K-band ILFT

The fabricated K-band ILFT starts to oscillate at a bias current of 0.79 mA from 1.5 V. The measured output spectra of the K-band ILFT versus the output frequency under free-running and locked conditions with probe and cable losses and input bias VBIAS of 0.56 V are shown in Figs. 2.22 and 2.23, respectively. The measured peak output power is –11.76 dBm at 26.32 GHz under free-running condition and –8.09 dBm at 26.32 GHz under locked condition with input power of 4 dBm, input bias VBIAS of 0.56 V, and 4.7-dB power loss from cable and probe. Because of the contribution of input power, the locked ILFT has a higher output power than the free-running ILFT.

The simulated and measured input power versus the output frequency with the input bias VBIAS of 0.56 V and external tuning voltage VTUNE of 1.5 V is shown in Fig.

2.24. The upper and lower locking ranges are labeled as the maximum and minimum output frequencies under locked condition, respectively. The simulated and measured locking ranges versus input power are shown in Fig. 2.25 where the measured locking range is from 156 to 567 MHz while the input power varies from –9 to –1 dBm. At an input power greater than 0 dBm, the locking range decrease slightly, as shown in Fig.

2.25. With small input power, the measurement result is close to the simulation result.

４０

With large input power, the measured locking range is smaller by 100 MHz. This is because the valid frequency range of the simulation model is not completely covered with the desired frequency range.

The locking range is mainly determined by two important factors. One is the nonlinear term a3 of the frequency pre-generator whereas the other is the nonlinear characteristic of the ILO. As input power is small, the linear model of the ILO is valid.

Thus, the locking range is dominated by the nonlinear term a3 as can be seen from (2.15). As the input signal is increased, the locking range is increased due to the increase of Vi and a3. If the input signal is increased to a moderate value which causes the conduction angle smaller than 250°, this leads to the large decrease of a3 as can be seen from Fig. 2.4. The locking range is, therefore, almost saturated.

The simulated and measured locking range versus the input bias voltage VBAIS of M1/M2 with input bias VBIAS of 0.65 V and tuning voltage VTUNE of 1.5 V are shown in Fig. 2.26. It can be seen from Fig. 2.26 that the locking range increases with a decrease in the input bias. This result can be explained by the fact that the lower input bias allows only a small current through M3/M4. Thus, the weaker negative-resistance generated from M3/M4 reduces the effective quality factor of LC-tank. Besides, the conversion gain of the frequency pre-generator is a function of input bias VBIAS. Therefore, the locking range is increased at the higher third-order harmonic current region as can be seen from Fig. 2.4.

The varactors C1/C2 are designed in the K-band ILFT. In Fig. 2.27, the total output frequency under locked condition is 3920 MHz as the varactors tuning voltage VTUNE varies from 0 to 1.5 V with a dc power consumption of 2.95 mW and an input power of 4 dBm. The output frequency range of the K-band ILFT under free-running condition is from 24.08 GHz to 26.27 GHz. With a 4-dBm input signal, the output

４１

frequency range of the K-band ILFT under locked condition is from 22.58 GHz to 26.50 GHz. Therefore, the output frequency range extends from 2190 MHz to 3920 MHz.

The measured phase noises of the reference input, free-running output, and locked output from 1 kHz to 10 MHz is shown in Fig. 2.28. It shows that the phase noise difference between the reference input and the locked output is 10.5 dB from 1 kHz to 1 MHz offset. The slightly larger output phase noise at a signal frequency higher than 1 MHz offset is due to excess noise from the internal circuit and output buffer. The spur at around 1MHz offset is from signal generator.

The measured output phase noise as a function of input power is shown in Fig.

2.29. At large input power levels, the measured phase noise of the locked output can approach the theoretical limit of 10log (3²) = 9.5 dB, as derived in Section 2.1. The phase noise degradation from the frequency pre-generator is 0.8 dB at 1-kHz offset and 1.5 dB at 100-kHz offset, respectively. In addition, the phase noise at small frequency offset can be close to the theoretical limit as compared to that at large frequency offset with the same input incident amplitude Vi due to the low pass frequency response.

The measured output spectrum is shown in Fig. 2.30 where the HRRs compared to the desired third-order harmonic are 22.65, 30.58, 29.29, 40.35 dBc for the first, second, forth, and fifth harmonics, respectively. The HRRs of even-order harmonics are 6.64-dB higher than those of odd-order harmonics because of the common-mode rejection capability of R1. In general, R1 does not affect the output performance for odd-order harmonics.

Finally, the measurement of reference input and locked output waveforms are

４２

also tested by the oscilloscope. The measured output waveform with cables and probe losses, input bias VBIAS of 0.65 V, and tuning voltage VTUNE of 1.5 V is shown in Fig.

2.31. Due to the phase shift from the cables, the phase relation between input and output signal as shown from oscilloscope is not exactly the same as those of K-band ILFT. It can be seen from Fig. 2.30 that the locked K-band ILFT can provide a stable output waveform with the three time frequency of input signal.

2.3.2 V-band ILFT

The V-band ILFT starts to oscillate at a bias current of 1.55 mA from 1.2 V. The measured output spectra of the V-band ILFT versus the output frequency under free-running and locked conditions with probe and cable losses are shown in Figs.

2.32 and 2.33, respectively. The loss from the external waveguide subharmonic mixer is de-embedded by the spectrum analyzer. The measured peak output power is –16.14 dBm at 60.025 GHz under free-running condition and –14.81 dBm at 60.025 GHz under locked condition with 4-dBm input power, a VBIAS of 0.55 V, and 9.6-dB power loss from cable and probe.

The measured input power versus the output frequency when the input bias VBIAS

is set at 0.55 V are shown is Fig. 2.34. It can be seen from Fig. 2.35 that the locking range achieves 1422 MHz with 6-dBm input power and 1662 MHz with 9-dBm input power. As the input power is smaller than 1 dBm, the ILO stage is linear and a3 is nearly constant. Thus, the locking range is increased with Vi. With the input power greater than 1 dBm, the locking range is nearly saturated because of the large decrease of the nonlinear term a3. If the input signal is increased to be larger than 2 dBm, the ILO becomes nonlinear and (2.15) is not valid. Under this condition, the extra third-order harmonic is generated by the nonlinear ILO. Therefore, the locking range is increased instead of saturated.

４３

The measurement setup for output phase noise with external down-conversion mixer is shown in Fig. 2.36. A power splitter is required for the operation of instrument. The measured phase noises of the reference input, free-running output, and locked output from 1 kHz to 10 MHz is shown in Fig. 2.37. The noise contribution for external down-conversion mixer is not de-embedded. It shows that the phase noise difference between the reference input and the locked output is 10 dB from 1 kHz to 500 kHz offset. The slightly larger output phase noise at a signal frequency higher than 500 kHz offset is due to excess noise from the internal circuit.

Because the phase noise measurement with ultra low noise floor and a cross-correlation method can be provided by the signal source analyzer, the measured value of the output phase noise can be lower than –155 dBc as shown in Fig. 2.37.

The measurement of reference input and locked output waveforms are tested by the high-speed wideband sampling oscilloscope. The measured output waveform with cables and probe losses is shown in Fig. 2.38. Due to the phase shift from the high-speed cables, the phase relation between input and output signal as shown from oscilloscope is not exactly the same as those of V-band ILFT. It can be seen from Fig.

2.38 that the output waveform is similar to the simulated result as shown in Fig. 2.15.

Due to the limitations of the instruments currently available, the HRR can not be measured. From the simulation results, the HRRs are higher than 18.9 dBc for every undesired harmonics.

In Table 2.2, the recently published CMOS subharmonic ILFMs are compared with the proposed ILFTs. It can be seen that the proposed ILFTs, in contrast to the corresponding CMOS subharmonic ILFMs, can operate with lower dc power consumption. Moreover, this design is the first CMOS ILFT operated in the millimeter-wave band.

４４

As compared with the work in [23], the locking range is around six-times of the proposed work in this chapter. Because low quality factor of LC-tank is chosen and the output power in [23] is only one-third that of the proposed work, the locking range can be larger than the proposed work as shown in (2.15). In addition, the method for the generation of third-order harmonic signal is different. The characteristic of third-order harmonic generation devices can affect the output amplitude directly as shown in Fig. 2.39 where the signal vin is the input signal for quadrature signal generation.

The published bulk-CMOS VCOs worked at the K-band and V-band listed in Table 2.3 are compared with the proposed ILFTs. It can be seen that the locking range of the proposed ILFT is similar to the tuning range of a bulk-CMOS VCO. The proposed ILFT can provide similar output power with lower power consumption even when the input power PINJ is considered as compared with the corresponding bulk-CMOS VCOs.

The simulation and measurement results have shown that the proposed ILFTs can achieve high output power and low power consumption. However, the locking range of ILFT still can not be larger than 10-GHz even if the quality factor of LC-tank is decreased. The main reason is that the large parasitic capacitances between frequency pre-generator stage and ILO stage. The generated third-order harmonic signal is leaked to substrate. Thus, the locking range expressed in (2.15) should be considered the effect. To achieve larger locking range, the transformer can be designed to increase the injection current as can be seen from Fig. 2.40.

４５

2.4 SUMMARY

A millimeter-wave CMOS subharmonic ILFT with a triple-frequency pre-generator is proposed and analyzed. A model for the proposed ILFT is developed to calculate both the locking range and the output phase noise. Based on the model, the design guideline for the maximization of the locking range and the minimization of the output phase noise is developed. The quality factor of the LC-tank of the ILO stage and the conversion gain of the frequency pre-generator stage are maximized to obtain a wider locking range, higher output voltage, and lower output phase noise with low dc power consumption.

According to the developed design guidelines, both the K-band and V-band CMOS ILFTs have been designed and fabricated using 0.18-μm and 0.13-μm technologies, respectively. As seen from the measurement results, the fabricated CMOS K-band ILFT can achieve the locking range of 4.83 % with 4-dBm input injection power and 0.45-mW dc power consumption. Moreover, the locking range of 15.06 % is performed using varactors. The fabricated V-band CMOS ILFT has a locking range of 2.3 % with 6-dBm input injection power and 1.86-mW dc power consumption. The measurement results have verified the performance of the proposed ILFTs.

Since it is feasible to design a high-performance VCO at low frequency without the use of full-speed frequency dividers, the proposed CMOS ILFT offers great potential application in LO signal generators for frequency synthesizers in the millimeter-wave band or even in the sub-millimeter-wave band.

４６

Table 2.1

Dimensions of devices in (a) K-band ILFT and (b) V-band ILFT.

(a)

(b)

４７

Table 2.2

Comparison with published subharmonic ILFMs.

４８

Table 2.3

Comparison with published bulk-CMOS VCOs.

４９

Fig. 2.1 The model of the proposed ILFT.

Fig. 2.2 Simplified noise source model in the proposed ILFT.

５０

Fig. 2.3 The schematic of the proposed ILFT.

５１

Fig. 2.4 HSPICE simulated coefficient of output harmonic current as a function of conduction angle.

Fig. 2.5 HSPICE simulated HRRs for various value of R1 for K-band ILFT..

５２

Fig. 2.6 Simulated locking range as a function of input bias VBIAS for K-band ILFT.

５３

Fig. 2.7 The transient simulation of the free-running K-band ILFT.

Fig. 2.8 Simulated output spectrum of the free-running K-band ILFT.

５４

Fig 2.9 The transient simulation of the locked K-band ILFT with 4-dBm input power, 0.65-V VBIAS, and 8.48-GHz input frequency.

Fig. 2.10 Simulated output spectrum of the locked K-band ILFT with 4-dBm input power, 0.65-V VBIAS, and 8.48-GHz input frequency.

５５

Fig. 2.11 Simulated tuning voltage VTUNE versus output frequency with 0.65-V VBIAS

and 4-dBm input power for K-band ILFT.

Fig. 2.12 Simulated input power versus output frequency with 1.5-V VTUNE and 0.65-V VBIAS for K-band ILFT.

５６

Fig. 2.13 The transient simulation of the free-running V-band ILFT.

Fig. 2.14 Simulated output spectrum of the free-running V-band ILFT.

５７

Fig. 2.15 The transient simulation of the locked V-band ILFT with 6-dBm input power, 0.55-V VBIAS, and 20.3GHz input frequency.

Fig. 2.16 Simulated output spectrum of the locked V-band ILFT with 6-dBm input power, 0.55-V VBIAS, and 20.3-GHz input frequency.

５８

Fig. 2.17 Simulated input power versus output frequency with 0.55-V VBIAS for V-band ILFT.

Fig. 2.18 Simulated the phase noise of ILFT input and output for V-band ILFT.

５９

Fig. 2.19 Chip microphotograph of K-band ILFT (0.66 mm × 0.69 mm).

６０

Fig. 2.20 Chip microphotograph of V-band ILFT (0.59 mm × 0.66 mm).

６１

Fig. 2.21 Measurement setup for subharmonic ILFT testing.

６２

Fig. 2.22 Measured output spectrum of the fabricated K-band ILFT under free-running condition with probe and cable losses and VBIAS of 0.56 V.

Fig. 2.23 Measured output spectrum of the fabricated K-band ILFT under locked condition with probe and cable losses, VBIAS of 0.56 V, and input power of 4 dBm.

６３

Fig. 2.24 Simulated and measured input powers versus output frequency with 1.5-V VTUNE and 0.56-V VBIAS for K-band ILFT.

Fig. 2.25 Simulated and measured locking ranges versus input power with 1.5-V VTUNE and 0.56-V VBIAS for K-band ILFT.

６４

Fig. 2.26 Locking range as a function of input bias VBIAS with 4-dBm input power and 1.5-V VTUNE for K-band ILFT.

Fig. 2.27 Measured tuning voltage VTUNE versus output frequency with 0.65-V VBIAS

and 4-dBm input power for K-band ILFT.

６５

Fig. 2.28 Measured phase noise of reference input, free-running output, and locked output with 0.65-V VBIAS and 4-dBm input power for K-band ILFT.

Fig.2.29 Measured phase noise characteristics of locked output as a function of input power with 0.65-V VBIAS for K-band ILFT.

６６

Fig 2.30 Measured output power spectra of first, second, third, fourth, and fifth harmonics with 0.65-V VBIAS and 4-dBm input power for K-band ILFT.

Fig 2.31 Measured input and output waveforms with cables and probe losses, 0.65-V VBIAS, and 1.5-V VTUNE for K-band ILFT.

６７

Fig 2.32 Measured output spectrum of the fabricated V-band ILFT under free-running condition with probe and cable losses.

Fig. 2.33 Measured output spectrum of the fabricated V-band ILFT under locked condition with probe and cable losses and input power of 4 dBm.

６８

Fig. 2.34 Simulated and measured input power versus output frequency for V-band ILFT.

Fig. 2.35 Simulated and measured locking ranges versus input power for V-band ILFT.

６９

Fig2.36 Measurement setup for output phase noise with external down-conversion mixer.

Fig. 2.37 Measured phase noise of reference input, free-running output, and locked output with 6-dBm input power for V-band ILFT.

７０

Fig. 2.38 Measured input and output waveforms with cables and probe losses for V-band ILFT.

Fig. 2.39 Circuit diagram of the subharmonic ILFT in [23].

７１

Fig. 2.40 Transformer-based ILFT.

７２

７３

CHAPTER 3

60-GHZ CMOS PHASE-LOCKED LOOP WITH