EEE 3300 Electronics 1

Lecture Notes
Unit 29
MOSFET DC Circuits
 The following sequence of examples follows
(with a few additional explanations) the
examples in Sedra and Smith.
 A DC analysis of MOSFET transistors is
considerably more involved than that of
BJT transistors, and the results are less
intuitive or “obvious-by-inspection”.
 Students may be comforted to know that, on
the other hand, small-signal MOSFET
amplifier analysis is very similar to that of
BJT amplifiers, and is actually slightly simpler
than that of BJT.

Introductory Comments
 ID versus VDS (NMOS) or ID versus VSD
(PMOS)
Example 1
 Consider the following
circuit employing an
enhancement-type NMOS.
 The transistor’s
parameters are given as
follows:
 Vt = 2V , μnCox =
20[μA/V2], L = 10μm
and W = 400μm.
 We want ID = 0.4mA and
VD = 1V.
 What should be the two
resistors RD and RS ?
Example 1: Where do we start?
 We see that the NMOS
gate is connected to
ground (VG = 0).
 How do we know which
equation to use – Triode
Mode or Saturation Mode?
 For an NMOS we need to
check the condition:
?
VDS  VGS  Vt
Example 1: Is MOSFET in
Saturation Mode?
?
VDS  VGS  Vt
 If TRUE then NMOS is in
Saturation. If FALSE then
NMOS is in Triode Mode.
Example 1: Is MOSFET in
Saturation Mode?
?
VD  VG  Vt
 The value of VD is desired
and the values of VG and
Vt are given.
 VD = 1V, VG = 0, Vt = 2V
 1 > 0 – 2 = -2.
 Indeed the NMOS must be
in Saturation to meet the
VD requirement.
Example 1: MOSFET is in Saturation
Mode  ID Equation of Saturation
Mode

 So let us now use the

Saturation Mode ID
equation:

1 W
I D  ( nCox ) (VGS  Vt ) 2 
2 L
3 1 6 400
 0.4 10   (20 10 )   (VGS  2) 2
2 10
 Everything is given,
except for VGS.
Example 1: ID Equation of Saturation
Mode and Units Consistency

 Note that one needs to be

careful to work with a
consistent system of
units. In the expression
below all currents are
expressed in [mA].

1 W
I D  (  nCox ) (VGS  Vt ) 2 
2 L
1 3 400
 0.4   (20 10 )   (VGS  2) 2
2 10
Example 1: Solving for VGS from
the ID Equation of Saturation Mode
1 W
I D  (  nCox ) (VGS  Vt ) 2 
2 L
1 3 400
 0.4   (20 10 )   (VGS  2) 2

2 10

 The equation for VGS

simplifies into:

1  (VGS  2)  1  VGS  2
2
Example 1: Solving for VGS from the ID
Equation of Saturation Mode 
Multiple Solutions

1  (VGS  2)  1  VGS  2
2

 There are two solutions:

 VGS,1 = 1V or VGS,2 = 3V.
Which one is correct?
 This is where we use the
fact that the transistor is
enhancement-type with Vt
= 2V.
Example 1: Solving for VGS from the ID
Equation of Saturation Mode 
Weeding Out the Wrong Solution

 Obviously solution VGS,1 is

false.

 For VGS = 1V < Vt = 2V

there is no channel and the
NMOS is in Cutoff Mode.
Example 1: Solving for VGS from the ID
Equation of Saturation Mode  The
Correct Solution

 Therefore:
 VGS = 3V
 This is the value of VGS that
is consistent with the
required value of ID = 0.4mA
(according to the Saturation
Mode current equation).
 Now:
 VGS = VG – VS = 0 – VS = 3V
 VS = -3V.
Example 1: Solving for VGS from the ID
Equation of Saturation Mode 
Completing the Design

 Completion of the design

requires now no more than
Ohm’s Law:
VS  VSS  3  (5)
RS    5K
ID 0.4
VDD  VD 5  1
RD    10 K
ID 0.4
 Done.
 The need to solve a quadratic equation is
something that we have not seen in BJT
DC circuits.
 There is always just one good solution.
Screening out the false solution may not
always be obvious.
 If all MOSFET node voltages are
known then determining at which
mode the MOSFET is trivial. However,
in many circuits some of the node
voltages may be unknown.

General Observations About

MOSFET DC Circuits
Triode Mode
confirmed
Example 2: MOSFET in which
Gate is connected to Drain
 The transistor’s
parameters are given as
follows:
 Vt = 2V , μnCox =
20[μA/V2], L = 10μm and
W = 100μm.
 Find R such that ID =
0.4mA.
Example 2: MOSFET in which Gate is
connected to Drain  Saturation by
Structure
 What does the “structural
condition” VDG = 0 mean?
? ? ?
VDS VGS  Vt  VD VG  Vt  0  Vt
 Obviously, for
enhancement-type NMOS
(in which Vt > 0) the
above condition means
that the transistor is
always at Saturation
Mode!
 We may conclude that for enhancement-
type NMOS, a short-circuit between drain
and gate means “Saturation by structure”.

General Observations about

VDG=0
Example 2: MOSFET in Saturation
 ID Equation Solution
1 W
I D  (  nCox ) (VGS  Vt ) 2 
2 L
1 3 100
 0.4   (20 10 )   (VGS  2) 2
2 10
 which gives two solutions
VGS,1 = 0 or VGS,2 = 4V,
and of course the second
solution (VGS = 4V) is the
correct one.
Example 2: MOSFET in Saturation
 ID Equation
 VGS = VG – VS = VG – 0 =
4V  VG = VD = 4V.
 Therefore:

VDD  VD 10  4
R   15K
ID 0.4
 Done.
 Do BJT DC circuits also exhibit similar
structural properties?
 For instance, if VBC = 0, is the BJT always
in active mode?
 Yes, but in such a case the BJT behaves
like a diode.

General Discussion
Example 3
 Here VG = VDD
 It is given that Vt
= 1V and kn’W/L =
1 [mA/V2].
 What should be RD
so that VD = 0.1V?
Example 3 Solution
 Here VDD = VG and
VS = 0. Therefore
VGS = VDD – VS = 5
– 0 = 5V.
 We want VD =
0.1V, or VDS =
0.1V.
 Again, all the
voltages that are
needed to check
the conduction
mode are known.
Example 3: Which mode?
? ?
VDS  VGS  Vt  VD  VG  Vt
?
 0.1  5  1  4V
 So, obviously, the
transistor must be in the
Triode Mode
 We could have suspected
that due to the very small
value of VDS
Example 3: Triode Mode by
Structure?
 Is this “Triode
Mode by
structure”?
 Not at all.
 For instance, had
we required VD =
4.5V, the transistor
would have to be
?
VD  VG  Vt
?
 4.5  5  1  4V
in Saturation
Mode.
Example 3: ID Equation
W 1
I D  k n ' [(VGS  Vt )VDS  VDS ] 
2

L 2
1
1 [(5  1)  0.1   0.01  0.395mA
2

VDD  VD 5  0.1
RD    12.4 K
ID 0.395

Done.
Example 3: Triode Mode  MOSFET
as a Variable Resistance
 So, under these
conditions, what is the
value of the voltage-
controlled resistance rDS ?
VDS 0.1
rDS    253
ID 0.395
 Because VDS = 0.1V is
small this resistance is
almost-linearly dependent
on VGS.
 The first three examples were simple design
examples (funny that we start right away
with design…). All the necessary data to
determine at which mode the transistor
needs to be was essentially given. It just
needed to be slightly organized.
 In the next set of examples we shall study
another type of DC analysis problems, in
which we will have to assume a mode, and
then follow-up on the assumption, and either
justify it or discover a contradiction.

General Comments
Example 4
 Consider the following
standard DC analysis
problem - the circuit is
given, along with the
MOSFET parameters.
 We need to find the
currents and voltages.
 The MOSFET
parameters are Vt = 1V
and kn'W/L =1mA/V2
Example 4 Solution Beginning
 For DC we know that
IG = 0.
 Here the two resistors
RG1 and RG2 form a
perfect voltage divider.
 The value of VG is
therefore known.
Example 4 Solution Challenge
 The value of VG is
therefore known:
RG 2 10
VG  VDD  10   5V
RG1  RG 2 10  10
 The current ID is
unknown, and as a
result we do not yet
know the voltages VD
and VS.
 At which mode is the
MOSFET ?
Example 4: Assume Saturation
Mode
 Of course, we will have
to justify the
assumption.
 It is a good idea to
start with Saturation
assumption simply
because the
equation is way
simpler than that of
the Triode Mode.
1 W 1 W 1 W
ID  kn ' (VGS  Vt ) 2  kn ' (VG  VS  Vt ) 2  k n ' (VG  I D RS  Vt ) 2
2 L 2 L 2 L
Example 4: Expressing MOSFET
Node Voltages in Terms of ID
 VS = RSID = 6∙ID
 In the above the units of
the coefficient 6 are KΩ.
 If [VS] = V then [ID] = mA
 Likewise:
 VD = VDD – RDID = 10-6∙ID
 We’ll need VD later to
verify Saturation Mode
Assumption.
1 W 1 W 1 W
ID  kn ' (VGS  Vt ) 2  kn ' (VG  VS  Vt ) 2  k n ' (VG  I D RS  Vt ) 2
2 L 2 L 2 L
Example 4: Setting the ID
Equation  VS = RSID = 6∙ID
 In the above the units
of the coefficient 6 are
KΩ.
 If [VS] = V then [ID] =
mA
 This is a quadratic
equation in which ID is
the unknown (let us
take the units of ID as
mA):
1
I D  1 (5  I D  6  1)  18I D  25I D  8  0
2 2

2
Example 4: Solving the ID
Equation
 There are two solutions:
ID1 = 0.89mA and ID2 =
0.5mA. Which one is
correct?
 Again, we need to make an
assumption, and follow
through:
 Let us assume that ID =
0.89mA. Then VS = 6ID =
6·0.89 = 5.34V > VG =5V
1
I D  1 (5  I D  6  1)  18I D  25I D  8  0
2 2

2
Example 4: Solving the ID
Equation  There are two solutions:
ID1 = 0.89mA and ID2 =
0.5mA. Which one is
correct?
 Let us assume that ID =
0.89mA. Then VS = 6ID =
6·0.89 = 5.34V > VG =5V
 This is of course a
contradiction, as for a
negative VGS the
transistor is in Cutoff.
1
I D  1 (5  I D  6  1)  18I D  25I D  8  0
2 2

2
Example 4: Solving the ID
Equation  There are two solutions:
ID1 = 0.89mA and ID2 =
0.5mA. Which one is
correct?
 So we retract one step
and take the other
solution for ID: ID =
0.5mA.
 Now: VS = 6ID = 6·0.5 =
3V < VG =5V.
 Furthermore, VGS = 2V >
Vt = 1V.
1
I D  1 (5  I D  6  1)  18I D  25I D  8  0
2 2

2
Example 4: Verifying
Saturation Mode Assumption
 We need now to find VD,
so that we can check if
indeed the transistor is in
Saturation Mode:
 VD = VDD - IDRD = 10 -
6·0.5 = 7V
 Check:
? ? ?
VDS VGS  Vt  VD VG  Vt  7  5  1
 Indeed Saturation Mode -
assumption is justified.
 The MOSFET parameters are Vt = 1V
and kn'W/L =1mA/V2
 The MOSFET parameters are Vt = 1V
and kn'W/L =1mA/V2
Example 5 (Follow-up to
Example 4)  What do we do if
we discover that
the Saturation
Mode assumption
is not satisfied?
 Then we need to
assume Triode Mode,
and start all over
again.
 Let us construct an
example in which
something like that
happens.
Example 5 Solution Beginning
 Let for instance RD=15KΩ
(with everything else
unchanged from the above
example).
 The solution of the new
example follows the tracks
of the previous solution, to
the point in which the
Saturation Mode check
fails:
 VD = VDD - IDRD = 10 -
15·0.5 = 2.5V
Example 5 Solution Beginning

 RD=15KΩ
 VD = VDD - IDRD = 10 -
15·0.5 = 2.5V
 Check:
? ? ?
VDS VGS  Vt  VD VG  Vt  2.5  5  1

 Not a Saturation Mode -

assumption is not
justified.
Example 5 Triode Mode
Assumption
 We now assume Triode
Mode and use the
relationships VD = 10-
15ID and VS=6ID:

W 1
I D  k n ' [(VGS  Vt )VDS  VDS 2 ] 
L 2
 I D  1  [( 5  6 I D  1)  (10  15I D
1
 6 I D )  (10  15I D  6 I D ) 2 ] 
2
 94.5I D 2  65I D  10  0
Example 5 Triode Mode
Solution
94.5I D  65I D  10  0
2

The solutions are ID1 =

0.455mA and ID2 = 0.232mA.

Trying the first one: VD = 10 -

0.455· 15 = 3.175V , VS =
6·0.455 = 2.73V.

Then: VGS = 5 - 2.73 = 2.27V >

Vt =1V. Okay, so far.
Example 5 Triode Mode
Solution VD = 3.175V < VG -Vt = 5 -1 = 4V.

Okay too - it is a Triode Mode.

Check that the other ID solution
fails to justify the Triode Mode
assumption.
 Why are the Gate resistors so large?
 Keep in mind that in a MOSFET amplifier the
signal comes in through the Gate, and the
amplified voltage is taken off the Drain.
 We take these resistors to be as large as we
can to increase the small-signal input
resistance.
 In a BJT amplifier we could not take the Base
resistors to be too large because of the Q-
Point thermal stability condition.

Discussion about Gate

Resistors
 1.1) RD=10kΩ
Assume Saturation vG  2.5V  vGS
Check: vGS  VTH  2.5V  1.8V  yes
1 ' W 
iD  kn  vGS  VTH 2
2 L
(100e  6)102.5V  1.82  0.254mA
1
iD 
2
VD  VDD  I D RD  5V  (0.254mA)(10k)  2.55V
Check: vD  vG  VTH  2.55V  2.5V  1.8V  0.7  yes
Saturation mode confirmed

1.2) RD=20kΩ
Assume Saturation
VD  VDD  I D RD  5V  (0.254mA)( 20k)  0.1V
Check: vD  vG  VTH  0.1V  2.5V  1.8V  0.7  no
Triode mode confirmed
W 1
I D  kn ' [(VGS  Vt )VDS  VDS 2 ]  (100e  6)(10)(( 2.5  1.8) VDS  0.5  (VDS ) 2 )
L 2
Example 6: PMOS Circuit
 An enhancement-type
PMOS has the following
parameters:
 Vt = -1V
 Kp'W/L = 1mA/V2.
 We need to choose the
resistors RG1, RG2 and RD, so
that the PMOS is in
Saturation Mode, and has ID
= 0.5mA and VD = 3V.
Example 6 Solution
 Since we need the
transistor to be in
Saturation Mode, let us use
the Saturation Mode current
equation:
1 W
I D  k p ' (VGS  Vt ) 
2

2 L
1
 0.5  1 (VGS  (1)) 
2

2
 (VGS  1) 2  1
Example 6 Solution
(VGS  1)  1
2

 Solutions are: VGS,1 = 0 and

VGS,2 = -2V.
 The first solution (VGS = 0 )
is false because it is larger
(more positive) than Vt = -
1V, and in such a condition
there would be no channel.
 Therefore VGS = -2V is the
correct solution.
Example 6 Solution
 Because VS = 5V, we have
that VG = VS + VGS = 5 - 2
= 3V.
 Now:
VG RG 2 5 RG1 RG1 2
   1  
VDD RG1  RG 2 3 RG 2 RG 2 3

 So we can simply take RG1

= 2MΩ and RG2 = 3MΩ.
Example 6 Solution
 Using the desired VD value,
we can now find the resistor
RD: VD 3
RD    6 K
I D 0.5

 Let us not forget to check if

the MOSFET is indeed in
? Saturation? Mode: ?
VDS VGS  Vt  VD VG  Vt  3  3  (1)  4V

 Indeed Saturation Mode.

Example 6 Discussion
 What if the problem had required
Saturation Mode with ID = 0.5mA
and VD = 4.5V?
 Design would then become
impossible, indicating that the
specifications (a) Saturation Mode
and (b) ID = 0.5mA and VD = 4.5V,
are incompatible and thus
contradictory.
 Situations of inconsistent
specifications may happen in design
problems. The designer then needs
to decide how to modify the
specifications.