I have a design that uses the AD8224 Instrumentation Amp.
I am recording the outputs using a MSO4034 Tektronix scope and then analysing them in Matlab.
The gain is approximatly 125 with DC input signals approxiatly 100mV. The AC signal being detected is approximatly 2mV.
The input to the Instrumentation amp is the output from a LTC6244 operating as a transimpendience amplifier (current to voltage converter).
I obtain the expected result.
However, when I use the scope to measure the inputs to the intrumentation amp and attempt to replicate the operation in of the instrumentation amp in Matlab I just get a noisy signal.
You probably have one of two problems: the probe is injecting noise into the input or the DC resistance of the probe affects the biasing. 2mV is fairly small and with a typical x10 'scope probe you will then only get 20mV/division, so I am not sure you will get a usable result anyway by probing the input.
Yes, I think you are just trying to probe too small a signal. You could try a x1 'scope probe.
I do a lot of low noise/low signal work so have a low noise FET amp which gives me x10 & x100 gain. With a x1 'scope probe I get 20uV per division which is handy, but not always enough!