[R-390] Two caps in parallel

[John Kolb]

On Fri, 29 Mar 2002, Scott, Barry (Clyde B) wrote:

> Ok, so let's say I put a 1mfd cap and a 1000pfd cap in parallel
and plot frequency vs. reactance.  Will the plot look differently
than a single 1.0001mfd cap (if one existed)?  From the responses,
I assume the plot will be different.
>
> Barry(III) - N4BUQ

Yes. Any component is not a pure cap, resistor, whatever, but
has other characteristics also. A resistor will also have series
inductance (very small) from the lead lenght, and capacitance
across the leads and resistive element. A coil will have C from
any one turn to all others, and series R. A cap will have series
L and R. Electrolytics have lots of series inductance and
resistance compared to other types of caps, ceramic and mica
being among the lowest.

If you take a 0.1 uF disc cap of the old 1/2" round style
and short directly across the leads at the body, you create
a resonant tank circuit that you can test with a grid dip
meter seeing a resonance at 4.5 to 5 MHz. (the small .1's
used widely for bypassing on computer boards are probably
mych higher) Above 5 MHz the cap is no longer an effective
bypass, looking more like an L than a C.

If you measure and plot the Z of the large and small cap
in parallel, it will start high at low frequencies, fall
to a minimum value at some frequency, and start going back
up again. At a still higher freq, it will start going down
again as the smaller cap starts to become effective, eventually
bottoming out and starting up again if the measurements are
carried to a high enough freq.