AW[n]: Re[2]: LF411 Op-Amp dilemma

Haible_Juergen#Tel2743 HJ2743 at
Thu Jan 23 22:11:00 CET 1997

I have to admit that at first I found it a little annoying when the
'ol opamp topic came up again, but now I really enjoy it.
There's so many details all of you contribute - it's really interesting
and I have to try different opamps a lot in the future.

from Gene:
> For example, Electronotes was very fond of
>      using the 748 in comparator applications like square and pulse wave
>      outputs, and darned if I can't find a chip (even the TL082) that 
>      snappier rectangles.

I have never had one of these in my hands. Built comparators from
TL071's, LM311's, even discrete transistors, but never from this old
748. Definetly something I have to make up for!

> The LF devices all use feedforward compensation and so
> are considered generally crap for audio

What exactly do you mean by "feedforward"? I only know about
the common technique of using the miller effect on a cap
between the b and c of a transistor in some gain stage of
an opamp. Do you mean *this* being crap audio, or are you
speaking of something else? And how's the compensation
done in the "better" opamps?

> In the TL081 this comes out to about 13nV/^Hz vs
> about 3nV/^Hz in the 5534.

The TL071 has better noise than the TL081, costs the same,
and should be otherwise identical. Or is there any other
difference between 071 and 081, other than noise?

from JDM:
> I KNOW the 411s work.

This is the most important point. We have a solution that works
without problems. You will never know what side effects there
will be with other solutions. So I encourage everybody who wants
to experiment, try everything and report the results to DIY, but
if you want a quick solution that works, go and get these 411's.

I still don't know why I need a 1456 opamp in my keyboard scanner.
The Oberheim schematics show this one. A very cheap garden
variety (dual) opamp, but unfortunately I didn't have one at home,
and it was weekend. But with any other opamp, I got an overshoot
in the staircase voltage at the reset of the counter, resulting in a
wrong voltage for the lowest "C" key. I tried faster opamps
(LF353, TL072, don't know what else), other slow opamps (LM356),
capacitors across the feedback resistor (even a cap trimmer; at
certain settings I got a certain improovment), but nothing
worked perfectly. Only when I called up my friend in the night and
borrowed a dirty-cheap-old-1456, the circuit worked immediately.
Must be a different behaviour in settling after a voltage step
for sure, but don't ask me *why*.

Opamps always tend to do *not* what you expect them to do
 - at least to some degree. That's why I go discrete more and more
where ever I can.
Remember the thread about low slewrate and its effect on large-
level signals? Sure it changes something, but will the distortion
*always* be unpleasant when just a mix of two full scale saw waves
is processed? (No doubt that it is unpleasant on "normal" audio
I hope I will soon know about the effects of the 741's in the SEM
filter. In my new board I have included a socket for the 741 that
buffers the BP output. I'll try several opamps there, and a short
wire in comparison, of course.
Same for the 301 in the SEM filter.

Alot of different topics in one mail - hope someone found it interesting.


More information about the Synth-diy mailing list