A_Comprehensive_Analysis_on_Comparator_Encoder_Circuit_to_Develop_Flash_Type_ADC_Based_on_Different_Implementation_Routes
A_Comprehensive_Analysis_on_Comparator_Encoder_Circuit_to_Develop_Flash_Type_ADC_Based_on_Different_Implementation_Routes
Abstract— The present paper describes the design and than others, although optimization of different comparator
performance evaluation of two different flash-type analog-to- and encoder circuits can reduce the power dissipation.
digital converter (ADC) circuits, targeting applications towards Analog-to-digital converters (ADCs) are a crucial part of
high speed and low power consumption. The first considered
analog-to-digital communication blocks in addition to digital
flash-type ADC circuit comprised of a sample and hold circuit
signal processing systems [5]. The flash type ADC is highly
2024 IEEE Silchar Subsection Conference (SILCON) | 979-8-3315-4082-1/24/$31.00 ©2024 IEEE | DOI: 10.1109/SILCON63976.2024.10910918
Authorized licensed use limited to: Green University of Bangladesh. Downloaded on May 11,2025 at 06:26:48 UTC from IEEE Xplore. Restrictions apply.
Fig. 1. Block diagram of flash-type analog-to-digital converter (ADC) circuit
Authorized licensed use limited to: Green University of Bangladesh. Downloaded on May 11,2025 at 06:26:48 UTC from IEEE Xplore. Restrictions apply.
Fig. 3. Schematic circuit diagram of the clamped push-pull output comparator made in cadence virtuoso (Case 1)
For generating the input signals in the form of pulse voltage was designed using CMOS technology varying the W/L ratio
source (means logic high and logic low) for the 8:3-bit of every PMOS and NMOS as needed, the aspect ratio of the
priority encoder circuit, seven comparator blocks were PMOS (M3 and M4) is 4.5µm/1µm, while the W/L ratio of
joined parallelly from which the output has been generated in the M6 PMOS is 38µm/1µm, and for PMOS M8 it is
the manner of pulse voltage source and this developed 4.5µm/1µm. On the other hand, the aspect ratio of the NMOS
output was worked as the input of the priority encoder circuit (M1 and M2) is 3µm/1µm, while for NMOS M5 and M9, it
(port D1 to D7, as mentioned in the priority encoder circuit is 4.5µm/1µm, and for NMOS M7 it is 35µm/1µm. The DC
part), which in turn generates binary bits from the encoder. current source used in the circuit is 10µA [12-14]. The second
In this paper, two types of comparator circuits have been comparator is a two-stage, open-loop comparator, as shown
made and analyzed in cadence virtuoso to develop the whole in Fig. 4, whose design is mostly similar to the two-stage,
flash-type ADC architecture. The clamped push-pull operational amplifier. It was designed using CMOS
comparator offers high speed and strong drive, but consumes technology varying the aspect ratio of every PMOS and
more power. On the other hand, the two-stage open-loop NMOS as needed, the W/L ratio of the PMOS (M3 and M4)
comparator provides better gain and noise immunity, is 4.5µm/1µm, while the W/L ratio of the M6 PMOS is
operates at low power, and may be slower. Both the types 38µm/1µm. On the other hand, the aspect ratio of the NMOS
were chosen here for their balance of speed, power (M1 and M2) is 3µm/1µm, while for NMOS M5 and M8, it
efficiency, and simplicity, making them suitable for high- is 4.5µm/1µm, and for NMOS M7 it is 35µm/1µm. The DC
speed, high-resolution Flash ADCs. The first comparator is current source used in the circuit is 30µA [15-16].
a clamped push-pull output comparator shown in Fig. 3. It
Fig. 4. Schematic circuit diagram of the two-stage open-loop comparator made in cadence virtuoso (Case 2)
Authorized licensed use limited to: Green University of Bangladesh. Downloaded on May 11,2025 at 06:26:48 UTC from IEEE Xplore. Restrictions apply.
For generating the binary bits (A, B and C) corresponding to A = D4+D5+D6+D7
the input analog signal for the flash-type ADC block, an
B = (D2+D3) D4' D5'+D6+D7
encoder block is needed, which converts the comparator-
generated pulse signal into the corresponding binary number. C = [D1D2' D3' D4'+D3D4'+D5] D6'+D7
In this study, priority encoder circuit have been made and V = D0+D1+D2+D3+D4+D5+D6+D7
analysed in cadence virtuoso to develop the whole flash-type
ADC architecture. expressed above by which the thermometer code is
An 8-bit to 3-bit priority encoder was made using basic logic transferred into binary numbers. In a priority encoder, if more
gates within the cadence virtuoso environment, shown in Fig. than one input is active, the output will represent the highest-
5. The whole design process involves a meticulous process numbered active input. For instance, if inputs D7 to D0 are
of circuit design and simulation. In digital systems, the present, and both D2 and D1 are active, the output will
priority encoder outputs the binary code of the highest represent D2, as it has a higher priority [9]. The logic for this
priority input line that is currently in use. An 8-to-3-bit encoder can be expressed using boolean algebra, and the
priority encoder translates eight input lines into three output corresponding logic diagram can be meticulously crafted
lines, with the output line representing the input line with the using the gates mentioned. To confirm the circuit's operation,
highest order of precedence. With cadence virtuoso, the simulations were done. With the tools provided by cadence
design process began with the schematic capture, in which virtuoso, designers can test their circuits under a variety of
the encoder circuit was formed by connecting logic gates. scenarios and make sure they function as intended. The
The three fundamental gates that make up a digital circuit are outcomes of the simulation aid in verifying the priority
AND, OR, and NOT gates. To make sure that in the priority encoder's truth table, guaranteeing that the highest priority
encoder, while many inputs are active, only the input with the input is accurately encoded into a 3-bit binary output.
highest priority is taken into the account, multi-input AND In Fig. 6 the working principle of a flash-type analog-to-
gates were employed. NOT gates were used to invert signals digital converter has been demonstrated without considering
when needed, OR gates combined the outputs of the AND the type of comparator and encoder used in the circuit to
gates. Mathematical expressions for every binary bit are check the results at a specific time instant, when the signal
Fig. 5. Schematic circuit diagram of the 8:3-bit priority encoder made using logic gates in cadence virtuoso
Authorized licensed use limited to: Green University of Bangladesh. Downloaded on May 11,2025 at 06:26:48 UTC from IEEE Xplore. Restrictions apply.
voltage is 0.8 times of the reference voltage, which is a is preferable compared to the first one. Noise margin is also
constant DC voltage. As, for every type of comparator block an important factor in analyzing the performance of any
like two-stage open-loop comparator or clamped push-pull circuit. It depicts how much the circuit is perfect to
output comparator and for every priority encoder made by demonstrate the output from the input signal in the
any means, the working principle of the flash-type ADC interference of noise.
remains the same. Now, the analog input signal having III. RESULT AND DISCUSSION
amplitude 2.5V (full swing is 5V) and 500Hz frequency has
passed through the sample and hold circuit. The output The power consumption of the flash-type ADC (case 1)
waveform from the sample and hold circuit is gone into one and flash-type ADC (case 2) obtained are as follows:
of the inputports of every comparator joining in parallel and 31.47mW and 31.53mW respectively. The power
the other input port is filled with the reference voltage, which consumption of the flash ADCs has been calculated in the
is divided at every comparator using a certain value of resistor Cadence Virtuoso platform using the calculator tool from the
by the voltage divider rule. Applied 5V reference voltage is power graph by using the average function from the function
divided at the ports a, b, c, d, e, f, and g. At port a, the voltage panel in the tool, of that particular circuit. The propagation
is 7/8Vref, at port b, that is 6/8Vref. In that way, at port g, the delay also has been calculated using the delay function in the
voltage is 1/8Vref. So, all the comparator gives the output as cadence virtuoso platform in the calculator tool for every
logic low (0) except the topmost comparator in Fig. 6. It gives flash-ADC circuit. Noise analysis also has been done in the
high logic (1). Now, the generated thermometer code by the cadence platform using the noise analysis tool. Flash ADCs
comparator block generates the binary bits after the priority are extremely fast but come with limitations such as high-
encoder block. A thermometer code is a unary code for power consumption, exponential circuit complexity, and
integers in which each successive digit is a 1. Thus, the limited resolution (typically up to 8 bits). Even, comparator
number 3's thermometer code is “1110”. It is often used in mismatch and high input capacitance also affect accuracy and
applications involving gradual changes in value. The output speed.
port V is the indication bit that is if anyone of the input ports A. Flash-type ADC-I (Case-1)
of the encoder block is at logic high, then only the outport V
In this type of flash analog-to-digital converter, a sample and
gives logic high value, otherwise it gives logic low value [17].
hold circuit, a clamped push-pull output comparator and a
Propagation delay, power consumption and noise analysis
priority encoder made using basic logic gates like multi-input
have been done for the above-mentioned two types of flash
AND, OR and NOT gates have been used to design. In Fig.
ADC circuits to analyse the performances of these circuits.
7, the output binary bits (A, B and C) against the sinusoidal
Propagation delay = ½ {(propagation delay for the signal
input signal having 2.5V amplitude and 500Hz frequency
goes high to low) + (propagation delay for the signal goes low
have been shown. The analog-to-digital converter (Case-1)
to high)}. Power consumption is also one of the most
has been simulated for 20ms. There 2 binary stages or
important parameters to judge any circuit based on its
numbers have been detected i.e. 000 and 111. For high values
performance. If any circuit consumes more power than the
of analog signal, it gave the value 111, and for low values it
other circuit of the same functionality, then the second circuit
gave 000.
Authorized licensed use limited to: Green University of Bangladesh. Downloaded on May 11,2025 at 06:26:48 UTC from IEEE Xplore. Restrictions apply.
B. Flash-type ADC-II (Case-2) power consumption was carried out and found to be 31.47mW
In this type of flash analog-to-digital converter, a sample and for case 1, while in case 2 it was 31.53mW. The noise analysis
hold circuit, a two-stage, open-loop comparator and an 8:3-bit of the circuits has also been performed and the respective
priority encoder made up using basic logic gates like multi- parameters have been computed.
input AND, OR and NOT gates, have been used to design. In
Fig. 8, the output binary bits (A, B and C) against the ACKNOWLEDGMENT
sinusoidal input signal having 2.5V amplitude and 500Hz We extend our appreciation to the Institute of
frequency have been shown. The analog-to-digital converter Engineering and Management (IEM), Kolkata for providing a
(Case-2) has been simulated for 20ms. There 2 binary stages conducive environment for academic exploration and
or numbers have been detected 001 and 110. For high values fostering an atmosphere of learning.
of analog signal, it gave the value as a pattern of 110 and for
low values, it gave 001 pattern. REFERENCES
[1] A. Kumary and S. Rao, “Design Techniques of Flash ADC: Review”,
In Table I, we have analyzed the power consumption for the Advances in Communication, Signal Processing, VLSI, and Embedded
two flash-type ADC circuits. The propagation delay and noise Systems, vol. 614, pp. 123-129, 2019.
[2] L. Mendonca and R. Bhandarkar, “Implementation of SAR ADC for
power spectral density (p.s.d) for each output line (A, B, and Biomedical Applications—A Review”, Advances in Communication,
C) of both flash ADC circuits have been analyzed for input Signal Processing, VLSI, and Embedded Systems, vol. 614, pp. 181-
189, 2019.
frequency of 500Hz. Propagation delay of flash ADC-II is [3] K. Khatun, V. R. Vakacharla, A. R. Kizhakkan and A. K. Rathore,
283.5 , which is greater than flash ADC-I. Basically, “Small-Signal Analysis and Control of Snubberless Naturally Clamped
Soft-Switching Current-Fed Push–Pull DC/DC Converter”, IEEE
propagation delay denotes how quickly the output responds to Transactions on Industry Applications, vol. 56, pp. 4299-4308, 2020.
the input changes. Here, in between two different [4] D. Chen, X. Cui, Q. Zhang, D. Li, W. Cheng, C. Fei, Y. Yang, “A
Survey on Analog-to-Digital Converter Integrated Circuits for
comparators, as we used the two-stage open loop comparator Miniaturized High Resolution Ultrasonic Imaging System”,
Micromachines, vol. 13, no. 114, 2022.
of flash ADC in case 2, thus the propagation delay is higher [5] S. S. Kumar, V. Goel, D. Prasad, and V. Nath, “A 0.48 mW High
there as compared to the other [15]. The noise analysis of the Performance 4-Bit Flash ADC for System-on-Chip Applications in 90
nm CMOS Technology”, Proceedings of the Fourth International
circuits has also been shown in Table I. Moreover, flash Conference on Microelectronics, Computing and Communication
ADCs are essential for real-time, high-speed applications like Systems (MCCS), vol. 673, pp. 939-957, 2019.
[6] A. Esmailiyan, F. Schembari, R.B Staszewski, “A 0.36-V 5-MS/s
radar systems, high-speed communication (e.g., 5G), Time-Mode Flash ADC With Dickson-Charge-Pump-Based
oscilloscopes, medical imaging (MRI, CT scans), and Comparators in 28-nm CMOS”, IEEE Transactions on Circuits and
Systems–I, vol. 67, no. 6, 2020.
LIDAR. They mostly excel where fast data acquisition and [7] A. Soumya, B. S. Sucharita, and R. Alekya, “Analysis of an Efficient
low-latency processing are critical, but they are less suitable Modular Adder Designs Based on Thermometer & One-Hot
Encoding”, Alochana Chakra Journal, vol. 9, Issue. 4, pp. 3679-3683,
for high-resolution or low-power applications. 2020.
[8] M. K. Ray, S.K. Surshetty, V. Goel, D. Prasad, and V. Nath, “Design
TABLE I. COMPARATIVE ANALYSIS OVER TWO TYPES OF of 4-Bit Multiplexer-Based Encoder for Analog to Digital Converter”,
Proceedings of the Fourth International Conference on
FLASH ADC CIRCUITS Microelectronics, Computing and Communication Systems (MCCS),
vol. 673, pp. 959-966, 2019.
Parameters Flash ADC-I (Cs.-1) Flash ADC-II (Cs.-2) [9] K. Datta, D. Prasad, S. K. Surshetty, V. Goel, and V. Nath, “Design of
a 4-Bit Wallace Tree Encoder for Flash ADC in 90 nm CMOS
Technology”, Proceedings of the Fourth International Conference on
Power 31.47mW 31.53mW Microelectronics, Computing and Communication Systems (MCCS),
vol. 673, pp. 967-974, 2019.
Delay-A 108.7 s 283.5 s [10] I. Maity, A. Majumdar, “Computation of key analog circuit parameters
employing different topological configuration of current mirror circuits:
Delay-B 108.7 s 283.5 s A comprehensive analysis of gain, noise, and power consumption”, RP
Current Trends in Engineering and Technology, vol. 3, pp. 23–29,
2024.
Delay-C 108.7 s 283.5 s [11] D.R. Oh, J.I. Kim, D.S. Jo, W.C. Kim, D.J. Chang, “A 65-nm CMOS
6- bit 2.5-GS/s 7.5-mW 8× Time-Domain Interpolating Flash ADC
Noise p.s.d-A 8.311 nV/sqrt (Hz) 14.18 nV/sqrt (Hz) With Sequential Slope Matching Offset Calibration”, IEEE Journal of
Solid-State Circuits, vol. 54, 2019.
Noise p.s.d-B 14.18 nV/sqrt (Hz) 14.18 nV/sqrt (Hz) [12] S. Chinthala and M. A. B. M, “High Throughput Circuit Design of
Flash Type Analog to Digital Converter”, 5th International
Conference on Recent Advances in Information Technology (RAIT),
Noise p.s.d-C 8.311 nV/sqrt (Hz) 8.311 nV/sqrt (Hz) Dhanbad, India, pp. 1-6, 2023.
[13] I. Maity, “Cadence Virtuoso based circuit simulation of universal logic
gates: A board tutorial”, RP Current Trends in Engineering and
IV. CONCLUSION Technology, vol. 3, pp. 1–7, 2024.
[14] K. R. Scott and S. P. Khatri, “A Flash-based Digital to Analog
This paper has found out the optimized and best structure of Converter for Low Power Applications”, IEEE 40th International
Conference on Computer Design (ICCD), Olympic Valley, CA, USA,
a flash-type ADC concerning the power consumption, pp. 1-8, 2022.
propagation delay and noise margin with the frequency of [15] I. S. A. Halim, S. L. M. Hassan, N. D. b. M. Akbar, and A. A. A.
Rahim, “Comparative study of comparator and encoder in a 4-bit Flash
analog input signal by interchanging different components ADC using 0.18μm CMOS technology”, International Symposium on
like the clamped push-pull output comparator, two-stage Computer Applications and Industrial Electronics (ISCAIE), Kota
Kinabalu, Malaysia, pp. 35-38, 2012.
open-loop comparator and 8:3- bit priority encoder blocks [16] Y. Zheng, Y. Zhao, N. Zhou, H. Wang and D. Jiang, “A short review
made using basic logic gates like AND, OR, NOT gates in of some analog-to-digital converters resolution enhancement
methods”, Measurement, vol. 180, 2021.
two different cases of flash-type ADC circuits. The first [17] P. J. Lim, and B. A. Wooley, “A High-speed Sample-and-Hold
considered flash-type ADC circuit was made up of a track and TechniqueUsing a Miller Hold Capacitance”, IEEE Journal of Solid-
State Circuits, vol. 26, pp. 643-651, 1991.
hold circuit, a clamped push-pull output comparator and an [18] S. B. Mashhadi, and R. Lotfi, “Analysis and Design of a Low-Voltage
8:3-bit priority encoder designed utilizing basic logic gates Low-Power Double-Tail Comparator”, IEEE Transactions on Very
Large Scale Integration (VLSI) Systems, vol. 22, pp. 343-352, 2014.
like AND gate, OR gate and NOT gate (Case 1). While, the
second flash-type ADC circuit was made up of open-loop
two-stage comparator and an 8:3-bit priority encoder circuit
(Case 2). The propagation delay with respect to the output line
A for case 1 flash ADC has come at a value of 108.7µs while
for case 2, it has come at a value of 283.5µs. Evaluation of
Authorized licensed use limited to: Green University of Bangladesh. Downloaded on May 11,2025 at 06:26:48 UTC from IEEE Xplore. Restrictions apply.