EEVblog Electronics Community Forum
Electronics => Projects, Designs, and Technical Stuff => Topic started by: RoyFokker on May 25, 2018, 01:00:42 am
-
xd
-
How are you measuring the low frequency noise and what kind of levels are you seeing?
-
Have you checked that the output of the first stage is near 0V. The input bias currents of the AD797 are too high to use 20M for the biasing, try 1M, same as the feedback resistor - for now. They'll still be too high, but should give some idea whether or not the X100 gain is needed.
-
In both cases the input bias current and input current noise is too high for good results. For good low frequency performance with a piezoelectric transducer, a low input bias current amplifier is required.
-
Worst case output offset for the instrumentation amp from offset current is 35e-9*1e6*20, that's not going to make it clip.
Not that it's relevant ... you are using BJT based amplifiers on an extremely high impedance source, don't do that. Use an AD8220 or something.
-
On the AD797 one, C7 16uF is the wrong way round. :)
-
On the AD797 one, C7 16uF is the wrong way round. :)
I like to destroy virtual components!
Also C6 and quite a bit more obvious.
JS