A Novel Realization of Low-Power and Low-Distortion Multiplier Circuit with Improved Dynamic Range

THEORETICAL AND APPLIED ELECTRICAL ENGINEERING VOLUME: 15 |NUMBER: 5 |2017 |DECEMBER A Novel Realization of Low-Power and Low-Distortion Multiplier Circuit with Improved Dynamic Range Ali NADERI SAATLO1, Abolfazl AMIRI 2 1Department of Electrical-Electronics Engineering, Faculty of Engineering, Urmia Branch, Islamic Azad University, Urmia, Iran 2Department of Electronics Engineering, Faculty of Engineering, IAU University of Bushehr, Alishahr, Bushehr, Iran [email protected], [email protected] DOI: 10.15598/aeee.v15i5.2433 Abstract. A novel topology of four-quadrant analog multiplier circuit is presented in this paper. The voltage mode technique is employed to design the circuit in CMOS technology. The dynamic input and output ranges of the circuit are improved owing to the fact that the circuit works in the saturation region not in weak inversion. Also the proposed multiplier is suitable for low voltage operation and its power consumption is relatively low. In order to verify the performance of the proposed circuit, performance of the circuit affected by second order effects including transistor mismatch and mobility reduction is analyzed in detail. It will be shown that any conceivable mismatch in the transistor parameters leads to second harmonic distortion. Additionally, the effect of mobility reduction in the third harmonic distortion will be computed. In order to simulate the circuit, Cadence and HSPICE software are used with TSMC level 49 (BSIM3v3) parameters for 0.18 µm CMOS technology, where under supply voltage of 1.5 V, total power consumption is 44 µW, the corresponding average nonlinearity remains as low as 1 %, and the input range of the circuit is ±400 mV. Keywords CMOS design, four quadrant, low distortion, modulation, multiplier circuit. 1. Introduction In recent years, analog multipliers are widely used in many applications such as phase-locked loops, adaptive filters, modulators, automatic gain controlling, image processing, artificial neural networks and fuzzy integrated systems [1], [2], [3] and [4]. Different methods of implementation of this building block have been recently presented based on the use of bulk driven MOS [5], Floating Gate MOS (FGMOS) [6] and class-AB mode [7]. In the past decade, the demand for portable operation of electronic systems has led to the trend of designing circuits to be featured with low power consumption and operate for low supply voltages. One possible technique to design the low-power dissipation multiplier circuit is to use MOSFETs in sub-threshold region [8], [9] and [10] in which most of them follow the Gilbert cell topology and modified Gilbert cell [11]. The drawback of designs in this region has been referred to poor dynamic range, limited bandwidth and low voltage swing. Another approach of designing low power multiplier circuits is to use the translinear principle of MOS transistors operated in the weak inversion [12] and [13]. Although this approach has the advantage of low power consumption, the dynamic range of these circuits is very small and operation speed is slow. On the contrary, presented multipliers based on the translinear loop in saturation region exhibit wider bandwidth, higher dynamic range and lower distortion and thus they are more preferred than those operating in weak inversion [14]. Nonetheless, the channel length modulation and body effect are the important issues in the circuits based on translinear loop principle. Another salient feature of the circuits is the four-quadrant operation capability, an important asset very useful in various applications [15] and [16]. Some of the wellknown multiplier circuits operate only in one [17] and [18] or two quadrants [19] and [20], which was discussed in [21] and not suitable for many of mentioned applications. c 2017 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 780 THEORETICAL AND APPLIED ELECTRICAL ENGINEERING VOLUME: 15 |NUMBER: 5 |2017 |DECEMBER Another factor, which is important in the multiplier circuit, is non-linearity factor, because of the fact that the multiplication operator is a linear map between input and output. Therefore this factor is a serious challenge in the multiplier circuits, which is commonly affected by body effect, mobility reduction and mismatch in the circuit devices. In some existing analog multipliers, the effects of these non-idealities were properly studied and a few techniques were proposed in order to reduce the non-linearity [22] and [23]. However, they suffer from low accuracy and/or low bandwidth. Moreover, single supply voltage circuits are preferred to those in dual mode [24], where the multipliers reported in [8] and [20] require dual supply voltage. As such these circuits are not suitable for today’s world of portable equipment. In this paper, a novel design of four quadrant analog multiplier is presented which benefits from advantages of differential output topology. The dynamic input and output ranges of the circuit are significantly improved. High linearity and high accuracy are further advantages of the circuit. Also the proposed multiplier is suitable for low voltage operation and its power consumption is relatively low. The performance of the proposed multiplier is characterized using HSPICE with TSMC in 0.18 µm CMOS technology. The paper is organized in 5 sections: The proposed circuit is presented in Sec. 2. , followed by the performance analysis in Sec. 3. In Sec. 4. , HSPICE simulation results of proposed multiplier circuit are presented to prove the efficiency of the design. Finally, Sec. 5. concludes the most important achievement of the proposed circuit. 2. The Proposed Multiplier The proposed four-quadrant multiplier circuit is shown in Fig. 1, which is based on the square-difference algebraic identity as: (x+y)2−(x−y)2= 4xy. (1) According to this, to realize this equation, two squaring functions should be designed in which their outputs need to be subtracted. Let us consider the proposed circuit of Fig. 1. Assume that all of the transistors operate in saturation region (except for M17 and M18), thus the drain current of transistors by neglecting the second order effect such as mobility reduction and channel-length modulation can be expressed as: ID=K(VGS −VT)2,(2) where K= 0.5µ0COX (W/L)is related to transconductance parameter, VGS is gate-to-source voltage and VTrepresents the threshold voltage of MOS transistor which can be affected by body effect. The body effect refers to change in the transistor threshold voltage resulting from a voltage difference between the transistor source and substrate, which can be characterized by: VT=Vt0+γhpVSB +|2ϕF| − p|2ϕF|i,(3) where Vt0is the zero-bias threshold voltage, γis the body-effect coefficient and ϕF is the Fermi potential. Considering the figure, two squaring circuits are shown in left half and right half of the structure. Focusing on the left side squaring circuit, since transistors M1 and M2are biased in the saturation region and also ID1=ID2, the relationship can be written as: KN(Vin −V1−VTN)2=KN(V1−VTN)2.(4) Simplifying equation above we have: V1=2Vin1VTN −V2 in1 (−2Vin1+ 4VTN)=Vin1 2.(5) One can find the voltage of V2at the same way as: V2=−Vin1 2.(6) The voltages of V1and V2are utilized to turn on M9 and M10 transistors, respectively. In this case, their Vin1 _ Vin1 +Vin2 _Vin2 + Vout _ + M1 M2 M9M10 M13 M3 M4 M14 M16 M11 M12 M7 M5 M15 M6 M8 VDD V1V2V4 V3 VBp M17 M18 Fig. 1: The proposed four-quadrant analog multiplier circuit. c 2017 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 781 THEORETICAL AND APPLIED ELECTRICAL ENGINEERING VOLUME: 15 |NUMBER: 5 |2017 |DECEMBER currents are added together and flow to transistor M13: ID13 =ID9+ID10 = KP(VDD −V1− |VTP|)2+ (VDD −V2− |VTP|)2. (7) Replacing Eq. (5) and Eq. (6) in Eq. (7); after few mathematical manipulations we have: ID13 = 2KP"Vin1 22 + (VDD − |VTP|)2#.(8) It can be clearly seen from Eq. (8) that the current ID13 is the square of the input voltage plus some constant voltages. The same procedure can be followed for the right half of the circuit to obtain ID15: ID15 = 2KP"Vin2 22 + (VDD − |VTP|)2#.(9) The currents of ID13 and ID15 are transferred to the output through transistors M14 and M16, respectively. Transistors M17 and M18 are biased in the triode region (by setting VBp =−1) and perform as the resistors in which their resistance values can be represented by: Rn≈[µnCox(W/L)n(VSGn − |VTP|)]−1.(10) By setting R17 =R18 =R, the output voltage of proposed circuit can be derived as: Vout =R(ID14 −ID16) = = 2RKP"Vin1 22 −Vin2 22#.(11) According to Eq. (11), by establishing Vin1=Vx+Vy and Vin2=Vx−Vythe ultimate voltage is eventually what would be expected as follows: Vout = 2RKP(VxVy).(12) Take notice that summation of the signals is provided by series connection of the voltage sources (Vx and Vy). Also subtraction of the input signals was realized in the same way except for changing the polarity of Vy, which were performed using a well-known inverting amplifier. Also, there is no need subtraction at the output node, because the output is differential. 3. Performance Analysis In this section, performance of the circuit affected by second order effects including transistor mismatch and mobility reduction is analyzed in detail. It will be shown that any conceivable mismatch in the transistor parameters leads to second Harmonic Distortion (HD). Additionally, the effect of mobility reduction in the third harmonic distortion will be computed. Following that, the effect of corresponding parameters derived in each section as well as improvement methodology will be thoroughly discussed. 3.1. Second HD Due to the Mismatch In Sec. 2. , the well-matched parameters including trans-conductance and threshold voltage of the transistors were assumed to obtain the output of the circuit. Considering Eq. (5), due to the fact that the voltage of V1is resulted by supposing these matched parameters, any possible mismatch in the proposed circuit will affect the voltage of this node. Similarly, the voltages of V2,V3and V4get affected by the mismatch accordingly. Since these voltages have direct proportion to Vin1and Vin2, consequently the total mismatch is referred to the input signals and can be modeled as: V1=Vin1 2+ ∆vin1Vin1,(13) V2=−Vin1 2−∆vin1Vin1,(14) V3=Vin2 2+ ∆vin2Vin2,(15) V4=−Vin2 2−∆vin2Vin2,(16) where ∆vin1and ∆vin2are mismatch percentages of Vin1and Vin2, respectively. By applying Vin1=Vx+Vy and Vin2=Vx−Vyto the multiplier circuit, the output voltage is given by: Vout = 2RKPVxVy+ (2∆v2 in1)(Vx+Vy)2. . . . . . −(2∆v2 in2)(VxVy)2.(17) It can be clearly seen that the terms of ∆v2 in1and ∆v2 in2are very small (because ∆vin1and ∆vin2<1), therefore the resulted error will be negligible. It is worthwhile to calculate the harmonic distortion of the circuit at the output considering the method presented in [25], if one of the inputs (Vx) is kept constant and the other one is sinusoidal in the form of Vy=bvmsin t, second harmonic distortion can be derived as follows: HD2=∆v2 in1−∆v2 in2 2Vx(4∆v2 in1+ 4∆v2 in2+ 1)bvm.(18) The equation implies that when the mismatch percentage of ∆v2 in1and ∆v2 in2increases, second harmonic distortion decreases. Also, it decreases with decreasing Vx as well. c 2017 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 782 THEORETICAL AND APPLIED ELECTRICAL ENGINEERING VOLUME: 15 |NUMBER: 5 |2017 |DECEMBER 3.2. Effect of Mobility Reduction in Third HD If the mobility reduction is taken into calculations, the drain current of a MOS transistor operated in saturation is given by [26]: ID=K(VGS −VT)2 1 + θ(VGS −VT),(19) where θis the mobility degradation parameter which varies typically from 0.001 to 0.1 V−1. This equation may be expanded in a Taylor series: ID=K(VGS −VT)2· ·1−θ(VGS −VT) + θ2(VGS −VT)2+. . ..(20) To simplify the calculations, just the first order of θ is used, and the higher-order terms are ignored. Replacing the expansion in Eq. (4), one can reach V1and V2as: V1≈Vin1 2+θVin1(V2 in1−3Vin1VT P + 3V2 T P ) 4VT P ,(21) V2≈ −Vin1 2+θVin1(V2 in1+ 3Vin1VT P + 3V2 T P ) 4VT P .(22) The same procedure can be followed to obtain V3and V4. In this case, the output of the multiplier circuit can be represented as follows: Vout ≈RKPV2 in1−V2 in2+. . . . . . θVDD(2V3 in2−2V3 in1−3V2 in2VTP + 3Vin1VTP) VTP . (23) By applying Vin1=Vx+Vyand Vin2=Vx−Vy, the final output will be obtained. Since the output voltage includes third-order of the inputs, third harmonic distortion is achieved by keeping one of the inputs (Vx) as a constant and the other one as sinusoidal. Again using the method presented in [25] we have: HD3=θVDD 2VxVT P −4θ(3VTP −V2 x+ 2VxVTP)bv2 m.(24) 4. Post Layout Simulation Results In this section, simulation results are presented using HSPICE with TSMC level 49 (BSIM3v3) parameters for 0.18 µm CMOS technology so as to verify the performance of the proposed circuit. The simulation results are carried out after extracting the layout, which is drawn by Cadence software using single poly and two metals (Metal1 and Metal2). Figure 2 shows the full layout of the circuit, in which the area is 66.35 µm×58.2 µm. The aspect ratio of transistors is given in Tab. 1 and the supply voltage is 1.5 V. Considering the condition of triode region for PMOS transistors of M17 and M18, choosing VBp =−1V guaranties that these transistors operate in the triode region and work as the active resistances. DC transfer characteristic of the circuit over a considerable range of the inputs is shown in Fig. 3, in which one of the inputs (Vy) is kept constant and the other one (Vx) swept from −400 mV to +400 mV. By changing the constant voltage of Vyand then sweeping of Vx, desired outputs will be obtained. Within this range, the average of measured nonlinearity error is 0.94 %. Tab. 1: Transistor aspect ratios. Transistor name W/L (µm/µm) M1-M810/0.18 M9-M12 12/0.18 M13-M16 4/0.18 M17-M18 15/0.18 Fig. 2: Layout of the proposed multiplier circuit. 0 50 100 150 200 250 -250 -200 -150 -100 -50 0-100-200-300-400 100 200 300 400 Vout (mV) Vx (mV) Vy (mV) 400 -400 0 Fig. 3: Simulation result for DC transfer characteristic. c 2017 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 783 THEORETICAL AND APPLIED ELECTRICAL ENGINEERING VOLUME: 15 |NUMBER: 5 |2017 |DECEMBER Figure 4 shows the multiplier being used for balance modulator as well as the error quantity. Vxand Vyare 500 kHz and 50 kHz, 800 mVP-P sinusoidal carrier and modulation signals, respectively fed to inputs of the proposed multiplier. Also Fig. 5 demonstrates how the multiplier circuit can be employed as a frequency doubler. In this simulation, if both frequencies of the input voltage are 500 kHz, the figure shows the corresponding output waveform with double frequency of 1 MHz. Frequency response in Fig. 6 shows that bandwidth of the circuit is 196 MHz when the input signal is applied to Vx, and Vy= 400 mV. The same result is obtained for constant value of Vxand AC signals for Vy. 400 -400 0 250 -250 0 5 -5 0 s Output Signal (mV) Error (mV) 0 10 20 30 40 Time (µs) Input Signals (mV) Fig. 4: Proposed multiplier as an amplitude modulator. 500 kHz carrier sinusoid and 50 kHz modulating signal (upper waveform); AC modulated output (middle waveform); Error measurement (lower waveform). 400 -400 0 250 -250 0 5 -5 0 0 10 20 30 40 Time (µs) Output Signal (mV) Fig. 5: The proposed multiplier as a frequency doubler, input signals (upper waveform); output signal (middle waveform); Error measurement (lower waveform). 0 -2 -4 -6 -8 -10 -12 -14 10 0 10 2 10 4 10 6 10 8 10 10 Frequency (Hz) Output Voltage (dB) Fig. 6: Frequency response of the circuit. The Total Harmonic Distortion (THD) versus input signal at 100 kHz and 1 MHz is shown in Fig. 7. THD simulations are carried out for both of Vxand Vy, when one of them is constant and another one is sinusoidal. In the worst case, an input signal of 1 Vp-p at a frequency of 1 MHz resulted in a THD of less than 1.2 %. 0 0.2 0.4 0.6 0.8 1 Vx P-P and Vy P-P (V) 0 0.2 0.4 0.6 0.8 1 1.2 1.4 THD (%) 1 MHz (Vy=const.) 100 kHz (Vy=const.) 1 MHz (Vx=const.) 100 kHz (Vx=const.) Fig. 7: Relation between THD, Vxand Vy. In order to evaluate the robustness of the circuit against the process variation, the Monte Carlo analysis with 100 samples is performed by applying ±5% Gaussian distribution at ±3σlevel in the variation of all transistors aspect ratio and threshold voltage simultaneously. Two sinusoidal signals with the frequencies of 500 kHz and 1 MHz and also 400 mVp-p and 800 mVp-p amplitudes are applied to the circuit under the aforesaid variations and then the outputs are compared with the ideal values. The average of error in each sample is considered as the relative error. The c 2017 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 784 THEORETICAL AND APPLIED ELECTRICAL ENGINEERING VOLUME: 15 |NUMBER: 5 |2017 |DECEMBER -4 -3 -2 -1 0 1 2 3 4 Relative error (%) 0 2 4 6 8 10 12 Number of samples Fig. 8: Monte Carlo analysis of the circuit for ±5% mismatch in threshold voltage and transistors aspect ratio. result is shown in Fig. 8, in which 68 % of the total samples occurred with the relative error of less than ±1%. To analyze the performance of the proposed circuit regarding temperature variations the simulations are carried out in different temperatures. The threshold voltage is the most important parameter in the analysis of temperature dependence of CMOS circuits [27]. Therefore, a small variation in threshold voltage causes a large change in the output. Although single-ended output of the squaring circuits (see Eq. (8) or Eq. (9)) includes the threshold voltage, the output of the complete circuit (see Eq. (12)) does not depend on the threshold voltage, therefore no remarkable change occurs at the final output. Figure 9 shows the relative error of the circuit in different temperatures, where the maximum error occurred at −40 ◦C with 1.18 %. In this simulation, the obtained output at the temperature of 25 ◦C is considered as the reference value (relative error = 0), then the resulted outputs in other temperatures are compared with that value and the relative error is computed. It should be pointed out that the input signals are the same as the signals that were applied in the Monte Carlo analysis. The characteristics of the circuit are -40 -20 0 20 40 60 80 Temperature (°C) 0 0.2 0.4 0.6 0.8 1 1.2 Relatice error (%) Fig. 9: Relative error of the circuit versus different temperatures. summarized in Tab. 2 and compared with the former works to prove the efficiency of the circuit. 5. Conclusion A new CMOS voltage-mode analog multiplier circuit was presented in this paper. The key features of the circuit are its high accuracy and high linearity as well as its body effect-free operation, owing to the fact that the circuit was designed based on a new symmetrical configuration. Compared to the previously reported works, the dynamic input and output ranges of the circuit are considerably improved, since the circuit works in the saturation region not in weak inversion. To prove the efficiency of the proposed circuit, it was employed as a modulator and frequency doubler, and the simulation results were compared with ideal performance of these applications. The performance of the proposed multiplier was characterized using HSPICE with TSMC level 49 (BSIM3v3) parameters for 0.18 µm CMOS technology. Tab. 2: Comparative parameters of the proposed multiplier with other recent works. [9] [10] [22] [24] This work Power supply (V) 0.5 1.5 1.4 1.5 1.5 Input range (mV) ±80 ±120 ±560 ±200 ±400 Output range (mV) ±10 ±150 ±10 µA±2±250 Power consumption (µW) 0.714 6.7 72.6 32 44 THD (%); 100 kHz, 400 mV 4.11 4.2 1.3 1.7 0.58 Nonlinearity (%) 5.6 3.2 1.9 5.3 0.94 −3 dB bandwidth (MHz) 0.221 0.268 249 1980 196 Tech. (µm) 0.18 0.35 0.25 0.35 0.18 c 2017 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 785 THEORETICAL AND APPLIED ELECTRICAL ENGINEERING VOLUME: 15 |NUMBER: 5 |2017 |DECEMBER Acknowledgment The authors would like to thank Prof. Khayrollah Hadidi for his valuable remarks and fruitful discussions in improving the presentation of the paper. References [1] BLAKIEWICZ, G. Analog multiplier for a lowpower integrated image sensor. In: 16th International Conference Mixed Design of Integrated Circuits and Systems. Lodz: IEEE, 2009, pp. 226– 229. ISBN 978-1-4244-4798-5. [2] NADERI, A., A. KHOEI, K. H. HADIDI and H. GHASEMZDEH. A new high speed and low power four-quadrant CMOS analog multiplier in current mode. International Journal of Electronics and Communications. 2009, vol. 63, iss. 9, pp. 769–775. ISSN 1434-8411. DOI: 10.1016/j.aeue.2008.06.002. [3] CHOI, J., J. PARK, W. KIM, K. LIM and J. LASKAR. High multiplication factor capacitor multiplier for an on-chip PLL loop filter. Electronics Letters. 2009, vol. 45, no. 5, pp. 239–240. ISSN 0013-5194. DOI: 10.1049/el:20092874. [4] SAATLO NADERI, A. and I. S. OZOGUZ. Design of high-linear, high-precision analog multiplier free from body effect. Turkish Journal of Electrical Engineering &Computer Sciences. 2013, vol. 24, iss. 1, pp. 820–832. ISSN 1300-0632. DOI: 10.3906/elk-1307-159. [5] BHAWNA, A. and G. MANEESHA. Low-voltage bulk-driven class AB four quadrant CMOS current multiplier. Analog Integrated Circuits and Signals Processing. 2010, vol. 65, iss. 1, pp. 163–169. ISSN 0925-1030. [6] KELES, S. and H. KUNTMAN. Four quadrant FGMOS analog multiplier. Turkish Journal of Electrical Engineering &Computer Science. 2011, vol. 19, no. 2, pp. 291–301. ISSN 1300-0632. DOI: 10.3906/elk-1001-377. [7] OLIAEI, O. and P. LOUMEAU. Four-quadrant class AB CMOS current multiplier. Electronics Letters. 1996, vol. 32, iss. 25, pp. 2327–2329. ISSN 0013-5194. DOI: 10.1049/el:19961553. [8] CHANG, C. C. and S. I. LIU. Weak inversion fourquadrant multiplier and two-quadrant divider. Electronics Letters. 1998, vol. 34, iss. 22, pp. 2079– 2080. ISSN 0013-5194. DOI: 10.1049/el:19981496. [9] PANIGRAHI, A. and K. P. PRASHANTA. A novel bulk-input low voltage and low power four quadrant analog multiplier in weak inversion. Analog Integrated Circuits and Signals Processing. 2013, vol. 75, iss. 2, pp. 237–243. ISSN 0925-1030. DOI: 10.1007/s10470-012-9951-3. [10] LIU, W. and S. LIU. Design of a CMOS low-power and low voltage 4-quadrant analog multiplier. Analog Integrated Circuits and Signals Processing. 2010, vol. 63, iss. 2, pp. 307–312. ISSN 0925-1030. DOI: 10.1007/s10470-009-9382-y. [11] COUE, D. and G. WILSON. A 4-quadrant subthreshold mode multiplier for analog neuralnetwork applications. IEEE Transactions on Neural Networks. 1996, vol. 7, iss. 5, pp. 1212–1219. ISSN 1045-9227. DOI: 10.1109/72.536315. [12] RUIQI, W. and X. JIANLI. MOS Translinear Principle Based Analog Four-Quadrant Multiplier. In: IEEE International Conference on Anti-Counterfeiting, Security and Identification (ASID). Taipei: IEEE, 2012, pp. 1–4. ISBN 978-1-4673-2144-0. DOI: 10.1109/ICASID.2012.6325304. [13] GRAVATI, M., M. VALLE, G. FERRI, N. GUERRINI and L. REYES. A novel current-mode very low power analog CMOS four quadrant multiplier. In: Proceedings of the 31st European SolidState Circuits Conference (ESSCIRC). Grenoble: IEEE, 2005, pp. 495–498. ISBN 0-7803-9205-1. DOI: 10.1109/ESSCIR.2005.1541668. [14] MENEKAY, S., R. C. TARCAN and H. KUNTMAN. Novel high precision current-mode multiplier/divider. Analog Integrated Circuits and Signal Processing. 2009, vol. 60, iss. 3, pp. 237–248. ISSN 0925-1030. DOI: 10.1007/s10470-009-9289-7. [15] RYAN, C. Applications of a four-quadrant multiplier. IEEE Journal of Solid State Circuits. 1970, vol. 5, no. 1, pp. 45–48. ISSN 0018-9200. DOI: 10.1109/JSSC.1970.1050065. [16] PROMMEE, P., M. SOMDUNYAKANOK, K. ANGKAEW, A. JODTANG and K. DEJHAN. Single Low-Supply and Low-Distortion CMOS Analog Multiplier. In: International Symposium on Communications and Information Technologies. Beijing: IEEE, 2005, pp. 12–14. ISBN 07803-9538-7. DOI: 10.1109/ISCIT.2005.1566843. [17] DE LA CRUZ-BLAS, C., A. LOPEZ-MARTIN and A. CARLOSENA. 1.5-V MOS translinear loops with improved dynamic range and their applications to current-mode signal processing. IEEE Transaction on Circuits and Systems II: Analog and Digital Signal Processing. 2003, c 2017 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 786 THEORETICAL AND APPLIED ELECTRICAL ENGINEERING VOLUME: 15 |NUMBER: 5 |2017 |DECEMBER vol. 50, iss. 12, pp. 918–927. ISSN 1057-7130. DOI: 10.1109/TCSII.2003.820230. [18] MENEKAY, S., R. TARACAN and H. KUNTMAN. Novel high precision current mode circuits based on the MOS translinear principle. AEU - International Journal of Electronics and Communications. 2009, vol. 63, iss. 11, pp. 992–997. ISSN 1434-8411. DOI: 10.1016/j.aeue.2008.08.010. [19] LOPEZ-MARTIN, A., C. DE LA CRUZ BLAS, J. RAMIREZ-ANGULO and R. CARVAJAL. Current mode CMOS multiplier/divider circuit operating in linear/saturation regions. Analog Integrated Circuits and Signal Processing. 2011, vol. 66, iss. 2, pp. 299–302. ISSN 0925-1030. DOI: 10.1007/s10470-010-9552-y. [20] SIRIPRUCHYANUN, M. and W. JAIKLA. A current-mode analog multiplier/divider based on CCCDTA. AEU - International Journal of Electronics and Communications. 2008, vol. 62, iss. 3, pp. 223–227. ISSN 1434-8411. DOI: 10.1016/j.aeue.2007.03.009. [21] AMIRI, A. and A. NADERI SAATLO. Voltage mode implementation of highly accurate analog multiplier circuit. In: 23rd Iranian Conference on Electrical Engineering. Tehran: IEEE, 2015, pp. 1059–1062. ISBN 978-1-4799-1972-7. DOI: 10.1109/IranianCEE.2015.7146368. [22] CHAISAYUN, I., S. PIANGPRANTONG and K. DEJHAN. Versatile analog squarer and multiplier free from body effect. Analog Integrated Circuits and Signals Processing. 2012, vol. 71, iss. 3, pp. 539–547. ISSN 0925-1030. DOI: 10.1007/s10470-011-9701-y. [23] IBARAGI, E., A. HYOGO and K. SEKINE. A CMOS analog multiplier free from mobility reduction and body effect. Analog Integrated Circuits and Signals Processing. 2000, vol. 25, iss. 3, pp. 281–290. ISSN 0925-1030. DOI: 10.1023/A%3A1008377914605. [24] CHUNHONG, C. and L. ZHENG. A low-power CMOS analog multiplier. IEEE Transactions on Circuits and Systems II: Express Briefs. 2006, vol. 53, iss. 2, pp. 100–104. ISSN 1549-7747. DOI: 10.1109/TCSII.2005.857089. [25] TOUMAZOU, C., F. J. LIDGEY and D. G. HAIGH. Analogue IC design: the current-mode approach. London: The Institution of Electrical Engineers, 2008. ISBN 978-0863412974. [26] WONG, S. and C. A. SALAMA. Impact of scaling on MOS analog performance. IEEE Journal of Solid-State Circuits. 1983, vol. 18, iss. 1, pp. 106–114. ISSN 0018-9200. DOI: 10.1109/JSSC.1983.1051906. [27] FILANOVSKY, I. M. and A. ALLAM. Mutual compensation of mobility and threshold voltage temperature effects with applications in CMOS circuits. IEEE Transactions on Circuits and Systems I: Fundamental Theory and Applications. 2001, vol. 48, iss. 7, pp. 876–884. ISSN 1057-7122. DOI: 10.1109/81.933328. About Authors Ali NADERI SAATLO was born in Urmia, Iran, in 1982. He received his B.Sc. degree in Communication Engineering from Urmia Azad University, in 2005, the M.Sc. degree in Electrical Engineering from Urmia University, Urmia, Iran in 2008, and the Ph.D. in Electronics Engineering from Istanbul Technical University, Istanbul, Turkey in 2014. Since 2011, he has been a faculty member of electrical engineering department of Urmia Azad University. His research interests are analog and digital integrated circuit design for fuzzy applications, fuzzy sets and systems, high performance analog circuits, and digital signal processing. He is the author or coauthor of more than 30 peer-reviewed papers in international and national journals and conference proceedings. Abolfazl AMIRI was born in Borazjan, Iran, in 1987. He received his B.Sc. and M.Sc. degrees in Electronics Engineering from the department of Electrical and Electronics Engineering, Bushehr Azad University, Bushehr, Iran, in 2009 and 2012, respectively. He is currently a Ph.D. student of Islamic Azad University of Tehran, Iran. His areas of interest include current-mode circuits design, low power VLSI circuits, and CMOS analog integrated circuit design. He is the author of 15 technical papers in electronics. c 2017 ADVANCES IN ELECTRICAL AND ELECTRONIC ENGINEERING 787