[sdiy] If you thought the LM324 was slow ...

[jhaible at debitel.net]

The PS-synths use LM324s in their VC-ADSRs, as a buffer for the
integration capacitor, and as a schmitt trigger / flipflop for the
attack phase. 
It's a decent solution, especially as single supply operation is required,
and the input voltage goes down to GND. No major speed issues here.
(0.5V/us slew means 30us for 15V - fine for an envelope.)

So I ordered a bunch of these, and I got "something similar" (as
usual, when I buy from Reichelt). I thought you cannot possibly
go wrong with a 324 replacement, could you?

Now I know it is possible. I got "LP324"s instead. And I got a big overshot
of the Attack peak at short Attack times. 

So I finally downloaded a data sheet and saw it's a "low current version" of the
LM324. Which is an improovement in most cases, *except* when
slew rate is critical. The slew rate of the LP324 is "50V/ms".
ms, not us !

The best part, quoting directly from the data sheet:
"This reduction [of slew rate] can cause reduced performance
in AC applications where the LM324 is being replaced by 
an LP324. Such situations usually occur when the LM324
has been operated near its power bandwidth."

Right they are. And I will certainly look at the good old LM324 with
higher respect in the future. (;->)

JH.