PCIe-6351 Voltage Drop in Analog Output

[asun]

Hi,

I'm running into a very peculiar problem with my PCIe-6351 DAQ card, and I was wondering if anyone had an idea of what might be going on.

For my setup, I'm using a 192061B-02 (SHC68-68-EPM) cable to connect the DAQ card to my pcb, which has the following connector attached: http://www.mouser.com/ProductDetail/TE-Connectivity-AMP/2-5174225-5/?qs=YQ8j8X//jHAhEXN0OFWZwg==

I power the board with the +5V from the DAQ. I can toggle the circuitry on my pcb on and off with a digital output from the DAQ (essentially the +5V is attached to the source of a PMOST and the digital output is attached to its gate). I am also applying a constant analog output voltage (100mV, -5 to +5 range, AO1) from the DAQ to the pcb. For the sake of debugging, I have removed every component connected to AO1 such that it is now an open circuit (I am applying the 100mV to just a metal trace). I am doing this all with the NI MAX test panel.

Here's the problem. When I power my board off (transistor is off) and measure AO1, I get 100mV as expected. However, when I power it on, I get 97mV (as read by a multimeter). Given the 16bit DAC and 10V range I would expect at max only about ~0.15mV of difference. So I thought maybe I'm drawing more current than the DAQ can supply.

I then turn the board off and apply +5V from a sourcemeter to power my pcb externally. I measure 100mV at AO1. There's no drop. The source meter tells me that my PCB is drawing a little less than 200mA. According to the datasheet, I should be able to draw 590mA without the disk drive power connector (I later added the disk drive power connector and I still have the same problem).

If I use a BNC-2110 and measure AO1 (without the pcb), I also get 100mV. Using the same setup, I also tried placing a 200mA load between +5V and gnd and measuring AO1, and I still get 100mV.

Is there something that I am missing here? Any help would be appreciated.

Thank you!

[ChrisR.]

It sounds like some of your +5V current is returning to the board through a ground shared with the analog output.  Do you have your AOGND tied to digital ground anywhere?  Since 200mA is inducing a 3mV offset, it appears you have 15mΩ (3mV/200mA) of shared impedance between your analog and digital circuits.

You are measuring AO1 with respect to AOGND, right?

[asun]

Hi Chris,

Thank you for responding.

All the AO GND and D GND (and AI GND) pins from the DAQ are tied together. Therefore, the analog and digital circuits share the same ground (the digital signals are static when my analog circuitry is running, hence I'm not too worried about switching). This also means that, yes, I am measuring AO1 with respect to AO GND.

I should also mention that these GND pins are shorted to the ground plane as soon as they reach the board, so I'm not sure how such a large resistance could exist.

[ChrisR.]

The resistance is in the cable.  You're sending 200mA back to the board through all your ground lines.  That's why you don't see the offset when you use an external power source; the power in that case doesn't flow in the SHC68-68 EPM cable.

If it's not too difficult at this point, you could try disconnecting the AOGND lines from the other grounds, and then measure your AO signal with respect to them.  I suspect the error would be gone.