Differential comparator with reference comparison voltage level

[Junus2012]

Dear friends,

I am interesting in this circuit of comparator, as you see it has three inputs, the comparator compare both inputs to a reference voltage and based on the comparesion it generates either 0 or 1 , I never had a chance to deal with three input comparator, may be it looks near to common mode feedback circuit but the later one is analog,



can you please refer me to some material about this circuit so I can design it


below a part of the comparator description by the author



and if you like to have more details this is the paper link

https://ieeexplore.ieee.org/document/7987597

Thank you in advance

 

[wwfeldman]

i'm going with it is a window comparator built from 2 comparators

like
https://www.ti.com/lit/ug/tidub01/tidub01.pdf

not 1 comparator with three inputs

 

[vivekroy]

I never had a chance to deal with three input comparator, may be it looks near to common mode feedback circuit but the later one is analog,

If it was a common mode feedback circuit, it would have compared (voutp+voutm)/2 with VR. From the description by the author it seems like it compares (voutp-voutm) to VR.
To implement this, you would need a difference circuit (ref: http://www.learningaboutelectronics.com/Articles/Difference-amplifier-circuit.php) followed by a comparator.

 

[FvM]

I have presently no access to IEEE, thus I can't review the paper.

Basically it's a simple function: comp = sign(Voutp-Voutm-Vr)

I can imagine many different implementations depending on the available technology and design requirements, thus I don't see much sense in putting the question in such a general way. I presume the paper describes an implementation, what's unclear about it?

 

[vivekroy]

I presume the paper describes an implementation, what's unclear about it?

The paper references to Razavi's textbook for the comparator xD

 

[Junus2012]

Dear friends

Thank you you for your reply

The author didn't give a descreption about the paper nor any citation about it from different people, the author referred to Behzad Rezavi book but not specifically for this circuit. For some reasons the author didn't show the details of this circuit.

And one important interesting thing, the author is trying to adjust calibrate the offset of the main amplier by using this comparator, but what about if this comparator has an offset ? may be the comparator has an offset worse than the main amplifier itself

- - - Updated - - -

I have presently no access to IEEE, thus I can't review the paper.

Basically it's a simple function: comp = sign(Voutp-Voutm-Vr)

I can imagine many different implementations depending on the available technology and design requirements, thus I don't see much sense in putting the question in such a general way. I presume the paper describes an implementation, what's unclear about it?

Could you please refer me to at least one of these possible solutions ? I am using CMOS 0.35 µm technology,

- - - Updated - - -

If it was a common mode feedback circuit, it would have compared (voutp+voutm)/2 with VR. From the description by the author it seems like it compares (voutp-voutm) to VR.
To implement this, you would need a difference circuit (ref: http://www.learningaboutelectronics.com/Articles/Difference-amplifier-circuit.php) followed by a comparator.

Thank you for your answer,
it makes sense, but right now I am thinking how big the circuit will be, and as I have commmented before, of the added amplifier has an offset it will be added to the main amplifer