A New Translinear-Based Dual-Output Square-Rooting Circuit

Abstract

A new wide input range square-rooting circuit is presented. The proposed circuit consists of a dual translinear loop, an absolute value circuit, and current mirrors. A current-mode technique is used to provide wide input range with simple circuitry. The output signal of the proposed circuit is the current which is proportional to the square root of input current. The proposed square-rooting circuit was confirmed by using PSpice simulator program. The simulation results demonstrate that the proposed circuit provides the excellent temperature stability with wide input current range.

Recommended by Krishnamachar Prasad

Faculty of Engineering, King Mongkut's Institute of Technology Ladkrabang, Bangkok 10520, Thailand

Received 25 June 2008; Accepted 30 September 2008

1. Introduction

A square-rooting circuit is widely used in analog instrumentation and measurement systems. For example, it is used to linearize a signal from a differential pressure flow meter, or to calculate the root mean square value of an arbitrary waveform [1]. Typically, voltage-mode square-rooting circuits can be realized by using operational amplifiers (op-amps) and can be attached to passive and active elements such as an analog multiplier to form squarer and resistors [2], the BJTs to form a log and antilog amplifier [3], and MOS transistor operating in triode region [4]. However, op-amp-based square-rooting circuit has the high-frequency limitation due to the finite gain bandwidth product (GBW) of the op-amps, and some of them are not suitable for IC implementation. Second-generation current conveyors (CCIIs) are useful in analog signal proposing circuits. Since the gain bandwidth product of an op-amp is finite, the higher the gain it realizes, the less bandwidth it possesses. In the past, the realization of square-rooting circuit using CCIIs has been proposed in the technical literature [5, 6]. Lui [5] proposed square-rooting circuit based on the use of the CCIIs connected with nonsaturated MOS transistors, op-amps, and resistors. The high-frequency limitation of this circuit is due to the finite GBW of the op-amp and the MOS transistor operated in nonsaturation. Moreover, the use of op-amps and floating resistors makes this circuit not ideal for IC fabrication. Differential difference current conveyors (DDCCs-) based square-rooting circuit is proposed by Chiu et al. [6]. However, the disadvantage of this circuit is the same as the proposed square-rooting circuits of Lui [5]. The square-rooting circuit is realized by using bipolar junction transistors (BJTs), based on the current-mode technique, which have been reported as being a quite attractive feature of wide bandwidth and suitable for implementing in monolithic form [7]; but only positive input current range can be applied into the circuit. The current-mode square-rooting circuit based on MOS transistors operating in class AB has also been reported [8].

This paper, a new BJT wide input range current-mode square-rooting circuit, is introduced. It consists of a dual translinear loop, an absolute value circuit, and current mirrors. The proposed square-rooting circuit is operated in current mode that has the following advantages.

The square-rooting circuits of Filanovsky and Baltes [4], Liu [5], and Chiu et al. [6] are limited for high frequency due to the finite GBW of op-amps and MOS transistor operated in nonsaturation. While the proposed circuit has no restriction, therefore, the proposed square-rooting circuit operates more high-frequency than that of the previous square-rooting circuits.

(i) The proposed square rooting is suitable for bipolar IC technology.

(ii) The proposed square rooting provides wide input current range.

(iii): The proposed square rooting provides excellent temperature stability.

(iv) It possesses high output impedance.

2. Circuit Description

Figure 1 shows the dual translinear loop of the proposed square-rooting circuit; here _I1 , _I2 ,_I3 , and _I4 are the currents taken as the collector currents of _Q1 ,_Q2 , _Q3 , and _Q4 , respectively. Neglect the based currents and assume that the four transistors are identical. Summing the based-emitter voltages around the closed-loop containing _Q1 , _Q2 , _Q3 , and _Q4 , gives by [9] [figure omitted; refer to PDF] Substituting for the relationship between collector currents _IC and base voltages _VBE [10] yields [figure omitted; refer to PDF] from which [figure omitted; refer to PDF] Let _I1 be the constant current source that provides the bias current for the circuit. When the input signal current _Iin is applied to the circuit, then the relationship of the currents _I2 , _I4 , and _Iin , since _I2 is equal to _I4 , can be expressed as [figure omitted; refer to PDF] It means that the currents _I2 and _I4 are a square root of the input current _Iin with the current gain equal to _I1 . In addition, the temperature effect in terms of the thermal voltage is compensated. It can be noted from Figure 1 that only positive input signal current _Iin can be applied to the circuit. To achieve wide input current range that can apply both negative and positive signal currents, the absolute-value circuit is required.

Figure 1: Dual translinear loop.

[figure omitted; refer to PDF]

Figure 2 shows the absolute-value circuit. The transistors _Q1 -_Q4 and_Iq function as a current-mode full-wave rectifier [11]. The current source _Iq provides the biasing current for the circuit. The current mirrors, _Q5 -_Q6 and _Q7 -_Q8 , are supplied by the current source _IB1 , which ensures that the two current mirrors are continuously on, thereby, improving frequency response and linearity overall system. The current source _IB2 is used to eliminate the DC current offset of the output current. The output current _Io of circuit can be expressed as [figure omitted; refer to PDF]

Figure 2: Absolute value circuit.

[figure omitted; refer to PDF]

Figure 3 shows the proposed square-rooting circuit using a dual translinear loop, an absolute-value circuit, and current mirrors. In this circuit, the input is a current, and the output is also the dual currents which are proportional to the square root of the input current. Using (4) and (5), the output current can be expressed as [figure omitted; refer to PDF] From (6), it means that the output current _Iout is a square-root of the input current _Iin , with the current gain equal to _Io . It is also shown in (6) that the output current is not sensitive to temperature. It is noted that the proposed square-rooting circuits in Figure 3 provide the output current which is proportional to the square root of the input current at high output impedance. Hence, it can be directly connected as the load. The proposed circuit in Figure 3 can easily be modified to be as voltage-in current-out or voltage-in voltage-out circuits by using the converting resistances. If the square-rooting circuit with voltage-in current-out circuit is desired, the new input voltage can be applied to the node Y of absolute-value circuit and disconnect grounded resistor; while its nodeX is terminated with grounded resistor. If the square-rooting circuit with voltage-in voltage-out circuit is continually desired, the additional grounded resistor is required to connect at nodes Z for operating as current-to-voltage conversion.

Figure 3: Proposed dual-output current-mode square-rooting circuit.

[figure omitted; refer to PDF]

In the practical realization, the device mismatch between NPN and PNP bipolar transistors groups of _Q11 -_Q14 function is the major factor that contributes to the errors from the ideal performance. The output current error can be expressed as [figure omitted; refer to PDF] where _βN and _βP are the current gains of NPN and PNP bipolar transistors, respectively, and _Io is the bias current of the circuit. If _βN =137.5 , _βP =110 , _Io =50 μ A, and _Iin =1 mA, then the resulting output current error is equal to 2%.

3. Simulation Results

The square-rooting circuit in Figure 3 is simulated by using the PSpice simulator program. The proposed square-rooting circuit is simulated based on the model parameters of the AT&T ALA400-CBIC-R [12]. The supply voltages are chosen as _VCC =2.5 V and _VEE =-2.5 V. The current supplies are _Iq =8 μ A, _IB1 =100 μ A,_IB2 =116 μ A, and _Io =50 μ A.

Figure 4 shows the simulated DC transfer characteristic for the input current _Iin of the proposed square-rooting circuit in Figure 3. The simulation of transfer curve is compared with the calculated value. This result demonstrates that the proposed square-rooting circuit yields the operating current range from <-1 mA to >1 mA of the input current. At _Iin =1 mA and -1 mA, it also shows that the difference of the output current between simulation value and calculated value is 10.11 μ A (4.52%) and 13.88 μ A (6.18%), respectively. The amplitude error of the output current signal more than 2% may be resulting from the error of the absolute-value circuit which is neglected.

Simulated DC transfer characteristics of proposed square-rooting circuit: (a) positive input; (b) negative output.

(a) [figure omitted; refer to PDF]

(b) [figure omitted; refer to PDF]

Figure 5 shows the operation of proposed square-rooting circuit in Figure 3 while applying the 2 m_AP-P triangle wave with 100 kHz frequency at the input. The input and output waveforms are shown in Figures 5(a) and 5(b), respectively. Again, a 2 m_AP-P sinusoidal signal with 100 kHz frequency is applied to the proposed square-rooting circuit in Figure3. The input and output waveforms are shown in Figures 6(a) and 6(b), respectively. Figure 6 is confirmed while the input is nonlinear, as the output corresponds proportionally to the square root for the input. The simulated output waveforms are also compared with the calculated values. The simulated frequency response of the proposed circuits has been done as shown in Figure 7. It should be noted that the bandwidth is about 30 MHz. This simulation, the power consumption, is approximated to 15 mW.

Operation of circuit for the 100 kHz input triangular signal: (a) input waveform; (b) output waveforms.

(a) [figure omitted; refer to PDF]

(b) [figure omitted; refer to PDF]

Operation of circuit for the 100 kHz sine wave input signal: (a) input waveform; (b) output waveforms.

(a) [figure omitted; refer to PDF]

(b) [figure omitted; refer to PDF]

Figure 7: Simulated result for frequency responses.

[figure omitted; refer to PDF]

To demonstrate the performance of the proposed square-rooting circuit, Figure 8 shows the simulated output waveform for the cases of the 1 MHz frequency triangle wave input signal and for 2 m_AP-P amplitude. From Figure 8, it is shown that the proposed square-rooting circuit provides the good output waveform at 1 MHz. Figure 9 shows the output current of proposed square-rooting circuit at 50^° C, 75^° C, and 100^° C temperatures while applying the 100 kHz frequency triangle wave with 2 m_AP-P amplitude at the input of the circuit. From the simulation result in Figure 9, it is obviously shown that the proposed square-rooting circuit provides the excellent temperature stability; this result can be confirmed as in (6).

Operation of circuit for the 1 MHz input triangular signal: (a) input waveform; (b) output waveform.

(a) [figure omitted; refer to PDF]

(b) [figure omitted; refer to PDF]

Figure 9: Output waveforms at different temperatures at 100 kHz frequency input signal.

[figure omitted; refer to PDF]

4. Conclusions

In this paper, a new current-mode square-rooting circuit is presented. The proposed circuit employs a dual translinear loop, an absolute-value circuit, and current mirrors. Simulation results show that the proposed square-rooting circuit provides the wide input current range with excellent temperature stability. Better performance can be expected by using the bipolar transistors and the parameters of complementary high performance processes which were not available to the authors. The proposed square-rooting circuit is suitable for IC fabrication because of the absence of the external resistor.

References

[1] E. O. Doebelin Measurement Systems: Application and Design , McGraw Hill, New York, NY, USA, 2004.

[2] P. E. Allen, D. R. Holberg CMOS Analog Circuit Design , Oxford University Press, New York, NY, USA, 2002.

[3] J. Millman, A. Grabel Microelectronics , McGraw Hill, New York, NY, USA, 1992.

[4] I. M. Filanovsky, H. P. Baltes, "Simple CMOS analog square-rooting and squaring circuits," IEEE Transactions on Circuits and Systems I , vol. 39, no. 4, pp. 312-315, 1992.

[5] S.-I. Liu, "Square-rooting and vector summation circuits using current conveyors," IEE Proceedings: Circuits, Devices and Systems , vol. 142, no. 4, pp. 223-226, 1995.

[6] W. Chiu, S.-I. Liu, H.-W. Tsao, J.-J. Chen, "CMOS differential difference current conveyors and their applications," IEE Proceedings: Circuits, Devices and Systems , vol. 143, no. 2, pp. 91-96, 1996.

[7] C. Toumazou, F. J. Lidgey, D. G. Haigh Analogue IC Design: The Current-Mode Approach , Peter Peregrinus, London, UK, 1990.

[8] V. Riewruja, K. Anuntahirunrat, W. Surakampontorn, "A class AB CMOS square-rooting circuit," International Journal of Electronics , vol. 85, no. 1, pp. 55-60, 1998.

[9] B. Gilbert, "Translinear circuits: a proposed classification," Electronics Letters , vol. 11, no. 1, pp. 14-16, 1975.

[10] P. Gray, P. J. Hurst, S. H. Lewis, R. G. Meyer Analysis and Design of Analog Integrated Circuit , John Wiley & Sons, New York, NY, USA, 2001.

[11] S. J. G. Gift, "New precision rectifier circuits with high accuracy and wide bandwidth," International Journal of Electronics , vol. 92, no. 10, pp. 601-617, 2005., [email protected]

[12] D. R. Frey, "Log-domain filtering: an approach to current-mode filtering," IEE Proceedings G: Circuits, Devices and Systems , vol. 140, no. 6, pp. 406-416, 1993.