Are decoupling capacitors necessary for op-amp if DC output?

[sebmadgwick]

I am designing a circuit containing a couple of op-amp circuits, one is a simple battery simulator and the other is an electronics load.  Both circuits will regulate DC.  I do not see the role of decoupling captors for an IC that has no internal switching, clocks, or AC signals.

Am I missing something?

I've included the schematics below for reference.

[dmendesf]

Just the noise can make it oscillate if there's not enough decoupling.

[xani]

It's about signal changes, not signal being AC or DC. Your circuit definitely does have those . It also decouples your opamps from potential interference from other parts of the circuit.

So no, do not save that $0.01 and just put few decoupling caps in, even if chance is small you don't want to be debugging issues caused by lack of one

[srb1954]

I am designing a circuit containing a couple of op-amp circuits, one is a simple battery simulator and the other is an electronics load.  Both circuits will regulate DC.  I do not see the role of decoupling captors for an IC that has no internal switching, clocks, or AC signals.

Am I missing something?

Yes.

You might be only thinking of DC signals but the op-amp still has gain up to several hundred kHz and can potentially oscillate at high frequencies depending on gain and phase margins of the circuit. The impedance of unbypassed supplies can provide an additional spurious feedback path and degrade those circuit gain and phase margins making the circuit more prone to oscillate.

[Siwastaja]

Your theory would work if all signals, including those internal to the opamp, are limited in frequency.

Most "DC" applications have edge case transients (minimally, power-up) that matter, too.

With slow opamps (say, GBW significantly below 1MHz) you may be able to get away without power decoupling. Does it make sense, though? But thinking about it and questioning is a good idea.

This also depends on your power distribution network. Power is coming from somewhere and that somewhere typically has output capacitance already. How long and what shape is the wiring?

[TimFox]

Similarly, if you use a 78xx regulator to reduce the DC voltage from a "clean" source to a "stable load", so that nothing is happening in the time domain, it may well oscillate if you do not have the magic >330 nF capacitor (as specified in the data sheet) from the input terminal to the common terminal.  I once dealt with a colleague's design where he was only looking at the output of a low-noise HF amplifier powered this way on a spectrum analyzer (not an oscilloscope) and didn't recognize the effect of a roughly 100 kHz oscillation in the regulator feeding the amplifier.

[Etesla]

I think the most realistic answer anyone can give is this:
It might have problems, it might not. Because there might be problems, its better to be safe and add the decoupling capacitors for a couple of cents than it is to not add them and risk having problems.