I appolagize if this has been covered, but does paralleling capacitors effectivly lower the esr. When modeling a power supply, can you estimate the same effect by adding a single large capacitor with the lower resistance of all the caps. Does it work much same as paralleling resistors?
Now to the actual design I was thinking of. First, this is to use up some parts I have, and may not be ideal. Would it be a good idea to have the capicitor filter work by having two large capacitors of say 20-40,000uf's, followed by a 4700uf cap of lower esr, followed by a 220uf cap of even lower esr/higher quality, then a .1uf film cap, or possibly bypassing all of the caps with the .1uf MKP caps. Will this create a lower esr for the power supply then simply using a couple of large capacitors, or say a series of 4-8 10,000uf caps?
Now to the actual design I was thinking of. First, this is to use up some parts I have, and may not be ideal. Would it be a good idea to have the capicitor filter work by having two large capacitors of say 20-40,000uf's, followed by a 4700uf cap of lower esr, followed by a 220uf cap of even lower esr/higher quality, then a .1uf film cap, or possibly bypassing all of the caps with the .1uf MKP caps. Will this create a lower esr for the power supply then simply using a couple of large capacitors, or say a series of 4-8 10,000uf caps?
Paralleling capacitors lowers the ESR just as one would expect. It's very common in switching supplies, where high ESR at high frequencies is a serious problem. Grab my impedance converter in the Everything Else forum and use the parallel button to get an idea of what happens. You'll find that lowering ESR by adding films will only work at very high frequencies, higher than you might think, because the basic impedance of the small values is too high to begin with. Pay attention to Z and R. It only works if the electrolytic is doing a poor job at the higher frequencies, and they're often better than commonly acknowledged. Note that all electrolytics are not created equal, and the low ESR ones made for HF work really are much better. Oscons are almost unbeatable on DF, but they have high DA and don't come in high voltages. Lately I've been pondering the idea of high speed regulators and no electrolytics on the output, or series "clean-up" circuits. I used to think it was nonsense, but given the flaws of electrolytics, now I'm not so sure.
Thanks for the reply, that was very helpful. Sounds like paralleling caps is a good thing, but does sound like lots of smaller ones would work better than one or two large ones. Otherwise the effect is less beneficial.
you need to look at your power distribution Z in series with the ps C too - wire adds R and L
too often low esl/esr film caps paralled to big electros are totally wasted in the ps iteslf when the "faster" caps could actually have helped if properly placed on the amplifier board next to the "demand"
at the ps electro terminals the low esl/esr film caps are on the far end of the ps wire series inductance which directly "cancels" the effect of the C
too often low esl/esr film caps paralled to big electros are totally wasted in the ps iteslf when the "faster" caps could actually have helped if properly placed on the amplifier board next to the "demand"
at the ps electro terminals the low esl/esr film caps are on the far end of the ps wire series inductance which directly "cancels" the effect of the C
Conrad - Where did you get the information that Oscons have high DA? I'd be interested in that work if you don't mind. I just got done doing some work with Oscons and had what I thought was a positive result, or at least better than what was in there to begin with. If I can do better I'd certainly be open to it.
Before this thunderstorm knocks me off line...
I measured a variety of electrolytics a couple weeks ago, and the DA of the Oscons was worse than most. DF is excellent, and I really like them for their HF performance, but if DA is a problem in the signal path, and I'm not saying it is or isn't, I'd avoid the Oscon. I'm more into measurement than having the golden ears to hear the more subtle stuff
I measured a variety of electrolytics a couple weeks ago, and the DA of the Oscons was worse than most. DF is excellent, and I really like them for their HF performance, but if DA is a problem in the signal path, and I'm not saying it is or isn't, I'd avoid the Oscon. I'm more into measurement than having the golden ears to hear the more subtle stuff
- Status
- Not open for further replies.