Application Note 5098

Transcription

1 LO Buffer Applications using Avago Technologies ABA-3X563 Silicon Amplifiers Application Note 5098 Introduction An oscillator or a voltage-controlled oscillator (VCO) is usually buffered with an external amplifier prior to the LO port of a mixer as in Figure 1. This buffer amplifier is used to prevent a shift on the oscillating frequency, or a frequency pull, caused by the load variation at the oscillator s output. It provides good reverse isolation and delivers a constant power level to the LO port. The Avago Technologies ABA-3X563 RFICs are a series of silicon broadband amplifiers designed to operate from a 3 V voltage supply. This series offers flat gain (>20 db) across a wide frequency band (up to 3.5 GHz for the ABA and 2.5 GHz for the ABA-32563), good input and output return loses, unconditional stability, good reverse isolation (30 db) and flat output power characteristics. These characteristics make this amplifier suitable for LO buffer applications. This note will demonstrate the characteristics of the ABA-3X563 series for LO buffer applications at 2 GHz. Avago Technologies Broadband Amplifier, ABA-3X563 series The ABA-3X563 is designed to be close to 50 W at the input and output pins. As a result, this device requires few additional passive components to operate in a 50 W system. The wideband characteristic of this device is achieved by means of internal feedback that also limits variation in gain over. Using the components listed in Figure 3, the output versus input power characteristic was measured. As seen in Figure 2, these devices exhibited almost constant output power across a wide range of input power. Operating the amplifier in the saturated region improves AM suppression and phase noise. For instance, the ABA showed a constant output power of +9.8 dbm with a variation of less than 0.2 db across -5 dbm to +5 dbm input power. In addition to ABA-3X563 series that operate from 3 V voltage supply, Avago Technologies also offers ABA-5X563 series that operate at 5 V. Table 1 summarizes the performance of the ABA-3X563 and ABA-5X563 series using the schematic in Figure 3. Oscillator Attenuator 3 V LPF Attenuator Mixer / Modulator 0 dbm +/-3 dbm 0 dbm Nominal ABA-3X563 Figure 1. Example of LO buffer application using the ABA-3X563

2 20 ABA-3X563 : Pout vs. Pin at 2 GHz 15 Pout (dbm) dbm +4.9 dbm 0 ABA ABA Pin (dbm) Figure 2. Output versus input power curve for ABA-3X563 series at 2 GHz Table 1. Demonstration board performance comparison at 2 GHz using the schematic and components in Figure 2 Symbol Unit ABA ABA ABA ABA ABA ABA Vc V Icc ma P1dB dbm OIP3 dbm Gp db NF db VSWR in VSWR out V 390 pf (opt) 18 pf 18 pf ABA-3X nh input 18 pf 18 pf output Figure 3. Schematic and components for 2 GHz application board 2

3 ABA-3X563 performances This note will demonstrate the characteristics of the ABA-3X563 device based on the following conditions. 1. Input versus output power over 2. Output power versus frequency over 3. Isolation versus frequency over 4. Input and output return loss versus frequency over 5. Current versus input power over 6. Second harmonic versus frequency 7. Phase noise measurement The measurements were done using devices soldered on the test board DEMO-ABA-3X563 with the components listed in Figure 3. The test board and the ABA-3X563 samples can be ordered from our web site: 1. Input versus output power over These devices exhibited constant output power over a wide range of input power. Between -10 dbm and +0 dbm input power, the ABA delivered about +4.8 dbm to +4 dbm output power at. At, the output power degraded by +0.5 db while at, the power increased by 0.5 db. Between -5 dbm to 5 dbm, the ABA delivered about +9.4 dbm to +10 dbm output power at 25 C. At 85C, the output power degraded by +0.5 db while at, it increased by 0.5 db. Pout (dbm) Figure 4a. ABA output versus input power over Pout (dbm) ABA at 2 GHz ABA at 2 GHz Pin (dbm) Pin (dbm) Figure 3b. ABA output versus input power over

4 2. Output power versus frequency over Using an input power of 0 dbm, the output power was measured against the frequency at three s (-55, 25 and ). The ABA series can easily deliver a constant output power over frequency with acceptable amplitude variation. For instance, the ABA delivered about +4.8 dbm to +5.2 dbm of power across a 400 MHz frequency bandwidth from 1800 MHz to 2200 MHz. The ABA delivered about +9.5 dbm to +9.8 dbm of power across the above frequency range. The variation with is also relatively small. Pout (dbm) ABA : Pout vs. Frequency (Pin = 0 dbm) 3 Figure 5a. ABA output power versus frequency over 11 ABA : Pout vs. Frequency (Pin = 0 dbm) 10 Pout (dbm) 9 8 Figure 5b. ABA output power versus frequency over

5 3. Isolation versus frequency over Another important characteristic of a buffer amplifier is the reverse isolation. The following figures show the measured isolation of the ABA-3X563 series. For the ABA-31563, the isolation is 33 db across 1800 MHz to 2200 MHz at an input power of 0 dbm. The isolation for ABA is around 31 db across the frequency range. The variation of isolation over is very small. Isolation (db) ABA : Isolation vs. Frequency (Pin = 0 dbm) -37 Figure 6a. ABA isolation versus frequency over -28 ABA : Isolation vs. Frequency (Pin = 0 dbm) -30 Isolation (db) Figure 6b. ABA isolation versus frequency over

6 4. Input and output return loss versus frequency over -4 ABA : Input RL vs. Frequency (Pin = 0 dbm) As the device is operating in the saturation region, the input and output return losses need to be understood. Using 0 dbm input power, the input and output return losses were measured over. The ABA s input return loss is better than 11 db and the output return loss is better than 13 db from 1800 MHz to 2200 MHz frequency range. The ABA s input return loss is better than 12 db and the output return loss is better than 14 db at that frequency range S11 (db) Figure 7a. ABA input return loss versus frequency over -8 ABA : Input RL vs. Frequency (Pin = 0 dbm) S11 (db) Figure 7b. ABA input return loss versus frequency over

7 -6 ABA : Output RL vs. Frequency (Pin = 0 dbm) S22 (db) Figure 8a. ABA output return loss versus frequency over -6 ABA : Output RL vs. Frequency (Pin = 0 dbm) S22 (db) Figure 8b. ABA output return loss versus frequency over

8 5. Current versus input power over The current is another important factor that needs to be considered when operating at high input power. The following figures illustrate the change of current with input power over using a fixed supply voltage at 3 V. Icc(mA) ABA at 2 GHz Pin (dbm) Figure 9a. ABA Icc versus input power over 41 ABA at 2 GHz Icc (ma) Pin (dbm) Figure 9b. ABA Icc versus input power over

9 6. Second harmonic versus frequency As the amplifier is operating at higher input power, the second harmonic performance may become important. Using an input power of 0 dbm, the second harmonic was investigated across frequency 1.6 GHz to 2.4 GHz. For the ABA-32563, the second harmonic at 2 GHz is about 22 dbc below the fundamental. This is translated to an output power of about -12 dbm at 4 GHz with +9.8 dbm at 2 GHz. Use of a low-pass filter (LPF) as shown in Figure 1 will eliminate the potential harmonic problem in the full design application. 2nd Harmonic (dbc) Second Harmonic vs. Frequency at 0 dbm ABA ABA Frequency (GHz) Figure 10. ABA-3X563 second harmonic versus frequency

10 7. Phase noise performance The phase noise of these devices was measured using Avago Technologies HP4532 VCO and PLL analyzer. Two signal generators with low noise option were used in the test setup. A low noise option of Avago Technologies signal generator 8664A was used as a reference signal source for the analyzer. The phase noise was first measured directly from a signal generator 83711B. Then, it was measured again with the ABA device inserted between the 83711B and the analyzer. Theoretically, the phase noise at the noise floor region would degrade with the presence of a buffer amplifier, which mainly attributed by the amplifier s noise figure. Following are the phase noise curves measured at 2 GHz. It is noted that the phase noise curve with the buffer amplifier did not change much compared to without the buffer, especially at the frequency far away (more than 1MHz) from the carrier frequency. Phase Noise (dbc/hz) ABA Phase Noise vs. Frequency at 2 GHz freq (Hz) Figure 11a. ABA phase noise measurement at 2 GHz ABA Phase Noise vs. Frequency at 2 GHz ABA31_0 dbm SG_4.8 dbm ABA32_0 dbm SG_9.8 dbm -110 Phase Noise (dbc/hz) freq (Hz) Figure 11b. ABA phase noise measurement at 2 GHz

11 Conclusion Result shows that all four ABA-3X563 devices show good performance under saturation conditions and are well suited for LO buffering applications. Higher reverse isolation can be achieved by cascading two or more ABA-3X563 devices appropriately. Although this work covers buffer applications at 2 GHz, the ABA-3X563 devices will show similar behavior over their whole operating frequency range. Summary Table The following summary was obtained from small samples of ABA-3X563 devices. The results can be used as a reference on the performance but shall not be used as final design values. Table 2. Performance summary at ABA ABA Input Power Range -10 dbm to 0 dbm -5 dbm to +5 dbm Output Power 2 GHz +4.2 dbm to +4.8 dbm +9.4 dbm to dbm Icc 2 GHz 17 ma to 17.5 ma 38.5 ma to 39.5 ma Freq Pin=0 db 1.8 GHz to 2.2 GHz 1.8 GHz to 2.2 GHz Output Power Pin=0 dbm +4.8 dbm to +5.2 dbm dbm to dbm Pin=0 dbm 33.5 db 31 db Input Pin=0 dbm >11 db >11 db Output Pin=0 dbm >12.5 db >14 db 2 nd Pin=0 dbm >8 dbc >20 dbc References 1. Information on ABA-3X563 devices can be found at our web site: 2. Avago Technologies Application Note 5043: ABA GHz Broadband Silicon RFIC Amplifier. 3. Avago Technologies Application Note 5042: ABA GHz Broadband Silicon RFIC Amplifier. 4. Avago Technologies Application Note 5052: ABA High P1dB Broadband Silicon Amplifier. 5. R.W Rhea Oscillator design and computer simulation, 2 nd ed., Noble Publishing Corporation, Georgia, 1995, pp For product information and a complete list of distributors, please go to our web site: Avago, Avago Technologies, and the A logo are trademarks of Avago Technologies in the United States and other countries. Data subject to change. Copyright Avago Technologies. All rights reserved EN - September 1, 2010