#Pi filter design for power supply series#

Even in those demanding applications where an extremely low noise supply is required, there is probably a switching circuit somewhere upstream in the power tree. It has been shown that in many applications an appropriately filtered switching converter can replace a linear regulator for production of a low noise supply. This has kept them out of high performance analog circuits where linear regulators have ruled the roost. However, they have the major drawback in that their outputs can be noisy due to the high switching transients. They are valued for their small size, low cost, and efficiency.

These days switching power supplies are nearly ubiquitous and used throughout every electronic device. I caught a few typos and oddities, but can't claim that I got them all.Designing Second Stage Output Filters for Switching Power Supplies Clear your mind and the problem will solve itself.)ĭISCLAIMER: The young 'uns are being particularly demanding. (Meaning if you think it's hard, it will be. Quote: Nothing is hard, but thinking makes it so. You can always add an L or an R later if you think you need it.

If you already had perfect DC, you wouldn't need the extra filtration. 5 Ohms, then in that limited sense they will behave the same.at DC. 5 Ohms and an inductor has a DC resistance of. Inductors attenuate high frequencies more than low frequencies. DC (aka 0Hz) gets attenuated exactly the same as 120Hz, which gets attenuated exactly the same as RF. Basically, resistors are neutral as far as frequency goes. resistors, this is covered ad nauseum in every electronics text that's ever been written. Or, as a fellow I know used to say: If some is good, then more is better, and too much is just enough.Īs for inductors vs. One general rule of thumb applies: Make it big. They always love it, even if it sounds like fingernails on a blackboard. Just ask anyone who's ever built a speaker. If they used it and it worked out okay, to them it's the best thing in the world. This is one of those human psychology things. If someone wants to put it higher up in the hierarchy, I won't put up much of a fight.īut so-and-so used this filter and what's-his-face used the other filter and they both say their way is the best and I just don't know what to do. Given that capacitance multipliers are generally used in combination with other things, it gets complicated.

I did so because a capacitance multiplier acts like an arbitrarily large capacitor in principle, you could duplicate the effect with a really, really big cap. Also, my inclusion of Capacitance multipliers with straight C filters is rather arbitrary. For instance you could have something like a CRCLC-Capacitance multiplier-C. Pro-good isolation from line problems, very quietĬon-lack of control over rail voltage, bulky, heavyĬon-lack of control over rail voltage, not as good at eliminating high frequency aberrations as CLCĬon-lack of control over rail voltage, not as good at eliminating high frequency aberrations as CRC Pro-excellent isolation from line problems, absolute control over rail voltageĬon-heat, complexity, potential for transient problems (ringing) if not executed properly, potential for dynamic restriction, potential for noise Here's the hierarchy from most best to least best: ).įor such a simple subject, this generates far more than its fair share of angst. And, yes, I know I'm doing a lot of talking without much action, but I'm still buying parts (and I'm a real slow-starter at the best of times - just ask the missus. In my case, looking at a ZV6 with a regulated power supply (like the one described in ZV5), I probably won't bother with the extra space/cost vs performance that the four inductors would give me. So the question, what's more useful in general, CLC, CRC or just plain C?

#Pi filter design for power supply series#

Nelson Pass' post in the second thread especially seems to indicate this.Īnd one other thing I've been wondering is, how much of the effect of the filter is caused by the series resistance of the inductor used in the filter, and how much is due to the inductance? Most of the inductors of the 2mH variety would run between 0.2 and 0.5 ohms, which is almost exactly the value of the resistor that you would use if your were going to use a CRC type filter. There appear to be varying ideas on the usefulness of such filters for the power supply, especially for a class A, balanced amplifier with regulation on the power supply. I've been reading various items on this, including the following threads from this esteemed forum: PI-Filter, PS question - CLC vs C filtering, First Pass project, need Aleph-M PSU advice please and proper design of pi-filter for Aleph 30.