how to determine the dynamic range of an active RC filter?

a écrit dans le message de news: snipped-for-privacy@e56g2000cwe.googlegroups.com...

That's funny how people are able to _design_ such things and can't answer such elementary questions. Wondering what definition of designing they have.

Bwahahahaha! They don't _design_, they "plug-in" ;-)

But this one sounds like homework.

...Jim Thompson

My reading was that he picked a cookbook-style op-amp active filter, plugged in some numbers, and got some component values and randomly picked an op-amp off the shelf.

Yeah, some people call that "designing", but realistically "design by cookbook" happens much more often. I rarely design my own op-amps out of grains of silicon myself anymore, and whatever filter topologies I've invented, I can guarantee you they all oscillated or motorboated whether I wanted them to or not!

Tim.

The biggest signal is somewhat smaller than the rails, limited by the rails. The smallest signal is indeed the noise.

Rene

The worst case situation for an an active filter is the response to a step input. Calculate (or measure) the step response. The overshoot value must be less than the output swing capability of the op amp. This is what determines the maximum input voltage Vinmax. If you have designed a multi-stage filter, put the lowest Q sections first. This will minimize the step response overshoot of the first stage. The lowest Q section is the one in which the ratio of the real to imaginary components of the poles is the largest. ~ The smallest signal that the filter can process depends on the minimum acceptable S/N ratio. Let the minimum acceptable S/N ratio = N dB. Let the maximum possible output = Vmax. Then the minimum signal you can process is calculated by Vinmin = Vmax/[10^(N/20)]. ~ The dynamic range is then 20log(Vinmax/Vinmin) Regards, Jon

It will depend on the filter architecture. For example, a state-variable filter will generally have internal nodes which are sq-rt(2) times higher than the input signal at fo.

My "p+1" architecture never exceeds the size of input signal.

...Jim Thompson

Thanks for your detailed explanation,Jon, I would try what you' ve told me! sorry for my poor knowledge in filter and opamp, I only did this reseach for 4 months. But I really designed the opamp my self for a structure of folded cascode. Would you be so kind to reconmend me some books that teach dynamic range and THD of filter Thanks very much!

I wonder what determines the "minimum acceptable S/N ratio" The filter is a continuous-time filter, so if the smallest signal that the filter can process is equal to the noise signal? looking forward to reply! thanks!