Design and Implementation of High Speed and Low Power 12-bit
SAR ADC using 22nm FinFET
1VASUDEVA G., 2UMA B. V.
1Research Scholar, Electronics and Communication Engineering Department,
RV College of Engineering, Bengaluru,
Affiliated to Visvesvaraya Technological University,
BELAGAVI, KARNATAKA, INDIA
2Professor and Dean Student Affairs, Electronics and Communication Engineering Department,
RV College of Engineering, Bengaluru,
Affiliated to Visvesvaraya Technological University,
BELAGAVI, KARNATAKA, INDIA
Abstract: Successive Approximation Register (SAR) Analog to Digital Converter (ADC) architecture
comprises of sub modules such as comparator, Digital to Analog Converter and SAR logic. Each of
these modules imposes challenges as the signal makes transition from analog to digital and vice-versa.
Design strategies for optimum design of circuits considering 22nm FinFET technology meeting area,
timing, power requirements and ADC metrics is presented in this work. Operational Transconductance
Amplifier (OTA) based comparator, 12-bit two stage segmented resistive string DAC architecture and
low power SAR logic is designed and integrated to form the ADC architecture with maximum sampling
rate of 1 GS/s. Circuit schematic is captured in Cadence environment with optimum geometrical
parameters and performance metrics of the proposed ADC is evaluated in MATLAB
environment. Differential Non Linearity and Integral Non Linearity metrics for the 12-bit ADC is
limited to +1.15/-1 LSB and +1.22/-0.69 LSB respectively. ENOB of 10.1663 with SNR of 62.9613 dB
is achieved for the designed ADC measured for conversion of input signal of 100 MHz with 20dB noise.
ADC with sampling frequency upto 1 GSps is designed in this work with low power dissipation less
than 10 mW.
Key Words: SAR ADC, 22nm FinFET, Low Power, High Speed, Folded Resistive String,
Operational Transconductance Amplifier.
Received: February 15, 2021. Revised: October 3, 2021. Accepted: November 29, 2021. Published: January 3, 2022.
1. Introduction
The links between analog world of transducers
and digital world of signal processing and data
handling are Analog to digital converters(ADC)
and Digital to analog converters(DAC).The
A/D interface to reside on the same silicon with
large DSP or digital circuit is due to increasing
trend of integration level of integrated circuits.
Scaling in transistor geometries have resulted in
low voltage references for operation of digital
circuits. ADCs that are interfaced with DSPs
blocks also needs to operate at low voltages.
Among the most popular Nyquist rate ADCs
SAR ADCs are important because of high
speed conversion and low voltage operation.
Compared with other types of ADC, SAR
ADCs exhibit the best energy efficiency for
medium to high speed, moderate resolution and
low power applications. Aili Wang et al.[1]
have designed a 10-bit SAR register ADC with
SNDR of 59.59 dB and power dissipation
limited to 41.3 μW operating at maximum
frequency of 50 MS/s using 14 nm SOI FinFET
technology. The number of capacitors in the
ADC is reduced using segmented architecture.
Aligned switching with Skip (ASS) logic is
used to save power dissipation. The SAR logic
is based on VCM-based MCS Scheme which
saves power dissipation further by 85%. Dual
mode power is used to further reduce power
dissipation. Lukas Kull et al.[2]presented an 8-
bit Time-interleaved ADC that is designed to
operate in the 24-72-GS/s. The SNDR at low
frequency is 39 dB and 30 dB at Nyquist
frequency. The Time-interleaved ADC uses 64
asynchronous SAR ADCs to perform the
conversion. 14-nm CMOS FinFET technology
is used in the design that requires 0.15 mm2
area. James Hudner et al.[3] in their work,
proposed SAR ADC that operates at 56GS/s
and is based on time interleaved logic using
16nm FinFET technology. The power
dissipation is limited to 475 mW. Sung-En
Hsiehet al.[4] presenteda 11-bit SARADC that
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.1
Vasudeva G., Uma B. V.
E-ISSN: 2224-2856
1
Volume 17, 2022
uses semi resting DAC and cascaded input
comparator. The designed ADC operating at
600-kS/s with 0.3 V supply voltage requires
187 nW of power dissipation. The SFDR is 73
dB at 9.46 bits. Keisuke Okuno et al.[5] in their
work reported on 8-bit SAR ADC using 16 nm
FinFET. The time interleaved ADC is designed
to operate at 800 MS/s with ENOB of 6.53 bits.
The results of designed ADC are compared
with several other ADCs demonstrating
superiority in performances. Ashish Joshi et
al.[6] have designed ADC using 45nm FinFET
models based on SAR logic that is designed to
operate at 909 kS/s with 9 μW of power
dissipation. The ADC uses switch capacitor
DAC and opamp based comparator circuit.
Ewout Martens et al., [7] have designed ADC
based on SAR logic using 16nm CMOS
FinFET models that operates at 300 MS/s with
ENOB of 11.2 bit. The power dissipation is 3.6
mW with 76 dB harmonic distortion. Zhiliang
Zheng et al.,[8] presented a new technique for
error cancellation in SAR ADCs. In this
technique, the conversion time is increased by
50% to cancel the first order capacitor
mismatch error, where typically 100%
additional conversion time is needed compared
to the ideal operation.Gilbert Promitzer
[9]presented a paper on non calibrating SAR
ADC with fully differential switched capacitor
for low power with 12 bit resolution. The power
consumption is reduced because of cancellation
of the VCM buffer and by implementing self
timed comparator. To achieve lower power
consumption and higher sampling rate the
resolution was improved from 10 bit to 12 bit.
Eric Fogleman et al.,[10] presented a technique
to reduce in band noise of DAC block in ADC
architecture. Using a second order 33-level tree
structured mismatch shaping DAC an audio
ADC Delta-Sigma modulator is designed. The
prototype modulator is implemented in a
standard
0.5 - 3.3V single-poly CMOS fabrication
process. All 12 of the fabricated prototypes
achieve a 100dB peak signal-to noise and
distortion ratio (SINAD) and 102dB dynamic
range over a 10–20 kHz measurement
bandwidth. John McNeill et al.,[11] presented
the “Split ADC” architecture. The “Split ADC”
architecture uses two independent and identical
ADCs to sample the same input, Vin. The two
independent outputs are averaged to produce
the ADC output code. The difference of the two
outputs provide information for the background
calibration process.Mi-rim Kim et al.,[12]
presented a 12-bit SAR ADC with hybrid RC
DAC. The hybrid RC DAC is employed to
reduce the size and improve energy efficiency
by reducing the total number of capacitors. The
prototype ADC fabricated in a180 nm CMOS
and occupies 0.25 mm2 active die area. The
measured DNL and INL are +0.47/-0.48 LSB
and +0.75/-0.76 LSB respectively. The ADC
shows the maximum SNDR of 64.2 dB and
SFDR of 80.4 dB with a 2.8V Supply
consuming 1.16mW.Dragisa Milovanovic et
al.,[13] presented second order sigma-delta
modulator in CMOS 0.35 µm technology for
audio applications. In this they presented a
technique to improve the swing, dynamic range
and stability analysis of the second order sigma-
delta modulator by scaling the gain of the
integrators. he area occupied by this design is
0.57 mm2.Oguz Altun et al.,[14] presented the
multi rate multi bit sigma-delta modulator for
low power implementation in 90 nm for
wireless application. This design achieves 71.4
dB SNDR in 200 kHz GSM band and dissipates
1.1 mA of total current from a 1.5 V Supply.
Victor Aberg[15] in his master’s thesis in
Embedded Electronic system have presented a
design of SAR ADC using 28 nm FD-SOI
CMOS technology that comprises of scaled
Capacitive DAC. The ADC has been designed
to operate at 800 MS/s with SNDR of 38.4 dB
and consumes power less than
1.1 mW.
In this work a 12-bit SAR ADC is designed and
implemented for high speed and low power
using 22 nm FinFET technology. This paper is
organized as follows: Section 2 describes SAR
ADC architecture. Section 3 provides design of
SAR ADC. Section 4 presents FinFET and
Section 5 presents Design of SAR Logic Block
which includes power estimation, DAC design
and Comparator design. Section 6 describes the
implementation of SAR logic block, DAC and
Comparator. Section 7 presents Results and
Discussion. Section 8 concludes the paper.
2. SAR ADC
The block diagram of SAR ADC is as shown in
Figure 1. In a Successive Approximation
Register Analog to Digital converter, the input
signal Vin is sampled at the beginning of each
conversion cycle. The process of conversion
starts by comparing the input signal Vin and the
half reference voltage Vref /2 to determine the
MSB of Vin and also determines the search
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.1
Vasudeva G., Uma B. V.
E-ISSN: 2224-2856
2
Volume 17, 2022
region for the second MSB. The binary search
algorithm is allowed to approximate the actual
Vin, and the reference voltage is used for the
MSB will be divided by 2.The result is added or
subtracted from the previous reference voltage
which delimitates the new binary search
regions.
Fig-1. Block diagram of the SAR A/D
converter[8]
Reference voltage updated and Vin, each
comparison between them generates N-bits of
SAR ADC and one bit of Vin which needs N
Comparisons.
3. Design of SAR ADC
In this work SAR ADC is designed and
implemented. The design is divided into sub
blocks like DAC, comparator, SAR logic. The
schematic of the sub blocks is designed using
Virtuoso schematic editor. The simulations for
the individual blocks were carried out using
ELDO simulation tool. Integration of the sub
blocks was done using virtuoso schematic
editor. The specifications for SARADC
(Table 1) is identified considering the reference
design [16].
Table-1. Design specifications of SAR ADC
Resolution
12-bit
Technology
22 nm
Clock Rate
1GHz
Area
<1mm2
Power Supply
1.8V
Operating Current
1mA
Input Range( Vin)
0.2V-1.4V
Power Dissipation
<10 mW
INL
+/- 2 LSB
DNL
+/- 2 LSB
SNR (dB)
64 dB
SNDR (dB)
60 dB
SFDR (dB)
72 dB
ENOB
9.5
Several factors are considered for ADC design.
The ratio in between the maximum input signal
level and the minimum detectable input signal
that the modulator can handle is called the
Dynamic Range(DR).The ratio in power
between the input sine wave and the noise of the
converter from DC to Nyquist rate[15] is called
Signal to Noise Ratio(SNR). SNR is expressed
in decibels and the equation is given in (1).
Pnoise
Psignal
SNR log10=
………….. (1)
The ratio of the signal power to the total noise
and harmonic power at the output is called the
Signal to Noise Plus Distortion
Ratio(SNDR)[15].Harmonic content is present
in SNDR which is not there in SNR, Otherwise
they are similar. Distortion is not important for
small signal levels. As the signal level increases
and distortion degrades SNDR will be less than
SNR. Equation of SNDR is given in (2).
nPdistortioPnoise
Psignal
SNDR +
=log10
.......... (2)
The ratio of the power value of the input sine
wave with a frequency f(IN) to the power value
of the peak spur observed in the frequency
domain is called Spurious Free Dynamic
Range(SFDR)[15].The ratio between the
sampling frequency fs and the Nyquist rate
f(IN) is called the Over Sampling Ratio(OSR).
Since f(IN) =2fB, the over sampling ratio
equation is given in (3).
fB
fS
OSR 2
=
………(3)
Due to the periodical charge/discharge of load
capacitances we get dynamic power dissipation
and it is shown in equation (4).
OSRfNCsVDDP = 2
…... (4)
Where Cs is sampling capacitor, fN is system
Nyquist rate, OSR is over sampling ratio and α
is the average change of the voltage on the
sampling capacitor. Static power dissipation
Pstat is computed considering the current flow
during static state of the circuit that is
considered as leakage current and the power
Pstat is given as in equation (5).
Vs
Ci
Cs
gmVDDIstatVDDPstat ==
… (5)
The power dissipation in ADC is either due to
higher sampling frequency or due to the power
dissipation due to charging and discharging of
the capacitors. Based on the factors that
constitute power dissipation in ADC it is
required to design the sub systems in ADC with
low power strategies.
4. FinFET
Double Gate FinFET device shown in Figure 2
(a) and its characteristics demonstrating the
increased current flow in the channel by
controlling with two gate voltage is presented
Digital
Output
V(K)
Vin
Comp
Sample
& Hold
SAR
DAC
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.1
Vasudeva G., Uma B. V.
E-ISSN: 2224-2856
3
Volume 17, 2022
in [17]. Small signal model for the DG FET is
presented in Figure 2(b). Cgd, Cgs and Cds are the
parasitics in FinFET that limits the device
operation at high frequencies and Rgd, Rgs, Rds
and Rsub limits the device for low power
operations.
Fig-2. FinFET[17]
Fig-2(a). FinFET device structure of DGFET [17]
Fig-2 (b). FinFET Device Small signal model [17]
Predictive Technology Model (PTM)
parameters for FinFET is presented in Table-2
and the corresponding model files are
considered for design of OTA and OTA based
comparator. Considering structural and
electrical parameters of FinFET device
modeling is carried out for analysis of input and
output characteristics. The theoretical and
practical mismatches are identified based on
simulation results and the appropriate geometry
settings for FinFET are identified for maximum
frequency of operation and low power
dissipation [18].
Table-2. FinFET Device Parameters
Parameter
Value
Channel length
22 nm
Oxide thickness 1
2.5 nm
Oxide thickness 2
2.5 nm
Gate length
22 nm
Source/drain extension
length
50 nm
Gate to source/drain
overlap
2 nm
Work function
4.6 eV
Source/Drain doping
1 x 1019cm-3
Dielectric constant of
channel
11.7
dielectric constant of
insulator
3.9
Bandgap
1.12 eV
Affinity of channel material
4.05 eV
Mobility of electrons
1400 cm2/Vs
saturation velocity
1.07e+07
cm/s
The building blocks of SAR ADC are designed
using FinFETs and the design procedures are
discussed in detail.
5. Design of SAR logic block
It is required to turn ON/OFF ADC reference
current source switches, depending on
comparator output at every clock till the end of
conversion and generate the ADC data output.
This logic is implemented in RTL. Block
diagram of the SAR logic block is shown in the
Figure-3.
Fig-3. Symbol of SAR logic block
All the flip-flops in this module will be reset
asynchronously when reset_n is active, the
entire module works on the positive edge of
clock(clk). In reset deactivate condition, the
module monitors Start of Function (SOF)
signal. On detecting the transition on SOF
signal from low to high, the conversion process
starts and continues for 14 cycles. During the
conversion phase, depending on the output of
the comparator, for every clock cycle of
comparison (from 2nd to 13th) a 12-bit register
(Digital Output) is updated. At the end of
14thcycle, an End of Function (EOF) signal is
set, and the 12-bit register values are latched on
to the data_out<11 : 0> (output) bus. During the
next clock cycle (15thcycle), the SOF and EOF
signals are reset. The logic output of SAR is
interfaced with the DAC input using buffer
modules and to transfer the SAR output to the
output pins of ADC module.
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.1
Vasudeva G., Uma B. V.
E-ISSN: 2224-2856
4
Volume 17, 2022
Fig-4. Timing diagram used for SAR logic
All the flip-flops in this module will be reset
asynchronously when reset_n is active. The
entire module works on the positive edge of
clock clk. The timing diagram determined to
design the SAR logic block is shown in Figure
4. In reset deactivate condition, the module
monitors sof signal, the conversion process
starts on detecting the transition on sof signal
from low to high. The input of the SAR logic is
the output of the comparator which generates
Vcmp out as in Eq. (6),
0, Vdac Vin
Vcmp (compare _ in) 1, Vdac Vin
=
...(6)
Considering the output of comparator
(compare_in), the conversion logic is as
follows:
• 1st clock cycle: The s1 signal is set to 1 and
bit_sel[11 : 0] is set to 0 00
• 2nd clock cycle: The s1 signal is cleared, s2
signal is set to 1 and set bit_sel[11] to 1if
compare_in is 1
• 3rd clock cycle: Retain the bit_set[11] as 1
if compare_in is 0, else clear bit_set[11]
and set bit_set[10] to 1
• and this process is continued upto 14th
cylce
• 14th clock cycle: Retain the bit_set[0] as 1
if compare_in is 0, else clear bit_set[6],
clear signal s2, set signal eof and register
the bit_set[11:0] into ADC output
data_out[11 : 0]
• 15th clock cycle: Clear eof and monitor of
sof and start from 1st cycle.
The current bias for ADC design required are
1i, 2i, 4i, 8i and 16i units that need to be derived
from the current references. The reference
current is designed using current mirrors from
which all other current bias is derived. To
produce the desired output current, the current
source transistor and its cascaded neighbour
needs to be biased and is designed using current
mirrors.
5.1 Power estimation
The total power consumption can be estimated
as given in equation (7).
Ptotal = total number of current cells *Iref *Vdd ...(7)
In this design, Vddof 1.8 V is selected and
therefore, it is necessary to minimize the total
current in the design to reduce the total power
consumption. There are three building blocks in
ADC with DAC requiring higher current
references. Comparator and SAR logic is set to
operate at maximum current of 50 µA and DAC
is set to operate at maximum of 100 µA, the
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.1
Vasudeva G., Uma B. V.
E-ISSN: 2224-2856
5
Volume 17, 2022
total current required by the ADC is 200
µA.For analysis of transistor matching
pelgrom’s paper [19] has become the standard
source and his formula for the standard
deviation of saturation current for identically
sized devices was used for the design.Mismatch
causes time independent random variations in
physical quantities of identically designed
devices,which means each current source in the
matrix generates a current that varies slightly
from the desired current Iref.To see that random
variations do not degrade the performance of
the circuit below its specifications, the current
sources have to be designed.The formulas are
given in equations (8) and (9).
( ) ( )
( )
( )
2 2 2
d TO
2
22
dGS TO
I 4 V
IVV
=+
−
…….(8)
where
( )
2
2VTO
TO
A
VWL
=
and
( )
2
2
2
A
WL
=
.(9)
The dependant area parameters are AVTO and
Aβ. Note that variation with spacing is
neglected. The expressions for W2 and L2 are
derived from equations (8) and (9).W2 and L2
are used to design the width and length of the
transistors of the biasing circuit.
( ) ( )
2
2
2LSB VT
2 2 2
GS T GS T
2I 4A
A
W
V V V V
I
* co x I
=+
−−
..(10)
( )
( )
2
2 2 2
GS T VT
2
LSB
*Cox
L A * V V 4A
I
2I I
= − +
..(11)
5.2 DAC Design
In this work a new architecture for Digital to
Analog (DAC) is proposed, designed, modelled
and evaluated for its performance [20]. The
proposed 12-bit DAC block diagram is shown
in Figure 5. The DAC structure is split into two
groups of 6-bits. The first stage generates Vout1
corresponding to
6 MSBs and the second stage generates Vout2 for
6 LSBs. The output of two-stage DAC Vout1 and
Vout 2 are accumulated in the adder circuit to
generate the final analog output VOUT.
Fig-5. Proposed 12-bit DAC structure[20]
The 12-bit DAC is designed using two stages of
6-bit DAC [20]. Each of the 6-bit DAC consists
of two step voltage divider type DAC and
folded resistive string network. Device
mismatches and area optimization is achieved
with folded resistive string approach. Also,
improvement in resolution is achieved with
coarse and fine voltage generation logic from
the two step voltage divider method. Schematic
capture is carried out using Cadence tool. From
the simulation studies, it is observed that the
designed DAC has a maximum operating
Bandwidth of 100 MHz and the gain at 3 dB is
41.86 dB. The power dissipation of designed
DAC has been calculated at 4.33 mW and hence
suitable for high speed ADC. The INL and
DNL of the DAC design has been calculated as
+0.034 V to -0.001 V and +0.06 V to -0.05 V.
The performance is accomplished with a design
area of 450 µm2.
5.3 Comparator design
Comparator which is another sub block of SAR
ADC circuits is designed using Operation
Transconductance Amplifier (OTA) [18]. The
OTA circuit is realized using eleven transistors
-
+
VREF2
VREF
VOUT
6-bit two-step DAC
MSB 6-bit two-step DAC
LSB
b1
b2
b3
b4
b5
b6
A1
-
+
A2
-
+
b7
b8
b9
b10
b11
b12
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.1
Vasudeva G., Uma B. V.
E-ISSN: 2224-2856
6
Volume 17, 2022
as shown in Figure6. The transistors F1 and F2
are the differential pair and forms the
transconductance cell that converts the input
voltage V+in and V-in (differential input
voltages) to current. The differential current
output (Iout) of the differential pair is converted
to single ended current at the output by using
the current mirrors F3 to F8, F10 and F11. F9
transistor is used to bias the differential pair and
is used as current sink circuit. The cut-off
frequency of the OTA is decided by setting the
appropriate bias current and the load
capacitance of the OTA. The transconductance
gain gm of the OTA is controlled by setting the
current that enters the transistor F9 and the gate
voltage Vb is appropriately set. Design of OTA
is primarily identifying the transistor
geometries such that the simulation results are
matching the hand calculations. Design
methodology based on gm/ID method is the most
popular approach that identifies the transistor
geometries based on data sheets and simulation
results. In this method of design of OTA
circuits based on datasheet several design
variables that were required for the design were
assumed without clear rules. The design
specifications meeting input range, common
mode rejection and noise parameters were not
considered in this approach. Even the channel
length variations with regard to gm/ID were not
considered in the design process.
Fig-6. OTA based comparator
The OTA is designed using FinFET transistors
and the comparator circuit is designed
integrating the sub circuits with OTA. The
building blocks of the comparator design such
as input level shifter, differential pair with
cascode stage and class AB amplifier for output
swing are designed and integrated. The gain of
the comparator is 103dB, with phase margin of
650, CMRR of 76 dB and output swing from rail
to rail of 0 to 1.8 V is achieved. The circuit
provides unity gain bandwidth of 5 GHz and
DC Open Loop voltage gain of 90 dB.
6. Implementation
The SAR ADC primarily consists of 3 main
blocks. SAR Logic Block, a resistive string
folded segmented DAC and OTA based
Comparator with Offset- Cancellation feature.
6.1 The SAR Logic Block Implementation
SAR logic design is presented in terms of Finite
State Machine (FSM)and is illustrated for its
logic function as in Table 3. Each conversion
requires 14 clock cycles with the first clock
being reset mode and in this mode all outputs
are set to zero. From the 1st clock to 13th clock
data is converted sequentially and the SAR
output is generated. In the 14th clock cycle the
data generated is stored in the output register for
conversion to analog data by the DAC circuit.
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.1
Vasudeva G., Uma B. V.
E-ISSN: 2224-2856
7
Volume 17, 2022
Table-3. SAR logic conversion table using FSM
Cycle
Sample
B11
B10
B9
B8
B7
B6
B5
B4
B3
B2
B1
B0
Comp.
0
1
0
0
0
0
0
0
0
0
0
0
0
0
-
1
0
1
0
0
0
0
0
0
0
0
0
0
0
a11
2
0
a11
1
0
0
0
0
0
0
0
0
0
0
a10
3
0
a11
a10
1
0
0
0
0
0
0
0
0
0
a9
4
0
a11
a10
a9
1
0
0
0
0
0
0
0
0
a8
5
0
a11
a10
a9
a8
1
0
0
0
0
0
0
0
a7
6
0
a11
a10
a9
a8
a7
1
0
0
0
0
0
0
a6
7
0
a11
a10
a9
a8
a7
a6
1
0
0
0
0
0
a5
8
0
a11
a10
a9
a8
a7
a6
a5
1
0
0
0
0
a4
9
0
a11
a10
a9
a8
a7
a6
a5
a4
1
0
0
0
a3
10
0
a11
a10
a9
a8
a7
a6
a5
a4
a3
1
0
0
a2
11
0
a11
a10
a9
a8
a7
a6
a5
a4
a3
a2
1
0
a1
12
0
a11
a10
a9
a8
a7
a6
a5
a4
a3
a2
a1
0
a0
13
0
a11
a10
a9
a8
a7
a6
a5
a4
a3
a2
a1
a0
-
Figure 7 presents the SAR logic circuit diagram
using D-flip flops. The SAR logic is realized
using ring counter and a code register. In the
first clock i.e., clock zero all the outputs of D-
flip flop are reset to zero. In the first clock the
MSB bit is set to logic ‘1’ and this logic data is
shifted through the shift registers.
Fig-7. SAR logic realized using ring counter and code register
The Verilog code is developed for SAR logic
and is verified for its functionality in Cadence
environment and the verified code is
synthesized to generate the netlist. The netlist
is imported into Virtuoso schematic
environment and is integrated with analog
block of ADC structure.
6.2 DAC Implementation
The Digital to Analog (DAC) converter is
basically Current-Steering Mode - Segmented
architecture. The current sources are designed
as 1i, 2i, 4i, 8i and 31*16i. The virtuoso
schematic is as shown in Figure 8.
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.1
Vasudeva G., Uma B. V.
E-ISSN: 2224-2856
8
Volume 17, 2022
Fig-8. Virtuoso opus schematic of DAC block
The designed 12-bit DAC has to work
according to its functionality. Once the netlist
of spice is generated the simulations results are
produced using ELDO, then add all the required
input voltages such as input bits and bias
voltage. The input reference currents were also
provided. The power supply is given as 1.8V.
6.3 OTA Based Comparator Implementation
Comparator schematic captured in Virtuoso is
presented in Figure 9. The OTA circuit is
presented in Figure 9(a) and Figure 9(b)
presents the OTA used as comparator.
The simulation of the comparator was done
with a test case where a PWL signal was
provided as the Analog input to the comparator.
The circuit is operated at 1.8V and 1 GHz clock
frequency.
Fig. 9 Comparator schematic
Fig-9(a).OTA sub circuit
Fig-9(b). OTA as comparator
B1
1
B6
B5
B4
B3
B2
B1
V
VL
VRef
.
VS
S
V
H
_
+
VL
_
+
DA
C
(LS
B)
Vout
B1
2
B1
0
B9
B8
B7
VREF
L
VREF
H
DAC
(MSB)
3:8
Decoder
Folded
Resistor
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.1
Vasudeva G., Uma B. V.
E-ISSN: 2224-2856
9
Volume 17, 2022
7. Results & Discussion
Figure 10 presents the simulation result of SAR
ADC where it shows 6-bit (MSB) digital output
such as B12, B11, B10, B9, B8, B7 along with
its analog input (analog_in) and DAC output
(DAC Out). An input signal of frequency of 100
MHz is considered in this test case and the
amplitude of the signal is set between 0 to 2 V.
12-bit ADC operating at sampling frequency of
1 GHz, the 6 LSB bits will have very fast
switching process and hence MSB bits are
captured and presented for verification. An
input signal of 10 kHz is considered as input
and the voltage level is between 0 to 2 V is
sampled at 100 MHz. The digital output of
ADC is captured considering the 6 LSB bits to
check for logic correctness of the ADC.
Fig-10. Simulation result of SAR ADC (LSB 6-bits)
Fig-11.Simulation results of SAR ADC (MSB 6-bits)
MSB ‘B7’
Bit‘B8’
Bit ‘B9’
Bit‘B10’
Bit‘B11’
LSB ‘B12’
DAC Out
Analog_in
LSB ‘B6’
Bit‘B5’
Bit ‘B4’
Bit‘B3’
Bit‘B2’
MSB ‘B1’
S/H Out
Analog_in
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.1
Vasudeva G., Uma B. V.
E-ISSN: 2224-2856
10
Volume 17, 2022
Figure 11 illustrates about the simulation result
of SAR ADC 6-bit (LSB) where the results
obtained for analog input (analog_in) , digital
to analog converter output (DAC Out), and
digital output bits such as B6, B5, B4, B3, B2,
B1. From this simulation result, it is observed
that the output of all the digital bits has been
high when analog input voltage is equal to its
reference voltage or equal to VDD voltage.
The power dissipation simulations were done
on the extracted layout net list including
parasitic resistance and capacitance of the
power buses, nets used for signal routing using
ELDO simulator. The power supply is 1.8V.
Figure 12 illustrates the active current across
multiple cycles. The power dissipation results
are inclusive of SAR Logic block. The
maximum switching current is captured from
the SAR logic switching states and it is
observed that the maximum switching current
of 0.1806 mA occurs whenever the MSB bits is
activated. Considering the maximum switching
current, the average power dissipation is
computed.
Fig-12.Active current plots of ADC
7.1 ADC metrics
Input signal with maximum frequency of 100
MHz is sampled at 1 GHz sampling frequency
with amplitude of 2 V. The input signal is added
with random noise and multiple harmonics that
occur in the band 10-40 kHz, 10-40 MHz and
70-100 MHz. The noise power is also varied to
estimate the ADC metrics. Figure 13presents
the input data with noise and harmonics. Figure
14 presents the FFT spectrum of input data with
noise and harmonics.
Fig-13. Input data with 20dB noise and with 3rd order harmonics
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.1
Vasudeva G., Uma B. V.
E-ISSN: 2224-2856
11
Volume 17, 2022
Fig-14. Frequency spectrum of input data
The ADC schematic is simulated in Virtuoso
environment and the corresponding outputs are
imported into MATLAB environment to
compute ADC metrics such as ENOB, SNDR,
SFDR and SNR. Figure 15 presents the ADC
metrics plot for input signal of 1.2Vwith higher
noise power defined in the range of 40 to 50 dB.
The ADC metrics are ENOB = 4.8449, SNDR
= 30.9261, SFDR = 55.6228,
SNR = 30.9261.
Fig-15. ADC metrics with noise power in 40 – 50 dB harmonics in 10 kHz to 40 kHz
Fig.-16. ADC metrics with noise power in 40 – 50 dB harmonics in 10 MHz to 40 MHz
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.1
Vasudeva G., Uma B. V.
E-ISSN: 2224-2856
12
Volume 17, 2022
The frequency spectrum depicting ADC
metrics is presented in Figure 16. ADC metrics
are ENOB = 4.8392, SNDR = 30.8923, SFDR
= 54.9778, SNR = 30.8923 for noise power
in 40-50 dB and harmonics in 10 MHz – 40
MHz range.
Fig. 17 ADC metrics with noise power in 10 – 20 dB harmonics in 10 MHz to 40 MHz
The frequency spectrum depicting ADC
Metrics is presented in Figure 17for noise
power in 10-20dB and harmonics in 10 MHz to
40 MHz range.ADC Metrics are ENOB =
8.1610, SNDR=50.8893, SFDR = 75.4777 and
SNR = 50.8999.
Fig. 18 ADC metrics with noise power in 10-20 dB harmonics in 70 MHz to 100 MHz
The frequency spectrum depicting ADC
metrics with noise power in 10 dB, harmonics
in 70 MHz to 100 MHz is presented in Figure
18.ADC metrics are ENOB=10.1663,
SNDR=62.96 dB, SFDR = 87.54 dB and
SNR=62.9613 dB. It is observed from the
measurement results that the change in
harmonics intensity for the ADC operating at
70MHz to 100MHz band width resulted in
improvement in ENOB, SNDR and SNR.The
highest INL error measured is +1.22/-0.69 LSB.
Figure 19 presents the DNL and INL
measurements for 12-bit ADC.
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.1
Vasudeva G., Uma B. V.
E-ISSN: 2224-2856
13
Volume 17, 2022
Fig-19. DNL and INL measurements for 12-bit ADC
Table 4-Comparison table
Parameters
Existing
Design[16]
Proposed Design performance
parameter and its values
Resolution
12-bits
12-bits
Technology
40 nm
22nm
Clock frequency
200 MHz
1 GHz
Power Supply
1.2 V
1.8 V
Operating Current
15 mA
1 mA
Input voltage range
-
0.2 V-1.4 V
Power Dissipation
8 mW
7 mW
INL
1 LSB/-1.2 LSB
+1.22/-0.69 LSB
DNL
0.7 LSB/-0.3
LSB
+1.15/-1LSB
SNR (dB)
-
62.9613
SNDR (dB)
-
62.96
SFDR (dB)
67.9
87.54
ENOB
9.85
10.1663
Novelty in Proposed Design
-
OTA Comparator Two Step DAC
8. Conclusion
This work involved in designing a FinFET
based 12-bit SAR ADC for suitable for high
speed and low power applications. The design
consists of sub blocks like DAC, comparator,
SAR logic. The OTA based comparator is
designed to achieve gain of 100 dB with phase
margin of 65° and UGB of 5 GHz. DAC circuit
designed using two stage folded resistive
structure which can operate at maximum
frequency of 10 GHz with power dissipation of
less than 4 mW. Also, DAC performs the
conversion of data with minimum noise
measured considering INL and DNL of +0.034
V to -0.001 V and +0.06 V to -0.05 V
respectively. Thermometric decoder and shift
registers based SAR logic is used along with
and FSM based approach which reduces glitch
in the SAR output. The sub blocks are
integrated to achieve SAR ADC and its
performance is analysed considering noise
sources and harmonics. From the results
obtained the ADC structure is able to work at
1GSps with effective conversion metrics
suitable for high speed and low power
applications.
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.1
Vasudeva G., Uma B. V.
E-ISSN: 2224-2856
14
Volume 17, 2022
References:
[1] Aili Wangand C.-J. Richard Shi, “A 10-bit
50-MS/s SAR ADC with 1 fJ/Conversion in
14 nm SOI FinFET CMOS,” Integration, the
VLSI journal, vol.62,pp.246-257, June 2018.
[2] L. Kull et al., "A 24–72-GS/s 8-b Time-
Interleaved SAR ADC With 2.0–3.3-
pJ/Conversion and >30 dB SNDR at Nyquist
in 14-nm CMOS FinFET," in IEEE Journal
of Solid-State Circuits, vol. 53, no. 12, pp.
3508-3516, December. 2018
[3] J. Hudner et al., "A 112Gb/s PAM4 Wireline
Receiver Using a 64-Way Time-Interleaved
SAR ADC in 16nm FinFET," IEEE
Symposium on VLSI Circuits, pp. 47-48,
2018.
[4] S. Hsieh and C. Hsieh, "A 0.44-
fJ/Conversion-Step 11-Bit 600-kS/s SAR
ADC With Semi-Resting DAC," IEEE
Journal of Solid-State Circuits, vol. 53, no. 9,
pp. 2595-2603, September. 2018
[5] K. Okuno, K. Obata, T. Kato and K.
Sushihara, "An 800-MHz 8-bit High Speed
SAR ADC in 16nm FinFET process," IEEE
International Meeting for Future of Electron
Devices, Kansai (IMFEDK), pp. 24-25,2017.
[6] A. Joshi, S. K. Sharma, S. K. Manhas and S.
Dasgupta, "Design and Analysis of Low
Power and Area Efficient Single Capacitor
DAC Based Successive Approximation ADC
Using 45 nm FinFET," Fifth International
Conference on Communication Systems and
Network Technologies, pp. 792-796,2015.
[7] Ewout Martens, Benjamin Hershberg, and
Jan Craninckx, “A 69-dBSNDR 300-MS/s
Two-Time Interleaved Pipelined SAR ADC
in 16-nm CMOS FinFET. With Capacitive
Reference Stabilization,” IEEE Journal of
Solid-State Circuits, vol.53, no.4, April 2018.
[8] Zhiliang Zheng, Un-Ku Moon, Jesper
Steensgaard, Bo Wang, and Gabor C. Temes
, “Capacitor Mismatch Error Cancellation
Technique for a Successive Approximation
A/D Converter,” IEEE Journal of Solid-State
Circuits, vol.33, pp. 326-329, July 1999.
[9] Gilbert Promitzer, “12-bit Low-Power Fully
Differential Switched Capacitor
Noncalibrating Successive Approximation
ADC with 1 MS/s,” IEEE Journal of Solid-
State Circuits, vol.36,no.7, pp. 1138-1143,
July 2001.
[10]Eric Fogleman, Jared Welz and Ian Galton,
“An Audio ADC Delta–Sigma Modulator
with 100-dB Peak SINAD and 102-dB DR
Using a Second-Order Mismatch-Shaping
DAC,” IEEE Journal of Solid-State Circuits,
vol. 36, no. 3, pp.339-348, March 2001.
[11]John McNeill, Michael Coln and Brian
Larivee, “A Split ADC architecture for
deterministic digital background calibration
of a 16-bit 1-MS/s ADC,” IEEE Journal of
Solid-State Circuits, vol. 40, no. 12, pp.2437-
2445, December 2005.
[12]Mi-rim Kim , Young-Ouk Kim, Yong-Sik
kwak and Gil-Cho Ahn, “ A 12-bit 200-KS/s
SAR ADC with Hybrid RC DAC,” IEEE
Conference 978-1-4799-5230-4/14,pp.185-
188,2014.
[13]Dragisa Milovanovic,Milan Savic and Miljan
Nikolic, “Second-Order Sigma-Delta
Modulator in Standard CMOS Technology,”
Serbian Journal of Electrical
Engineering,vol.1,no.3,pp.37- 4,November
2004.
[14]Oguz Altun, Jinseok Koh, Phillip E.Allen “
A 1.5V Multirate Multibit ΣΔ Modulator For
GSM/WCDMA In a 90nm Digital CMOS
Process,” IEEE Conference pp.5577-
5580,2005.
[15]Victor Aberg “Design of 28 nm FD-SOI
CMOS 800 MS/s SAR ADC for wireless
applications,” Master’s thesis in Embedded
Electronic System Design, Chalmers
University of Technology, University of
Gothenberg,Sweden,2016.
[16]Junjie Wu and Jianhui Wu, “A 12-Bit 200
MS/s Pipelined-SAR ADC Using Back-
Ground Calibration for Inter-Stage Gain,”
Electronics 2020, 9, 507
[17]A. Amara, O. Rozeau and Editors “Planar
Double-Gate Transistor: From Technology to
Circuit,” Springer,pp.1-20.
[18]Vasudeva G and Uma B V, “Design of OTA
based Comparator for High Frequency
Applications using 22nm FINFET
Technology,” IJECE Journal,vol.12, no.2,
April 2022.
[19]Marcel. J. M. Pelgrom, A. C. J. Duinmaijer
and A. P.G. Welbers, “Matching properties
of MOS transistors,” IEEE Journal of Solid
State Circuits, vol. 24, no. 5, pp. 1433-1440,
October 1989.
[20]Vasudeva G and Uma B V, “Two-Stage
Folded Resistive String 12-Bit Digital to
Analog Converter using 22nm FinFET,”
ICSCN Conference, Springer 2021.
Creative Commons Attribution License 4.0
(Attribution 4.0 International, CC BY 4.0)
This article is published under the terms of the Creative
Commons Attribution License 4.0
https://creativecommons.org/licenses/by/4.0/deed.en_US
WSEAS TRANSACTIONS on SYSTEMS and CONTROL
DOI: 10.37394/23203.2022.17.1
Vasudeva G., Uma B. V.
E-ISSN: 2224-2856
15
Volume 17, 2022
