Design of CMOS Fully Differential Operational Transconductance Amplifier
Design of Operational Transconductance
Transcript of Design of Operational Transconductance
Design of Operational Transconductance
Amplifiers for Filtering Applications
A Project Report Submitted
in Partial Fulfillment of the Requirements
for the Degree of
Bachelor of Technology
in
Electronics and Communication Engineering
by
Monalisha Saikia (Gau-C-15/L-286)
Sonu Kumar (Gau-C-15/012)
Amit Anand (Gau-C-15/013)
Jyotishna Kurmi (Gau-C-15/017)
Binita Brahma (Gau-C-15/032)
Under the Guidance of
Mr. Antaryami Panigrahi
Department of Electronics and Communication Engineering
CENTRAL INSTITUTE OF TECHNOLOGY-KOKRAJHAR,
ASSAM, INDIA
May-2019
ABSTRACT
In the wireless transreceiver, the analog baseband section is responsible for adjacent
channel selectivity, anti-aliasing and dynamic range maximization. A channel filter
with high linearity, tunable Band Width, low noise etc. is important for the performance
of the whole RF front-end.Frequency and gain programmability, as well as power
efficiency and silicon area are key aspects for the baseband filter. In this thesis, a detail
comparison of various OTAs with respective small signal models are presented and
simulated in 250nm CMOS technology. A 3rd order passive Buttorworth filter has been
implemented using OpAmp-RC and Gm-C integrators using Ladder synthesis and
biquad respectively. 2nd Order biquads are cascaded to the 1st order sections to make
3rd filter. 3rd order filter using Opamp-RC integrators are simulated in Multisim
Environment, transistor level simulation based on Gm-C biquad has been implemented
in Tanner v.15 Software. Opamp-RC based filter offers a band width of 425 kHz, pass
band gain of 0 dB, where as Gm-C integrator based filter offers 85 MHz, pass band
gain of 0 dB . Over-all power dissipation of the 3rd order Gm-C filter is 4.3mW with
1.8V DC Supply.
ACKNOWLEDGEMENT
We would like to express our profound sense of gratitude to the Head, department of
Electronics and Communication Engineering, Central Institute of Technology,
Kokrajhar for the providing the access to the facilities, tools and all the resources for
successful completion of the project work. We wish to convey our gratitude to all the
faculties of department of Electronics and Communication Engineering who have
enlightened us during entire career of our technical studies. The cooperation received
from the technical staff of Electronics and Communication Engineering is thankfully
acknowledged. Lastly, we would like to thanks the researchers of various articles and
books that we have referred for making our knowledge finer and stronger.
Monalisha Saikia (Gau-C- 15/L-286)
Sonu Kumar (Gau-C-15/012)
Amit Anand (Gau-C-15/013)
Jyotishna Kurmi (Gau-C-15/017)
Binita Brahma (Gau-C-15/032
LIST OF FIGURES
1. Figure 1 : Interferer levels in received GSM Signal. Pg.1
2. Figure 2: OTA symbol representation and equivalent
model.
Pg.4
3. Figure 3: Different types of Transconductors. Pg.5
4. Figure 4: Small signal model of different OTAs for
calculating Gm (left), Rout (Right): (a), (b), (c).
Pg.6
5. Figure 5: Miller OTA and its small signal equivalent. Pg.8
6. Figure 6: Telescopic Cascode OTA and its small signal
equivalent.
Pg.8
7. Figure 7: Folded Cascode OTA and its small signal
equivalent.
Pg.9
8. Figure 8: Folded Cascode OTA and its small signal
equivalent.
Pg.10
9. Figure 9: Differential OTA circuits in CMOS : (a) Single
ended output, (b)Feed forward OTA, (c) Fully differential
OTA
Pg.11
10. Figure 10: Telescopic Cascode OTA Schematic in S-Edit. Pg.15
11. Figure 11: AC Response of Telescopic Cascode in W-Edit. Pg.16
12. Figure 12: Folded Cascode OTA in S-Edit. Pg.16
13. Figure 13: AC Response of Telescopic Cascode in W-Edit. Pg.17
14. Figure 14: Schematic of Miller OTA in S-Edit Pg.17
15. Figure 15: AC Response of Miller Cascode in W-Edit Pg.18
16. Figure 16: Folded cascode OTA in s edit Pg.20
17. Figure 17: AC Figure 16: Response of folded cascode OTA in W-
edit Pg.21
18. Figure 18: Butterworth 3rd order LPF and its magnitude
response of various order Pg.22
19. Figure 19: Diagram for writing equations Pg.23
20. Figure 20: Gm-c implementation of 3rd order butter LPF Pg.24
21. Figure 21. implementing inductor using gyrator
Pg.25
22. Figure 22: The complete filter (3rd order Butterworth using
gyrator) Pg.25
23. Figure 23: signal flow diagram of a practical biquad (LPF) Pg.26
24. Figure 24: signal flow graph of the biquad using Gm-c
integrator for LPF Pg.27
25. Figure 25: 3rd order Butterworth LPF (biquad) Pg.28
26. Figure 26: circuit diagram of biquad using op-amp in Multisim Pg.29
27. Figure 27: 3rd order Butterworth LPF in s edit Pg.31
28. Figure 28: Ac response of 3rd order Butterworth LPF in w-edit Pg.32
29. Figure 29: Ac response of 3rd order Butterworth LPF due to
transconductance variations in w-edit Pg.32
30. Figure 29: Ac sweep of biquad of op-amp and Active RC
integrator Pg.33
31. Figure 30: DC sweep analysis in s edit Pg.34
32. Figure 31: 𝒊𝒐𝒖𝒕 𝒗𝒔 𝑽𝒊𝒅 Pg.34
33. Figure 32: Test bench for noise analysis folded cascode OTA Pg.35
34. Figure 33: Noise simulation folded cascode OTA Pg.35
LIST OF TABLES
Table
No
Description of Table Pg.No
1 : Summary of Expressions of gain different transconductors
(single ended)
7
2: Summary of the different differential OTA structures.
10
3: Comparison of Simulation Results of OTAs
19
4: Simulation Results of Folded Cascode (Differential) in Tanner 21
5: Equations describing the Passive RLC Network 23
6 Comparison of the Equations of 3rd Order Passive and Active
Filters
30
7 All Simulation Results of 3rd Order Butterworth Low Pass Filter. 36
TABLE OF CONTENTS
I. Abstract
II. Acknowledgement
III. List of Figures
IV. List of Tables
1 Motivation & Introduction ......................................................................................... 1
1.1 Filters in CMOS
Technologies………………………………………………………. 2
1.2 Scope of the
Report:…………………………………………………………………. 3
1.3 Organisation of the
Thesis:………………………………………………………….. 3
2 OTA and its architectures ......................................................................................... 4
2.1 Conventional Architectures of
OTA………………………………………………... 5
2.1.1 Small signal parameters ............................................................................... 7
2.2 Classification of different types of OTA 7
2.3 Nonideal perfomances of OTA 12
2.3.1 Harmonic Distortion ................................................................................... 12
2.3.2 Intermodulation Distortion ......................................................................... 12
2.3.2 Distortion in MOS Differential OTA .......................................................... 13
3 Comparison of various architecture of OTA ........................................................... 15
3.1 Telescopic Cascode Simulation and Results .............................................. 15
3.2 Folded Cascode Simulations and Results ................................................... 16
3.3 Miller OTA Simulation and Results ........................................................... 17
4 Folded Cascode OTA........................................................................................... .... 20
4.1 Folded Cascode Simulations and Results ................................................... 21
5 Filter........................................................................................... .............................. 22
5.1 Butterworth Filter ....................................................................................... 22
5.2 Ladder Synthesis of Active filter ................................................................ 23
5.3 Gyrator based Synthesis for Gm-C filter .................................................... 24
5.4 Synthesis using Biquad ............................................................................... 26
5.5 Biquad Implemetation using Active-RC and Gm-C integrators ................. 29
6 Simulations and Results........................................................................................... 31
6.1 AC Response of 3rd Order Butterworth Filter ............................................. 31
6.2 AC Simulation of Op-Amp based Biquad .................................................. 32
6.3 Transconductance simulations .................................................................... 33
6.4 Noise Analysis of the Biquad ..................................................................... 34
Conclusion .................................................................................................................. 36
References ................................................................................................................... 37
Appendix ..................................................................................................................... 38
CHAPTER 1 Introduction
In Modern day receivers, the received signal may contain other adjacent channels with
significantly higher power level than the desired channel. A typical GSM scenario is
shown below, where the specification for the possible power levels of the neighboring
channels as a function of frequency offset (In MHz) [1].
If the adjacent channel which is 600kHz offset from the desired channel must be
attenuated by 10 dB below the desired channel to detect the information, the adjacent
channel must be attenuated by 65 dB. Since the desired channel is -98 dBm, the power
of adjacent channel, -43 dBm, should be attenuated to -108 dBm. In the case of GSM,
the desired channel has a bandwidth of only 200KHz located around 900 MHz. For the
detection of the transmitted information, the required filter must have a quality factor
Q as; 𝑄 =𝑓0
𝐵𝑊=
900𝑀𝐻𝑧
200𝑘𝐻𝑧= 4500. [1] which is impossible to realize with modern
technology. So frequency translation (down conversion) is necessary for easier
detection with feasible filters. In modern receivers, detection of information is usually
done in low IF or zero IF. The analog filters in low IF or zero IF separate the desired
channel from undesired ones and interferers, which may be orders of larger than the
Figure 1 Interferer levels in received GSM Signal.
desired signal. Accomplishing the same job without analog filters would require analog
to digital converters (ADCs) with much larger number of bits to properly digitize and
process large interferers. Therefore, analog baseband filters are needed for the proper
detection of the desired signal in radio receivers. In the design of the high-performance
electronic circuits, the use of analog filters is unavoidable. The receiver’s ability to
suppress adjacent channel interference and in-band blocking signals is directly
dependent on the channel selection filtering. As pointed out in the previous section, the
received signal contains not only the desired channel but a multitude of neighboring
channels or other interferers which must be attenuated before the detection can be done.
1.1 Filters in CMOS Technologies:
In CMOS technologies, Active RC, MOSFET-C and Gm-C filters are the most popular
baseband filters for wireless communication.
Active RC filters usually apply lossy and lossless integrators to attenuate
interferers and are still widely used in low frequency application.
MOSFET-C filters have similar architecture as active RC filters but replace
passive resistors with a triode region transistor.
Gm-C filters use transconductors and capacitors to attenuate interferers.
Because of their open loop nature, Gm-C filters have been used for high
frequency application. Each filter mentioned has its own advantages and
disadvantages in terms of speed, linearity, and tunability.
1.2 Challenges in Implementations of Filters in CMOS Technology in CMOS
Technologies:
1. For the realization of fully integrated analog filters, there are many things
which must be considered. First, the size of capacitor and resistor should be
limited. Since integrated capacitors such as poly insulator poly capacitors
or metal insulator metal capacitors occupy a large area, the size of capacitor
must be limited.
2. Frequency response of the filter should be stable, since the quality factor of
the filter is highly affected by small errors in the phase and component
variation. This means that every element in the filter should have accurate
and stable values in the presence of fabrication tolerances and temperature
variations, but this is almost impossible in an integrated circuit.
3. Analog continuous time filters should be able to handle large signal. For a
wide dynamic range required in radio receiver, large signal swings are
essential. However, these are increasingly difficult to achieve as power
supply voltages are reduced and bandwidth is increased.
4. CMOS filters are often located with data converters and digital circuits on
the same chip. In this case, noise due to clock or switching is transferred to
the analog filters through substrate or power supply line. This noise highly
degrades dynamic range and signal to noise ratio of the analog filter.
Therefore, special design and layout techniques are required to minimize
the effect of the noise.
1.4 Scope of the Report:
In this report we discuss the various topologies of CMOS OTAs for the application in
filters. Individual performance of the OTA are analysed using their small signal
parameters i.e. 𝐺𝑚 and 𝑅𝑂𝑈𝑇, Band width , Linearity. Each topology is designed in
250nm CMOS technology and their simulation results are presented.
1.5 Organization of this Thesis:
Chapter 2: Architectures of OTA in CMOS technology.
Chapter 3: Simulation and discussion of different OTAs in 250nm CMOS
Technology.
Chapter 4: Summary and Conclusion: summarizes the report and gives
concluding comments on the presented concept.
Chapter 5: Synthesis of Active filters from the Passive RLC section
Chapter 6: Simulation of 3rd order Gm-c Filter in Tanner EDA.
CHAPTER 2 An OTA is a voltage controlled current source (VCCS) type of amplifier where output
current is proportional to the input voltage. The basic difference of an OTA from an
Op-Amp is that output node of the OTA provide very high output resistance, where as
Op-Amp gives low output impedance. OTAs are used in open loop manner, where as
Op-Amps are mostly used in closed loop [2]. More comparative features are listed in
the Appendix.
Some of the key features of the OTA are their fast speed in comparison with the
conventional low output impedance op-amp and their transconductance tunability
(Since 𝐺𝑚 ∝ √2. 𝐼𝐵𝐼𝐴𝑆). For a filter to be designed for multi-standard wireless
applications tunability of the filter is highly desired which is achieved easily through
the tunable 𝐺𝑚.The Internal nodes of an OTA i.e. nodes except input and output nodes
offer low impedance, so it can operate upto very high frequencies (upto GHz), however
these low impedance and parasitic capacitance cause a nonzero phase shift known as
excess phase shift (𝜑).Out of all these merits, OTAs offer limited linearity performance
compared to Op-Amps as is a since the transconductors operate in an open loop.
The main characteristics of practical OTA are:-
1. Limited linear input voltage range.
2. High bandwidth.
3. Finite signal to noise ratio (S/N) ratio.
4. Very high output impedance.
V2
G m
V1 i abc
i o
GND1
V1
V2 G m ( V 1 - V2)
i o
Figure 2 OTA symbol representation and equivalent model
The transconductance bias dependence allows several decades of tuning for
transconductance with MOS transistor operating in weak inversion and about two
octaves for transistor operating in strong inversion. For open loop application OTAs
are used as multipliers, nonlinear circuits and for closed loop applications discrete-time
based (switched-capacitors) circuits [2].
2.1 Conventional Architectures
An ideal transconductance amplifier is an infinite bandwidth voltage controlled current
source , with an infinite input and output impedance [2], however the CMOS OTAs
provide limited output impedance due finite value of 𝑟𝑑𝑠 = (1𝜆. 𝐼𝐷𝑆
⁄ ) 𝑜𝑟 [𝑔𝑑𝑠1 =
𝜆. 𝐼𝐷𝑆] and limited value of transconductance 𝐺𝑚 limited by power dissipation (𝐼𝐵𝐼𝐴𝑆)
and size of the chip (W/L of all MOSFETs). We will now discuss some of the basic
OTAs implemented in CMOS and compare their performances as in the figures shown
in below (Fig. 3).
ioIb
M1vin
io
M1
M2
Ib
vb
vin
io
M1
M2
Ib
vb
vin
A
vo1
(a) (b) (c)
ioM1
M2
Ib
Ib2
vin vb
1:1io
Ib
(d) (e)
Figure 3 Different types of Transconductors [2]: (a) Single transistor (Inverting ), (b) Cascoded, (c) Gm-
boosted Cascode, (d) Folded Cascode, (e) Current Mirror Cascode.
To compare the performances of the different OTAs shown above, we take the help of
the small signal models to calculate the short circuit transconductance parameter and
output resistance. In Figure 4 (a), small signal model for single transistor
transconductor in Figure 3. (a) is shown, where the left part of the figure shows the
method of calculating the short-circuit transconductance and right part shows the
method of deriving the output resistance [4].
(a)
(b)
Figure 4 Small signal model of different OTAs for calculating Gm (left), Rout (Right): (a), (b), (c)
(c)
Small signal parameters:-
To find the transconductance , we evaluate the ratio of small signal output current to
that of the small signal input voltage applied and can be expressed as;
𝐺𝑚= 𝑖𝑜𝑢𝑡
𝑣𝑖𝑛|
𝑂𝑢𝑡𝑝𝑢𝑡 𝑛𝑜𝑑𝑒 𝑖𝑠 𝑠ℎ𝑜𝑟𝑡𝑒𝑑 𝑡𝑜 𝑔𝑟𝑜𝑢𝑛𝑑 (1)
Similarly the output resistance (𝑅𝑂𝑈𝑇) is derived by connecting a voltage source 𝑉𝑋 at
the output node by deactivating all the independent small signal sources and measuring
the current 𝐼𝑋 flowing into the output node.
𝑅𝑂𝑈𝑇 =𝑉𝑋
𝐼𝑋|
𝑖𝑛𝑑𝑒𝑝𝑒𝑛𝑑𝑒𝑛𝑡 𝑠𝑚𝑎𝑙𝑙 𝑠𝑖𝑔𝑛𝑎𝑙 𝑠𝑜𝑢𝑟𝑐𝑒𝑠 𝑑𝑒𝑎𝑐𝑡𝑖𝑣𝑎𝑡𝑒𝑑 (2)
Following these methods and taking the help of respective figure shown in Figure 4,
we have summarised the expressions for all the basic architectures.
Table 1 Summary of Expressions of Gain of Transconductors
Structure/Fig. 𝑹𝒐𝒖𝒕 Gm Gain(Gm × 𝑹𝒐𝒖𝒕)
Simple/3(a) 1/gds1 𝑔𝑚1 𝑔𝑚1
𝑔𝑑𝑠1
Cascode/3(b) 𝑔𝑚2
𝑔𝑑𝑠1𝑔𝑑𝑠2 𝑔𝑚1
𝑔𝑚1𝑔𝑚2
𝑔𝑑𝑠1𝑔𝑑𝑠2
Enhanced/3(c) 𝐴𝑔𝑚2
𝑔𝑑𝑠1𝑔𝑑𝑠2
𝑔𝑚1 (1+A) 𝐴 𝑔𝑚2𝑔𝑚1 (1 + A)
𝑔𝑑𝑠1𝑔𝑑𝑠2
Folded/3(d) 𝑔𝑚2
𝑔𝑑𝑠1𝑔𝑑𝑠2 𝑔𝑚1 𝑔𝑚2𝑔𝑚1
𝑔𝑑𝑠1𝑔𝑑𝑠2
In fig 1a, the single input real transconductor is a MOS driver transistor M1 operating
in saturation region and it has a drawback of low output impedance .
2.2 Classification of different types of OTAs:
Group of transconductance amplifier is discussed below.
Miller OTA:
The circuit diagram is shown in Fig. 6 in which NMOS M1 & M2 are acting as a input
of the differential pair and the PMOS transistors M2 and M3 forming a current mirror
circuit acting as active load. The main reason behind this current mirror load instead of
resistor is to reduce chip area. The tail current is used to control the biasing.
Telescopic OTA:
Telescopic OTA shown in Figure 7 comprises of 8 MOSFETS and a tail current source,
out of which M1,M2,M3,M4 are pMOS and M5,M6,M7,M8 are nMOS. It consists of
M7 and M8 as the driver transistors which carry the differential input signal. The
Telescopic OTA is biased with 𝐼𝐵𝐼𝐴𝑆 current.Vbias2 and Vbias3 are biasing voltages
for the pMOS transistors and Vbias4 is used for nMOS pair M5 and M6.
Folded Cascode OTA (Figure 8):
It consists of two differential input pairs with the same type of transistors, M1-M2. The
amplification can be obtained with signal applied to M1 and M2 of the OTA ensuring
M1
M2
GND2
Vin+
Vout
Iout
Vin-
Itail
gm1vgs1 r02
GND1
1/gm3
GND2
+
-
Vin
vgs2
r01
vx
Figure 6 Miller OTA and its small signal equivalent.
Vout-
Vbias3
GND4
Vbias2
Vbias4
Vbias1
Vout+
Vin+Vin-
Telescopic cascode
M5 M6
M7
M4
M1
M3
M2
M8
M9 gm7vgs7 r07
GND1
gm5vgs5 ro5
GND2
+
-
Vin
vgs1
Isc
X
y
Figure 7 Telescopic Cascode OTA and its small signal equivalent to calculate the Gm.
all transistors in the saturation region . Transistors MP3 and MP4 work as Common
Gate amplifiers which are of PMOS type and folding is achieved with the current
source implemented using MP1 and MP2. It is cascade of a differential
transconductance stage which is designed by connecting the output of a Common
Source stage to a Common Gate stage.
Since the folded-cascode OTA presents a single parasitic pole and relatively large DC
gain, it is commonly used for high-frequency applications. For high speed applications,
fast settling OTAs are required, which demand both high UGB and Power.
The small signal model of the left half of the circuit is shown below in Figure 9.
MP1 MP2
MP3 MP4
M3 M4
M5 M6
M1 M2
M0
GND1GND2
Vbias1
Vbias2
Vbias3
Vbias4
Vout-
Vin- Vout+Vin+
Vb
P-1
Figure 8 Folded Cascode OTA
Table 2 Summary of Expressions for the differential OTA structures
Topology Output resistance Gm Voltage Gain
Miller (𝑔𝑚2𝑟02𝑟01)||
1
𝑔𝑚3
𝑔𝑚2 𝑔𝑚2(𝑔𝑚2𝑟02𝑟01)||
1
𝑔𝑚3
Telescopic
cascode
(𝑔𝑚5𝑟05𝑟07)||
(𝑔𝑚3𝑟03𝑟01)
𝑔𝑚7 𝑔𝑚7(𝑔𝑚5𝑟05𝑟07)||
(𝑔𝑚3𝑟03𝑟01)
Folded
cascode
(𝑔𝑚3𝑟03𝑟05)||
(𝑔𝑚𝑝3𝑟0𝑝3(𝑟0𝑝1 ||𝑟01))
𝑔𝑚2 𝑔𝑚2(𝑔𝑚3𝑟03𝑟05)||
(𝑔𝑚𝑝3𝑟0𝑝3(𝑟0𝑝1 ||𝑟01))
gm2vgsro2 ro p1
GND1
gmp3vgs
ro p3
Isc,out
GND2
+
Vgs
-
+
Vgs
-
ro1
GND3
2
r0p3
R0p1||r02
GND6
gm3Vgs3 r03
GND7
r05
vx
GND8
P
Q
vs1
vs2
Ix
gmp3Vgsp3
Figure 9 Folded Cascode OTA and its small signal equivalent to find Gm and Rout.
(b)
(c)
a. Simple differential OTAs
b. Balanced OTAs
c. Conventional fully differential OTA
M1
M2
GND2
Vin+
Vout
Iout
Vin-
Itail
(a) (b)
(c)
Figure 9 Differential OTA circuits in CMOS : (a) Single ended output, (b)Feed forward OTA, (c) Fully
differential OTA
2.3 Non-Ideal Performances of OTAs:
2.3.1 Harmonic Distortion (HD)
If the fully differential transconductor with inputs; 𝑉2 = 𝑉1+ and 𝑉1 = 𝑉1−presents
nonlinear characteristics, the output current, , can be expressed by the following
expansions:
𝑖𝑜− = 𝑖𝑜1= 𝐼𝐵1
+ 𝑎1(𝑉1 − 𝑉2) + 𝑎2(𝑉1 − 𝑉2)2 + 𝑎3(𝑉1 − 𝑉2)3 + … … … … … .. (3)
𝑖𝑂+ = 𝑖𝑜2= 𝐼𝐵1
+ 𝑎1(𝑉2 − 𝑉1) + 𝑎2(𝑉2 − 𝑉1)2 + 𝑎3(𝑉2 − 𝑉1)3 + … … … … … .. (4)
Where 𝐼𝐵1 is the amplifier bias current. It is evident from these expressions that the
even harmonic distortion components appear at the output with the same amplitude and
same phase, and they ideally cancel each other when the differential output current is
processed. The main advantage of fully differential topology are due to
symmetry,making the structure less sensitive to common mode signals.
The 𝑛𝑡ℎ harmonic distortion (𝐻𝐷𝑛)component in the output current can be defined as
the ratio of the component of frequency n𝜔 to the one at the fundamental 𝜔.
𝑯𝑫𝟐 =𝟏
𝟐
𝒂𝟐
𝒂𝟏 𝑼 (5)
𝑯𝑫𝟑 =𝟏
𝟒
𝒂𝟑
𝒂𝟏𝑼𝟐 (6)
It is important to note that HD is proportional to U, where U is the peak or maximum
value of the signal applied and𝐻𝐷3 to 𝑈2. These relationships hold true for all values
of U, which are not too large and this is the region where the so-called low-distortion
conditions are valid. For even larger values of U, the values of 𝐻𝐷2and 𝐻𝐷3flatten off
with increasing U.
2.3.2 Inter-Modulation Distortion (IMD):
Application of the sum of two cosine (sine) waveforms of frequencies 𝜔1 and 𝜔2which
are close enough in the frequency spectrum and both of same amplitude U at the input
gives rise to output signal components at all combinations of 𝜔1, 𝜔2 and their
multiples.
Second-order intermodulation distortion (𝐼𝑀2) is defined by the ratio of the component
at frequency 𝜔1 ± 𝜔2 to the one at 𝜔1 or 𝜔2. Under low-distortion conditions:
𝑰𝑴𝟐 =𝒂𝟐
𝒂𝟏 𝑼 (7)
Third-order intermodulation distortion (𝐼𝑀3) can be detected at the frequencies
2𝜔1 ± 𝜔2 and2𝜔2 ± 𝜔1. It is given by the ratio of the component at frequency
2𝜔3_𝜔1 (or one of the other three frequencies), which is 3𝑎3𝑈3, to the fundamental,
which is 𝑎1U, as given by
𝑰𝑴𝟑 =𝟑
𝟒.
𝒂𝟑
𝒂𝟏.𝑼𝟐 (8)
Comparison of the four equations above shows that:
𝑰𝑴𝟐 = 𝟐𝑯𝑫𝟐& 𝑰𝑴𝟑 = 𝟑𝑯𝑫𝟑 (9)
For a MOS transistor, the transfer characteristic 𝐼𝐷𝑆~𝑉𝐺𝑆 is quadratic and not
exponential unlike BJT. Less distortion is thus expected. The drain current 𝐼𝐷𝑆 and gate
to source voltage 𝑉𝐺𝑆 of a MOS transistor is written as;
𝐼𝐷𝑆 = 𝛽(𝑉𝐺𝑆 − 𝑉𝑇𝐻)2 (10)
in which 𝛽 is the process parameter, which includes thesize(𝑊
𝐿), and µ𝑛. 𝐶𝑜𝑥 and
𝑉𝑇𝐻is the threshold voltage.The transistor is biased at a specific dc value of 𝑉𝐺𝑆 .
A small variation of this voltage causes a small variation in drain current. Therefore,
the small signal current 𝑖𝑝 can be written in terms of its Taylor's Series expansion around
its DC current 𝐼𝐷𝑆 as;
𝑖𝑝 =𝐼𝑑𝑠𝑝
𝐼𝐷𝑆=
2𝑉𝐺𝑆𝑃
𝑉𝐺𝑆−𝑉𝑇+
1
4(
2𝑉𝐺𝑆𝑃
𝑉𝐺𝑆−𝑉𝑇)
2 (11)
Comparing this equation with equation(5),we get
𝑰𝑴𝟐 = 𝟐𝑯𝑫𝟐 =𝟏
𝟐.
𝑽𝑮𝑺𝑷
𝑽𝑮𝑺−𝑽𝑻=
𝑰𝑷
𝟒and, IM3=0 (12)
2.3.3 Distortion in MOS Differential OTA:
The transfer characteristic of a differential pair with MOST is symmetrical around
the origin. No second-order distortion can thus occur. Since a single MOST amplifier
does not generate third-order distortion, it will be interesting to examine what distortion
performance can be obtained with a MOST differential amplifier. The differential
output current 𝑖0𝑑is again twice the ac current in each transistor. Thus ,the relative
current swing is given by;
𝑖𝑝 =𝐼𝑜𝑑𝑝
𝐼𝐵=
𝑉𝑖𝑑𝑝
𝑉𝐺𝑆−𝑉𝑇−
1
8(
𝑉𝑖𝑑𝑝
𝑉𝐺𝑆−𝑉𝑇)
3
(13)
Comparing this with equation(2),we getthe following expression;
𝑰𝑴𝟑 = 𝟑𝑯𝑫𝟑 =𝟑
𝟑𝟐(
𝑽𝑰𝒅𝒑
𝑽𝒈𝒔𝟏−𝑽𝑻)
𝟐
=𝟑
𝟑𝟐𝒊𝒑
𝟐 (14)
All the above equations give the relation between the bias parameters and the signal
amplitude levels. These relations can be applied to the individual amplifiers shown
above to get the Harmonic and Intermodulation distortion terms. In this report we focus
on the small signal performance of the above topologies of differential input OTAs in
CMOS, so small signal equivalent of the respective OTAs are shown below.
CHAPTER 3
Comparison of various Architectures of OTA:
In order to understand the theoretical expressions and have a fair comparison between
the different OTA topologies simulation are performed using the Tanner S-EDIT
simulator and BSIM3v3 transistor model(level 49) for the AMS 0.25μm CMOS
technology.
3.1 Telescopic cascode Simulation:
Figure 10 Telescopic Cascode OTA Schematic in S-Edit
AC ANALYSIS:
3.2 Folded cascode:
3.4 AC ANALYSIS
Figure 11 AC Response of the TelescopicCascoded OTA
Figure 12 Folded Cascode OTA in S-Edit
3.4 Miller’s OTA:
AC Analysis of Miller OTA:
Figure 13 AC Response of Folded Cascode OTA in w-edit.
Figure 14 Schematic of Miller OTA in S-Edit
Discussions:
Figure 11, 13 and 14 shows the frequency response of 3 different topologies of the
OTA The DC gain of folded cascode is (shown in Fig 13) 33.4 dB with unity Gain
Bandwidth of 255.99 MHz for a load capacitor value of 1pF, -3dB Bandwidth of
360 MHz while achieving the phase margin of 82º. The DC gain of telescopic cascode
shown in Figure 11 15 dB with unity Gain Bandwidth of 159 MHz for a load capacitor
value of 10 pF, -3dB Bandwidth of 245 MHz while achieving the phase margin of
95º.By increasing the value of (W/L) ,we can further increase the gain of these OTA
with the low power consumption area will be increased and increased parasitic
capacitance will also cause the fall in BW.
From these simulations we can observe that the telescopic cascode OTA provide high
output resistance and high band width, but it requires the double the transistors to that
of a Miller OTA. So area of the OTA will increase and also voltage swing will be less.
Folded cascode can operate with higher swing and same gain but Power dissipation
increases due to the folding node. Miller OTA gives lesser gain , but has a simpler
design. The linearity aspect of these OTA would be studied in future.
Figure 15 AC response of miller OTA in W-edit
Table 3 Comparison of Simulation Results
Parameters Miller OTAs Telescopic OTAs Folded OTAs
Unity gain
frequency (MHz)
401.13 159 255.99
Phase margin (º) 71 95 82
Gain (dBV) 9.4 14.14 33.4
Power (mW) 23 8 0.739
CHAPTER 4
Folded Cascode OTA
The folded Cascode OTA is in a way a compromise between the miller OTA and the
telescopic Cascode OTA. It permits low supply voltage, still having a rather high output
voltage swing and the input and output common mode levels can be designed to be
equal. Its gain is higher than for the miller OTA and its speed is lower than for the
telescopic Cascode OTA, which makes it a good compromise between these two
OTAs.
The applied voltage is 1.8 v (vdd) which having power consumption of 739.08𝜇𝑤. The
input voltage which have applied in folded cascode OTA is vdd/2 ie; 0.9 v.Load
capacitor used is 1pf.
Figure 16 Folded Cascode OTA in S-Edit
Table 4Simulation Results of Folded Cascode (Differential) in Tanner
Parameters Descriptions
Technology 250nm CMOS
Unity gain frequency 255.99 MHZ
Phase margin 𝟏𝟖𝟎° − 𝟗𝟖° = 𝟖𝟐°
-3db frequency 10 MHZ
Power(mW) 739.08µW
Figure 17 AC response of Folded cascode OTA in W-edit
CHAPTER 5
5.1 Butterworth filter
Butterworth filters [5] are suitable for applications in which ripples in the pass and stop
band are intolerable. For this reason, the Butterworth filter is also called a “maximally
- flat amplitude” filter. It exhibits a nearly flat passband with no ripple. The roll-off is
smooth and monotonic, with a low-pass or high pass roll-off rate of ±20 dB/decade (±6
dB/octave) for every pole. Thus, a 5th-order Butterworth low-pass filter would have an
attenuation rate of 100 dB for every factor of ten increase in frequency beyond the
cutoff frequency. It also has a reasonably good phase response.
5.2 Ladder synthesis [6]:-
Using MOSFET-C integrators, the ladder simulation method for the MOSFET-C
approach is entirely similar to the active-RC. Ladder synthesis is realized by the LC
ladder structure with known component values. The LC active simulation proceeds by
deriving the signal flow graph equation or by writing the branch currents and node
equations of the ladder along with the respective V-I relationships. The scaling factor
in this case is the design Transconductance 𝑔𝑚. The source and the load resistors of the
LC ladder network have the value of 1
𝑔𝑚. The equations are implemented with the
integrators in fully differential form and all the integrators capacitors grounded. All the
Figure 18 -Butterworth 3rd Order Low Pass Filter and its Magnitude Response for various Orders
signals in the MOSFET-C circuits are voltages, which in turn produce current summed
at the op-amp inputs, and that the integration constant must be time rather than
capacitance. This is achieved scaling the equation by a resistor R. An often cited
advantage of the MOSFET-C techniques is the reduced sensitivity to parasitic
capacitor. MOSFET-C circuits promise to become increasingly attractive in the future.
Table5 Equations describing the Passive RLC Network
Parameter Scaled equation Remarks
𝐼𝐿1 =𝑉𝑆 − 𝑉2
𝑠𝐿1 𝑉𝐿1 =
𝐼𝐿1
𝑔𝑚=
𝑔𝑚. 𝑉𝑆 − 𝑔𝑚. 𝑉2
𝑠𝐿1𝑔2𝑚
𝑔𝑚𝑣 form is easy for
transconductor implementation.
𝐼𝐿3
=𝑉2 − 𝑉𝑜𝑢𝑡
𝑠𝐿3
𝑉𝐿3 =𝐼𝐿2
𝑔𝑚=
𝑔𝑚. 𝑉2 − 𝑔𝑚. 𝑉𝑜𝑢𝑡
𝑠𝐿3𝑔2𝑚
Current integrated to voltage
through integrator (𝑔𝑚 − 𝑐).
𝑉2 =𝐼𝐿1 − 𝐼𝐿3
𝑠𝐶2
𝑉2 =𝑔𝑚 (
𝐼𝐿1
𝑔𝑚−
𝐼𝐿3
𝑔𝑚)
𝑠𝐶2
=𝑔𝑚. 𝑉𝐿1−𝑔𝑚. 𝑉𝐿3
𝑠𝐶2
No internal nodes are available in
current form.
𝑉𝑜𝑢𝑡 = 𝐼𝐿3. 𝑅𝐿 𝑉𝑜𝑢𝑡 =
𝐼𝐿3
𝐺𝐿=
𝑉𝐿3. 𝑔𝑚
𝐺𝐿
Resistor 𝑅𝐿 is using 𝑔𝑚 block in
unity feedback.
The implemented Gm−C filter from the LC ladder (shown above in Fig. 1) is shown in
Fig. 2. The first gm yields a current gm.(Vs−V2) which gets integrated through the
capacitor of value L1.gm2 as written in Table, whereas the last gm works as a resistor
Figure 19 Diagram for Writing Equations
1. 𝑰𝑳𝟏 =𝑽𝑺−𝑽𝟐
𝒔𝑳𝟏
2. 𝑰𝑳𝟑 =𝑽𝟐−𝑽𝒐𝒖𝒕
𝒔𝑳𝟑
3. 𝑽𝟐 =𝑰𝑳𝟏−𝑰𝑳𝟑
𝒔𝑪𝟐
4. 𝑽𝒐𝒖𝒕 = 𝑰𝑳𝟑. 𝑹𝑳
ofvalueRL.
5.3 Gm-C filter synthesis using gyrator:-
We can have another simpler method of implementing LC filter, by replacing passive
inductor with active inductors which are also called gyrators. A gyrators is simply
represented as two transconductor connected back to back. A gyrator-C network is
said to be lossless when both the input and output impedances of the transconductor of
the network are infinite and the transconductances of the transconductors are constant.
When one port of the gyrator is connected to a capacitor, the network is called the
gyrator –C network[8]. The problem of these active inductor is that they have a limited
frequency range in which they can behave as inductors. In addition to this, they are
thought as linear, which they are not. Non idealities of the transconductors also affect
the performance of inductive behavior i.e. noise, Band width, finite output impedance,
etc.
- -
---
-
gm
gm gm
gm
gm
L1.gm2
L3.gm2
Vs
Vout
VL1 VL3
V2
C2
Figure 20 Gm-C implementation of 3rd Order Butterworth Low Pass filter
The inductors L1 and L3 in Fig 1 can be replaced as capacitively loaded gyrators
(Active Inductors). It is obtained by first converting the voltage source to a current
source (Norton transformation, which also converts series source resistor R into shunt
resistor. The first OTA (Gm block) in Figure 4 performs V-I conversion, where as last
one implements a resistor R.
+
- gm1
+
-gm2
C
VX
IX
ZIN
L= C/gm1.gm2VX
Vout
Figure 21 Implementing Inductor using Gyrators
-
+
+
-
-
+
+
-
-
+
+
-
-
+
+
-
-
+
+
-
-
+
+
-
-
+
+
-VS
L1.gm2
2C2
2C2
L1.gm2
L3.gm2
L3.gm2
Voutgm
gm gm gm gm gm gm
Figure 22 The complete filter (3rd Order Butterworth) using 2 gyrators
5.4 Synthesis using Biquad:-
The biquad is second-order Quadratic equation described generally as;
H(s) =𝑎1.𝑠2+𝑎2.s+𝑎3
𝑏1.𝑠2+𝑏2.𝑠+𝑏3 (5.1)
The coefficients of numerator a1, a2, a3 can be chosen to yield a low-pass, band-pass,
or high-pass filter responses. i.e. if a1, a2 both become 0, then a low pass filter (LPF)
transfer function is produced, if a2, a3 becomes 0, then a Band pass filter function is
realized, and for a1, a3= 0, an HPF is realized. To realize higher order filters, biquad
sections can be cascaded [4], [7]. i.e. a 3rd order filter can be implemented by making
cascade connection of a biquad (2nd order) and a 1st order filter. And a 5th order filter
can be made by cascading 2 biquads and a 1st order filter. Our focus to implement a
3rd order Butterworth LPF can also be achieved using a biquad and a first order
integrator.
By looking at the above signal flow graph we can have the expression for the
Output at 1st summing point as;
𝑣𝑥 = 𝑣𝑖𝑛 (𝑘𝑜
𝜔2𝑜) − 𝑣𝑜𝑢𝑡 (5.2)
, Output at 2nd summing point is given by;
𝑣𝑦 = (𝑣𝑖𝑛 (𝑘𝑜
𝜔2𝑜) − 𝑣𝑜𝑢𝑡)
𝜔𝑜
𝑠−
𝑣𝑜𝑢𝑡
𝑄 (5.3)
and the Output at last block is given by;
Vin
Vout
K0/w0
2w0/s
-1/Q
w0/s
-1
Fig 23 Signal flow diagram of a Practical Biquad (LPF)
biquad
𝑣𝑜𝑢𝑡 = {(𝑣𝑖𝑛 (𝑘𝑜
𝜔2𝑜) − 𝑣𝑜𝑢𝑡)
𝜔𝑜
𝑠−
𝑣𝑜𝑢𝑡
𝑄}
𝜔𝑜
𝑠 (5.4)
(𝑣𝑖𝑛𝑘𝑜
𝑠𝜔𝑜−
𝑣𝑜𝑢𝑡𝜔𝑜
𝑠−
𝑣𝑜𝑢𝑡
𝑄)
𝜔𝑜
𝑠= 𝑣𝑜𝑢𝑡 (5.5)
𝑣𝑖𝑛𝑘𝑜
𝑠𝜔𝑜−
𝑣𝑜𝑢𝑡𝜔𝑜
𝑠−
𝑣𝑜𝑢𝑡
𝑄=
𝑠𝑣𝑜𝑢𝑡
𝑄 (5.6)
𝑣𝑖𝑛𝑘𝑜
𝑠𝜔𝑜= 𝑣𝑜𝑢𝑡 (
𝜔𝑜
𝑠+
1
𝑄+
𝑠
𝜔𝑜) (5.7)
𝑣𝑖𝑛𝑘𝑜
𝑠𝜔𝑜= 𝑣𝑜𝑢𝑡 (
𝑄𝑠2 + 𝑠𝜔𝑜 + 𝜔𝑜2𝑄
𝜔𝑜𝑄𝑠)
𝑣𝑖𝑛𝑘𝑜 =𝑣𝑜𝑢𝑡
𝑄(𝑄𝑠2 + 𝑠𝜔𝑜 + 𝜔𝑜
2𝑄) (5.8)
𝒗𝒐𝒖𝒕(𝒔)
𝒗𝒊𝒏(𝒔)|
𝒃𝒊𝒒𝒖𝒂𝒅=
𝒌𝒐𝑸
𝑸𝒔𝟐+𝒔𝝎𝒐+𝝎𝒐𝟐𝑸
=𝒌𝒐
𝒔𝟐+𝒔𝝎𝒐
𝑸+𝝎𝒐
𝟐 (5.9)
For calculating the transfer function of Bi-quad LPF, we can write the currents for the
transconductors as;
𝑖1 = 𝑣𝑜𝑢𝑡(−𝑔𝑚3) = −𝑔𝑚3𝑣𝑜𝑢𝑡 (𝑖1 is the current at the output of 𝑔𝑚3 ) (5.10)
𝑖2 = 𝑣𝑖𝑛(−𝑔𝑚1) = 𝑔𝑚1𝑣𝑖𝑛 (𝑖2 is the current at the output of 𝑔𝑚1 ) (5.11)
At the output node applying KCL, we can have;
𝑖𝑠𝑢𝑚1 = 𝑖1 + 𝑖2 = 𝑔𝑚1𝑣𝑖𝑛 − 𝑔𝑚3𝑣𝑜𝑢𝑡 (5.12)
gm2
-
gm4
-
gm3
+gm1
+
-
-
+
Vout
Vin
isum1
i1
i2
i5
i4
isum2
Vint2+
C1
C2
Fig 24 Biquad using Gm-C integrators for LPF
So now we can we can pass this current through an integrating capacitor C1 to have an
equation given by;
𝑣𝑖𝑛1 =𝑖𝑠𝑢𝑚1
𝑐1𝑠=
𝑔𝑚1𝑣𝑖𝑛−𝑔𝑚3𝑣𝑜𝑢𝑡
𝑐1𝑠 (5.13)
(𝑖4 is the current at the output of 𝑔𝑚2)
𝑖4 = 𝑣𝑖𝑛1(𝑔𝑚2) = 𝑔𝑚2𝑣𝑖𝑛1 (5.14)
(𝑖5 is the current at the output of 𝑔𝑚4 )
𝑖5 = 𝑣𝑖𝑛2(−𝑔𝑚4) = 𝑔𝑚4𝑣𝑖𝑛2 (5.15)
Applying KCL at node2;
𝑖𝑠𝑢𝑚2 = 𝑖4 + 𝑖5 = 𝑔𝑚2𝑣𝑖𝑛1 − 𝑔𝑚4𝑣𝑖𝑛2 (5.16)
(voltage across output capacitor C2) 𝑣𝑖𝑛2 = 𝑣𝑜𝑢𝑡 =𝑖𝑠𝑢𝑚2
𝑐2𝑠 (5.17)
𝑣𝑜𝑢𝑡 =𝑔𝑚2𝑣𝑖𝑛1−𝑔𝑚4𝑣𝑖𝑛2
𝑐2𝑠=
𝑔𝑚2𝑣𝑖𝑛1−𝑔𝑚4𝑣𝑜𝑢𝑡
𝑐2𝑠 (5.18)
𝑣𝑜𝑢𝑡 (1 +𝑔𝑚4
𝑐2𝑠) =
𝑔𝑚2𝑣𝑖𝑛1
𝑐2𝑠=
𝑔𝑚2(𝑔𝑚1𝑣𝑖𝑛 − 𝑔𝑚3𝑣𝑜𝑢𝑡)
𝑐1𝑐2𝑠2
𝑣𝑜𝑢𝑡(𝑐1𝑐2𝑠2 + 𝑔𝑚4𝑐1𝑠 + 𝑔𝑚2𝑔𝑚3) = 𝑔𝑚1𝑔𝑚2𝑣𝑖𝑛
𝑣𝑜𝑢𝑡(𝑠)
𝑣𝑖𝑛(𝑠)=
𝑔𝑚1𝑔𝑚2
𝑐1𝑐2𝑠2+𝑔𝑚4𝑐1𝑠+𝑔𝑚2𝑔𝑚3 (5.19)
We can simplify above equations as;
gm2
-
gm4
-
gm3
++
-
Vin
isum1
i1
i2
i5
i4isum2
Vint2+
C1
C2
+gm5
gm6
--
+
Vout
C3
-
V
V
V
gm1
2nd
Order (Biquad) 1st Order
Figure 25 3rd order Butterworth low pass filter (biquad)
𝑯(𝒔)|𝒃𝒊𝒒𝒖𝒂𝒅 =𝒈𝒎𝟏𝒈𝒎𝟐
𝒄𝟏𝒄𝟐⁄
𝒔𝟐+𝒔𝒈𝒎𝟒
𝒄𝟐+
𝒈𝒎𝟏𝒈𝒎𝟑𝒄𝟏𝒄𝟐
(5.20)
5.5 Biquad using Op-Amp and Active-RC
Integrator [7]:
Circuit Diagram for Opamp based biquad which is cascaded with first order filter to
get a 3rd order LPF circuit is shown below in Fig. 24. Here three Inverting Amplifiers,
two Summing Amplifiers and two Integrators are used for Biquad filter (based on
signal flow diagram of Biquad in fig 21) and one integrator at last is working as first
order filter.
Figure 26 circuit diagram of 3rd Order filter using Op-Amp-RC biquad and 1st order integrator in MULTISIM
Table 6 Comparison of the Equations of 3rd Order Passive and Active Filters
FILTERS TF(S) PASS BAND GAIN 𝒇𝒄
PASSIVE 𝑹
𝒔𝟑(𝑳𝟏𝑪𝟐𝑳𝟑) + 𝒔𝟐(𝑳𝟏𝑪𝟐𝑹) + 𝒔(𝑳𝟏 + 𝑳𝟑) + 𝑹
𝑹𝑹⁄ 251
MHZ
GM-C (𝑔𝑚1𝑔𝑚2𝑔𝑚5 𝑐12𝑐2⁄ )
𝑠3 + 𝑠2 (𝑔𝑚4
𝑐2+
𝑔𝑚6
𝑐1) + 𝑠 (
𝑔𝑚4𝑔𝑚6
𝑐1𝑐2+
𝑔𝑚3𝑔𝑚4
𝑐1𝑐2) +
𝑔𝑚4𝑔𝑚3𝑔𝑚6
𝑐12𝑐2
𝒈𝒎𝟏𝒈𝒎𝟐𝒈𝒎𝟓
𝒈𝒎𝟒𝒈𝒎𝟑𝒈𝒎𝟔
85M
HZ
ACTIVE
- RC
−𝑹𝟗𝑹𝟏𝟑𝑹𝟏𝟏𝑹𝟐𝑹𝟒𝑹𝟓𝑹𝟖𝑹𝟏𝟔
(𝟏 + 𝑺𝑹𝟏𝟔𝑪𝟑)(𝑹𝟑𝑹𝟏𝑹𝟏𝟓)(𝑺𝟐𝑪𝟏𝑪𝟐𝑹𝟏𝟎𝑹𝟏𝟑𝑹𝟏𝟏𝑹𝟒𝑹𝟔𝑹𝟐𝑹𝟖 + 𝑺𝑪𝟏𝑹𝟏𝟏𝑹𝟒𝑹𝟔𝑹𝟕 + 𝑹𝟏𝟑𝑹𝟖)
−𝑹𝟗𝑹𝟏𝟑𝑹𝟏𝟏𝑹𝟐𝑹𝟒𝑹𝟓𝑹𝟖
𝑹𝟑𝑹𝟏𝑹𝟏𝟑𝑹𝟖
424K
HZ
CHAPTER 6 Simulations and Results: 6.1 AC Analysis
In figure 27, the Butterworth LPF (biquad) response for 1.8v supply voltage and 𝑐𝑙 =
1.35𝑝𝐹 is shown. Here the open loop gain is approximately 0dB and attenuation rate
is 51dB/decade In this circuit it consists folded cascode OTA. Each symbol represents
a folded cascode OTA.
Figure 27 3rdorder Butterworth LPF in S-edit
Fig 28 are the frequency response of 3rd order Butterworth LPF which having -3dB
frequency of 85 MHz.
6.2 SIMULATION OF OP-AMP BASED BI-QUAD:
The simulation of 3rd order LPF using Op-amp is shown above. After simulation we
Figure 28 AC response of 3rd order Butterworth LPF in W-edit
Figure 29 AC sweep of Biquad of Op-Amp and Active RC Integrator
got Cut Off frequency(fc) of about 424 kHz, Stopband of 1.67 MHz and Attenuation rate is -60 dB/decade.
6.3 DC SIMULATION In order to know the output current without the presence of signal we give only one
input of 0.9v to M5 and 0v to M7 .The output is taken from drain terminal of the
PMOS3 and NMOS2.An Ammeter is connected at the output which, we measure the
output current of 200µA.
Figure 30 DC sweep analysis in S-Edit
6.4 NOISE SIMULATION Here noise simulation is done to measure the input referred noise in the band of interest
i.e. upto 250MHz. The maximum rms noise voltage obtained from the simulation is is
8.31nv/√𝐻𝑧 at 7 GHz.
Figure 31 𝑰𝒐𝒖𝒕 Vs Vgs
Figure 32 Test bench for Noise analysis folded cascade OTA
Table 7 All Simulation Results of 3rd Order Butterworth Low Pass Filter.
TECH ACTIVE(GM-C) PASSIVE
PASS BAND EDGE 85MHz 251MHz
STOP BAND
EDGE
660MHz ----
ATTENUATION
RATE
-60 dB/decade -60 dB/decade
POWER
DISSIPATION
4.37𝑚W NA
INPUT REFERRED
NOISE
8.31 (𝑛𝑉√𝐻𝑧
⁄ ) ----
Figure 31 Noise Simulations folded cascade OTA
CONCLUSION
In this report we have studied various architectures of OTA and filter synthesis
methods. A passive 3rd order Butterworth low pass filter based on typical specs for
MHz band i.e. 250MHz is taken from literature and its active synthesis using various
methods is studied. Biquad based synthesis is taken up for the implementation of the
filter. The 3rd order Butterworth LPF is designed using a 2nd order biquad and a first
order integrator. The biquad using Op-Amp RC integrator and Gm-C integrators are
also studied. Simulations for the Op-amp-RC based filter is performed in the NI
Multisim environment which shows the cut-off frequency 424kHz. The filter using
Gm-C biquad is simulated at transistor level in Tanner S-Edit using 250nm CMOS
which achieves a pass band of 85MHz. The Gm-C filter has an attenuation rate of
60dB/decade. The power consumed for the fiilter is 4.37mW at 1.8V dc supply. The
biquad using Op-amp RC integrators, Gm-C integrators are, Passive 3rd order filters
have been compared.
REFERENCES
[1] S. J. Yoo,‘Design of Analog Baseband Circuits for Wireless Communication
Receivers’, Ph.D Thesis, The Ohio State University, 2004.
[2] E. Sanchez- Sinencio and J. Silva-Martinez, "CMOS transconductance amplifiers,
architectures and active filters: tutorial", IEE Proceedings - Circuits, Devices and
Systems, Vol. 147, Issue. 1, pp. 3-12, Feb 2000.
[3] W. Sansen, "Distortion in Elementary Transistor Circuits", IEEE Transactions on
Circuits and Systems II: Analog and Digital Signal Processing, Vol.46, Issue 3, pp.
315-323, Mar 1999.
[4] B. Razavi, Fundamentals of Microelectronics, 2nd Edition, John Wiley, 2008
[5] A. S. Sedra, K.C. Smith, Microlectronics Circuits, Oxford Publications, Ch. 11.
[6] W. K. Chen, Passive, Active, and Digital Filters, CRC Publications, Ch. 16.
[7] B. Razavi, ‘A Circuit for All Seasons: The Biquadratic Filter’, IEEE Solid-State
Circuits Magazine, Spring 2018.
[8] A Panigrahi, "Notes on Design of Gm-C Filters for Channel Selection”. 2019
APPEDNIX
Comparison of OTA and Op-Amp:
Features of OTA- OTAs are simple Voltage input Current output (VCCS) building blocks useful for
tunable or controllable analog operations performed usually in open loop, as their
transconductance can be tunable by varying the Bias current.
OTA are good at driving capacitive loads i.e. modern CMOS circuits.
They can be used in tunable signal processing e.g. tunable filters, Automatic gain
control etc.
They can be used as Voltage controlled Resistors (Tunable resistors) with unity
(negative) feedback around it.
Both Gain and Bandwidth is tunable. Behave like first order systems so stable and
no compensation is necessary unlike Op-amp.
Since they are designed to produce current output, so output impedance is usually
very high.
Since operated in open loop, they are prone to linearity problems.
Features of Op-amp: -
OpAmps are simple Voltage input voltage output (VCVS) building blocks useful
for many analog operations performed usually by adding some components
around it using feedback.
Opamp can drive high capacitive loads and resistive loads.
Since they are designed to produce voltage output, so output impedance is usually
very small.
Opamps offer limited Band width i.e. upto few 10's of MHz and might require
some compensation for stability.
Limited Slew Rate.