MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET...

20
Basics of the MOSFET The MOSFET Operation The Experiment MOSFET Characteristics- Theory and Practice Debapratim Ghosh [email protected] Electronic Systems Group Department of Electrical Engineering Indian Institute of Technology Bombay 20 September 2012 Debapratim Ghosh Dept. of EE, IIT Bombay 1/20

Transcript of MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET...

Page 1: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

MOSFET Characteristics- Theory and Practice

Debapratim Ghosh

[email protected]

Electronic Systems GroupDepartment of Electrical EngineeringIndian Institute of Technology Bombay

20 September 2012

Debapratim Ghosh Dept. of EE, IIT Bombay 1/20

Page 2: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

MOS StructureMOS Structure Operation

Introduction- the MOSFET

Metal Oxide Semiconductor Field Effect Transistor

The name describes nearly everything about the device itself.

The first three words Metal Oxide Semiconductor describes the layer-wisestructure of the device.

The last three words Field Effect Transistor describes the principle ofoperation.

MetalOxide

Semiconductor

Debapratim Ghosh Dept. of EE, IIT Bombay 2/20

Page 3: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

MOS StructureMOS Structure Operation

MOS Structure Physics

MOS transistors can be of two types- NMOS and PMOS.

An NMOS has a lightly doped p-substrate (where there is scarcity ofelectrons).

The metal terminal is called the Gate.

The oxide layer (usually SiO2) is an insulator.

The p-type substrate is grounded while the gate voltage VG is varied.

We will see how the MOS structure behaves as VG is varied.

p-sub

VG

Debapratim Ghosh Dept. of EE, IIT Bombay 3/20

Page 4: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

MOS StructureMOS Structure Operation

MOS Structure Physics- Accumulation

Let us apply a negative gate voltage i.e. VG < 0.

This negative VG sets up an electric field through the oxide.

The electrons (minority carriers) are pushed away towards ground, and theholes (majority carriers) are pushed towards the oxide.

The region below the oxide is now devoid of n-type charge carriers.

This region of operation is called accumulation region.

--------

+++++++++++++++ E-field

VG < 0

Debapratim Ghosh Dept. of EE, IIT Bombay 4/20

Page 5: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

MOS StructureMOS Structure Operation

MOS Structure Physics- Depletion

Let us apply a small positive gate voltage.

This small VG sets up a weak electric field though the oxide.

The holes are now pushed away from the oxide, deep into the substrate.

However, the electric field is too weak to pull all the minority electronstowards the oxide.

At this time, the immediate region below the oxide is devoid of any mobilecharges (electrons or holes).

This region of operation is called depletion region.

Weak E-field

VG = 0+

Accumulated carriersRegion devoid of carriers

Debapratim Ghosh Dept. of EE, IIT Bombay 5/20

Page 6: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

MOS StructureMOS Structure Operation

MOS Structure Physics- Inversion

Let us now increase the gate voltage VG .

As VG increases, the electric field becomes stronger, and the minorityelectrons accumulate below the oxide.

At a certain value of VG , the concentration of moblie electrons becomes sohigh that the region just below the oxide becomes as n-type as the rest of thesubstrate is p-type.

This region of operation is called inversion region.

These accumulated electrons can now be used to generate a current.

----------------

E-field

VG ≫ 0

Debapratim Ghosh Dept. of EE, IIT Bombay 6/20

Page 7: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

MOS StructureMOS Structure Operation

MOS Threshold Voltage

To accumulate mobile electrons below the oxide, the gate voltage VG has to besufficiently high to cross the “threshold”. This is governed by a number of factors.

The work-function difference between the gate and the silicon substrate (thisleads to the “flatband” voltage).

The gate voltage component required to bring about surface inversion(surface just below the oxide).

The concentration of acceptor ions in the substrate.

The concentration of trapped charges inside the oxide.

The substrate voltage (so far weve assumed it to be ground).

Note:

The threshold voltage VTN for NMOS is positive.

The threshold voltage VTP for PMOS is negative.

Debapratim Ghosh Dept. of EE, IIT Bombay 7/20

Page 8: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

The MOS TransistorOperating Regions of the MOSFET

The MOS Transistor

Once the threshold has been crossed, we need to make the electrons move,i.e. set up a current.For this, we need two more terminals- Source (S) and Drain (D), and apotential across them to control the flow of electrons.The drain and source are heavily-doped n-type regions.We now have a 4-terminal device- drain, source, gate and body.The drain and source can be interchanged!

Source (S)

Gate (G)

Drain (D)

Body (B)

p-substrate

n+n+

Debapratim Ghosh Dept. of EE, IIT Bombay 8/20

Page 9: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

The MOS TransistorOperating Regions of the MOSFET

MOS Transistor Characteristics- Linear Region

Assume that VG > VTN and VGS − VTN > VDS .The device is on as the threshold has been crossed. The inversion layer (fullof electrons) is now a connecting path between the two n+-type source anddrain regions.Due to a nonzero VDS , electrons flow from the drain to the source via theinversion layer. The inversion layer is now called a channel.The current flowing in the channel is called the drain current (ID). For thisbias condition, ID is given by

ID =kn

2(2(VGS − VTN)VDS − V 2

DS) (1)

VS = 0

VG

VD

n+ channel

n+n+

Debapratim Ghosh Dept. of EE, IIT Bombay 9/20

Page 10: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

The MOS TransistorOperating Regions of the MOSFET

MOS Transistor Characteristics- Linear Region (cont’d...)

Based on our discussion so far, try to do the following exercises.

For the above biasing, plot a graph of ID v/s VGS as you increase VGS ,starting from 0V. You may assume that VDS is small (though not necessary).Now you know why this is called the linear region!

Now for a given VGS , plot a graph of ID v/s VDS as you increase VDS ,starting from 0V. At what value of VDS is the ID maximum?

Debapratim Ghosh Dept. of EE, IIT Bombay 10/20

Page 11: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

The MOS TransistorOperating Regions of the MOSFET

MOS Transistor Characteristics- Saturation Region

We have seen that the ID reaches a maxima when VDS = VGS − VTN .

At this time, we see that the VGD = VTN . At this time, the channel depth at thedrain-substrate interface is zero. This is called pinch-off.

When VDS is increased further, VGD < VTN and the pinchoff point shifts towardsthe source.

The ID is now very weakly dependent on VDS . The channel voltage is equal toVDS,sat = VGS − VTN . The rest of the drain bias voltage is across the pinched-offregion.

Substituting VDS = VGS − VTN in equation (1), we get

ID =kn

2(VGS − VTN)

2 (2)

VS = 0

VG

VD

n+n+

Debapratim Ghosh Dept. of EE, IIT Bombay 11/20

Page 12: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

The MOS TransistorOperating Regions of the MOSFET

The Saturation Region (cont’d...)

From equation (2) we see that ID is now independent of VDS .

The plot of ID v/s VDS in the saturation region is a straight line parallel tothe VDS axis.

That does not happen practically. If the effective length after pinch-off, i.e.L′ , is significantly less, ID does change with VDS !

We know that, kn = µnCox(WL). After pinch-off, we have kn = µnCox(

WL′).

Assume L′ = L−∆L. Substitute this in (2).

VS = 0

VG

VD

L′

L

n+n+

Debapratim Ghosh Dept. of EE, IIT Bombay 12/20

Page 13: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

The MOS TransistorOperating Regions of the MOSFET

The Saturation Region (cont’d...)

We now have

ID =

(

1

1− ∆LL

)

µnCox

2

(

W

L

)

(VGS − VTN)2 (3)

It can also be shown that, ∆L ∝

VDS − VDS,sat . Using power series, we get

1−∆L

L= 1− λVDS

Assuming λVDS ≪ 1, equation (3) now becomes

ID =µnCox

2

(

W

L

)

(VGS − VTN)2(1 + λVDS) (4)

Clearly, the decrease in channel length causes ID to be linearly varying with VDS !This is called channel length modulation, and is a critical issue in IC design.

Debapratim Ghosh Dept. of EE, IIT Bombay 13/20

Page 14: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

The MOS TransistorOperating Regions of the MOSFET

BJT and MOSFET- A Comparison

Bipolar Junction Transistor MOS Transistor

1. Current-controlled current source 1. Voltage-controlled current source

2. Current flows due to both electronsand holes (bipolar)

2. Current flows due to one type ofcarrier (unipolar)

3. No two terminals are interchange-able.

3. Source and Drain can be inter-changed.

4. No two terminals are strictly iso-lated.

4. Gate is isolated by means of aninsulator.

EB

C

An n-p-n BJT

S

G

D

B

An n-channel MOSFETDebapratim Ghosh Dept. of EE, IIT Bombay 14/20

Page 15: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

Threshold VoltageOutput DC CharacteristicsInput Characteristics in SaturationOutput Small Signal Characteristics

Experiment- Part 1

In this part, we will measure the NMOS threshold voltage. We will use the ICCD4007.

Connect the NMOS substrate to ground, and the PMOS substrate to VDD .

We will operate the NMOS in the linear region. Apply a small VDS of around0.25 V and keep it constant for a set of ID v/s VGS readings.

Vary VGS from 0 to VDD and note ID .

+−

+−+

-

+

-VGS

VDDVGG VDS

IDRG

RD

Debapratim Ghosh Dept. of EE, IIT Bombay 15/20

Page 16: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

Threshold VoltageOutput DC CharacteristicsInput Characteristics in SaturationOutput Small Signal Characteristics

Experiment- Part 1 (cont’d...)

We expect to see an ID v/s VGS plot like this.

VGS

ID

Extrapolating the linear portion of the plot to find the intercept on the VGS axisgives us VTN .

Q: Why do we need to extrapolate?

Debapratim Ghosh Dept. of EE, IIT Bombay 16/20

Page 17: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

Threshold VoltageOutput DC CharacteristicsInput Characteristics in SaturationOutput Small Signal Characteristics

Experiment- Part 2

In this part, we investigate the ID − VDS characteristics.

The circuit to be used is the same as in Part 1.

For a fixed value of VGS , vary VDS to get different values of ID .

The expected ID v/s VGS plot is as shown.

VDS

ID

This plot will help you find the Early Voltage VA. How?

Debapratim Ghosh Dept. of EE, IIT Bombay 17/20

Page 18: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

Threshold VoltageOutput DC CharacteristicsInput Characteristics in SaturationOutput Small Signal Characteristics

Experiment- Part 3

In this part, we look at the ID − VGS relationship for an NMOS in thesaturation region.

We make VGS < VDS . This ensures that VDS > VGS − VTN .

We know that in the saturation region, ID = kn2 (VGS − VTN)

2.

What kind of an ID v/s VGS plot do you expect?

+−+

-

VGS

VDD

IDRG

RD

Debapratim Ghosh Dept. of EE, IIT Bombay 18/20

Page 19: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

Threshold VoltageOutput DC CharacteristicsInput Characteristics in SaturationOutput Small Signal Characteristics

Experiment- Part 4

In this part, we will measure the small signal transconductance gm of the NMOS,defined as

gm =∂id∂vgs

VDS

(5)

Here, id and vgs are small-signal quantities. Measure gm using the circuit shown.

Bias the NMOS in saturation with VGS = 2V and VDS = 5V.

Now apply a sine wave and find out the voltage gain Av = Vout/Vin. Also,Av = gmRD . From this you can find the unknown gm.

VDD

Vout

Vin

RG RD

Debapratim Ghosh Dept. of EE, IIT Bombay 19/20

Page 20: MOSFET Characteristics- Theory and Practicedghosh/mosfetSlides.pdfBasics of the MOSFET The MOSFET Operation ... Electronic Systems Group Department of Electrical Engineering Indian

Basics of the MOSFETThe MOSFET Operation

The Experiment

Threshold VoltageOutput DC CharacteristicsInput Characteristics in SaturationOutput Small Signal Characteristics

Food For Thought

* Bring out some differences between a BJT and a MOSFET, other than theones mentioned in this document.

* Rather than using a metal for the gate, the contemporary VLSI industry usesa material called polysilicon. What is it and what advantages does it offerover using a metal for the gate?

* Suppose we take a different approach to measuring gm than the oneexplained. Partially differentiating equation (2) w.r.t VGS , we find

gm = kn(VGS − VTN)

Eliminating kn, we get

gm =2ID

(VGS − VTN)

Is this procedure correct? What differences do you expect to find between thevalues of gm calculated by these 2 methods?

Debapratim Ghosh Dept. of EE, IIT Bombay 20/20