[sdiy] Why do we need the buffers? - gm-C filter question

[René Schmitz]

Hi again,

René Schmitz wrote:
> Your BJT OTA inputs aren't really hi Z in the same sense that opamp 
> inputs are.
> Basically the input current is a function of the input differential pair 
> transistors beta and the amplifier control current Iabc. Ib=Ic/ß. ->
> So assuming a beta of say 200 and 2*Ic = Iabc = 0.5mA, you'll get 
> 12.5uA... 

Crap. There was missing a 10e-3 there... But anyway the datasheet says 
2uA @0.5mA. So ß is more like 125.

If you compare that to the maximum Iout for a 10mV swing at full 
control, its really huge still.

> But a collorary 

make that corrolary...

question could be to ask, why not to use buffers, and
> just divide down at the input and then boost back at the output. Running 
> all stages in between at the same voltage level.
> I'd say the answer to that is its been traditional folklore to do it the 
> other way. You know, an engineer finds this one pole circuit on the 
> datasheet. Cut and paste 4 times. :-)

That assuming the buffers were ideal of course.

Cheers,
  René

-- 
uzs159 at uni-bonn.de
http://www.uni-bonn.de/~uzs159