Re: caps (was: 8250LP....)

From: silverdr_at_wfmh.org.pl
Date: Mon, 21 Aug 2017 20:32:39 +0200
Message-Id: <9312AC08-E260-48AD-8B3D-EA6826185CDB@wfmh.org.pl>
> On 2017-08-21, at 18:38, Francesco Messineo <francesco.messineo@gmail.com> wrote:
> 
> 
> 100nF or 47nF ceramic are ubiquitous close to each digital IC as
> supply bypass, but every board has usually a few electrolytics on the
> supply rails scattered around or just at the power supply connection
> if the board is small. Those are also called "bypass" capacitors (they
> just bypass a larger part of the board, not just the IC surroundings).

I see. It's me that I tend to call "bypasses" only those dedicated to particular components like ICs.

> Tantalum don't leak, but have some other problems (sometimes they
> explode or burn for example,

This might happen quickly when you "overdrive" them or put them the wrong way but exploding (to a degree) and burning is certainly a domain of leaking aluminium ones, even if used correctly :-)

> and cost more than the MLCCs) and their
> ESR is usually quite different from the same capacity aluminum
> electrolytic. So you can't just replace a tantalum capacitor for a
> regular aluminum one unless you understand that ESR/ESL don't matter
> in that application, but when you understand that, then MLCCs are
> cheaper and smaller :)

Unless you need the capacitances to match the bigger electrolytic ones. Especially at higher voltages. For (relatively small) bypasses they're obviously the best choice but I wouldn't be so sure to recommend ceramics as a general replacement for electrolytic caps. For (your) example if we talk about the bigger ones like those used as board power rails filters then you can quickly need to get into mF range. I just checked with a big supplier here and the biggest MLCC I can get is 22µF and is almost twice the price of a 1000µF tantalum and only about 30% cheaper than the 470µF tantalum-polymer one.

-- 
SD! - http://e4aws.silverdr.com/


       Message was sent through the cbm-hackers mailing list
Received on 2017-08-21 19:00:03

Archive generated by hypermail 2.2.0.