Please recommend a single op amp to implement a second-order or higher-order high-gain (32dB) active low-pass filter solution

向阳xl

Please recommend a single op amp to implement a second-order or higher-order high-gain (32dB) active low-pass filter solution [Copy link]

Origin of the problem: In order to simplify the circuit and reduce costs, it is hoped to realize an active second-order low-pass filter powered by a single power supply to supply AD acquisition to the microcontroller.

Using a 25m ohm current sampling resistor, the maximum voltage range obtained is 0-75mv.

In order to achieve amplification and filtering simultaneously, an amplifier with a gain of 32dB (40 times) is required. At the same time, active filtering is desired to achieve good filtering effects.

Moreover, the input must be directly into the single-chip AD, and the circuit structure must use the same-phase input.

Design questions:

I have a shallow understanding of analog electronics. I have read the forum's talk about active filters, and most of the information I read is about the two structures of voltage-controlled voltage source and multi-path feedback, most of which are designed as Butterworth and Chebyshev types.

The voltage-controlled voltage source has a common-phase input and positive feedback, but its gain cannot be too large (less than 3).

The multi-path feedback structure is unstable, and since it is input from the inverting terminal, it cannot amplify the positive signal under a single power supply, or a bias must be added.

I looked up the information and found an elliptical filter design. The second order plus the two first orders at the input and output ends form a fourth order filter with a gain of 32dB and a cutoff frequency of 10HZ. (As shown in the figure below) I simulated it with multisim and the effect was ideal. I plan to adopt this solution. (In practice, I removed the first order filter at the output part, which is actually a third order. The simulation did not change much)

I would like to ask you whether this design is feasible. Is there any reliable and simpler solution that you can recommend?

maychang

As long as the op amp input and output selected by the OP are rail-to-rail and the offset voltage can meet the measurement accuracy requirements, this circuit does not seem to have any major problems.

maychang

According to the original poster's request, it seems that active filtering is not necessary. Can third-order or even fourth-order passive filtering be used, plus a stage of 40x broadband amplification?

向阳xl

maychang posted on 2019-11-27 15:19 According to the original poster's request, it seems that active filtering is not necessary. Can we use third-order or even fourth-order passive filtering, and add a 40-fold broadband amplification?

OK, thank you for your replies and suggestions on the first and second floors.

Indeed, if we consider simplification, first- or second-order passive filtering plus an amplifier is also possible.

I want to use fewer op amps, but I think active filtering is more reliable. My mindset is to implement such a filtering and high-gain circuit.

(In fact, it was mainly because my boss suggested that we combine the passive filter and the first-stage amplifier into an active filter amplifier when I designed the passive filter before. But the difference here is that we don’t need such a high gain, so we could use a voltage-controlled voltage source circuit to solve it.)

If the goal of saving components is not achieved here, then we can actually consider passive filtering and amplification, right?

向阳xl

maychang posted on 2019-11-27 15:19 According to the original poster's request, it seems that active filtering is not necessary. Can we use third-order or even fourth-order passive filtering, and add a 40-fold broadband amplification?

Why are the common second-order filter circuit structures voltage-controlled voltage source and infinite gain?

Is this elliptical filter design unreliable in practical applications?

I made this circuit for my graduation project, aiming to design a controller.

maychang

Xiangyangxl posted on 2019-11-27 15:39 Why are the common second-order filter circuit structures voltage-controlled voltage source and infinite gain? Designing this elliptical filter is actually...

Why are the common second-order filter circuit structures voltage-controlled voltage source and infinite gain?

Is this elliptical filter design unreliable in practical applications?

Elliptic filter refers to a filter whose transfer function conforms to the characteristics of elliptic function. Parallel to elliptic filter are Butterworth, Kuibyshev...

I don't think there's anything unreliable about elliptic filters.

If you have further questions, you can ask Teacher gmchen.

gmchen

Xiangyangxl posted on 2019-11-27 15:39 Why are the common second-order filter circuit structures voltage-controlled voltage source and infinite gain? Designing this elliptical filter is actually...

Common second-order filters use voltage-controlled voltage sources and infinite gain because the circuit structures of these two filters are relatively mature, there are more reference materials, and the design calculations are relatively simple. However, the defects or limitations of these two structures are also very obvious, so in some occasions with special needs, circuits with other structures are often used.

In addition, an elliptical filter refers to a filter whose filtering characteristics are approximated by an elliptical function, and a voltage-controlled voltage source or infinite gain refers to a circuit structure, and there is no necessary connection between the two.

The main difference between the elliptic filter and the Butterworth filter and the Chebyshev filter is that the former has a zero point in the stop band, while the latter two are all-pole filters. Due to the existence of zero points in the stop band, the transition band of the elliptic filter is steeper than that of the latter two, but its time domain characteristics or phase-frequency characteristics are worse than those of the latter two.

gmchen

Theoretically, a 4th order filter cannot be converted into a 3rd order filter by simply removing one section, because the zeros and poles of each filter section affect each other, and directly removing one filter section will change the frequency characteristics of the entire filter. So the best approach is to redesign it. However, if the requirements for frequency characteristics are not high, then simulation is sufficient to meet the requirements.

向阳xl

gmchen posted on 2019-11-28 13:40 Theoretically, a 4th-order filter cannot be converted into a 3rd-order filter by simply removing one section, because the zeros and poles of each filter section affect each other. Simply removing one section...

OK, thank you very much for your explanation. I will go deeper to understand the principle and calculation.

向阳xl

gmchen published on 2019-11-28 11:36 Common second-order filters use voltage-controlled voltage source and infinite gain because the circuit structures of these two filters are relatively mature and there are many reference materials, ...

Regarding other circuit structures, do you have any recommendations or reference materials?