In some circuits you can see several - I found up to 4 - small (e.g.200 nF) caps in parallel with the power supply electrolytic cap wired between + and gnd. What´s the use of them :?:
Thanks,
Markus
hifreq bypass......
just ordinary "engineering practice".
t
They're called "decoupling caps" and they're just there to eat noise on the power rails.
It's not uncommon to see multiple values in parallel like 0.01uF + 0.1uF + 10uF. The larger the cap the lower the frequency range it takes care of.
Large caps don't do well for high frequency so the little ones "help out". :D
Do not mistake the location of these caps in the drawn schematic for their location on the board itself.
There could be any number of sources of outside interference that might create small-but-objectionable spikes along the power lines between where the on-board power originates from and where it gets connected to the semiconductor device in question. Consequently, one may often see small-value caps huddled up close to each one of the chips on a board to provide "bulletproof" power conditioning/decoupling for each semiconductor on board. A nice illustration might be any old (i.e., pre Pentium) mobo or <2meg video card where you'll see nice little rows of DRAM chips, each with their own little 100nf cap.
Note that this cap-per-chip action is different in kind that the use of multiple cap values (e.g., 220uf + 10uf + 0.1uf) at a common source, though both are intended to keep the power supplied to devices smooth and spike free.
Thanks for the clarification. I remember seeing the setting Mark described with 4 200 nF caps distributed over the board. Although I do not remember their precise location most likely they were in the vicinity of chips.
Two more questions:
1. Why do large caps
Quotedon't do well for high frequency
? Is simply because they are too slow or, I don´t know the right word, lethargic?
2. I use a switch mode power supply and I usually get a high tone when I connect it to a TS, Great Cheddar or Fuzz Face. No problems so far with envelope filters. So my assumption is, that this whine just occurs with high gain boxes. Up to now I got rid of this with a lowpass filter in series with the positive power supply line. Normally I use a 10-50 ohm resistor and a large (1000-2200 µF) cap. My question: would I achieve the same effect with the setting Peter mentioned or even with just a e.g. 22 µF and a 100 nF cap??
3. I know how to calculate simple low or high pass filters. Is there any formula for calculating the capacity/frequency dependency for the setting Peter described (without adding a resistor in series)?
zc=1/2*pi*f*c
zl=2*pi*f*l
i think.