AVR-Chat

Zack,

Yes, it is extensively used in digital circuits too, but it's a bad idea.  The analysis that proved to me that it was a bad idea was done in the context of using surface-mount ceramic chip capacitors in the bypassing of digital circuits.  I don't know if there are other considerations for UHF and microwave.  Also, years ago you could only get small values in the smallest, low inductance, packages, so there was a counter-argument for paralleling different values.

A modern ceramic chip capacitor has a parasitic inductance of no more than the equivalent-sized piece of wire.  You can get 0.1uF and higher values in 0402 size.  The effectiveness of such a bypass capacitor is usually determined by how you connect it to the device pins and power and ground planes.

Graham.

--- In AVR-Chat@yahoogroups.com, Zack Widup <w9sz.zack@...> wrote:
>
> Hmm ... this technique is extensively used in UHF and microwave circuits.
> You will often see two or three bypass capacitors of different values at
> each point where a power lead or trace approaches an active device.
> 
> Zack
> 
> On Thu, Oct 11, 2012 at 7:59 AM, bayramdavies <Yahoo37849@...>wrote:
> 
> > **
> >
> >
> > --- In AVR-Chat@yahoogroups.com, Zack Widup <w9sz.zack@> wrote:
> >
> > > .. bypass the power ... with several
> >
> > > capacitors in parallel - maybe a few
> > > microfarads in parallel with a 1000
> > > pF and a 10 pF.
> >
> > Using different values of capacitors in parallel for high-frequency bypass
> > is not a good idea. The smaller capacitors can form parallel resonant
> > circuits with the parasitic inductance of the larger capacitors and
> > increase the impedance at high frequency. The way to do it is to use one
> > sufficiently large capacitor with the lowest practical parasitic impedance
> > and pay close attention to parasitics in routing.
> >
> > You can, of course, add an electrolytic somewhere or other to help with
> > low-frequency stability.
> >
> > Graham.