0
\$\begingroup\$

enter image description hereI am new in designing the active filters though I designed Active low, high and bandpass and notch filter successfully (using Opamp, R and C). But these filters are not actually designed by me :-( I am using the online tool to design them.

When initially I designed a normal active low pass filter by seeing its general diagram and then calculating the value of R and C as per my cut off frequency. This filter works fine when I am seeing its result by using function generator as input. But I wanted to design it for an audio signal so when I gave an audio signal as input then my whole signal is vanished out. Because of this, I moved to the online tool.

So I wanted to know the impact of the resistance and capacitor on the input/output signal. What are the criteria for choosing the values of the components? For Eg. A cutoff frequency of 500Hz has many pairs of R and C which give 500Hz.

It will be great to hear your thoughts on this. As I wanted to get deep knowledge in the filter design as these are one of the important circuits in the electronics field.

I am attaching this schematic of low pass filter for 500Hz cutoff frequency which I designed by the online tool. But when I take some other random values of R and C then this circuit not works.

\$\endgroup\$
5
  • \$\begingroup\$ Where is your schematic? \$\endgroup\$
    – winny
    Oct 17 '17 at 5:45
  • \$\begingroup\$ Actually, I want the general idea behind selecting the R and C values. So that I can apply it in general for any filter. So I did not add the schematic but now I added it. So please share your thoughts about it. Thank you. \$\endgroup\$
    – user152036
    Oct 17 '17 at 6:05
  • 2
    \$\begingroup\$ Note that this is not an active filter. The filtering function does not rely on the opamp making it a passive filter (R1 and C2). The opamp, R3 and R5 form a non-inverting amplifier with a voltage gain of around 10. \$\endgroup\$ Oct 17 '17 at 7:30
  • \$\begingroup\$ Ok I also tried the sallen key filter for band pass filtering which is purely active filter, I am just wanted to know how to calculate the values of componets. What are the parameters that I have to consider before selecting any value. \$\endgroup\$
    – user152036
    Oct 17 '17 at 7:38
  • \$\begingroup\$ Why do you want a bandpass filter? To cut out all lows so that men talk like chipmunksand music sounds tinny? Then you cut out all important high frequency consonant sounds in speech and make music sound like an old AM radio? For music I boost the lows a little because I do not use a sub-woofer and I leave the highs flat with no cut. My hifi speakers all have passive LC crossover filters, no active filters. \$\endgroup\$
    – Audioguru
    Oct 27 at 21:19
0
\$\begingroup\$

As you should know phase and impedance ratios of each ratio serve to negate or amplify the input. Since Zc(f) changes 20dB +/- 1 decade relative to R, sensitivity and group delay is low. The power of many C’s combined with optimal parameters and configurations trade off sensitivity with Q of each stage with steepness of skirts (Chebychev) or group delay (Bessel) or low inter-symbol-interference ,ISI Raised-Cosine) or high sensitivity with high Q. www.ti.com tools allow choice of tolerances to best understand. Try an 8th order filter in a quad Op Amp

\$\endgroup\$
5
  • \$\begingroup\$ Thanks @tony Stewart for your help. Actually I tried the tool of ti.com webench design only. In that I have to select the BW and cutoff freq.They show different circuits for same cut off frequency with different Q value and unity gain BW of OPamp. Phase and impedance ration of what !! please clarify me, I am not getting it o_0 \$\endgroup\$
    – user152036
    Oct 17 '17 at 7:43
  • \$\begingroup\$ Atop filters must be defined this way with bandpass f and gain, then Bandstop f, dB. But it does not choose optimum R for low current, so scale every value to >>10k and reduce all C by same then add R for balanced input voltage or null DC offset. \$\endgroup\$ Oct 17 '17 at 7:51
  • \$\begingroup\$ That low current point should be followed in designing all the filter!! And every time I take R > 10k, that's an efficient way to design a filter?? \$\endgroup\$
    – user152036
    Oct 17 '17 at 7:54
  • \$\begingroup\$ It depends. Sometimes I have used 100k range as optimal. But large signals or even DC with Rail-Rail output must use this. \$\endgroup\$ Oct 17 '17 at 8:04
  • \$\begingroup\$ Alright, that means it is better to take R in kilos for low current. I will try it out and do the calculation then revert back to you \$\endgroup\$
    – user152036
    Oct 17 '17 at 8:06
0
\$\begingroup\$

"I am attaching this schematic of low pass filter for 500Hz cutoff frequency which I designed by the online tool. But when I take some other random values of R and C then this circuit not works. "

Increasing the value of R and decreasing the capacitor C by the same factor (or vice versa) should have no effect on the filter as long as you are not approaching "exotic values" like 1 GOhm or 2 pF.

"As I wanted to get deep knowledge in the filter design as these are one of the important circuits in the electronics field."

Some general remarks:

It is not easy to gain "deep knowledge" in filter design because there are many, many alternative ciruits and many design strategies. It is a very challenging task to find the "best" (appropriate) circuit for a specific application.

In most cases, filters of higher orders (n>3) are composed of a series combination of filter stages n=2 (and n=1). However, this is not an absolute requirement. There are other design strategies based on passive RLC structures, which then are transferred into active realizations.

But - as said - in most cases the series approach is used and the following steps are performed:

1.) Filter specification based on typical requirements (path region, damping region and damping requirements),

2.) Selecting a corresponding transfer function (order, suitable approximation - Butterworth, Chebyschev, Besssel,...),

3.) Selecting one of the many available filter topologies (Sallen-Key, MFB, Integrator-stages, GIC-blocks,...),

4.) Using filter tables for finding the pole data for each second-order stage (pole frequency and pole-Q),

5) Using design formulas (available for the various topologies) for calculating the parts values,

Fazit: I think, it is clear now why it may be advantageous to use filter design programs. Otherewise you MUST consult a good text book on active filter design.

\$\endgroup\$
0
\$\begingroup\$

It’s perhaps useful to consider the energy involved in different RC values - for low voltages, capacitances Ike’s than about 1uF are convenient and for higher voltages somewhat smaller values are desirable if you want a low-cost and small-footprint device. For high frequencies you’ll want much smaller capacitances otherwise the power requirements will be high. As with all circuit designs there are trade offs between power consumption, parts cost, physical size and performance (noise immunity and output impedance particularly). So it’s a difficult question to answer without knowing all the parameters.

\$\endgroup\$
0
\$\begingroup\$

Lot of good suggestions here. Please also keep in mind that the output impedance of the driving source affects some filters more than others. For example, in your circuit, the output resistance of the source adds to R1, thereby lowering the corner frequency.

The same circuit has a nice, low output impedance because it's taken directly from the opamp. Though if the next stage has a low impedance input, you may have unexpected results.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.