Decoupling Capacitors

[JDemo]

Below I've attached an image of a basic power supply schematic (Img C) I found on the internet. It pretty much gives a good general idea to my question. My question is for the traces from the regulator, would the output trace (VCC33 in this case) come directly from the regulator (Img A) or would it have to branch off of the trace going into the 2nd decoupling capacitor (Img B) or another option which I'm unsure of?

[wraper]

Layout on the picture is bad but the worst part is that AMS1117 shouldn't be used with ceramic cap on the output due to too low ESR and it may oscillate. It should be used with electrolytic or tantalum cap or ceramic cap with added series resistor. You should take output through the output cap, branching defeats the purpose of the cap.

[JDemo]

Layout on the picture is bad but the worst part is that AMS1117 shouldn't be used with ceramic cap on the output due to too low ESR and it may oscillate. It should be used with electrolytic or tantalum cap or ceramic cap with added series resistor. You should take output through the output cap, branching defeats the purpose of the cap.

Thanks for the info on the caps. Basically light the image I've attached now (But with electrolytic or tantalum caps)? *Ignore the shitty traces, just for example*

[wraper]

Yes, like this. Though place capacitors close to the IC terminals. Probably best to rotate them by 90 degrees.

[JDemo]

So is this ok? Also on an unrelated/semi-related note, would this work as a valid 5v -> 3.3v power source?


Also ignore all the random planes and vias Im just testing some things which was for an unrelated problem where I couldnt for some reason output CAM files when using a ground plane. Turns out for some god awful reason my projects switch units from mils to inches and I type in the measurement for mils into a project using inches and it breaks things

[T3sl4co1l]

I'm going to be bold and say,

It doesn't matter.

Tim