--- In
Homebrew_PCBs@yahoogroups.com, "kilocycles" <kilocycles@...> wrote:
> Which type of film capacitors would be used at RF frequencies (say,
> 1.8 to 30 MHz)?
At the low end of that range many film caps are fine, as long as they
are low inductance construction. At the top and of that frequency
range, I think mylar film (the most common and lowest cost film caps)
start to show the losses of mylar, and polystyrene, polypropylene, and
teflon are better, but larger and more costly. And at all
frequencies, these dielectrics have better temperature stability.
So the big concerns for film (and really, for any dielectric)
capacitors for RF applications is the losses and stability of the
dielectric, and the inductance and series resistance of the
construction. To compare a pair of cases, fairly, you realy need to
pour over the data sheets. If the data sheet doesn't mention
dielectric losses, stability, inductance or resistance, assume the
news is not good or do your own testing.
> I recently had the opposite problem...finding a low
> value 150 pF cap to use in an audio bandpass filter circuit in an Op
> Amp (NE5532 used as the audio final amp). I had to use a silver mica,
> figuring it would be better than a ceramic disc.
Ceramic covers a lot of variation. The low K (NPO COG) dielectric is
low loss and high stability, as well as having capacitance that is
pretty stable with respect to applied voltage. The high K types (Z5U
Y5V) are not only variable with temperature and voltage, but loose
capacitance over time and often have high resistivity metalization (in
the multi layer, monolythic types). The middle K types (X5R X7R) are
intermediate in character (pretty good, pretty cheap and pretty
small). I have given up on the really high K types, and choose
between the low k and mid K types, depending on capacitance and loss
requirements.
Good silver micas are very good, and very expensive. The cheapest
micas have the annoying characteristic of producing small charge steps
when voltage is applied. It is a very specific noise for the type. I
think it has to do with the mica being made up of many distinct
layers, and charge may get trapped in or released from any given layer.