hi,
This is what I see, how does it compare with your plot.?
E

Yes... It's exactly what I'm getting. Which I assume... it's not correct at all

 

hi,
A quick check shows the +Vin and -Vin are 'almost' in phase.?
The actual difference gives only a +/-1mV signal.

 

hi,
A quick check shows the +Vin and -Vin are in phase.?
E

Yes, they are, my bad. I tried something different and forgot to change them.

 

OK, repost the Vin changes and I will edit my asc file.
E

 

https://www.google.com/search?q=design+cmos+"rail+to+rail"+"op+amp"+input+non-linearity
versus
https://www.google.com/search?q=design cmos "rail to rail" "op amp" input non-linearity&tbm=isch&tbs=ic:gray
+
https://www.google.com/search?q=design+cmos+"rail+to+rail+I/O"+"op+amp"+input+non-linearity
https://dl.acm.org/doi/pdf/10.1145/1284480.1284501 , https://docs.rs-online.com/3526/0900766b8122e846.pdf

 

I don't think it's working,

LTspice isn't known for it's ability to draw easily readable schematics, but yours is awful. I'm signing off...

 

Ok... so I have this, a differential input. But it's not actually what I was hoping to obtain, what should I change to get a proper output (I mean by that a rail to rail output).... the gain?

I'll try a new schematic with real current sources and I'll work a bit on that intermediate stage.

 

hi b,
One pointer to the design error is that Vout swings above zero volts when driven by a symmetrical input signal, check the internal biassing.
E

 

Try easier comparator architecture.

What is C3 and C4 for? I would delete them and add two more inverters at the output.

 

Try easier comparator architecture.

What is C3 and C4 for? I would delete them and add two more inverters at the output.

Well... I thought I should use them to get rid of the DC level.

I will try to use two inverters, but what are they for exactly, to have a better gain and the same phase?

 

Well... I thought I should use them to get rid of the DC level.

But what for?
What I understand, your aim is to create comparator with rail-to-rail input and rail-to-rail output. Is that correct?

I will try to use two inverters, but what are they for exactly, to have a better gain and the same phase?

If I remember correctly, you wrote somewhere that you are not satisfied with output value that should always be lowest or highest. More inverters in series should make output to be always on any rail. Phase depends on you, if you do not care you may have even number of inverters. More inverters, as you wrote, means more "gain".