Chip ID

[Haible Juergen]

	>I have the feeling that these strange things happen, because you  
	>do not operate the devices under the recommended conditions.
	>Anything can happen then. You must not rely on any odd "feature".

Well yes, I'm clearly talking about operation *outside* the specified
input common mode range. Early JFET opamp data sheets would just
tell you don't go there, or you'll be surprised what happens (I'm
paraphrasing,
obviously (;->) ). So we exceed the common mode range, and find
phase reversal.
But if a modern device is advertised as "not showing phase reversal",
this doesn't necessarily mean they have invented infinite common
mode range. They might have increased it a bit, and, if I got that right,
they are telling us now: "Don't go there, but if you do, at least what's
awaiting you is not the dreaded phase reversal". (It might be anything else,
like oscillation, though.)

>From a practical point of view, we will always face the problem of overdrive
to some extend. We *could* limit the input voltage range to the common
mode range (like +/-10V in a +/-15V system), but this means some extra 
circuitry. 
OTOH, nobody will expect "infinite" overdrive tolerance of a circuit without
external means of limiting.
But there's the very practical case of signals that are limited within the
supply rails, be it from a previous amplifier stage, or be it by limiting
external
signals to the supply rails with a pair of diodes. So IMO it makes a lot of
sense
to specify the behaviour of opamps under these conditions, even if they
slightly exceed the common mode range for "linear operation".

JH.