# Band Pass Filter Design Simulation

I'm trying to simulate a bandpass filter in ltspice. bandpass filter should be in the 2-5 Hz frequency range I calculated the capacitor and resistor value for this range. However, the simulation results are not good. I use the Sallen-switch bandpass filter design. I watched some videos about bandpass filter. But my frequency range is very narrow. In particular, there is a part of my project that needs to work in the 1-2.5 Milihertz (mHz) frequency range. So when I calculate the value of the components, the results are very different and it is difficult to find them at the right value in the market.

Part of this circuit will be a total of 9 bandpass filter structures. These structures are, for example; Since it is around 1-2.5mHz, 6-16mHz or 270-750mHz, I cannot get a good value in the simulation.

when I calculate gain I use K determine order is that right?

Therefore, what can you recommend in the millihertz frequency range?

Does LTspice give accurate results in the millihertz range?

Can you recommend different sources for bandpass filter designs?

it has calculated by 2-5 Hz of component values

totally my circuit will look like this. I will change the topology

• What exactly are you doing that requires 2-5 Hz and also milliHertz work? Don't hold anything back. Everything matters. Just me, but I'm not in the slightest bit interested in helping with milliHertz problems without knowing every single exact detail about the project. Commented May 28 at 7:42
• @periblepsis This is related to earthquake signals, so it has very small values. If you want to tell me your opinion, I'm listening.
– AEY
Commented May 28 at 7:44
• Okay. You should include this in your question. And it now makes a lot more sense, at least. It's not an area I know much about. You should do your study of how such signals are currently processed and provide us with what you can find out. Also, you should specify things like the minimum and maximum amplitudes, frequency content from the sensors that need to be isolated, what kind of sensors (inertial, strain, etc), amplitude and frequency content of system noise, and an idea of how the measuring systems inherently alter/filter/respond (if one may wish to back out (IFFT) that bit.) Commented May 28 at 7:53
• Filters are NOT just simple mathematical devices. The theory is mathematical. But when you are talking about these frequencies a lot of difficulties intrude (1/f noise sticks out like a sore thumb.) So everything probably matters a lot. What are scientists doing right now? They certainly seem to be succeeding at this. Or are you looking to push the frontiers here? (And now my mind is drifting over to DAS -- fiber -- sensing.) Commented May 28 at 7:58
• @periblepsis okay, thanks for reply :)
– AEY
Commented May 28 at 8:13

If ever there was an application that should 'go digital', this is it.

Implement an analogue lowpass anti-alias filter to cope with the highest frequencies you are interested in, digitise with a delta sigma converter, and then implement all the lower frequency filters digitally.

With a sample rate in the Hz or 10s of Hz region, you don't even need a 'DSP' processor, you could implement the filters in almost any MCU, or even on your PC.

On the down side, if you're not familiar with digital filters, then there's a bit of learning to do.

• yes definitely I will get digital but after this firstly it will filter it after it will go to ADC. For example, if I low pass up to 6 Hz, would it make more sense to do the rest digitally? because the maximum frequency value is 5 Hz
– AEY
Commented May 28 at 8:25
• but I need to read some values ​​then what should I do? for example 270-750 milli hertz
– AEY
Commented May 28 at 8:27
• but I need to read some values ​​then what should I do? for example 270-750 milli hertz  I don't understand the question, you implement the filters digitally, you then have the values at the outputs of the filters. Commented May 28 at 9:11
• yes right I think we can read it that way. statistical signal processing
– AEY
Commented May 28 at 11:55
• @AEY what is 'statistical signal processing'? Commented May 28 at 12:32

I edited the circuit a little so that it gave good answers at the desired intervals. I plan to revise the circuit with the resistance and capacity values ​​currently available on the market.

What opamps can you recommend that are suitable for this circuit?

If I can get the circuit working properly, I can move on to digital design.