On/Off-state Measurements

For on/off-state measurement, the GSG-configuration RF pad for baseband data input was used as DC pads instead. The chip was mounted on PCB board with bond-wiring for DC applications (VG1, VG2 and VDD). Off-chip bypass networks were designed and implemented on the PCB board.

Fig. 3.39. Die photograph of the proposed OOK modulator.

Table 3.2. Bias conditions during on/off-state small-signal measurements.

An Agilent E8361C network analyzer was used for small-signal measurements up to 67-GHz.

Table 3.2 shows the bias conditions during small-signal measurements. Fig. 3.40 shows the measured and post-EM simulated S-parameters at on-state. The measurement result shows higher gain above 52 GHz but narrower bandwidth than the post-EM simulation. The measured gain at 60 GHz is 10.2 dB. Fig. 3.41 shows the measured and post-EM simulated S-parameters at off-state.

The measured isolation (-S21 in dB) agrees well with the post-EM simulation. The measured isolation at 60 GHz is 35.2dB. This results in a measured on-off isolation of 45.2 dB at 60 GHz.

Both the measured S11 at on- and off-states does not agrees well with the post-EM simulation. This narrows the issue down to the input matching network. Since the input matching network consists of a series capacitor and a short stub, undesired variation of the capacitance is most likely the cause of issue.

For large-signal measurements, an Agilent E8257D signal generator and an Agilent 4419B power meter were used. Fig. 3.42 to 3.50 show the measured and post-EM simulated large-signal performances at on-state from 50 to 66 GHz. Fig. 3.51 shows the measured and post-EM simulated large-signal performances versus frequency. The measurement results show an OP1dB of 7.0 dBm, Psat of 8.9 dBm, and peak PAE of 18.4% at 60-GHz. As with the small-signal gain, the measurement results show slightly better large-signal performances than simulation above 52 GHz, but the difference is much smaller. Across 56 to 64 GHz, the measurement results show OP1dB

above 6dBm, Psat above 7.8 dBm, and PAEpeak above 13.7%.

Fig. 3.40. Measured and simulated S-parameters of the proposed modulator at on-state.

Fig. 3.41. Measured and simulated S-parameters of the proposed modulator at off-state.

Fig. 3.42. Measured and simulated large-signal performances of the proposed modulator at on-state at 60 GHz.

Fig. 3.43. Measured and simulated large-signal performances of the proposed modulator at on-state at 50 GHz.

Fig. 3.44. Measured and simulated large-signal performances of the proposed modulator at on-state at 52 GHz.

Fig. 3.45. Measured and simulated large-signal performances of the proposed modulator at on-state at 54 GHz.

Fig. 3.46. Measured and simulated large-signal performances of the proposed modulator at on-state at 56 GHz.

Fig. 3.47. Measured and simulated large-signal performances of the proposed modulator at on-state at 58 GHz.

Fig. 3.48. Measured and simulated large-signal performances of the proposed modulator at on-state at 62 GHz.

Fig. 3.49. Measured and simulated large-signal performances of the proposed modulator at on-state at 64 GHz.

Fig. 3.50. Measured and simulated large-signal performances of the proposed modulator at on-state at 66 GHz.

Fig. 3.51. Measured and simulated large-signal performances of the proposed modulator at on-state versus frequency.

Fig. 3.52 shows the measured and post-EM simulated large-signal performances at off-state at 60-GHz. As can be seen, the measurement shows a constant Pout at Pin < 2.4 dBm, as opposed to the linear increase shown by the simulation. This is due to the limit in sensitivity of the Agilent 4419B power meter used for output power measurement. The Pout level (including the losses of the probe and cables) at Pin < 2.4 dBm is below the sensitivity of the meter, and therefore constant measured results are shown instead. Once Pout level is above the sensitivity of the meter at Pin > 2.4 dBm, the measurement and simulation start to show good agreement. The measurement shows isolation performance of > 30 dB for Pin < 5 dBm. Compared with the measured on-state performances shown in Fig. 3.38, this means that the modulator can be driven at high Pin levels for output power performances at on-state, with little compromises in isolation performance at off-state.

Fig. 3.52. Measured and simulated large-signal performances of the proposed modulator at off-state at 60 GHz.

Table 3.3. Bias conditions during on/off-state small-signal measurements (continued). poses burden on the baseband circuit and may compromises the maximum data rate. A lower VG2,on

not only ease the burden on the baseband circuit, but also reduce the power consumption at on-state.

Furthermore, a higher maximum data rate is also possible. Therefore, on-state performances of the proposed modulator under lower VG2,on = 1.6 V, 1.4 V, and 1.2 V were also measured.

Table 3.3 shows the bias conditions during the small-signal measurements. Fig. 3.53 to 3.55 show the measured and post-EM simulated S-parameters at on-state under different VG2,on. As with the case under VG2,on = 1.8 V, the measured gain at 60 GHz is higher than simulation and the disagreement in S11 remains regardless of VG2,on. Both the measured and simulated S-parameters show little difference between different VG2,on. Gain at 60-GHz decrease only slightly with VG2,on, but the measured gain still maintains at around 10 dB at 60 GHz under VG2,on = 1.2 V.

Fig. 3.56 to 3.58 show the measured and post-EM simulated large-signal performances at on-state under different VG2,on versus frequency. The measured and post-EM simulated output power level and efficiency decreases with VG2,on. At 60 GHz, the measured OP1dB is 7.0 dBm, 6.8 dBm, 5.6dBm, and 2.3dBm under VG2,on = 1.8 V, 1.6 V, 1.4 V, and 1.2V, respectively. The significant drop in output power level under lower VG2,on is due to the drop in device gm, as can be seen from Fig. 3.12.

Fig. 3.54. Measured and simulated S-parameters of the proposed modulator at on-state with V_G2,on

= 1.4 V.

Fig. 3.53. Measured and simulated S-parameters of the proposed modulator at on-state with V_G2,on = 1.6 V .

Fig. 3.56. Measured and simulated large-signal performances of the proposed modulator versus frequency with V_G2,on = 1.6 V.

Fig. 3.55. Measured and simulated S-parameters of the proposed modulator at on-state with V_G2,on = 1.2 V.

Fig. 3.58. Measured and simulated large-signal performances of the proposed modulator versus frequency with V_G2,on = 1.2 V.

Fig. 3.57. Measured and simulated large-signal performances of the proposed modulator versus frequency with V_G2,on = 1.4 V.