Low-Distortion Signal Generation for ADC Testing

Transcription

1 214 IEEE International Test Conference Low-Distortion Signal Generation for ADC Testing Fumitaka Abe, Yutaro Kobayashi Kenji Sawada, Keisuke Kato Osamu Kobayashi, Haruo Kobayashi Gunma University STARC

2 2/41 Research Objectives Use an Arbitrary Waveform Generator (AWG) as low-cost low-distortion signal source for ADC testing Validate our 3 rd -harmonic-cancelling phase-switching signal generation technique Perform analysis, simulation, and experiments using AWGs for ADC testing

3 Expected Advantage of Our Solution Conventional Method Filter A Tough LPF requirements Filter B 4dB 4dB f in 1kHz 3f in 3kHz f in 3kHz 3f in 9kHz Our Method Filter C Relaxed LPF requirements Filter C 1dB f in 1kHz 1 2 f s 15kHz 1 2 f s f in 14kHz f in 3kHz 1 2 f s f in 12kHz 1 2 f s 15kHz 3/41

4 4/41 Outline Background to this research Proposed solution Problems with proposed solution Remedy 1 Remedy 2 Experimental results Conclusions

5 5/41 Outline Background to this research Proposed solution Problems with proposed solution Remedy 1 Remedy 2 Experimental results Conclusions

6 6/41 Background to this Research ADCs are important part of mixed-signal SOCs. Need a low-cost, low-distortion signal source for ADC linearity testing. #1: Use existing AWGs in testers low cost # 2: Develop a technique that doesn t require identification of AWG nonlinearity, just a DSP program change.

7 7/41 ADC Testing with Sine Signal ADC test signal Test result AWG ADC Reference level Allowable fundamental Go ADC linearity test fundamental ADC HD3 No Go ADC HD3 measurement

8 Problem with Conventional Method ADC test signal AWG ADC Test result fundamental HD3 Test signal Generation AWG HD3+ADC HD3 fundamental ADC HD3 cannot be measured accurately. Low quality test AWG HD3 8/41

9 Research Objective ADC test signal AWG Test signal Generation ADC ADC test result fundamental HD3 only program change fundamental Only ADC HD3 High quality test AWG HD3 reduction 9/41

10 1/41 Outline Background to this research Proposed solution Problems with proposed solution Remedy 1 Remedy 2 Experimental results Conclusion

11 Power [db] 11/41 Conventional Signal Generation with AWG AWG DSP Din DAC A CLK Din X X X X AWG sampling frequency: fs(awg) = 1/Ts fin DAC has 3 rd order nonlinearity fin Normalized frequency f/fs

12 Power [db] Proposed Signal Generation with AWG AWG Θ = π/3 DSP Din DAC CLK Din X X1 X X1 X = 1.15Acos(2πf in nt s π/6) fs/2-fin X 1 = 1.15Acos(2πf in nt s + π/6) -1-2 fin fs/2-3fin HD3 is cancelled Normalized frequency f/fs 12/41

13 Phase Switching Conventional 3 rd order non-linear system Phase rotation by x3 Principle of 3 rd Harmonics Cancellation fundamental: fin 3 rd harmonics: 3 fin Θ = π/3 3Θ = π Two waves with phase difference π are cancelled 13/41

14 Voltage [V] Power [dbm] Voltage [V] Power [dbm] Conventional and Phase Switching Signals 1V Waveform Conventional Measured Power Spectrum fundamental:9.dbm HD3:-68.dBm 3 rd harmonic 3 rd harmonic Time [sec] Frequency [khz] Frequency [MHz] 1.15V Waveform Phase Switching fundamental:9.dbm HD3:-78.2dBm fs/2-fin :-.91dBm 3 rd harmonic 3 rd harmonic Time [sec] π/3 Frequency [khz] Frequency [MHz] 14/41

15 15/41 Outline Research background Proposed solution Problems of proposed solution Remedy 1 Remedy 2 Experimental results Conclusion

16 Power [db] Power [db] Power [db] Power [db] Power [db] Power [db] 16/41 Conditions of HD3 Cancellation in Cascaded System Input signal 3 rd nonlinearity 3 rd nonlinearity 3 rd 3 次 nonlinearity 非線形 -1 generation -1 generation -1 generation Conventional Normalized Frequency f/fs Normalized Frequency f/fs Normalized Frequency f/fs Phase switching -2-3 cancellation Normalized Frequency f/fs -2-3 cancellation Normalized Frequency f/fs -2-3 cancellation Normalized Frequency f/fs

17 Problem of Phase Switching Signal Generation Ideal Nonlinearity Model fundamental ADC fundamental HD3 Phase switching Spurious due to phase switching HD3 due to ADC nonlinearity fundamental ADC fundamental HD3 Reduction of HD3 due to AWG nonlinearity Reduction of HD3 due to ADC nonlinearity Big problem! 17/41

18 18/41 Model for Theoretical Analysis AWG Input with Phase Switching n: odd AWG Nonlinearity Model n: even ADC Nonlinearity Model For simplicity fs(awg)=fs(adc) w/o LPF AWG ADC

19 19/41 AWG Output signal large spurious small

20 Direct Application of Phase Switching Signal ADC output Z(n) components signal HD3 By calculation : : CANNOT measure ADC HD3. 2/41

21 21/41 Outline Research background Proposed solution Problems of proposed solution Remedy 1 Remedy 2 Experimental results Conclusion

22 AWG Output and Low Pass Filtering w/ LPF AWG LPF ADC 22/41

23 Application of Phase Switching and LPF ADC output Z(n) components signal HD3 By calculation : : ADC HD3 can be measured. 23/41

24 Spurious Attenuation Effect AWG output and low pass filtering q 24/41

25 図 2.1(B) ADC の HD3 3 次高調波の本来値からのずれ Measurement Error [%] [%] 25/41 LPF reduction ADC MATLAB Simulation Results Spurious reduction at fs/2-fin HD3 measurement Error 1-2 1dB 2dB 1%.1% fs/2-finスプリアスの減衰量 [db] Spurious reduction at fs/2-fin [db]

26 26/41 Outline Research background Proposed solution Problems of proposed solution Remedy 1 Remedy 2 Experimental results Conclusion

27 Voltage Codes ADC Analog input Inside ADC Digital output Reproduction of phase switching signal Time π/3 π HD3 cancellation Sample Number Similar ADC HD3 can NOT be measured accurately 27/41

28 Voltage Codes ADC folded Analog input Inside ADC Digital output Reproduction of phase switching signal 3 2 Time π/3 π HD3 cancellation Sample Number ADC HD3 can be measured accurately Different 28/41

29 29/41 Conditions for Accurate ADC HD3 Measurement LPF reduction ADC folded 1 2 Change sampling frequency by folding Accurate measurement of ADC output HD3 with phase switching signal

30 3/41 Outline Research background Proposed solution Problems of proposed solution Remedy 1 Remedy 2 Experimental results Conclusion

31 31/41 ADC 3 rd Harmonic Measurement Diagram Agilent 3322A Oscilloscope, Spectrum Analyzer AWG LPF ADC 14bit DAC Satisfying Remedy 1 digital output CLK ADI AD7356 MHz 12bit SAR ADC PC Evaluation board EVAL CED1Z FFT AD7356 output ADC test signal generation Conventional Phase switching Satisfying Remedy 2

32 Experimental Environment for ADC Testing PC1 for test signal program Spectrum Analyzer for test signal analysis fs(awg)=1mhz, fin=2khz fs(adc)= mhz Oscilloscope for test signal analysis AWG 3322A for test signal generation Common mode noise suppression by a choke coil PC2 for ADC output analysis EVAL-CED1Z for generating sampling clock, etc. ADC AD7356 (12bit) 6 samples Device Under Test Low Pass Filters 32/41

33 Analog LPF Design for ADC Testing Oscilloscope, Spectrum Analyzer DUT AWG LPF ADC digital output CLK PC Evaluation board FFT AD7356 output Agilent 3322A Design Implementation Evaluation ADI AD MHz EVAL CED1Z 2kHz HD3@6kHz removal 2kHz Phase switching attenuation fs/2-fin fc=25khz 5 th order LC Butterworth LPF fs/2-fin fc= 1MHz, 2MHz, 2.7MHz, 3.7MHz 4 th order LC Butterworth LPF 33/41

34 34/41 LPF Implementation Oscilloscope, Spectrum Analyzer DUT AWG LPF ADC digital output CLK PC Evaluation board FFT AD7356 output Agilent 3322A Design Implementation Evaluation ADI AD MHz EVAL CED1Z fc=25khz fc=1mhz fc=2mhz fc=2.7mhz fc=3.7mhz fc=1mhz fc=2mhz fc=2.7mhz fc=3.7mhz for HD3 reduction for fs/2-fin spurious reduction

35 Gain[dB] LPF Evaluation with Frequency Response Analysis Oscilloscope, Spectrum Analyzer DUT AWG LPF ADC Agilent 3322A Design Implementation Evaluation ADI AD7356 digital output CLK MHz PC Evaluation board EVAL CED1Z FFT AD7356 output HD Frequency[Hz] fc=25khz fc=1mhz fc=2mhz fc=2.7mhz fc=3.7mhz - 6kHz - 54 db@4.8mhz - 3 db@4.8mhz - 17 db@4.8mhz - 8. db@4.8mhz 35/41

36 Voltage [V] Codes Voltage [V] Codes Voltage [V] Codes Voltage [V] Voltage [V] Codes Voltage [V] Codes Measured Waveforms of ADC Input and Output Phase Switching Signal kHz Time LPF Phase switching spurious@4.8mhz attenuation fs/2-fin Conventional fs/2-fin Spurious -8dB -17dB -3dB -54dB Analog Input Time Time Time Time Time A/D Conversion Sample Number Sample Number Sample Number Sample Number 3 2 Digital output Sample Number Spectrum Analysis HD3 Measurement 36/41

37 power [dbfs] Power [dbfs] Power [dbfs] 37/41 Comparison of Conventional and Proposed Methods Conventional w/o LPF Phase switching w/ LPF (fc=1mhz) rd 3 rd LPF - 54dB 2-2+ Frequency [MHz] Frequency [MHz] Conventional w/ LPF (fc=25khz) True ADC HD3: -94.6dBFs 3 rd True ADC HD3 Conventional method Measured HD3: -88.1dBFs Error 6.8% AWG 由来 3 次高調波をフィルタによりカットした手法 Frequency [MHz] Proposed method Measured HD3: -92.6dBFs Error 2.1%

38 ADC 3 rd harmonics detection error [%] ADC 3 rd harmonics detection error [%] ADC 3 rd harmonics detection error [%] ADC 3 rd harmonics detection error [%] ADC 3 rd harmonics detection error [%] ADC 3 rd harmonics detection error [%] ADC HD3 Measurement Results Spurious Supurious@fs/2-fin@ attenuation [db] [db] conventional phase switching Supurious@fs/2-fin attenuation [db] attenuation [db] ADC (AD7356) HD3 measurement error reduction is verified 18 Sample 1 Sample 2 Sample Spurious attenuation attenuation [db] [db] Supurious@fs/2-fin fs/2-fin attenuation [db] attenuation [db] Supurious@fs/2-fin attenuation [db] [db] 18 Sample 4 Sample 5 Sample Supurious@fs/2-fin attenuation [db] attenuation [db] 38/41

39 39/41 Outline Research background Proposed solution Problems of proposed solution Remedy 1 Remedy 2 Experimental results Conclusion

40 Conclusions Low distortion sine signal generation Without AWG hardware modification Just AWG program change and a simple analog LPF Verified with AWG (Agilent 3322) 6 samples of 12bit SAR ADC (AD7356) Greatly improved quality of ADC Linearity testing at virtually no extra cost. 4/41

41 41/41 Future Work Generalization to other types of low distortion sinusoidal signal generation - HD2, HD2&HD3 cancellation for 1-tone - IMD3 cancellation for 2-tone We have partially verified these. Detailed theoretical analysis, simulations, and experiments are underway.