NRSE

[Ledang]

Hi,

I have a tomography experiment which require measurements of the surface potential. In the experiment, I need to inject current 10mA at 10kHz to a pair of electrode in a medium and measure the potential (V) of each electrodes with respect to the ground. I’m using NI USB 6218. From the literature, I identified the output signal from the electrodes are grounded signal source NRSE (non-referenced single ended). I made connections similar to the attached file (circuit#1.jpg) and run the VI. The waveform graph in the VI shows the output signal does not centered at 0V and has offset of several volts (VI#1.jpg). I would like to know if my connection to the DAQ is correct? If not please advice

Tq

[Henrik_Volkers]

Have a look at the spec and the manual.

I assume you have a quite high source impedance and you are using a multiplexed input DAQ.

Two points:

- Input bias current ( U offset = +-100pA * (R9 +Rx) , Rx is the impedance of your R8 to GND @ test position )

- settle time error due to the mux & input cap (see pictures page 2 in the spec)

I strongly recomment a input buffer circuit to drive the analog input of the 621x. While building it, I would go with a INA and measure differential with the buffer.

Since you provided schematics of your current driver , it looks like building a input buffer shouldn't be a big problem for you 😉

However even with your buffer you have to look at the input bias current!  As always reading the datasheet helps ....

As a quick try: use only one channel of you DAQ and add a 1M (or 10M) bias resistor from AI to GND .  Look at the Offset voltage.

[johnsold]

I agree with Henrik that you should use a buffer ampliifier.  Even though the input impedance of many DAQ devices is quite high, they do not do well when directly connected to sources with high internal impedance.  Get a good INA.  Look at bandwidth and DC input characteristics. Your results will be much better than trying to connect the DAQ device directly to the electrodes.

Lynn

[Ledang]

Hi Henrik,

Thank you for the advice. As suggested ,  I made a quick search for the INA device and I found out the device INA 129P. The device has high slew rate and good BW at 700kHz at G=10 which I think it is appropriate in my application. The problem now, I can only received the device by tomorrow. Attached is a new schematic which consists buffer and difference amplifier (INA129) based on your suggestions. In my application,  I'm a bit confusing on how to connect to the AISEN and AIGND and the schematic shows of what I had understand. Can this schematic works? Please advice. 

Thank you.

[Henrik_Volkers]

Some notes:

All Amps need buffer caps  close to the housing, and I would add 10 - 47 Ohm decoupling resistors maybe exept for the current drive amp as shown for the first input stage....

(Reading the datasheets helps)

Next the INA Ref pin 5 should be your AISEN

Good layout and a star ground  (YES, you can route gnd wires in parallel to the star ground 😉 and if you use a ground plane still look for the currents in the plane!  )

The second INA input  (well ... GND) should be as close to your DUT as possible. 

I suggest a bootstrapped BIAS current path ...   that way you keep the high input impedance  of the buffer amp. (BTW the INA has a higher impedance (1E10 comared to the LM6172 with 1E7)

And don't use dual OPs ....    look for a single buffer amp with high input impedance and GAIN1 stability....

finally I added a 1M input resistor ... shouldn't be a problem for the current control generator and avoids floating inputs 😉

[Henrik_Volkers · Accepted solution]

And you will measure an offset voltage with the input (R9) tied to GND ...   

compensation can done in software, but you should run the unit for some hours and monitor the offset voltage.