differential amplifier

[Chethan]

advantages of differential pair amplifier

i am designing a differential amplifier which shd have a high gain and a high PSRR. even the offset shd be low.
can someone tell me whether a PMOS differential pair or an NMOS differential pair will be better.what r the advantages and disadvantages of both. i am using a p-sub process.
thanx in advance>>.....

 

[judean]

+pmos differential amplifier

Hi,
I think for a good psrr u should go for a pmos DA.

 

[suria3]

Hi,

Using NMOS as a differential input will give advantage in term of high gain as gm of NMOS is 2 ~ 3 times higher than PMOS, thus it can give high CMRR & PSRR compared to PMOS as the differential gain is high. In term of noise, it is recommended to use PMOS as input stage as the flicker noise is lower than NMOS and for PMOS differential stage, it can be operated from the 0V. Thus i would say, it depend on what application you are desiging for.


Thanks,
Suria

 

[sunking]

the better is pmos + cascode

 

[Syukri]

PMOS use din diff amp to act like active load, so it can control the signal...

The Id1 can be determine by pMOS more effectively because pMOS is good on transfering voltage compare to nMOS which got loss

 

[pthoppay]

If u were amplifier spcifications has stringent noise figure than it would be better to have pmos diff amp than nmos amp.

Regards,
Thoppay

 

[Sezi]

Hello suria3,
what do you mean by you can operate from 0V? that the lower input limit with desired linearity is 0V? if yes, what advantage does it make and in which applications?

thanks
sezi

 

[suria3]

Hi Sezi,

I meant the 0V, because the input PMOS can be turn on at lower potential voltage compared to NMOS.Let say the gate voltage is 0V, the source is 1.5V, so it vgs≥vth. This options can be useful like taking the input of rail-rail which will have swing from VDD to GND, even if you design for a source follower, unity gain buffer, it can be operated from 0V up the certain voltage limitation and about the applications, im not very sure about that. You may refer to other sources from net.

Regards,
Suria

 

[Sezi]

but then if you have nmos pair, input voltage can go upto vdd while for pmos pair this is not possible. so why do we prefer that the input should be able to go down to ground rather then vdd? doesn't the nmos have the same advantage for vdd rail?

 

[yuyangjp]

For p-sub process, I perfer to PMOS as input pair.

 

[suria3]

What you are saying also is true but as said, it is mainly depend on the type of input range come in to the amplifier. For example, when it comes to interface like LVPECL or PECL (source followers) in the chip design level, the out put from them could be in the range of 0.3V to 1.0V (for example). So, in the next block of design, the amplifier must be able to take up this range of swing for amplification, thus in this case PMOS as a differential input is preferred. So, it is totally depend on the input range.