I am now evaluating the AD7176-2 using the daughter board of the kit. I can now program and read the data out without any problem.
However, a weird problem I found out when I try to apply a test signal to the ADC inputs, say, AIN0 and AIN4. If I connect a 1Vpp sine wave (from Agilent 33220A) signal directly to the inputs, the output is correct. However, if I apply the signals through buffers (JFET OPAMPs), then the output totally messed up. I figured out the problem arises from the node of the buffered ground. When the ground is disconnected from the ADC input, the voltage output is measured near zero potential. In contrast, the voltage output is very noisy if it is connected to AIN4. I did try to connect the buffered ground to the other inputs and the symptom repeats. I wonder if you have any experience tackling this problem? I'd appreciate it if you have any thoughts on this.
My measurement conditions are listed in the following:
On-board gain of 2 amplifier is bypassed.
+/-2.5V split power supply is used and -2.5V is applied externally.
Internal REF and oscillator are used.
anti-alias filters are not used between the buffers and the ADC inputs.