Differential configuration in NI USB 6356

[Amartansh13]

Dear all, 

I am using NI USB 6356 for the acoustic noise measurement. In the datasheet, it is written that out of the 8 ai BNC ports, ai0 to ai3 takes measures in differential mode (By default). I have few questions regarding this:  

1) Is the inner and outer terminal of the BNC connector marked as ai0 (or ai 1, 2, 3) corresponds to ai+ and ai-? Because mostly the outer ring in the BNC connector corresponds to device ground. 

2) If the BNC terminals are ai+ and ai-, then is it efficient way of measuring differential measurements? I have rarely seen a BNC terminal for differential measurements. In other devices like NI USB 6281, there are three different pin outs namely ai+, ai- and ai gnd. Whereas, in NI USB 6356, there is BNC connectors named as ai0 (or ai1,2,3) and there is a seperate pin out in the corner of the device named as ai gnd. 

3)In the NI documentation, it is suggested to connect ai+ and ai- to ground using bias resistor. Since, my USB 6356 board has BNC connectors, do I need to use bias resistors and connect the inner and outer terminals of the BNC conector to the ai gnd using bias resistor? Or there is already a internal connection inside the device? Doing this adds extra noise to the measurement because both the resistors are not exactly of same value (1% tolerance). And since I am using 100Kohm or 1Mohm resistor, even a small difference in resistors causes noise (I guess). Also, using more wires from the BNC connector catches more noise from the environment (especially 50Hz power line noise).

4) If taking differential measurements with NI USB 6356 is as simple as just connecting the BNC cable coming from the floating voltage source (hydrophone) to the ai0 BNC terminal, then why I am still seeing the strong 50Hz and harmonics presence in the data? Ideally, differential measurement should be able to remove the common mode voltage (specially 50Hz). Also, sometimes, there is weird drift problem in this connection i.e. the measured signal is either decaying or growing (see the video and pic attached) 

The acoustic noise is very small (order of few 100 of microvolts) which is near to the 16 bit ADC noise floor of DAQ device and hence I need to be very careful and make sure that I am measuring only acoustic noise and not any noise coming from the wires, connectors or grounding issues. 

[Alexander_Sobolev]

6356 description says it runs in differential mode only, all 8 channels. 

BNC connector outer ring is connected to ai-, center - ai+. There is a FS (floating source) / GS (grounded source) switch on the 6356BNC, in FS it connects ai- to ground through internal resistor and capacitor - see panel photo https://www.ni.com/en-us/shop/model/usb-6356.html.

1 mV signal with +/-1 V ADC range gives 5-6 bits ADC,  not 16 bit. I guess you need an amplifier on the hydrophone... What is the cable length?

>> Ideally, differential measurement should be able to remove the common mode voltage (specially 50Hz).

I am not sure, but always thought that both differential wires will get the same amount of noise that will be compensated after subtraction if they are the same. Something like twisted pair, not wire in the shield.

[Amartansh13]

Thanks a lot for your reply!

My transducer is like a microphone i.e. floating source. So, I am connecting it to the ai0 but I am still getting significant 50Hz noise and its harmonics. What can be the issue? And why the only ai- is connected to ground through resistor and capacitor(I know it for biasing and bypassing AC and DC components)? But if we are connecting only ai- to ground through a resistor and capacitor, then don't you think we unbalancing the differential measurement? Because now the ai+ and ai- will not have same nature of AC components. Ideally, both ai+ and ai- should be connected to ground through bias resistors (as shown Ni doc pdf attached in chapter 4, page 14 and 15).

My another Question is: That sometimes when I am using floating configuration in differential mode, I am getting weird drift problem in which my signal is either growing or decaying continuously. This problem is solved only when I am connecting a 100KOhm resistor between ai+ and ai-. But connecting resistor adds noise which I don't want. What can be the reason for this drift problem? (See the video and pic in zip file attached)

I have also attached the VI which I am using for noise measurements.