Low Voltage Low Power Folded Cascode OTA Design For RF Applications
Low Voltage Low Power Folded Cascode OTA Design For RF Applications
Low Voltage Low Power Folded Cascode OTA Design For RF Applications
4029-4034
© Research India Publications. http://www.ripublication.com
Low Voltage Low Power Folded Cascode OTA Design For RF Applications
Sawssen Lahiani, Houda Daoud, Samir Ben Selem and Mourad Loulou
4029
International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 13 (2017) pp. 4029-4034
© Research India Publications. http://www.ripublication.com
M7 M8
Where K is the Boltzmann’s constant and T is the
V2
temperature. gm7 is the transconductance of M7 transistor. The
input referred flicker noise voltage of the FC OTA can be
Vss
written as:
Figure 1: Folded cascode OTA circuit KF KF g m1,22 KF g m7,82
Vin,1/f 2 = 2 +2 +2
Cox WL 1,2 f Cox WL 7,8 f g 2 Cox WL 7,8 f g 2
m9,10 m9,10 (5)
The FC OTA has a differential stage consisting of PMOS
transistors M9 and M10. M11 and M12 devices provide
Where KF is the flicker noise coefficient of NMOS and PMOS
respectively the DC bias voltages to M1, M2, M3 and M4
transistors, f is the frequency and Cox is the oxide capacitance.
transistors.
W and L are the transistors sizes. gm defines the
The open-loop voltage gain and the dominant pole are given by
transconductance of transistors differential pair.
equations (1) and (2),
(1)
g m3 1 g 1 THE FOLDED CASCODE OTA OPTIMIZATION
A v g m9 ( . // m5 . ) In order to optimally size each component forming the OTA
g ds1 g ds3 g ds5 g ds7 g ds9
when satisfying constraints and performance functions, we
developed an algorithm that allows automating the task [11].
1 (2)
fdp This algorithm was programmed using C++ software. It is
g m3 1 1 g m5
2π( . // . )C based on the flow chart given by figure 2. It describes an
gds1 gds3 gds5 gds7 gds9 L
optimization tool based on Heuristic algorithm. The first step
in the optimization is the expression of the different criteria by
a technology dependent model. For accurate modeling, a small
Where, gds1, gds3, gds5, gds7 and gds9 are respectively the signal analysis of the OTA is carried out to explicit the
conductances of M1, M3, M5, M7 and M9 transistors. gm3, gm5 different characteristics intended optimized. From the OTA
and gm9 are respectively the transconductances of M3, M5 and structure presented in Figure 2, we opted in this step for the
M9 devices and CL is the capacitance at the output node. TSMC CMOS 0.18μm process.
4030
International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 13 (2017) pp. 4029-4034
© Research India Publications. http://www.ripublication.com
B. Folded Cascade OTA Optimization The proposed OT A cell is simulated using the Advanced
Design System (ADS) tool with TSMC 0.18µm CMOS
We used in the OTA optimization the following series of process parameters under 1.8V power supply. The NMOS and
criteria: the PMOS transistor threshold voltages are respectively 0.436
The gain Av is maximized. V and -0.438 V. The optimized value of bias current is set to
The dominant frequency (f-3db) is maximized. 50 μA. Current sources are implemented using cascode current
mirrors.
The input referred noise (Vin,th) is minimized.
The power consumption (power) is minimized.
OTA CIRCUIT SIMULATIONS
The silicon area is minimized.
The designed folded cascode OTA has a gain of 73.57dB, a
The objective function to maximize can thus be formulated as
large unity-gain frequency of 2.3GHz (Fig.3). Fig.4 presents
follows:
the phase margin of 52degrees.
a3 a4 a5
F = a1A v + a 2f-3dB + + +
Vin,th 2 power Wi Li (6)
m3
freq=2.282GHz
Where a1-5 are the positive coefficients used for normalization. dB((Vout1-Vout2)/(Vin1-Vin2))=0.003
dB((Vout1-Vout2)/(Vin1-Vin2))
Initialisation of the 60
parameters’
vectors 40
20
Random choice of
m3
Verification of the 0
preliminary the vector
conditions variables
-20
-40
No -60
Calculs of the
Correction objective function 1E5 1E6 1E7 1E8 1E9 1E10
freq, Hz
Yes
No
ADS simulation m4
freq=2.272GHz
Yes phase((Vout1-Vout2)/(Vin1-Vin2))=-128.0 deg
phase((Vout1-Vout2)/(Vin1-Vin2)), deg
0
Optimazed
structure
-50
4031
International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 13 (2017) pp. 4029-4034
© Research India Publications. http://www.ripublication.com
80
Supply Voltage 2 1.8 -- 1.8
70
Power consumption -- 0.45 1.2 0.5
60
(mW)
Vin1.noise, nV
50
CMOS Process (µm) 0.35 0.18 0.18 0.18
40
30 m1 FOM -- -- -- 2052
freq=2.282GHz (dB.MHz/V.mW)
20 Vin1.noise=2.611nV
10
m1
0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
APPLICATION
freq, GHz
The overall performance of a wireless device greatly depends
Figure 5: Input-referred noise curve on the performance of its transceiver [16-19]. RF filters are
the essential components of all wireless transceiver front ends
as shown in fig. 7. To design band pass filter with high quality
160
factor, inductors play an important role [20]. Higher cost that
140
came from large size of silicon area to entertain on-chip
120
passive spiral inductors is one of the prime disadvantages.
Vout1.noise, nV
100
Another disadvantage of on-chip spiral inductor is its
80 deficiency of tuning ability which makes the design a little bit
60 complicated [21]. As a result the usage of on-chip spiral
m2
40 freq=2.282GHz inductors is decreasing day by day for high frequency
Vout1.noise=6.234nV
20 applications [22].
m2
0
0.0 0.5 1.0 1.5 2.0 2.5 3.0
freq, GHz
4032
International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 13 (2017) pp. 4029-4034
© Research India Publications. http://www.ripublication.com
operational frequencies, which restricts this type of filter from this large cutoff frequency, the Gm-C filter can operate in RF
being used in gigahertz frequencies. application and modern communication receivers. This filter
is able to re-configure the bandwidth, gain or linearity in order
In our case and to demonstrate the feasibility and the
to fulfill the requirements of different standards subjecting to
performance of folded cascode OTA, we simulated several
wireless standards.
applications. We have treated current mode and voltage mode
filters. Recently, analog filters design using Gm-C integrators
10
has acquired a great popularity. Tranconductance cells are
dB((Vout1-Vout2)/(Vin1-Vin2))
relatively simple circuits which allow to operate for high 0 m1
frequencies. -10
m1
freq=737.1MHz
The reason behind the selection of Gm-C concept is that the -20
dB((Vout1-Vout2)/(Vin1-Vin2))=-3.003
easy tuning capability by varying the Gm value of the
transconductors. Also the Gm-C filter has a low noise floor but -30
Vin+
+ +
gm1
+
gm2
+ +
gm3
+ + + CONCLUSION
gm4
Vin-
- - - - - - - - In this paper, a design of folded cascode OTA circuit has been
C1 C2 Vout- reported. The approach of optimization OTA design has been
used to improve the gain, the unity gain frequency, the input
referred noise and the power consumption. Simulation results
Figure 8: 2nd order Gm-C low pass filter circuit prove the effectiveness of this optimization approach which is
a time consuming method. The optimized topology achieves a
The basic Gm-C filter, consisting of a capacitor and a good input range with a high DC gain of 73.57dB and a large
transconductance, is shown in Fig. 8 and the transfer function bandwidth. It consumes only 0.5 mW under 1.8V supply
of is given by [13]: voltage. Based on this circuit, we have implemented a Gm-C
low-pass filter where the cutoff frequency is 737MHz.
g m2 .g m3.g m4
Vout g m4
H(S ) = REFERENCES
Vin 2 g m2 g .g (7)
S + S + m3 m4
C1 C1C2 [1] Mariano Jime´ nez-Fuentes et al, "A tunable highly
linear CMOS transconductor with 80 dB of SFDR",
The common transfer function of 2nd order low pass filter INTEGRATION, the VLSI journal vol.42, 2009, pp.
(LPF) is 277–285.
[2] M. Ismail and T. Fiez, Analog VLSI Signal and
0 2 Information Processing. New York: McGraw-Hill,
H(S ) = (8)
0
S2 + S + 0 2 1994.
Q
[3] S. R. Zarabadi, M. Ismail, and C. C. Hung, “High
performance analog VLSI computational circuits,”
The corner frequency ω0 and the quality factor Q of the IEEE J. Solid-State Circuits, vol. 33, no. 4, pp. 644–
circuit can obtain as: 649, Apr. 1998.
[4] C. C. Hung, K. A. Halonen, M. Ismail, V. Porra, and
g m3.g m4 0 g (9)
02 m2 A. Hyogo, “A low-voltage, low-power CMOS fifth-
C1C2 Q C1
order elliptic GM-C filter for baseband mobile, wireless
communication,” IEEE Trans. Circuits Syst. Video
The simulation’s result of Gm-C low pass filter is shown in Technol., vol. 7, no. 4, pp. 584–593, Aug. 1997.
fig. 9. With the following passive elements C1=0.5pF and
C2=1pF, the obtained cutoff frequency is 737MHz. Thanks to
4033
International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 13 (2017) pp. 4029-4034
© Research India Publications. http://www.ripublication.com
[5] L T. Y. Lo and C. C. Hung, “A wide tuning range G -C 0.18 μm CMOS Technology,” Modern Applied
continuous time analog filter,” IEEE Trans. Circuits Science, vol. 7, p. p70, 2013.
Syst. I, Reg. Papers, vol. 54, no. 4, pp. 713–722, Apr. [17] F. B. Arith, M. Mamun, M. A. S. Bhuiyan, and A. A.
2007. A. Bakar, “Low Voltage Schmitt Trigger In 0.18 μm
[6] J. Galan, R. G. Carvajal, A. Torralba, F. Munoz, and J. CMOS Technology,” Advances in Natural and Applied
Ramirez- Angulo, “A low-power low-voltage OTA-C Sciences, vol. 7, pp. 33- 38, 2013.
sinusoidal oscillator with a large tuning range,” IEEE [18] J. Uddin, M. B. I. Reaz, M. A. Hasan, A. N. Nordin,
Trans. Circuits Syst. I, Reg. Papers, vol. 52, no. 2, pp. M. I. Ibrahimy, and M. A. M. Ali, “UHF RFID antenna
283–291, Feb. 2005. architectures and applications,” Scientific Research and
[7] J. van Engelen, R. van de Plassche, E. Stikvoort, and Essays, vol. 5, pp. 1033-1051, 2010.
A. Venes, “A sixth-order continuous-time bandpass [19] K. A. Rosli, R. M. N. H. R. Daud, M. Mamun, and M.
sigma–delta modulator for digital radio IF,” IEEE A. S. Bhuiyan, “A Comparative Study on SOI
Solid-State Circuits, vol. 34, no. 12, pp. 1753–1764, MOSFETs for Low Power Applications,” Research
Dec. 1999 Journal of Applied Sciences, Engineering and
[8] Voorman, H., Veenstra, H. “ Tunable high-frequency Technology, vol. 5, pp. 2586-2591, 2013.
Gm-C Filters,”IEEE Journal of Solid-State Circuits [20] J. Jalil, M. Bin Ibne Reaz, and M. A. M. Ali, “CMOS
35(8), 1097–1108, 2000. Differential Ring Oscillators: Review of the
[9] Kar, S., Sen, S. “Tunable square-wave generator for Performance of CMOS ROs in Communication
integrated sensor applications,” IEEE Transactions on Systems,” IEEE Microwave Magazine, vol. 14, pp. 97-
Instrumentation and Measurement 60(10), 3369–3375, 109, 2013.
2011. [21] W. Yixiao, Y. Le, L. Huailin, and H. Ru, “Cost-
[10] Roohollah Nakhaei, Alireza Mallahzadeh, “ A efficient CMOS RF tunable bandpass filter with active
Simulation-Based Optimization Of Operational inductor-less biquads,” Proceeding of IEEE
Transconductance Amplifier Using An Evolutionary International Symposium on Circuits and Systems,
Algorithm,” in Australian Journal of Basic and Applied 2012, pp. 2139-2142.
Sciences, 6(5): 88-96, 2012. [22] H. Aljarajreh, M. Rozi Rifin, M. Mamun, S. Amin, and
[11] M. Fakhfakh, M. Loulou, and N. Masmoudi, “An W. M. D. W. Zaki, “Low Noise Amplifiers in RFID
improved algorithm-driven methodology to optimize Systems,” Journal of Engineering and Applied
switched current memory cells by transistor sizing,” in Sciences, vol. 8, pp. 21-28, 2013.
The IEEE International Conference on Electrical, [23] Stehr U, Henkel F, DalligeL Waldow P,” A fully
Electronic and Computer Engineering (ICEEC’04), differential CMS integrated 4th order reconfigurable
Sept. 2004. Cairo, Egypt, 2004. GM-C low pass filter for mobile communication”, in
[12] E.Kargaran, M.Sawan, Kh.Mafinezhad, H.Nabovati,” IEE procInt on Electronics Circuits &
Design of 0.4V,386nW OTA Using DTMOS Systems,1(Sharjah, UAE) 2003, pp 144-147.
Technique for Biomedical Applications,” in the
proceeding of IEEE 55th International Midwest
Symposium on Circuits and Systems
(MWSCAS),USA, 2012
[13] Daoud.H, Salem.S.B, Zouari.S, Loulou, “Folded
Cascode OTA Design for Wide Band Applications”,
International Conference on Design and Test of
Integrated Systems in Nanoscale Technology, pp437-
440, 2010.
[14] Kar S, and Sen S, “Linearity improvement of source
degenerated transconductance Amplifiers,” Analog
Integr Circ Sig Process, pp. 399- 407, 2013.
[15] Roohollah Nakhaei, Alireza Mallahzadeh, “ A
Simulation-Based Optimization Of Operational
Transconductance Amplifier Using An Evolutionary
Algorithm,” in Australian Journal of Basic and Applied
Sciences, 6(5): 88-96, 2012.
[16] F. I. B. A. Aziz, M. Mamun, M. A. S. Bhuiyan, and A.
A. A. Bakar, “A Low Drop-Out Voltage Regulator in
4034