Amplify a 0-5 mV signal to 0-5 V using an AD620.
What size should the gain resistor be? To amplify 5 mV to 5V is x1000 so the resistor should be
$R_G = {49.4 k\ohm \over G - 1} = {49400 \over 999} = 49.44 \ohm $
This is the output with a 56 O gain resistor and an input signal of 5 mVpp @ 40 Hz. I have tested with the next bigger resistor in my inventory which is a 68 O but then the output goes above 5 V. Replacing with different 56 O ressitor did not vary much. I could use some parallel resitors to get closer to 5 V but I think the current range will provide enough resolution when converted to digital.
I realize I set my AWG to 5 mVpp @ 40 Hz but forgot to include an offset to bring it up to V~MIN~ = 0 V and V~MAX~ = 5 mV. I will need to use an offset of 2.5 mV. The signal when using that offset is between 0.96 V and clipped at 3.16 V.
With only 8 V I should bring virtual ground somewhat lower, instead of 4 V I should bring it down to 1.5 V which seemed to be about the lowest an AD620 can go to the bottom rail. For this I will change the resistors of the rail splitter. Maybe I can use this same voltage divider to get the output to always be above 0 V? I find that more important that to get the full 0-5 V range.
The optimal level for the virtual ground, accounting for the minimal distance to the rails for the AD620 and a battery of 8 V, would be halfway between -1.5 and 6.5 V at 2.5 V. What should the rail splitter resistors be set at? The ratio of R2 compared to R1 should be around 15:65. Resistor values of 1200 and 5600 come quite close.
Measuring the voltage drop across the resistors on the breadboard shows 1.66 V and 6.50 V which is pretty good. But what happens to the resulting waveform? Suddenly the gain drops massively to a smal waveform between 0 V and 72 mV. Is my = -1.66/+6.5 V not working for the AD620? Or maybe a loose connection somewhere? Maybe; the datasheet lists ├▒2.3 to ├▒18. No, it's definitely the resistors; I tested.
Maybe I should start a little 'easier'? Say a 0-1 V signal to 0-5 V.
$R_G = {49.4 k\ohm \over G - 1} = {49400 \over 4} = 12.35 k\ohm $
I used the above circuit except with a 12k resistor and an input signal of 500 mVpp (no offset). Why does my scope show the output is only 2.64 Vpp? Maybe it's the 10k resistor on the input, which I use for current limiting. Is it needed at all? In the datasheet I notice the AD620 has 400 O resistors on both inputs. I am not certain but I think I can safely remove the 10k. I tried but the result is the same.
Oops! For some reason I was expecting an amplification to 5 Vpp but of course with a gain of 5 the expected output would be around 2.5 Vpp. So disregard the above paragraph.
To remember: I noticed I should connect my scope's ground to the circuit. Even though its ground is also connected to my AWG, the signal on my scope is a lot cleaner now.
Okay, back to '0-1 V signal to 0-5 V'. Using a 12k resistor and setting my AWG to 1 Vpp then my scope shows a 5.2 Vpp output. This is good.
Now for the next step I will set the offset on my AWG to 500 mVdc so the sine wave will bottom out at 0 V. As expected, this produces a waveform on the output that is clipped at the tops.
To remedy this clipping and allow a higher amplitude I will move the virtual ground up. I will try and make Vs- at -2 V and Vs+ at 6 V (my Lilon battery is about 8.2 V). This accounts for the fact that de AD620 will not go to ground.
The ratio for R6:R7 (using the codes in the schematic above) should therefore be 3:1.
AWG is set to 1 Vpp with 0,5 V Offset. To move the virtual ground up I used an 6,8 kO and a 2,2 kO resistor. On my scope I am seeing a waveform that is pretty close to a sine wave but the tops are much wider then the bottoms. Also there is not a 5x amplification I expect but only to 2,44 Vpp. What could be the cause of the problem?
Relative to virtual ground, the supply voltages measure -1.93 and +6.30 V.
My lithium-ion battery provides a little more then 8 V. Using a voltage divider I want to get +6.5 V and - 1.5 V, accounting for the AD620 which does not go all the way to rail.
