Grumpy_Mike, wouldn't that do the opposite of what I want? That is the capacitor will dampen the 0.005v fluctuation that I'm trying to measure?
That capacitor will charge to 1.000v, correct? At the time when there are no voltage fluctuations. Then absorb, thus dampening any fluctuation which drives the Vin above 1.000v?
That is the capacitor will dampen the 0.005v fluctuation that I'm trying to measure?
Only if you put it from Vin to ground. Not if you put it from Vin to the voltage you are trying to measure. It is called AC coupling.
Grumpy_Mike, thanks, I read up on capacitive coupling and it makes sense now. But how large should the capacitor be if my AC frequency will be very low? Say 0.1 hz.
C = 1/ (2 x 3.14 x F x Rin ), for F = 0.1 Hz C = 1.59 / Rin. To keep C below 10 uF, R should be > 100 k
But how large should the capacitor be
I wouldn't disagree with the above.
majenko:
Keep your ref at 5V. Connect the incoming signal to a capacitor, then that to an Op-amp. Set the op-amp up so that it swings around a 2.5V virtual ground, and enough amplification so your tiny variations are now around +/- 2V.Simplest arrangement is an inverting amplifier. 50/50 voltage divider on to non-inverting. Signal via capacitor and resistor Rin to inverting. Output through resistor Rf back to inverting. The gain is -(Rf/Rin).
Note that the values then sampled by the ADC will be backwards, so subtract them from 1024 to get the real values.
I'm still not quite sure how this all comes together. So if I connect the signal via a capacitor, my inverting input will wobble very close to 0v. Then the non-inverting is 2.5v (50/50). How is this 2.5v differential going to be amplified?!
This schematic is not the best for your needs. If you put cap in input line, than OPA should have double power line (+ and -). And it will leave DC offset question unanswered. Easiest way in reply #17.
Magician, can you please elaborate why that op-amp circuit won't work? I already have generic op-amps in my bin, so it would be nice to just one of them.
I tested a schematic linked in #17 with LM358, works just fine. Do you have them at your hands?
Magician, I'm trying to understand how everything works, so I'm equally interested in the process as the end result. If you can, please help me understand why the source->cap->resistor->op-amp won't work...
I didn't say it won't, It will works with dual power line +V, GND, -V. Next, as it outputs AC, you will have add one more cap at the output, and two resistors to create DC offset required for arduino ADC be capable to process negative halfwave.
Since it seems as if you want fluctuation on a sine wave which peak to peak is the same, you can ignore the DC biasing on the Op Amp.
If you connect the Op Amp to VCC and GND with a non-inverting amplifier setup where the positive input is GND, you will get a half wave rectified output.
Thus if you are sending it a 0.005 sine wave it will get amplified to a 0.005*(gain) half wave and then you can read in the half wave via your ADC.
This takes out the trouble of DC biasing.
Mushfiq
Yes but you have to have a much more expensive rail to rail op amp. And then you need to protect he input of the op amp to prevent damage when the input goes negitave.
Grumpy_Mike, why won't something like an LM124 work? If it brings the voltage up above 2.5, the Arduino should see that as high, and it should be good enough, no?
the Arduino should see that as high,
I thought you wanted to measure the fluctuations not just detect them? After all it says measuring in the title.
If I can count the frequency of the wave, it will be a good start. 
You can measure just the frequency better using a comparator like an LM339, but you don't say what you expect it to be. My guess is that it is either at mains or it is random so it won't get you very far.
However I thought you said:-
as it's a sign wave. I would like to capture the wave with enough resolution to plot it graphically
Grumpy_Mike:
Yes but you have to have a much more expensive rail to rail op amp. And then you need to protect he input of the op amp to prevent damage when the input goes negitave.
A expensive Op Amp is not required. I am using an LMV321 from Texas Instruments right now to do this exact thing with AC 60Hz signals. The voltage inputs range from 20mV to 1V with a 5x gain.
It is working for me. The input will not go negative due to there being no negative voltage present for it to swing too.
I too think a regular op-amp should be ok in my low frequency problem. Comparator would be the right tool to detect the frequency indeed, but eventually I'd like to get the wave sampled.
So which direction should I look at if an op-amp like LM124 is the heart of the circuit? Should I look at a non-inverting configuration, with the inverting side at ground, then feed the signal through a capacitor into the non-inverting side?
MikeX,
When deciding on op amps, you need to check the Common Mode Input Voltage Range. This is a direct quote from Note 10 on the datasheet:
The input common-mode voltage of either input signal voltage should not be allowed to go negative by more than 0.3V (at 25°C). The upper end of the common-mode voltage range is V+ ? 1.5V (at 25°C), but either or both inputs can go to +32V without damage (+26V for LM2902), independent of the magnitude of V+.
This range determines how much you can swing safely. V+ is your postive input voltage. Thus, you can swing 0 to (V+)-1.5.
Keep that in mind.
