Do active filters even care about input/output impedances?

[billcalley]

Hi All,

I've been looking at a lot of active filter design software and
design formulas, and none seem to care too much about the input and
output impedances that the filter will want to "see" when placed in a
circuit. When designing *passive* filters, we would typically specify
these impedance values as a matter of course (normally 50 ohms). Why
don't active filter design programs even ask what the input and output
impedances are that the filter will have to work with, nor state what
it is after the circuit is synthesized? Does it even matter; or will
the frequency response and gain just not be affected by most normal
values? Or is it assumed that the active filter will be placed between
certain impedance values? If not, then how can I tell what the
"optimal" impedance values should be for an active filter? This has me
baffled!

Thanks,

-Bill

 

[Ban]

billcalley wrote:

Hi All,

I've been looking at a lot of active filter design software and
design formulas, and none seem to care too much about the input and
output impedances that the filter will want to "see" when placed in a
circuit. When designing *passive* filters, we would typically specify
these impedance values as a matter of course (normally 50 ohms). Why
don't active filter design programs even ask what the input and output
impedances are that the filter will have to work with, nor state what
it is after the circuit is synthesized? Does it even matter; or will
the frequency response and gain just not be affected by most normal
values? Or is it assumed that the active filter will be placed between
certain impedance values? If not, then how can I tell what the
"optimal" impedance values should be for an active filter? This has
me baffled!

Thanks,

-Bill

You have to consider that active filters are usually used for *low*
frequencies from DC to maybe 10MHz, where the wavelength is much longer than
the mechanical dimensions of the components. We do not have reflections and
the capacitors are much bigger than the parasitic circuit capacitances. That
is why it's possible to have relativly high impedances involved, in order to
decrease capacitor size.
Active filters do not use inductors (coils) as do passive filters.
It is also a convention that these filters are used with 0 ohms driving and

10k loading impedances. The low O/P impedance of an opamp stage will not
affect freq. response or gain, because usually the values are in the 3k3 to

100k region, but for a precision measurement device real values have to be
plugged in. Many programs not only neglect the in/out impedances, but are
based on an ideal opamp model, with infinite gain and node impedances.
Though a good simulation program like spice can take all these into account.
--
ciao Ban
Bordighera, Italy

 

[Bob Eldred]

"billcalley" <billcalley@yahoo.com> wrote in message
news:1123397528.175438.19540@o13g2000cwo.googlegroups.com...

Hi All,

I've been looking at a lot of active filter design software and
design formulas, and none seem to care too much about the input and
output impedances that the filter will want to "see" when placed in a
circuit. When designing *passive* filters, we would typically specify
these impedance values as a matter of course (normally 50 ohms). Why
don't active filter design programs even ask what the input and output
impedances are that the filter will have to work with, nor state what
it is after the circuit is synthesized? Does it even matter; or will
the frequency response and gain just not be affected by most normal
values? Or is it assumed that the active filter will be placed between
certain impedance values? If not, then how can I tell what the
"optimal" impedance values should be for an active filter? This has me
baffled!

Thanks,

-Bill

Most active filters are designed to be driven by constant voltage meaning
nearly zero impedance. This is the case when they are driven from an op-amp
in the circuit before the filter. It's output impedance is nearly zero ohms.
If there is much resistance or impedance in the driving circuit, it will
affect the response of the active filter and must be included in the
response calculations. Think of it this way: Suppose the input resistor on a
certain filter is 10K. Now suppose you drive it with a 5K source, the real
input resistance is now 15K not 10k of the original design and it affects
the response. Driving from the low impedance output of an op-amp alleviates
this issue.

The output side of active filters is usually the output of an op-amp and,
therefore, provides nearly zero impedance to the following circuits. In
other words the active amplifiers "buffer" the circuits from one another.

Passive filters usually do not have amplifier buffering and cannot be
designed for near zero impedance. Therefore, they are designed for a
specific impedance, say 50 ohms. But, because this is not zero ohms it must
be included in the filter's calculations.
Bob

 

[billcalley]

Thanks Bob and Ban for the great explanations! Now I understand much
more about what active filters like to see impedance wise.

Best Regards,

-Bill