Getting good matching offset in folded cascode OpAmp

[shock369]

Hi all,
Can anyone explain how design width and length of transistor in this topology or ratio (W/L) of transistor in differential pairs and mirror to get a good matching offset.
Target of design : Design of the low offset (max. 5mV) rail to rail operational amplifier in CMOS technology. OpAmp be use for measurement.

Vdd=5V
Mim. length 0.7 µm

Thanks

 

[DenisMark]

folded cascode+offset

See my post and reference links on:

The best case for matching is diff. pair and cascode MOSTs working in weak inversion region and mirror MOSTs working in strong inversion region. Be care with rail-to-rail input because offset voltage will change with input common mode voltage.

 

[shock369]

offset analysis of folded cascode

I am beginner in analog ic design. This materials is very difficult to understand. To date i think that all transistor in OpAmp must work in saturation region. If diff. pair transistors work in weak inversion region, Gm opapm be small?

 

[zyyang]

Re: Folded cascode OpAmp

i think the offset is the same as the noise, but for r2r opamp, the input pair may be nmos or pmos or both, so the offset voltage must not be linear, you can refer the followed text

 

[shock369]

Re: Folded cascode OpAmp

That I mast do to avoid structural offset in this configuration? Which transistor determine structural offset a how a eliminate it.

 

[DenisMark]

Folded cascode OpAmp

"structural offset"? I know only systematic and random offset, what do u mean?
MOST can operate in cut-off, triode, saturation, breakdown regions. During saturation it can operate at weak, moderate and strong inversion level. The transconductance effectiveness has maximum at weak inversion level. The keyword for better understanding this is "Gm/Id methodology".

Read this:

look for 1212king.pdf on



Assuming threshold voltage mismatch as the main source one can show that
sigma(Vos)^2=sigma(VthDiffPair)^2+sigma(VthMirrorMOSTs)^2*(GmMirrorMOST/GmDiffPair)^2
That is why better to increase transconductance of diff. pair MOSTSs regarding mirroring MOSTs.
Mismatch of cascode transistors has negligible effect on offset (if mirror MOSTs have enough Rds).

If u have questions it's easy to send me private message. I'm no able to discuss at work.

 

[shock369]

Re: Folded cascode OpAmp

I have one problem whitch tranzistors models support matching analysis tranzistors which wort in weak inversion?