I'm trying to use the AD8370 for a data acquisition application. It is connected like this:
VCCI and VCCO are both 3.0 V, which should work according to the datasheet, but it doesn't really specify what the maximum input/output signals are using a 3 V supply. Given that the max input p-p signal for 5 V is 2 V p-p, I was hoping that a 1 V p-p input signal would work. I searched the forums and found the question "AD8370 Max Input Voltage @ Vs = 3.3V", that refers to some additional data, but I cannot see it -- maybe it refers to some mail that isn't disclosed?
In the captures below, a signal generator generates a 1.0 V p-p signal at 48 kHz (low frequency since I wanted to check that this has nothing to do with high frequencies being used -- I first observed the problem for MHz range signals). In the first screenshot, the yellow (C1) and green (C2) traces show the signal at the INHI and INLO pins. The blue waveform is generated by the scope and is C1-C2. The second screenshot shows the signal at the output.
I'm using the LOW gain mode with GainCode 10. Theoretically, this means that the amplification should be about 0.56, but since the load impedance is significantly higher than 100 Ohms, the voltage amplification is a bit higher -- say around 1.0. So, the amplitude of the output signal is the right order of magnitude.
Another example, 0.7 V p-p input below:
Here, the derivative of the difference signal is also shown and it is clear that it is still asymmetric, i.e., the tops are rounded but the bottoms quite sharp.
I'm getting good results for inputs that are around 0.5 V p-p or so and below. Are these results to be expected, or can you spot something obvious I've done wrong? I did test a few modifications of my circuit in order to rule out stupid things:
- I disconnected C50 and C51 -- the output is more or less identical.
- I added a 28 Ohm resistor in series with C52, to make the "source" impedance for INLO the same as R26 in parallel with the source impedance. No difference. (To be exact, when doing that modificaition R26 should also be changed, but so slightly that I didn't care for a quick test.) To be honest, I'm not quite sure if that could affect anything, but it was an easy mod.
- I tried other gain codes (smaller), and also tried using the HIGH gain mode just to make sure that the gain settings work as expected. Everything seems fine. The output is distorted in the same way also for smaller gain codes, so it seems to be a problem on the input side.
Thanks for any ideas or more information about expected 3 V behavior! (I can probably get by by attenuating the input -- for smaller signals I can amplify them to full amplitude (2 V p-p) for the ADC without problems -- but I would have hoped to be able to handle at least 1 V p-p. I don't want to add an attenuation circuit if there is some other modification that could solve the problem...
Staffan
