Problem with a integrator in LTspice

[icaroM98]

Hello,

I'm trying to simulate an integrator circuit in LTspice, but it doesn't work.
By the transfer function, the circuit behaves like an integrator and like a low-pass circuit.

When I input a square wave, the desired input signal should be a triangle wave, but the output that the LTspice shows is a constant saturated 12V signal.

What could be wrong with this simulation? I have attached the simulation images.

Thank you very much in advance.

 

[FvM]

By using an additional inverter function, your circuit has positive instead of negative feedback. To make an integrator, you can omit U2 and connect C1 to U1 output.

Initial condition of the integrator however needs special attention. It can be achieved by an explicite .IC statement, using the UIC option in transient simulation or a reset circuit.

 

[icaroM98]

I don't understand why the circuit has positive feedback. Can you explain to me?

Also I tried using the 'uic' option in transient simulation.

As a reference, I have also attached my calculations for the expression of the circuit's output. Sorry for the poor image quality.

 

[FvM]

Feedback loop is constituted by two inverting amplifier stages, results in overall positive feedback. Unstable by design.

--- Updated Oct 14, 2021 ---

If you want a non-inverting integrator, search for deboo integrator circuit.

 

[icaroM98]

That Cool! It didn't cross my mind that this circuit was unstable. This other circuit you indicated helps me a lot.

 

[Audioguru]

The opamp has no DC negative feedback (a resistor parallel with the integrator capacitor) causing its DC gain to be almost 1 million. Then its output will probably be at the positive or negative supply caused by it amplifying its DC input offset voltage 1 million times.