Hi everyone, I'm working with an instrumentation amplifier and I'm trying to adjust the cutoff frequencies, because I just want to amplify centain frequencies (1kHz-20kHz), this is the circuit

The circuit has a gain of 1000, this is the gain for 1mVp input signal:

I think, It has some distortion because I used the SPICE model of LM224N, but it doesn't really matters.

THE PROBLEM:
This is the frequency response of the instrumentation amplifier:

I want the amplifier have a pass band behavior with cutoff frequencies of fc1=1 kHz & fc2=22kHz.

I tried to solve it, with a high pass filter at the imput and a low pass filter at output, but it doesn't really worked.

Something like this:

And this is the result:

there are two problems, the gain has decreased and the cutoff frecuency of the Low pass filter that I solved fc=20kHz is not working.

I would be very grateful to receive any advice, thanks.

 

There are no pictures in your post.

 

mmm... why can't you see it?

https://s11.postimg.org/sll7bll7n/Imagen1.png

https://s21.postimg.org/7wiq28guv/Imagen2.png

https://s14.postimg.org/p116nievl/Imagen3.png

https://s13.postimg.org/ji2i3ekqf/Imagen4.png

https://s17.postimg.org/ou8s6j9pb/Imagen5.png

 

There are no pictures in your post.

I'm seeing 2 schematics and 3 graphs.
It's a full instrumentation amp made of 2 op-amps, followed by exactly the same thing, followed by a 4 resistor differential amp to get a single ended output.
OP has placed a C-R on the input as a high pass filter and and R-C on the output as a low pass filter.

 

I see them now.
They didn't appear when I first looked at the thread, for some reason.

 

So, how do you think I can change the frequency behavior of the instrumental amplifier? I was reading in a book a topic called frequency compensation, but I'm not sure how to do that in this case

 

Your calculation for the low pass is incorrect. Change the capacitor to 6.8nF

 

1. Why are all of the resistances so low? An LM224 will have difficulty driving them.

2. Look on the datasheet for the gain-bandwidth product, sometimes abbreviated GBW. Note that it is a graph, sometimes called open-loop gain. Look on the graph for the gain at 20 kHz. Divide that number by 1000 (your required gain). If GBW is stated as a single number, divide it by 20,000.

Either way, that is the absolute maximum gain of an LM224 at that frequency. If the feedback loop is trying to set a higher gain, the loop will not close, the input impedance will decrease, the output impedance will increase, the "virtual ground" at the inverting input will not behave, and the filters will not perform as required. Depending on the simulation tool, it might not show any of this.

Bottom line, audio is hard. Out of the jillions of opamps out there, very few will provide significant gain across the audio bandwidth with enough feedback to stabilize a circuit. Consider the NE5532/34.

ak