[sdiy] IN your mind, what is ....

[Harry Bissell Jr]

Perhaps Bob's opinion is based on manufacturing
variations that might make any solution for one batch
of parts invalid with another.  THAT would make for a 
poor opinion as well.

Most of the apps I've played with used the LM3900
because of its dirt cheap price and used it as a poor
opamp... not for the magic tricks it is capable of.

H^) harry



--- jhaible at debitel.net wrote:
> > > Quoting the LM3900 designer once more:
> > > "Because application design approaches are very
> different
> > > from those used with standard opamps, the LM3900
> is usually
> > > not called an opamp to avoid confusion with
> standard
> > > opamps."
> > 
> > Indeed.  Technically the LM3900 is a Norton opamp.
> 
> Sure. Maybe it even was the other way round, and the
> LM3900
> first coined that word, Norton opamp?
> Are there any predecessors for such a current mirror
> input
> device?
> 
> 
> > But hey, the Philbrick "brick opamp" of the 1960s
> didn't have much gain 
> > above 100 either!  
> 
> Yes, that's technical progress ...
> Today's state of art is tomorrow's junk - or vintage
> device, depending
> on your application. For many audio applications,
> you don't need
> an "operational" amp at all - you just need an amp.
> Discrete, when
> I have the choice. And with a single transistor
> input stage, because
> the second half of a differential pair will add 3dB
> of noise. (;->)
> (But then again, try building a GIC or FDNR filter
> from a LM3900 or a 
> Philbrick opamp ...)
> 
> BTW, when Harry speaks of Bob Pease dissing that
> LM3900 (Hi Harry!)
> this is from _today's_ point of view. Back then, he
> wrote the
> foreword to that Fredriksen Book where the 3900 is
> presented as an
> important part of National's opamp history. And he
> gave me that book
> with the comment that it's still valuable reading
> today. (;->)
> 
> JH.
> 
> -------------------------------------------------
> debitel.net Webmail