shared with numerous other electrical and electronic devices in the whole building. Will this cause any noise in the measurement signals? I have been going through grounding and shielding articles but haven't found a solid answer yet.

[Duffy2007 · Accepted solution]

Hello Testeraj,

you have done the things that i often recommend to ground the chassi of the DAQ device and the connection terminal. You can additionally reduce the external influnce with shielded cabels and a differential measurement setup.

But still, if you are working near a frequency converter or a drive, it could be that this device has an influnce on your measurement, also a noisy earth ground or a unfilter power supply can have an influence on the measurement noise. This is the reason, why there ins no solution, which is working in all situations.

[Duffy2007]

Hello Testeraj,

additionally to which device is the BNC 2110 connected?
You are right the BNC 2110 get his ground from the device, to which it is connected.

[Testeraj]

Thanks for the reply duffy.

The 2110 is connected to NI PCI 6023E. We already use coax cables with BNC connectors in Differential mode for all our sensors. I am thinking about routing to separate power cables from signal cables and shortening the signal cables as far as possible. Do you think that would help?

My other dilemma is whether I should create a separate ground connection used only for the measuring devices. 

[Duffy2007]

Hello Testeraj,

yes, normally separate the cables help.

How looks your noise on the signal?
Is the earth ground connected to a GND pin of the 2110, if yes, did it help to disconnect this?
What do you measure, if you short shield and wire of a BNC cable?

Can you explain more, in which environment the hardware is in use.

[Testeraj]

I am especially concerned about the noise in my fuel flow sensor output.

We haven't connected anything to the ground terminals on the 2110 itself.

When I have the Flow meter turned on, at zero flow, I am getting random 5mV spikes (this is a large flow error in our case) with the power spectrum showing showing no consistent spikes.

When I take the BNC connector out at the sensor end and short the shield and wire, I am getting a very clean, stable 0v signal. Does that mean the noise is coming from the flow meter itself?

The hardware is used indoors, no extreme temperature/humidity/vibration etc. There is a lot going on electrically since we have the Computer, DAQ and about 6 different sensor signal conditioners/amplifiers on the same tower, pretty much stacked on top of each other and all connected to the same power strip.

[Duffy2007]

Hello Testeraj,

it could be that it come form the flow meter, or from the power supply that supplies the flow meter. Try to measure the voltage from the power supply and parallel to this measure the flow, if you see that the power supply and the measured flow show a peak at the same time then the power supply is the reason for the peak.

Additional you can measure a battery, or use a function generatior, if available, to test the card an wires.

[glasss]

Hello,

is it possible to generate an output current with daq 6363?

Thanks in advance

[Duffy2007]

Hello glasss

the AO provide only +-5mA, if this is enough you can try to set up a PWM with the expected voltage and over the duty cycle the current can be introduced.