Diode noise

[Martin Czech]

:::Harry:  I tried replacing the cap with a resistor (470K) and
:::got _nothing_.

Of course not, the dc bias simply punched the high gain amp into rails.
If you don't like the cap, you could design the high gain amp as summer
and add some voltage that is derived as the average of the module
output voltage. This has three advantages: 

1.)
The module output has no dc,
whatever the offset voltages of the high gain amp may do (of course, the
amp for the integration of the output offset should be OP07 or better,
but it needn't be high bandwidth).  

2.)
The low freq. corner can be very low,
since you are free in setting integration resistance.


3.) 
Electrolytics are crap for signal coupling, this could *perhaps* cause
a strange noise distribution while the diode gives pretty much perfect
white sizzling noise.

Of course, this is a servo loop, if not properly designed it will
oscillate, may be at a very low frequency... OTOH the poles are so much
spread, especially the integrators pole that proper "looping" should
beeasy to design.



High gain phono preamps were once done that way...

The 1meg current limiting resistor could be turned into a pot,
noise should change a bit with bias current.


m.c.