BNC-2110 Phase Change in 1st Channel Read

[Kevin_D]

Hello NI Forums:

I've been having some serious issues lately, and for the life of me I cannot figure out why. Here is the setup:

I have an array of hydrophones (a.k.a microphones) that produce a voltage response to pressure (sound) waves. Using a BNC-2110 through an NI USB-6251 DAQ box, I connect my setup through my laptop's LabVIEW program.

The program outputs a sine wave at 1000Hz to a loud speaker through channel AO1, and reads 5 BNC channels (ai1:ai5), with the 4 hydrophones connected to ai1:ai4, and the output signal read through ai5. Using an oscilloscope, I was able to determine the 4 signals going into the BNC-2110 DAQ are the correct signals, however when I read them through my LabVIEW program, I get issues. The first channel read (ai1) takes on a small magnitude gain and a phase shift (frequency dependent) while the remaining channels go unscathed. I was able to correct this issue by shorting channel a0 AND by simultaneously reading channel a0 through the Measurement and Automation task. If I don't read the initial channel (a0), the phase shift occurs, however as soon as I start reading it, it corrects itself. There are other things that cause the phase shift as well:

Not having channel a0 shorted.

Having a0 set to ground source instead of floating source.

Not reading channel a0 in the M&A task

Any combination of the above

I have tried changing which hydrophone is input into channel a1 (the channel that gets phase shifted), with the same results on the same channel - therefore it is not my input signal, rather a software/hardware problem with the M&A or BNC-2110, or NI USB-6251.

Attached are some pictures that describe the phenomenon I am seeing.

Note: all hydrophones (Plots 1-3, & hydrophone A) should be reading the same signal as in the 2nd attachment.

[Kevin_D]

bump

[tannerite]

Hi Kevin_D,

Do you know what the output impedance of each hydrophone? If the impedance is relatively high, there could be some charge (from parasitic capacitances with the ADC). The device you are verifying the correct behavior on may have one ADC per channel, while the 6251 card has a multiplexer on the front end of one ADC; therefore, each channel is not fundamentally isolated. If the input impedance of each hydrophone is very large, any charge that was on the front end of the multiplexer will have no where to go, possibly causing larger amplitudes. One way to combat this is by using a unity gain buffer in between the hydrophone and the connector block. A unity gain buffer has a very low output impedance, which masks the impedance of the hydrophones.

It makes sense that the input impedance may be fairly high because the multiplexer will not read the a0 until you need it to. When the ADC does read a0, it shorts all of the that random built up charge to ground.