I was trying to simulate some opamp circuit in LTSpice and was coming to a dead end everywhere. I did some trouble shooting and finally landed on a strange behaviour.
Attached is a simple asc file of an opamp with the + input tied to output, - input grounded.

Under this condition, afaik, the output should be at Vcc+ or Vcc-. But it is at 0 volts.

Can someone please tell me what is going on?

 

But it is at 0 volts.

Not quite 0, but a theoretically stable point. Try editing the simulation command by ticking the 'Skip initial operating point solution ' check box.

 

Under this condition, afaik, the output should be at Vcc+ or Vcc-. But it is at 0 volts.

Can someone please tell me what is going on?

I think @Alec_t nailed it.

Unless told to do otherwise, Spice begins every .TRAN and .AC analysis with a .OP analysis to determine a starting point for the circuit's operation. I've found that very often with bi-stable circuits such as the one you posted, Spice will perversely converge on some initial operating point wherein the circuit is "balanced on a knife edge" between its two stable states-- something that simply doesn't happen in real life due to natural random noise or outside interference acting to get things underway. Oscillator circuits that mysteriously refuse to start oscillating are another common result of this behavior.

To get around this I usually insert something into the circuit (like a PWL or PULSE voltage source) to force the starting state I want.

 

? what circuit model you use for the x741 chip
// -- /!\ the uAh741.asc is experimental transistor model of the thing not too much tested /!\
// -- i may have better v. of it but it's what i found now ... to be compared with the macro models

 

Here is the behavior of an operator with the model on transistors (two stable states):

 

See

 

Opamps in a simulation have no noise or input offset voltage to toggle the bistable action, a REAL circuit uses its noise or input offset voltage to toggle.

 

? what circuit model you use for the x741 chip
// -- /!\ the uAh741.asc is experimental transistor model of the thing not too much tested /!\
// -- i may have better v. of it but it's what i found now ... to be compared with the macro models
View attachment 216622

I used this.
https://forum.allaboutcircuits.com/threads/how-to-add-op-amp-741-to-lt-spice.139069/

And I get this.


Maybe, I'll have to look for other 741 models or use Bordodynov's LM358.

Thank you all.

 

Opamps in a simulation have no noise or input offset voltage

? really

 

Opamps in a simulation have no noise or input offset voltage to toggle the bistable action, a REAL circuit uses its noise or input offset voltage to toggle.

If it was so, then this oscillator should not work. Or am I missing something?

 

See

 

https://www.ti.com/lit/an/sloa059/s...67364&ref_url=https%3A%2F%2Fwww.google.com%2F
( https://e2e.ti.com/blogs_/archives/...et-voltage-and-open-loop-gain-they-re-cousins )
... another way is to simulate the AOL with different -VIO offsets the one with the highest gain is likely the VIO ... and it does exists for most Op Amp macromodels

 

Not quite 0, but a theoretically stable point. Try editing the simulation command by ticking the 'Skip initial operating point solution ' check box.

Thank you. This makes the difference in the given asc. Let me try this in my main simulation.

 

Not quite 0, but a theoretically stable point. Try editing the simulation command by ticking the 'Skip initial operating point solution ' check box.

So far so good!!

Thank you.