A Unique System Concept to Improve the Accuracy of Wafer-Level Flicker-Noise Characterization. Andrej Rumiantsev, Stojan Kanev

Transcription

1 A Unique System Concept to Improve the Accuracy of Wafer-Level Flicker-Noise Characterization Andrej Rumiantsev, Stojan Kanev

2 Content Motivation Challenges for 1/f Measurements New Probe System Concept Verification Results Conclusion A. Rumiantsev, S. Kanev, MOS-AK

3 SUSS Motivation Parameter extraction and statistical analysis are based on Accurate on-wafer measurement results: DC (IV/CV) parameter RF (S-) parameter Low frequency noise (1/f) SUSS has to provide an invisible 1 test environment to reach the best possible measurement Accuracy 1 Below the accuracy of the measurement instruments A. Rumiantsev, S. Kanev, MOS-AK

4 Our motivation is your motivation Measurement Accuracy means: more efficient model extraction faster turnaround of the device models less design iterations faster time-to-market better return on investment (ROI) SUSS helps you 1 to make more money 1 With the best possible measurement accuracy A. Rumiantsev, S. Kanev, MOS-AK

5 Content Motivation Challenges for 1/f Measurements New Probe System Concept Verification Results Conclusion A. Rumiantsev, S. Kanev, MOS-AK

6 Challenges for 1/f Measurements Noise is an unavoidable consequence of semiconductor technology Since devices are becoming smaller and smaller, flicker noise becomes more important Its impact on circuit performance is significant in today's lowvoltage, high performance, mixed signal, RF, deep submicron designs The ability to accurately characterize noise becomes a fundamental requirement for design modeling! A. Rumiantsev, S. Kanev, MOS-AK

7 Challenges for 1/f Measurements Accuracy Passive system, Efficient shielding Manual prober Expensive EMIshielded room vs. Productivity Big amount of data Automated system New technologies do not allow compromises A. Rumiantsev, S. Kanev, MOS-AK

8 Content Motivation Challenges for 1/f Measurements New Probe System Concept Verification Results Conclusion A. Rumiantsev, S. Kanev, MOS-AK

9 New Probe System Concept Available 1/f Instrumentation: ProPlus solution w. 9812B and NoisePro software* Solution based on SR SR570 (current) pre-amplifier and Agilent IC-CAP** *[NoisePro], ** [Blaum 2001] A. Rumiantsev, S. Kanev, MOS-AK

10 New Probe System Concept The benefits of the new concept: 1. Eliminates the need for the expensive EMI-shielded room 2. Improves measurement accuracy 3. Increases productivity A. Rumiantsev, S. Kanev, MOS-AK

11 1. Expensive EMI-Shielded Room Faraday cage for the DUT and all sensitive system components No mechanical feed-throughs Faraday Cage Improved shielding efficiency A. Rumiantsev, S. Kanev, MOS-AK

12 2. Measurement Accuracy Convert automated system into electrically passive manual station: Turn off motors during the measurement circle Lock the chuck position Improved grounding to avoid ground loops Minimized AC noise Picture: [Blaum, 2001] A. Rumiantsev, S. Kanev, MOS-AK

13 2. Measurement Accuracy (cont.) The use of special ultra low-noise cables Reduced cable length Critical Signal Paths PreAmp Optimized cabling D B G S A. Rumiantsev, S. Kanev, MOS-AK

14 3. Productivity Automated measurement and parameter extraction Integration with industry standard measurement and modeling software Measurement at different temperatures Automated thermal management A. Rumiantsev, S. Kanev, MOS-AK

15 EMI Shielding: ProbeShield EMC2 Guard DSB G 1. Pre-Amplifier inside the EMI-shielded environment 2. Shortest possible low-noise triaxial cables for the Drain & Gate connection A. Rumiantsev, S. Kanev, MOS-AK

16 EMI Shielding: ProbeShield SR570 Pre-Amplifier ProbeShield inside (battery driven): The same EMC 2 Guard Option different cables EMC 2 Guard Shortest possible low noise cables for the low-pass filter RF pad design is recommended A. Rumiantsev, S. Kanev, MOS-AK

17 Cabling : A Close Look Drain Gate Fully grounded probes are obligation Ground connecting close to the DUT Ultra low noise triaxial cables (part of EMC 2 Guard) A. Rumiantsev, S. Kanev, MOS-AK

18 Content Motivation Challenges for 1/f Measurements New Probe System Concept Verification Results Conclusion A. Rumiantsev, S. Kanev, MOS-AK

19 EMI-Shielded Effectiveness EMI-certified test room Rohde & Schwarz EMI analyzer Measurements for different conditions: Front Rear Side A. Rumiantsev, S. Kanev, MOS-AK

20 EMI-Shielded Effectiveness (cont.) Reference measurement Front measurement (1) Side measurement (2) Rear measurement (3) EMI-shielded effectiveness >25dB from 1kHz to 70kHz A. Rumiantsev, S. Kanev, MOS-AK

21 Experimental Setup SUSS PA300PS ProbeShield w. EMC 2 Guard option ProPlus solution Chamber Noise pmos FET SiGe HBT A. Rumiantsev, S. Kanev, MOS-AK

22 System Configuration DUT SUSS PA300PS ProbeShield Completely EMI Shielded Design 4x Triaxial or RF Probes D Noise floor ~10-24 A2/Hz 1 Hz khz (1MHz*) B G SMU s cable connectors Pre-Amplifier (Amplifier / Filter) Unit Semiconductor Parameter Analyzer** Keithley 4200 or Agilent 4156x S SUSS NOISEpro driver available Noise Output Dynamic Signal Analyzer Agilent 35670A or Stanford Research SR B Noise Analyzer Controller Unit GPIB Bus * If use Agilent Vector Signal Analyzers 89610A or 89410A PC with NOISEpro & ProberBench Software A. Rumiantsev, S. Kanev, MOS-AK 2008.dll LAN Interface to ProberBench Electronics 22

23 Chamber Noise Verification Chamber Noise Verification using NoisePro Test < -120dB A. Rumiantsev, S. Kanev, MOS-AK

24 Verification with the DUT DUT type: pmos FET Same bias conditions Same temperature conditions Old probe system configuration: Preamplifier outside the EMI-shielded environment New probe system concept: Preamplifier inside the ProbeShield EMC 2 chamber A. Rumiantsev, S. Kanev, MOS-AK

25 Verification with the DUT Old system concept (PreAmp outside ) New system concept (PreAmp inside ) A. Rumiantsev, S. Kanev, MOS-AK

26 Verification with the DUT DUT type: SiGe HBT New probe system concept Different bias conditions Different temperature conditions A. Rumiantsev, S. Kanev, MOS-AK

27 Verification with SiGe HBT Under different biases SUSS confidential Line TC =25 C (ON) V CE = 1.5V Bias 1 I B = 40nA Bias 2 I B = 500nA Bias 3 I B = 8µA A. Rumiantsev, S. Kanev, MOS-AK

28 Verification with SiGe HBT Under different biases and temperatures Ae/Ac/Ab/T=10/10/10/25 Sib (A²/Hz) 1.0e e e e e e e e e e e e e e e e1 1.0e2 1.0e3 1.0e4 1.0e5 Frequency (Hz) 1.0e C 150 C Sic (A²/Hz) 1.0e e e e e e e e e e e1 1.0e2 1.0e3 1.0e4 1.0e5 V CE = 1.5V Bias 1 I B = 10nA Bias 2 I B = 40nA Bias 3 I B = 400nA Bias 4 I B = 800nA Bias 5 I B = 2µA Bias 6 I B = 20µA A. Rumiantsev, S. Kanev, MOS-AK

29 Content Motivation Challenges for 1/f Measurements New Probe System Concept Verification Results Conclusion A. Rumiantsev, S. Kanev, MOS-AK

30 Conclusion New Probe System concept was introduced EMC 2 Guard Save investment costs Increased measurement accuracy More efficient modeling extraction System automation for unattended test Faster time to data Turn-key solution Seamless integration of the industry standard measurement and modeling solution (such as ProPlus) A. Rumiantsev, S. Kanev, MOS-AK

31 References [1] M. Valenza, A. Hoffmann, D. Sodini, A. Laigle, F. Martinez, and D. Rigaud, "Overview of the impact of downscaling technology on 1/f noise in p-mosfets to 90 nm," IEE Proc. Circuits, Dev. and Systems, vol. 151, no. 2, pp , [2] "NoisePro 1/f noise characterization and modeling system," Product Brochure, ProPlus Design Solutions, Inc., San Jose, CA, USA., [3] G. M. Balim, M. G. Levina, and S. S. Smakhtin, "Nonequilibrium 1/f noise and hardware reliability," in Proc. 1st IEEE Int. Conf. Circuits Systems Communications, 2002, pp [4] A. P. van der Wel, E. A. M. Klumperink, S. L. J. Gierkink, R. F. Wassenaar, and H. Wallinga, "MOSFET 1/f noise measurement under switched bias conditions," IEEE El. Dev. Lett., vol. 21, no. 1, pp , [5] A. K. Kirtania, M. B. Das, S. Chandrasekhar, L. M. Lunardi, G. J. Qua, R. A. Hamm, and L.-W. Yang, "Measurement and comparison of 1/f noise and g-r noise in silicon homojunction and III-V heterojunction bipolar transistors," IEEE Trans. El. Dev., vol. 43, no. 5, pp , [6] A. Blaum, O. Pilloud, G. Scalea, J. A. V. J. Victory, and F. A. S. F. Sischka, "A new robust on-wafer 1/f noise measurement and characterization system," in Proc. Int. Conf. Microelectronic Test Structures, 2001, pp [7] "SR560 - DC to 1 MHz voltage preamplifier," Data Sheet, Stanford Research Systems, Inc., Sunnyvale, CA, USA. [8] "SR570 - DC to 1 MHz current preamplifier," Data Sheet, Stanford Research Systems, Inc., Sunnyvale, CA, USA. [9] "Model 4200-SCS. Semiconductor Characterization System," Product Brochure, Keithley Instruments, Inc., Cleveland, Ohio,USA, [10] "SR khz 2-channel dynamic signal analyzer," Data Sheet, Stanford Research Systems, Inc., Sunnyvale, CA, USA, [11] EMI/RFI Shielding Effectiveness of SUSS ProbeShield Technology, Technical Brief, SUSS MicroTec Test Systems GmbH, Sacka near Dresden, Germany, A. Rumiantsev, S. Kanev, MOS-AK