Decoupling a Standalone Atmega328 and other components

Is additional decoupling required for a minimal-configuration Atmega328? I don't see anything in online tutorials, but that doesn't mean anything.

Additionally in my circuit, I have a properly decoupled MAX7219, and my Recom 5V switching regulator (here) is decoupled per their typical application between Vin and Gnd (.1uF that I'm not sure I need). Along with the 7219, I'm running a 74HC540 octal inverter, and I'm not sure if that needs decoupling, as it does have Vcc and Gnd.

I've been trying to use the rule of thumb to decouple everything, but I wonder if I overdo it sometimes.

So to reiterate:

Do minimum-configuration Atmega328s require decoupling between Vcc and Gnd?
Does my Recom converter need a capacitor at all if Vin = 12V?
Do 74HC540s require decoupling between Vcc and Gnd?

Always decouple electronic circuitry.

Edit
Always decouple electronic circuitry.
.

@Larry - should I decouple chips? :stuck_out_tongue:

Yes Chris, I should have mentioned that.

.

All chips need decoupling All the time.

0.1uf ceramic capacitor between Vcc and Gnd, and between AVcc and Gnd of the '328p is required. These should be right next to the chip. If you see a tutorial on a standalone arduino that omits these (there are a lot of bad tutorials that don't include them), stop reading and find a better one.

All digital IC's should have a 0.1uf decoupling cap right next to the chip - unless of course the datasheet specifies something different.

That switching regulator you have sounds like it doesn't need any additional capacitors - it almost certainly has some inside. You can get smoother output voltage (less ripple, better transient response) by putting a cap across the output - depending on the nature of your application this may or may not be an issue.

It's easier by far to use more decoupling caps than you might really need than it is to have to debug a problem that turns out to be insufficient decoupling.

Is there a point at which too many decoupling caps (say .1uF typically) can adversely affect a system? Granted, my design is hardly that complicated, but for my own knowledge. I'm an AE, not an EE.

Prop-Forge:
Is there a point at which too many decoupling caps (say .1uF typically) can adversely affect a system? Granted, my design is hardly that complicated, but for my own knowledge. I'm an AE, not an EE.

When the start up current surge becomes so large it trips overcurrent protection on the power supply. You'd probably have to get into multi-farad supercapacitor levels to have to start worrying about that though. Salting and peppering your power rails with normal ceramics and electrolytics is unlikely to cause a problem.

Do AE's get to learn about resonance in systems? If so at some point you may start seeing things like the Tacoma Narrows Bridge.

Or something like this old school video.

Anyway, power distribution systems (of any size) are also resonant systems and bypassing is a way to keep the gallopin to a tolerable level.

I was Aeronautics/Flight Controls, so we learned all about what happens when things depart (that bridge was a favorite example of several instructors); I just never thought about how under/overdamping could relate to circuit design. Unfortunately, in both EE courses I was required to take, we spent an inordinate time on theory and not nearly enough on practical.