I never said anything about your boards. I don't know what the PDS requirements of your boards are. Do you?
You can ignore PDS design on your boards and if they work, they work. But to say that engineering is "superstition" puts the bad light on you. This is not the first time you have described a technique that just "got the job done" without understanding what it was doing or why. That's not really engineering. It's not always bad, but it's not engineering. It's just pushing stuff around until it works.
IESL+ESR,
the
When you say the PDS looks like a good low-Q capacitor, what exactly were your requirements on the PDS? At what frequency does the impedance go down when you add caps? Do you know the frequency content of your power noise?
My point is you weren't doing engineering because you didn't know what you needed from the PDS and you didn't properly characterize it to meet requirements. You just winged it. Like I said above, that is not engineering. It's ok until it doesn't work anymore.
What bypassing scheme is that? I'm not proposing a bypassing scheme. I am saying that to fully engineer a PDS you need to evaluate your needs and then design the PDS to meet those requirements. If a single value of cap does the job, that's fine. But if your requirements are such that you need a specific impedance over a wide frequency range then you may need more than one value or package size cap to minimize the impedance peaks of the interaction between your cap and the power planes.
oscillations,
Ok, so your PDS requirements were very easy to meet. That doesn't mean the techniques used by others are "superstition".
Rick