> Kieran, to put a little more reality into the discussion, the > inductor acts similar to a large resistor at higher frequencies, > but a small resistor at DC. XL is inductive reactance [similar > to "impedance"]: Yeah - thanks for pointing this out. I guess I wasn't thinking logically through it. I'd just read an IEE paper on high frequency PCB design and decoupling (which, by the way, talks about using tantalum caps for power supply decoupling... [another thread]), and I think I was a bit blinded by it! I also posted this question to sci.electronics.basics, and got a nice reply from one of the veterans there... It's copied below. -- From: John Popelish Subject: Re: Decoupling? Newsgroups: sci.electronics.basics Think of an analog that substitutes a vibrating engine for the current generated by the sub circuit. The bypass caps are analogous to mounting the engine on a slab of granite, so that most of the vibration from the engine is used to accelerate this large mass, so the amplitude of the vibrations is reduced. But if the granite slab is sitting on the floor, even those smaller amplitude vibrations will still get coupled into the floor (which represents the common supply) so dishes on the table will still rattle a little (representing other circuits sitting on the same supply, ah, floor, though much less than if the engine had been just bolted to the floor. But if you put a layer of foam rubber under the granite, the slab will be able to vibrate as it absorbs momentum forces from the engine, without all of that vibrational motion being also sent out over the floor. The slab is a low impedance path for high frequency forces by virtue of its mass and rigidity (capacitance with low series resistance and inductance), while the foam is a high impedance path by virtue of its resilience and low mass (inductance with low parallel capacitance). So almost all of the vibrational force of the vibrating engine will be transferred to the slab, while very little will make it into the floor, and up the table legs, and the wine in the long stemmed glasses (sensitive amplifier) will show barely a ripple. These sorts of analogies help me picture how circuits work before I get into the mathematical analysis.