Cascode amplifier and miller effect

Thread Starter

yuanshikai

Joined Sep 17, 2011
49
Would a folded cascode amplifier be better at reducing the miller effect or would a simple nmos cascode be better?
 
Last edited:

Adjuster

Joined Dec 26, 2010
2,148
At the same time, you might want to think about what the Miller Effect is, how it occurs, and what qualities of circuits influence it.
This may make it easier to understand the merits of the different circuit structures.
 

Thread Starter

yuanshikai

Joined Sep 17, 2011
49
all i really know is that the input capacitance increases if the circuit for the input signal has a high output impedance and cascodes reduce this because the input and output are isolated from each other.. right? but how can different cascodes types differ from each other? will the folded cascode be better at reducing the miller effect because the input and output stages have more differing characteristics?
 

studiot

Joined Nov 9, 2007
4,998
A cascode is a two stage amplifier comprising a common source (FET) or common emitter (Bipolar) followed by a common gate (FET) or common base (Bipolar) amplifier. When the two stages have opposites polarities npn - pnp we say that it is a folded cascode.

A cascade is a two (or more) stage amplifier where all the stages are in the same mode (eg common emitter, base or collector).

In the cascode common source first stage the voltage gain is approximately 1 so there low signal voltage across its drain gate capacitance and little or no miller effect.

Does this help?
 

jimkeith

Joined Oct 26, 2011
540
In the cascode output transistor, capacitive current resulting from the miller capacitance is shunted to common at the base in the common base connection. The drive transistor, if common emitter operates at a constant Vce so it has no miller capacitance, but it has significant voltage gain. If the drive transistor is the opposite polarity as in a folded cascode, it operates in the common collector mode thus has a voltage gain of unity and has the potential of wider bandwidth.

It is my guess that a bipolar output stage would provide better performance than an enhancement mode MOSFET because the drive transistor would be operating with a lower dynamic Vce (that minimizes miller effect) when it drives a bipolar emitter.
 

Thread Starter

yuanshikai

Joined Sep 17, 2011
49
thanks for your ideas.. learned a lot..
but wouldn't it all be about controlling the gain to get as close as possible to unity and not about the type of transistor? for the drive stage, is pmos or nmos better at reaching unity gain? given that these are all in common gate config..
 
Last edited:

t_n_k

Joined Mar 6, 2009
5,455
It might be helpful if we were looking at circuits typical of what you are considering. Are you able to post a couple of schematics? Then we have some common ground for discussion.
 

Thread Starter

yuanshikai

Joined Sep 17, 2011
49
cascode amplifier cascodeamp.JPG
pmos-nmos folded cascode amplifier pmosfoldedcascodeamp.JPG
nmos-pmos folded cascode amplifier nmosfoldedcascodeamp.JPG

This isn't really homework. It feels like it though so i put it here in the homework help section. It's just lab work that i may or may not try to understand. I'd rather understand it though.

I also don't really know if the having two current sources affect the gain of the input transistor.
 

Ron H

Joined Apr 14, 2005
7,063
In the cascode output transistor, capacitive current resulting from the miller capacitance is shunted to common at the base in the common base connection. The drive transistor, if common emitter operates at a constant Vce so it has no miller capacitance, but it has significant voltage gain. If the drive transistor is the opposite polarity as in a folded cascode, it operates in the common collector mode thus has a voltage gain of unity and has the potential of wider bandwidth.
This doesn't make sense to me. In neither case is either transistor in common collector mode.
In the case of BJTs, given that the output transistors are run at identical currents (and therefore have identical transconductances), there should be little difference in performance between cascode and folded cascode. This assumes that both also have the same high frequency characteristics. In the real world, NPNs generally have better high frequency characteristics than PNPs.

For MOSFETs, PMOS transistors need to have approximately 3 times the W/L ratio of NMOS to get the same transconductance, so there are capacitance penalties to pay for using PMOS. One advantage of folded cascode is that the voltage between the rails can be lower.
 

Thread Starter

yuanshikai

Joined Sep 17, 2011
49
So, your implying that, the miller effect is reduced the same whether it's cascode or folded cascode given the same performance of the transistors and the only advantage of nmos is that it has better frequency characteristics than pmos. But a folded cascode will need less voltage between the rails..am i reading u right?
 

Ron H

Joined Apr 14, 2005
7,063
So, your implying that, the miller effect is reduced the same whether it's cascode or folded cascode given the same performance of the transistors and the only advantage of nmos is that it has better frequency characteristics than pmos. But a folded cascode will need less voltage between the rails..am i reading u right?
That's my take on it. Other members might be able to dispute this, or add to it.

EDIT: Folded cascode will generally require more current.
 

studiot

Joined Nov 9, 2007
4,998
There is one other thing to add.

You can view dual gate FETs as a direct development of the cascode circuit. That is you can view the double FET in a cascode as one device with two gates.

Top is a 100Mhz cascode amp

Bottom is a dual gate FET substituted into the same circuit.
 

Attachments

lloger

Joined Nov 27, 2011
4
all i really know is that the input capacitance increases if the circuit for the input signal has a high output impedance and cascodes reduce this because the input and output are isolated from each other.. right? but how can different cascodes types differ from each other? will the folded cascode be better at reducing the miller effect because the input and output stages have more differing characteristics?
who know this question,please tell me. thank you:)
 
Top