> If you have a circuit running at 20 MHz do you need to worry about a cap
> that responds at 100MHz?  I just am not so clear on this (frequency
> response of a cap).

A capacitor like any other electronic component is not just a capacitor,
it is mostly a capacitor. As such, the capacitance part of its
characteristics is the most important. It is also a resistor (the material
it is made of has resistance) and a coil (the leads and parts have
inductivity). So at HF you have to consider all three and what you have is
a RLC group. A RLC group is a resonator and will have a peak frequency at
which its impedance will be maximum (!). So each type of capacitor will
have its own maximum and if you want to be sure that there will be no
maximum you can put two in parallel, of different types. The resulting
impedance characteristic should resemble that of a camel's back (two small
humps), hopefully neither at a frequency where suppression is critical.

The interesting thing is, low ESR capacitors have a very pronounced hump
(the lower the ESR the taller the hump). I have found by experiment that
sometimes putting normal instead of low ESR caps in a smpsu reduces
emissions.

Real low peaking caps are coaxial types, they show a definite impedance at
the peak (works like a transmission line I think).

Peter

--
http://www.piclist.com hint: PICList Posts must start with ONE topic:
[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads