Synopsys, Inc. · NASDAQ:SNPS
Synopsys — Technical Presentation (2002)
28 slides from Synopsys's technical presentation, 2002
Slide 1: Experiences with SABER: Data Sheet Driven Dynamic Power Semiconductor Models
Experiences with SABER Data Sheet Driven Dynamic Power Semiconductor Models Dr.-Ing. Wolfgang Fischer University of Magdeburg, Germany Institute of Electric Power Systems (IESY) wfischer@ieee.org SNUG Saber 2002 Munich
Slide 2: Purpose of the work
Purpose of the work Design of modern electronic power sources for welding processes requires powerful circuit simulation tools Switching frequencies reach 100 kHz and above at 1...20 kW rated power Power semiconductors (MOSFETs, IGBTs, diodes) determine efficiency, E.M.I., control characteristics et
Slide 3: Outline
Outline 1 Introduction 2 Power Semiconductor Models 3 Static and Dynamic Parameterization of SABER MOSFET model mp1 4 Application example 5 Conclusion SNUG Saber 2002 Munich 3
Slide 4: Problems to be Solved in Design Process
1 Problems to be Solved in Design Process Switching power losses Transient overvoltages Parasitic effects Ringing effects - conducted noise (EMI) Line-current harmonics (Power factor) Driver design (optimization of switching losses) Semiconductors in parallel Resonant topologies Magnetics (transform
Slide 5: Simulation Software for Power Electronics
1 Simulation Software for Power Electronics SABER www.synopsys.com SIMPLORER www.ansoft.com PSpice www.pspice.com IsSpice www.intusoft.com PowerSim www.powersimtech.com .... SNUG Saber 2002 Munich 5
Slide 6: Power Semiconductor Models
2 Power Semiconductor Models Ideal switching behavior, logic control not suited for switching losses, switching transients very fast simulation, well suited for overview, control loop design Physical based physical parameters mostly unknown, high complexity Data sheet based behavioral input/output t
Slide 7: Information Available from Data Sheets
2 Information Available from Data Sheets Characteristic values – Absolute maximum ratings – Thermal data – Avalanche characteristics – Electrical characteristics • static (on, off) • dynamic (C) • turn-off characteristics • turn-on characteristics – Body diode characteristics (MOSFET only) Diagrams
Slide 8: SABER MOSFET Model mp1
3 SABER MOSFET Model mp1 Drain mp1 Gate Source Typical: 3 capacitances 10...500 A (CGS,CGD nonlinear 100...800 V CDS constant) Module package IMOS = f (VDS,VGS) internal RG SNUG Saber 2002 Munich 8
Slide 9: 10 Static Parameters
3 10 Static Parameters Parameter Description STE26NA90 vds at P1 (between quasi- vds1 27 V linear and saturated area) vgs1 vgs at P1 5V id1 id at P1 63 A vds at P2 (between quasi- vds2 2V linear and saturated area) vgs2 vgs at P2 4V id2 id at P2 4A vds at P3 (in quasi-linear vds3 18 V area) vgs3 vgs
Slide 10: Static Parameterization
3 Static Parameterization Data sheet STE26NA90 ID (A) ID (A) P3 P1 P2 VDS (V) Vt VGS (V) Output characteristics Transfer characteristic Problems: Temperature dependency Determination of data points (only a part is drawn, boundaries not clear) SNUG Saber 2002 Munich 10
Slide 11: Test Circuit mos1
3 Test Circuit mos1 Static Output Characteristics VGS VDS 1kV 3..15 V 50 ms 100 ms t 100ms 200ms t SNUG Saber 2002 Munich 11
Slide 12: Simulated Output Characteristics
3 Simulated Output Characteristics ID/A STE26NA90 VGS = 6...15 V VGS = 5 V Simulation RDSon = 0,25 Ohm VGS = 4,5 V @ 26 A Data sheet VGS = 4 V VDS/V SNUG Saber 2002 Munich 12
Slide 13: Test Circuit mos2
3 Test Circuit mos2 Static Transfer Characteristic VDS VGS 25 V 15 V t 5 ms 10 ms t SNUG Saber 2002 Munich 13
Slide 14: Simulated Transfer Characteristic
3 Simulated Transfer Characteristic ID/A Simulation Data sheet VGS(th) STE26NA90 VGS/V SNUG Saber 2002 Munich 14
Slide 15: 14 Dynamic Parameters mp1
3 14 Dynamic Parameters mp1 Parameter Description STE26NA90 crss1 reverse transfer cap. vds=vgs=0 3 nF rev. transf. cap. vds= vds4, crss2 270 pF vgs=0 CDS coss1 output cap. vds=vgs=0 7 nF coss2 output cap. vds= vds4, vgs=0 1.13 nF vds4 vds on crss curve 25 V q1 Qg at beginning of Miller-plat. 43 nC
Slide 16: Simulated Interelectrode Capacitances
3 Simulated Interelectrode Capacitances C/F STE26NA90 STE26NA90 mos3 VDS/V SNUG Saber 2002 Munich 16
Slide 17: Simulated Gate-Charge Plot
3 Simulated Gate - Charge Plot VGS/ V Simulation Data sheet STE26NA90 QG/As SNUG Saber 2002 Munich 17
Slide 18: Test Circuit mos5
3 Test Circuit mos5 Inductive turn-off modification SNUG Saber 2002 Munich 18
Slide 19: Simulated MOSFET turn-off
3 Simulated MOSFET turn-off VDS/ V STE26NA90 VGS/ V ID / A t/ s SNUG Saber 2002 Munich 19
Slide 20: Switching behavior: Data sheet vs. Simulation
3 Switching behavior: Data sheet vs. Simulation SABER SABER Parameter Description Data sheet mp1 mp1 (first) (adjusted) td(off) turn-off delay not given 199.3 ns 200 ns tr(Voff) off-voltage rise time 108...152 ns 197.2 ns 109.3 ns tf fall time 25...35 ns 30.1 ns 22.3 ns tc cross-over time 145...203
Slide 21: Iterative Adjustment
3 Iterative Adjustment Data sheet: Simulation test characteristics circuits: ratings Output characteristics Data sheet based diagramms Transfer characteristic model (mp1): Gate charge plot 10 static parameters Nonlinear capacitances Measurement: 14 dynamic parameters Turn-on (Sw. times, energy) stat
Slide 22: A Modern Arc Welding Power Source
4 A Modern Arc Welding Power Source Uncontrolled PFC Inverter M.F. Uncontrolled Low frequency line rectifier transformer ouput rectifier inverter (optional) 350...700 V d.c. 350/700 V to 100 V Input: Inverter: Output: 230V/400V 50/60 Hz 50...200 kHz • 10..500 A/12...40 V d.c. single-/three-phase IGB
Slide 23: Inverter Circuit Simulation Model
4 Inverter Circuit Simulation Model MOSFET +Bodydiode Driver T3 T1 d d vp s s Transformer vp vm vm Load p2 Rectifier p1 m2 Snubber p3 m1 m3 Parasitics d T2 d T4 s vp s vp vm vm SNUG Saber 2002 Munich 23
Slide 24: Simulation Results: Switching Waveforms
4 Simulation Results: Switching Waveforms VGS/V 10.0 0.0 -10.0 -20.0 transient overvoltage -30.0 ringing effects (EMI) VDS/V parasitics feedback effects 600.0 400.0 200.0 0.0 Poff ID/A 20.0 10.0 0.0 -10.0 secondary snubber t/s 210u 212u 214u 216u 218u 220u 222u 224u 226u 228u SNUG Saber 2002 Munich
Slide 25: Simulation Results: Power Loss Analysis
4 Simulation Results: Power Loss Analysis Element Power losses MOSFET (4) 157 W Rectifier diodes (4) 175.8 W Transformer 81 W Snubber (2) 4.2 W Bodydiode (4) 9.6 W Total 427.6 W Pin 4631.9 W Pout 4189.3 W Pin - Pout 442.6 W SNUG Saber 2002 Munich 25
Slide 26: Benchmarking
4 Benchmarking Circuit model: 4 dynamic MOSFETs with body OS/Hardware CPU time diodes Solaris 2 x 2 rectifier diodes with snubber Sun Ultra 1 77.7 sec 143 MHz, 256 MB RAM Phase-shift control (circuit level) WinNT Transformer with linear core and 300 MHz Intel PII 40 sec leakage inductance 256 MB RAM
Slide 27: Conclusion
5 Conclusion Data sheet based behavioral semiconductor models can be a (good) alternative to physical based models for solving special problems Data sheet based models give a higher flexibility with new and improved devices and increase efficiency of design process Todays data sheets give sufficient
Slide 28: Conclusion
5 Conclusion Data sheet models include thermal behavior, but do not consider thermal parameter change Some model characteristics should be a topic of further development Model documentation is still (very?) poor CPU time and file size are not longer a problem using modern hardware Modeling and simul