The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> gain boost
https://designers-guide.org/forum/YaBB.pl?num=1287565010

Message started by ontheverge on Oct 20^th, 2010, 1:56am

Title: gain boost
Post by ontheverge on Oct 20^th, 2010, 1:56am

Hi,

I know gain in gain-boosting stage is ~ gm1*ro1*gm2*ro2*av (or equivalently, Rout~=ro1*gm2*ro2*av. can anybody help derive it for me?

thanks,

Steve

Title: Re: gain boost
Post by raja.cedt on Oct 20^th, 2010, 5:59am

hi,
its difficult to explain like this, but still i will try.
gain=effective gm*Rout. Effective gm=gm1.
Rout=gm2*r02*r01 but due to gain boosting gm2 will be magnified by Av, so Rout=gm2*R02*Av*R01.
Hence total gain=gm1*gm2*r01*r02*Av

Why don't you try with small signal analysis..

Thanks.

Title: Re: gain boost
Post by anil reddy on Oct 20^th, 2010, 11:21am

hi,

Replace everything with their small-signal equivalent and try...

do separate analysis for Rout and Gm..then it is easier to understand why the gain turned out to be that value...

best regards...

Title: Re: gain boost
Post by ontheverge on Oct 20^th, 2010, 4:19pm

small signal analysis seems the way to go (unless by inspection?)

I was able to find component values, but not sure what KVL or KCL should we use to get it solved, any help?

Title: Re: gain boost
Post by anil reddy on Oct 21^st, 2010, 5:59am

hi,

first of all, i would suggest you to read analog cmos circuit design by razavi. there is a nice explanation in the book regarding gain boosted cascode...if you go through the amplifiers chapter you will come to know what should be applied to get it solved...

good luck...

best regards...

Title: Re: gain boost
Post by aamar on Oct 22^nd, 2010, 3:25am

you can simply understand this amplifier as 3 stages of amplification.

the first is the M1 transistor which sees at its drain a small signal voltage (e.g. V1) equal to the Vi multiplied by the M1's gain (A1) and is equal to gm1xro1.

This signal V1 is amplified again by Av and applied to the gate of M2, so the gate voltage of M2 is simply V1xAv

The third stage is M2 which has a Gate Source voltage equal to V2-V1, however V2 is much larger than V1 so that it can be ignored for simplification. Then the drain voltage of M2 in this case is V2 multiplied with M2's gain (A2) which is nothing else than gm2xro2.

As a result the Vo is then VixA1xAvxA2, and hence the A=Vo/Vi=A1xAvxA2 or gm1xro1xAvxgm2xro2

best regards,

aamar

Title: Re: gain boost
Post by amarnathdinamani on Jul 6^th, 2011, 2:42am

Look into Baker Li cmos circuit design. He has derived it.You can also do the same intuitively Its like a usual cascode but the cascode's impedence is A*r0 instead of the usual r0.

Amarnath

Title: Re: gain boost
Post by weber8722 on Jul 8^th, 2011, 4:08am

Hi,

a very nice tool for this is Symbolic Spice, SSPICE. It is an old DOS program and gives analythic results like this on any circuit.

:D :)

Bye Stephan

Title: Re: gain boost
Post by Larry_80 on Jul 9^th, 2011, 6:49am

Could someone pls explain to me AC wise how the feedback for the NMOS boosted transistor is negative and in the same token, if we had boosted a PMOS, we would have put the feedback inthe positive terminal of the ideal amplfier.

Title: Re: gain boost
Post by raja.cedt on Jul 9^th, 2011, 9:16pm

hi,
please refer basic micro-electronics books by razavi. Here FB works like if source of nmos M2 decreases then op-amp will decrease it's gate, BTW feedback polarity independent of nmos or pmos.

Thanks.

Title: Re: gain boost
Post by hsh22 on Aug 1^st, 2011, 9:25am

you can use the small signal analysis

Title: Re: gain boost
Post by Larry_80 on Aug 5^th, 2011, 11:56am

raja.cedt wrote on Jul 9^th, 2011, 9:16pm:

Hey, so is the polarity of the feedback for the gain boost diagram posted by author Positive?

Title: Re: gain boost
Post by raja.cedt on Aug 6^th, 2011, 6:06am

hello,
it's -ve fedback. Just cke what's the total phase shift around the opamp at DC.

Thanks.