Hi Steve,
Wanted to ask you about the lower-end of the frequency range for VRM measurements. On one of your EDN articles, you suggest 1kHz is a low enough frequency to cover most PDN measurements, but what about lower than that? Most of the VNA's I'm looking at (Copper mountain, Keysight, R&S, Bode etc...) all seem to be around 5kHz, apart from Bode and Keysight (they go to down to 1-5Hz). I want to make sure I cover all frequency ranges. I know the PMIC will cover from DC to a 10's-100's kHz. But if I have, for example, a sensitive circuit that's switching (transient current) lower than 5kHz, say 1kHz , should I be worried if the VNA only measures to 5kHz? I'd expect the PMIC to maintain a low impedance across the low frequency range and the impedance profile to be flat and then peak at some point following the output inductance. Is this point generally much higher than 5kHz? It's clear to me what the benefits are of measuring down to Hz. Is it mainly for passives? This was the article I read: https://www.edn.com/measure-pdn-on-a-budget/ Thanks for taking the time to read my list of questions.
The PMIC, depending on its design tends to be inductive. See my article The Inductive Nature of Voltage-Control Loops. There can be a resonance near the bandwidth if the control loop isn't supper stable. So we do want to measure that. The question is the frequency range of the load sensitivity. Measure as low as as you can, and below a couple of kHz look at J2113A instead of J2102B. J2113A goes down to DC.