MOS ladder for comparator

[Fabien]

Hi all,

I'm just wondering if it's possible to design a comparator with PMOS in saturation as ladder? The power consumption of the ladder can be quite low and the size of PMOS is smaller than Polyresistors.

This design seems to work, any suggestions?

Thanks

 

[artmalik]

Where is the feedback for the op-amp stability. This schematic wont work as you have drawn.

 

[Fabien]

Ok, but after an AC analysis, the OpAmp seems to be stable (unconditionally).
And when I ran the simulation of the comparator with MOS ladder, it's working.

Is it wrong?

Thank you.

 

[artmalik]

any op amp has to have a negative feedback for DC stability, so your AC sims will fine .... but DC feedback is a must for any op amp based designs. The above circuit will work in simulations but not in reality.

 

[erikl]

any op amp has to have a negative feedback for DC stability, so your AC sims will fine .... but DC feedback is a must for any op amp based designs. The above circuit will work in simulations but not in reality.

A well-compensated opAmp - as shown in the OP's ac analysis - doesn't need negative feedback for stability if used as a comparator. And DC feedback isn't necessary in this case, as the comparison input is well biased (Vref). Perhaps some positive feedback should be inserted to achieve a bit of hysteresis (could be verified with small input changes around Vref).

Otherwise I'd think the circuit will work fine - also on silicon.

 

[artmalik]

I agree if the design is being used as comparator then it will fine.

 

[erikl]

I agree if the design is being used as comparator ...

... as the title and the original posting clearly tells and shows ;-)

 

[Fabien]

Thank you both of you for your answers.

any op amp has to have a negative feedback for DC stability, so your AC sims will fine .... but DC feedback is a must for any op amp based designs. The above circuit will work in simulations but not in reality.

For sure, if the design is used as an amplifier, a negative feedback should be used.

... as the title and the original posting clearly tells and shows ;-)

The circuit actually needs 400mV of hysteresis. The main question is: MOS ladder vs resistor ladder. I've seen several comparator designs, every time designed with polyresistors. The silicon area is very huge (more than 20 000µm²) for Mohms resistors. The MOS ladder is very small and I ran some MC analysis, it seems to be quite accurate!
If it is a good way to do so, I'll use this design.

Thank you.

 

[erikl]

The MOS ladder is very small and I ran some MC analysis, it seems to be quite accurate!
If it is a good way to do so, I'll use this design.

Sure you can do this! Several times I've successfully used small W/L MOSFETs instead of high MOhm resistors. In your case you even don't have to mind about accuracy and temperature dependence. Just care for good layout matching of both transistors! Bonne réussite!