• 沒有找到結果。

A Direct Conversion Merged LNA-I/Q-Mixer With Noise Reduction Using Dual Cross Coupling for WiMAX/WiBro Applications

N/A
N/A
Protected

Academic year: 2021

Share "A Direct Conversion Merged LNA-I/Q-Mixer With Noise Reduction Using Dual Cross Coupling for WiMAX/WiBro Applications"

Copied!
3
0
0

加載中.... (立即查看全文)

全文

(1)

32 IEEE MICROWAVE AND WIRELESS COMPONENTS LETTERS, VOL. 22, NO. 1, JANUARY 2012

A Direct Conversion Merged LNA-I/Q-Mixer With

Noise Reduction Using Dual Cross Coupling

for WiMAX/WiBro Applications

Chia-Hung Chang and Christina F. Jou, Member, IEEE

Abstract—A direct conversion merged low noise amplifier (LNA)

and I/Q-mixer with new noise reduction technique is presented in this letter. The merged LNA-I/Q-mixer employs a dual capacitor cross coupling (CCC) technique to improve the circuit noise per-formance. The first CCC is applied at the transconductance stage of the LNA to increase the effective transconductance, which in turn lowers the noise figure and decreases the current consump-tion. The second CCC connection is applied at the cascode stage of the LNA, which can reduce the influence of the noise current from cascode transistors. The simulation shows the proposed combina-tion of the LNA and I/Q-mixer configuracombina-tion with the dual CCC technique can achieve NF 3 dB better than that of conventional merged topologies in a wideband application. The measured NF is lower than 4.55 dB from 2.3 to 2.7 GHz. In addition, low-power and low-cost circuit is realized due to current-reused and inductor-less design at the interface of the LNA and I-Q mixer. The core circuit only consumes 16.83 mW and occupies 0.16 mm2area.

Index Terms—Cross coupling, direct conversion, I/Q-mixer,

low noise amplifier (LNA), noise reduction, wireless broad-band (WiBro), worldwide interoperability for microwave access (WiMAX).

I. INTRODUCTION

T

HE IEEE 802.16e mobile Worldwide Interoperability for Microwave Access (WiMAX) and Wireless Broadband (WiBro) systems adopt the orthogonal frequency division mul-tiple access (OFDMA) modulation technique, and also provide high speed and high data rate application. The tendency to-ward low power, low cost, and low noise for a direct conver-sion front-end receiver is required in the WiMAX/WiBro appli-cations. The merged low noise amplifier (LNA) and I/Q-mixer is one of the solutions to meet these requirements. There had been several types of merged topologies. The first one stacks LNA and I/Q-mixer with the common source/emitter (CS/CE) input configuration [1] to save power and demonstrate a good noise figure (NF), while the topology shows a limited band-width and requires additional inductors. Although multi-ele-ment matching networks can perform a wider bandwidth, pas-sive components occupy a fairly large chip area. In contrast, the

Manuscript received November 16, 2010; revised June 26, 2011, October 20, 2011; accepted November 22, 2011. Date of publication December 14, 2011; date of current version January 11, 2012.

The authors are with the Department of Electrical Engineering, National Chiao Tung University, Hsinchu 300, Taiwan (e-mail: hung.cm93g@nctu. edu.tw; christ@cc.nctu.edu.tw).

Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org.

Digital Object Identifier 10.1109/LMWC.2011.2179023

combination of LNA and I/Q-mixer with the common gate/base (CG/CB) input configuration can be considered as a candidate for a wide input matching, low power, and low cost applica-tion. However, CG/CB amplifiers suffer from higher NF than the topology using CS/CE amplifier. Using folded technique to convert the stacked circuit into a cascaded one [2], [3] is the second means, which decreases the power consumption to ob-tain low voltage design. However, the loaded inductor increases overall chip size. A third way applies an on-chip transformer [4]. The conversion loss and larger chip size from the transformer limit its application. A BLIXER topology [5] with noise can-celing [6] is another solution. In this stacked configuration, un-balanced transconductance from differential transistors is nec-essary to lower noise contribution.

In this letter, the merged LNA and I/Q-mixer is proposed and the CCC technique is incorporated simultaneously in the transconductance and cascode stages, which reduces the shot noise in collector current and the equivalent noise current from main and cascode transistors, respectively.

II. MERGEDLNAANDI/Q-MIXER

The merged LNA and I/Q-mixer with the dual CCC config-uration applied at the transconductance and cascode stages can improve the noise performance, as shown in Fig. 1. The and are employed to allow differential input voltage to drop across the base and emitter terminals of each input transistor. Because the loop gain of the positive feedback path is smaller than unity, the condition of oscillation will not happen. More-over, The effective transconductance is boosted by means of the feed forward connection, thereby increasing the conversion gain and reducing the NF [7], [8]. If the input signals are sinusoidal signals, the overall conversion gain of the merged topology can be expressed as follows:

(1) where the coupling factor is approx-imately equivalent to 1 if the cross couple capacitance is larger than the junction capacitance , and is the effective transconductance of the transistor , and is the load resistor. To achieve input impedance matching, the boosted is designed as inverse of the source resistance ( ohm). The conversion gain can be expressed as . Since is only required to be to obtain 20 dB conversion gain, the entire headroom is adequate for voltage drop across the load resistor and stacking three

(2)

CHANG AND JOU: DIRECT CONVERSION MERGED LNA-I/Q-MIXER WITH NOISE REDUCTION 33

Fig. 1. Schematic of the merged LNA and I/Q-mixer with dual CCC.

transistors. In addition, gain-controlled mechanism is also incorporated in this merged topology. Four gain states can be performed by switching the load resistors and steering current path from the cascode stage. The highest gain mode operates at maximum load impedance and without steering current. By switching on only SW1 or on SW1 and SW2 simultaneously, shown in Fig. 1, the reduced load impedance causes the lower gain modes. The lowest gain mode is performed by reducing load impedance and by sharing current on transistors and

from 0 V to VDD simultaneously.

The noise performance is a significant factor for the merged LNA-I/Q-mixer. Assuming a noiseless switching-quad from mixer, Fig. 2 shows the small signal equivalent noise model of a CB cascode amplifier with a load resistor, and the corre-sponding NF is derived as follows [9]:

(2) where and are the base and emitter resistors, and are the transconductance from transistors and

and are the current gain, resonant frequency and transition frequency, respectively, and is the parasitic capacitance between and .

Equation (2) reveals that the noise contributions of the second, third, fifth, and last terms are negligible for its small value. Accordingly, the dominant contributions for a CB cas-code amplifier are determined by the fourth and sixth terms. To further improve the NF, the first CCC is employed at the transconductance stage to boost and then the fourth term can be expressed as , which is decreased by a factor of . The second CCC is utilized at the cascode

Fig. 2. Small signal equivalent noise model of a CB cascode amplifier.

Fig. 3. Comparison of simulated NF performance of the merged LNA and I/Q-mixer with and without the dual CCC technique.

Fig. 4. Chip photograph.

stage, which behaves like another CB input configuration to increase its effective transconductance. The noise contribution from the sixth term is therefore reduced if and are boosted and is minimized. The parasitic capacitance can be alleviated by compact layout of the transcon-ductance and cascode transistors [10]. The simulated NF of the merged LNA and I/Q-mixer with dual CCC mechanism, as depicted in Fig. 3, exhibits a 3 dB improvement from the conventional merged topology without dual CCC under the same bias condition. These simulated comparisons demonstrate that the proposed dual CCC can indeed improve the noise performance.

III. MEASUREMENTRESULTS

A direct conversion merged LNA and I/Q-mixer with LO transmit path for a WiMAX/Wibro front-end receiver is fabri-cated using TSMC 0.35 um SiGe BiCMOS technology. The test

(3)

34 IEEE MICROWAVE AND WIRELESS COMPONENTS LETTERS, VOL. 22, NO. 1, JANUARY 2012

Fig. 5. S in dB, gain response, and NF for IF = 10 MHz.

TABLE I

CHIPPERFORMANCE ANDCOMPARISONWITHOTHERFRONT-ENDRECEIVER

chip photograph is displayed in Fig. 4. Due to the inductor-less design, the chip only occupies 0.16 mm of chip area for the merged LNA-I/Q-mixer. The chip is measured by mounting the die directly on a four layer FR-4 PCB. The signal generator gen-erates an RF and LO signal, fed into the LNA inputs and an I/Q divider, respectively, by external 180 hybrid baluns. The power consumption of the core circuit is 5.1 mA.

Fig. 5 presents the full band frequency response of the simu-lated input return loss, front-end receiver gain, and NF in com-parison with the measurement results. The measured conversion gain is approximately 21.4 dB with an IF frequency of 10 MHz. The gain flatness of less than 0.3 dB can be achieved from 2.3 to 2.7 GHz. In addition, the merged LNA and I/Q-mixer performs four gain states which range from 21.4 dB to 5.4 dB. The mea-sured input return loss which exhibits a broadband characteristic is below dB at the desired frequency band and also remains around the same level in the different gain states because of the CB input configuration. The de-embedded NF of the front-end

receiver from the loss of the balun can be less than 4.55 dB due to the proposed dual CCC technique while the measured IF fre-quency is fixed at 10 MHz. For the measurement of linearity, the IP1dB is dBm and dBm at the highest and lowest gain state, respectively. The IIP3 achieves dBm and dBm, respectively.

Table I summarizes the measured performance of the receiver and compares it with the prior works [3]–[5]. Although the com-bined LNA-I/Q-mixer, current-reused, and inductor-less design are also employed in [5] which demonstrates an excellent per-formance on chip area and power dissipation, the NF based on the proposed dual CCC technique in this study shows an approx-imately 1 dB improvement. The configuration using folded [3] or transformer [4] indicates a larger chip area than the stacked ones. To evaluate the performance, a figure of merit (FOM) is described as follows:

(3)

IV. CONCLUSION

A merged LNA-I/Q-mixer for dual mode WiMAX and WiBro applications with noise reduction technique is proposed. The merged LNA and I/Q-mixer achieves low power and low cost as a result of compact and concurrent design. Furthermore, the noise performance is improved by the proposed dual CCC tech-nique which is applied at the transconductance and cascode stages of the LNA. To demonstrate the properties of the CCC network, the merged LNA and I/Q-mixer is analyzed, designed, and measured.

REFERENCES

[1] H. Sjoland, A. Karimi-Sanjaani, and A. A. Abidi, “A merged CMOS LNA and mixer for aWCDMA receiver,” IEEE J. Solid-State Circuits, vol. 38, no. 6, pp. 1045–1050, Jun. 2003.

[2] Y.-S. Hwang and H.-J. Yoo, “A low power folded RF front-end with merged LNA and mixer for ZigBee/bluetooth,” in Proc. Radio Wireless

Symp., Jan. 2007, pp. 85–86.

[3] A. Amer, E. Hegazi, and H. F. Ragaie, “A 90-nm wideband merged CMOS LNA and mixer exploiting noise cancellation,” IEEE J.

Solid-State Circuits, vol. 42, no. 2, pp. 323–328, Feb. 2007.

[4] H.-K. Chen, J.-R. Sha, D.-C. Chang, Y.-Z. Juang, and C.-F. Chiu, “5.8-GHz merged LNA-mixer with on-chip balun,” Microw. Opt.

Technol. Lett., vol. 48, no. 3, pp. 508–511, Mar. 2006.

[5] S. C. Blaakmeer, E. A. M. Klumperink, D. M. W. Leenaerts, and B. Nauta, “The blixer, a wideband balun-LNA-I/Q-Mixer topology,” IEEE

J. Solid-State Circuits, vol. 43, no. 12, pp. 2706–2715, Dec. 2008.

[6] R. Bagheri, A. Mirzaei, S. Chehrazi, M. E. Heidari, M. Lee, M. Mikhemar, W. Tang, and A. A. Abidi, “An 800-MHz-6-GHz soft-ware-defined wireless receiver in 90-nm CMOS,” IEEE J. Solid-State

Circuits, vol. 41, no. 12, pp. 2860–2876, Dec. 2006.

[7] W. Zhuo, X. Li, S. Shekhar, S. H. K. Embabi, J. P. de Gyvez, D. J. All-stot, and E. Sánchez-Sinencio, “A capacitor cross-coupled common-gate low noise amplifier,” IEEE Trans. Circuits Syst. II: Expr. Briefs, vol. 52, no. 12, pp. 875–879, Dec. 2005.

[8] X. Fan, H. Zhang, and E. Sánchez-Sinencio, “A noise reduction and linearity improvement technique for a differential cascode LNA,” IEEE

J. Solid-State Circuits, vol. 43, no. 3, pp. 588–599, Mar. 2008.

[9] P. R. Gray, P. J. Hurst, S. H. Lewis, and R. G. Meyer, Analysis and

De-sign of Analog Integrated Circuits, version 4th ed. New York: Wiley, 2001, pp. 762–802.

[10] H. Samavati, H. R. Rategh, and T. H. Lee, “A 5-GHz CMOS wireless LAN receiver front-end,” IEEE J. Solid-State Circuits, vol. 35, no. 5, pp. 765–772, May 2000.

數據

Fig. 3. Comparison of simulated NF performance of the merged LNA and I/Q- I/Q-mixer with and without the dual CCC technique.
Fig. 5. S in dB, gain response, and NF for IF = 10 MHz.

參考文獻

相關文件

After students have had ample practice with developing characters, describing a setting and writing realistic dialogue, they will need to go back to the Short Story Writing Task

Teachers may consider the school’s aims and conditions or even the language environment to select the most appropriate approach according to students’ need and ability; or develop

Robinson Crusoe is an Englishman from the 1) t_______ of York in the seventeenth century, the youngest son of a merchant of German origin. This trip is financially successful,

fostering independent application of reading strategies Strategy 7: Provide opportunities for students to track, reflect on, and share their learning progress (destination). •

Theorem 5.6.1 The qd-algorithm converges for irreducible, symmetric positive definite tridiagonal matrices.. It is necessary to show that q i are in

• If the cursor scans the jth position at time i when M is at state q and the symbol is σ, then the (i, j)th entry is a new symbol σ

• Thresholded image gradients are sampled over 16x16 array of locations in scale space. • Create array of

In this paper, by using Takagi and Sugeno (T-S) fuzzy dynamic model, the H 1 output feedback control design problems for nonlinear stochastic systems with state- dependent noise,