integrators or followers

[Grant Richter]

> Another fully valid reason is that this is the cheapest way of getting
> an inversion since you anyhow must buffer the output of the capacitor, so
the
> op-amp must be there anyhow, so just setting it up diffrently makes it do
the
> inversion for free.
> 
Bernie Hutchins always treated the inverting and non-inverting input
of the OTA as being equivalent. He argued that the 100K to 220 ohm
ratio was drastic enough to treat either node like a summing node.
That means you can get either polarity from the same OTA stage using
either input with or without an inverting stage.

I have done it both ways and it works with 3080s and 13700s.