Worst Case Circuit Analysis (WCCA) White Paper

  Worst case circuit analysis (WCCA) is a technique which, by accounting for component variability, determines circuit performance under a worst case scenario, i.e., under extreme environmental or operating conditions. Environmental conditions are defined as external stresses applied to each circuit component, and can include temperature, humidity or radiation. Operating conditions include external electrical inputs, but must also consider factors such as component quality level, interaction between parts, and drift due to component aging. The output of a WCCA allows an assessment of actual applied part stresses against rated part parameters. This can help ensure sufficient part stress derating to meet design requirements. WCCA should be considered for all circuitry that is safety and/or financially critical. Performance of a WCCA, and implementation of its results, can help identify design problems and alternatives that can reduce financial, legal and safety risks to the manufacturer, and help ensure satisfactory performance for the customer under virtually all operating conditions. The advantages and disadvantages of the three major WCCA methods are presented in Table 1. A capacitor example is shown in Table 2. The WCCA process is outlined in Table 3. One of the most critical steps involved in completing a meaningful WCCA is the development of a part characteristic database. This database contains a composite of information necessary for quantifying sources of component parameter variation. Once these sources have been identified, the database can be used to calculate worst case component drift for critical parameters. Quantifying the contribution of environmental effects on component variability (as will be illustrated in an example) is also a critical step in the development of a WCCA. A number of starting places can be used to establish random and biased contributions to variability. They maybe summarized as: Company data (historical test data from other products, or special test programs) Vendor data (documentation of test conditions, sample size, number of lots, etc., is needed) Military specifications (tend to be very conservative) Outside sources (e.g., Jet Propulsion Lab for radiation data) Actual field performance indicates that components tend to drift beyond initial tolerance levels. The magnitude of component tolerance variation is dependent on a variety of sources, as illustrated in Table 4. When conducting a worst case circuit performance analysis, the key elements to be examined within the system are dependent upon the intended function of the circuit. Critical timing of digital circuits, transfer functions of filtering networks, and characteristics of amplifiers are examples of circuit performance elements. Table 5 describes those parameters which should be analyzed in a worst case performance analysis for digital and analog circuits. 1 DIABLO MOUNTAIN RESEARCH, LLC Walnut Creek, CA 94597 925.788.4807 Worst Case Circuit Analysis (WCCA) (Cont'd) Methodology Part statistics are based on two types of component variation: random and bias. Random variation is not predictable in direction. Bias, however, is predictable given known inputs. All sources of component variation can be grouped into one of these effects. The effects are subsequently combined to give an overall indication of part variability. Addition of individual random and biased variables is as follows: Bias Effects Added Algebraically Random Effects Root Sum Squared (±3σ limits of a normally distributed population) Determination of the minimum and maximum limits of component value due to drift is as follows: Worst Case Minimum = Worst Case Maximum = Method Nominal Value - (Nominal Value x Σ | Negative Biases | ) Nominal +    Value x ∑ (Random  Effects) 2   Nominal Value + (Nominal Value x Σ | Positive Biases | )    Nominal Value x ∑ (Random  Effects) 2 Table 1. WCCA Analysis Methods Advantages   Extreme Value Analysis (EVA) Most readily obtainable estimate of worst case performance (best initial WCCA approach) Does not require statistical inputs for circuit parameters (easiest to apply) Database need only supply part parameter variation extremes (easiest to apply) If circuit passes EVA, it will always function properly (high confidence for critical production applications) Disadvantages Pessimistic estimate of circuit worst case performance If circuit fails, there is insufficient data to assess risk (modify circuit to meet EVA requirements, or apply RSS or MCA for less conservatism) Root-Sum Squared (RSS) More realistic estimate of worst case performance than EVA Knowledge of part parameter probability density function (pdf) is not required Provides a limited degree of risk assessment (% of units to pass or fail) Standard deviation (σ) of piece part parameter probability distribution is required Assumes circuit sensitivities remain constant over range of parameter variability Assumes circuit performance variability follows a normal distribution Monte Carlo Analysis (MCA) Provides the most realistic estimate of true worst case performance Provides additional information in support of circuit/product risk assessment Requires use of computer Consumes a large amount of CPU time Requires knowledge of part parameter pdf 2 DIABLO MOUNTAIN RESEARCH, LLC Walnut Creek, CA 94597 925.788.4807 Worst Case Circuit Analysis (WCCA) (Cont'd) Calculation of Capacitor Minimum and Maximum Values The following example illustrates a representative calculation for determining the worst case minimum and maximum values for a 1200 µF CLR capacitor. These parameters are used to determine the potential resultant effect of CLR capacitor drift on circuit applications. Table 2. Capacitor Example Bias (%) Parameters: Capacitance Neg. Pos. Random (%) Initial Tolerance at 25°C -- -- Low Temp. (-20°C) 20 28 -- High Temp. (+80°C) -- -- 17 Other-Envt's (Hard Vacuum) -- 20 -- Radiation (10KR, 1013 N/cm2 )-- -- 12 Aging-- ---- TOTAL VARIATION 10 48 29 (20)2 + (10)2 = 22.4 where: Worst Case Minimum = - 48 - 22.4 = -70.4% Worst Case Maximum = + 29 + 22.4 = +51.4% Worst Case Minimum Capacitance = 1200µf - 1200µf ( ) | -.48 | + .224 = 355.2µf Worst Case Maximum Capacitance = 1200µf + 1200µf ( ) | +.29 | + .224 = 1816.8µf Table 3. Worst Case Circuit Analysis Process Action Rationale Determine Analysis Approach Full analysis vs. circuit partitioning to critical functions Define the circuitry to be analyzed (analog, digital, etc.) Determine availability of component data Identify the tools available to support the analysis Obtain the Data Performance Requirements and Specifications Schematics and Block Diagrams Interconnection Lists and Wiring Diagrams Full Parts Lists Theory of Operation Product Operational Environments Product Operational Configurations Thermal Design Analysis Analyst must be knowledgeable of the circuit and parameters being examined, including timing diagrams where appropriate 3 DIABLO MOUNTAIN RESEARCH, LLC Walnut Creek, CA 94597 925.788.4807 Worst Case Circuit Analysis (WCCA) (Cont'd) Table 3. Worst Case Circuit Analysis Process (Cont'd) Action Rationale Plan the Analysis Develop a Functional Breakdown and Theory of Operation Establish the Analysis Scope Develop a Part Parameter Database (See Table 2, 4, and 5) Identify Method of Analysis (See Table 1) Functional breakdown should include block diagrams that outline major sub functions Block diagrams should detail all functional interactions (inputs and outputs) Primary considerations include analysis of worst case circuit performance, and analysis of proper part application (i.e., derating) Secondary considerations may include circuit shielding, noise, EMI, ground loops and temperature control Critical part parameters include environmental (ambient and self heating temperature, vibration, humidity, radiation) and component aging characteristics Extreme Value Analysis (EVA) is easiest to apply. Analysis of a given circuit/product under simultaneous worst case parts limits. Results are conservative. Root-Sum-Squared (RSS) is a statistical approach. Results are more realistic than EVA, but methodology is more labor intensive. Monte Carlo Analysis (MCA) randomly selects part parameters and analyzes resulting system performance. Many simulations (typically 1,000 to 50,000 runs) must be made. Piece-part parameter statistical distributions must be known (or assumed normal). Perform the Analysis Worst Case Stress Analysis Worst Case Performance Analysis Document the Results Identifies components that are overstressed under worst case conditions (or exceed product derating guidelines) Analyze circuit to determine if performance under worst case conditions is achieved (simultaneous change of environmental conditions and part parameters to their worst realizable extremes) For circuit partitioning, minimize the active components to as few logical blocks as possible Include circuit and timing diagrams, as appropriate, in the analysis Compare WCCA results with the specifications, documenting all inconsistencies Analysis results should be verifiable (document or reference all information used to develop the analysis - circuit equations, part data sources, and circuit simulation tools used) If problems are found during the analysis, develop and propose potential fixes and alternative solutions 4 5 Worst Case Circuit Analysis (WCCA) (Cont'd) Table 4. Effected Parameters vs. Source of Variation Component Type Environmental Source of Variation Parameter Effected Bipolar/Field Effect Transistors Temperature HFE (Bias) , VBE (Bias) ICBO (Bias) RDSON (Bias) , VTH (Bias) Radiation HFE (Bias) , ICBO (Bias) VCE (Saturation) (Random & Bias) VTH (Bias) Rectifiers/Switching Diodes Temperature VF (Bias) , TS (Bias) , IR (Bias) Radiation IR (Bias) , VF (Bias) Zener Diodes Temperature VZ (Bias) , (Sometimes Random) ZZ (Bias) Resistors Temperature Resistance (Bias & Random, Random) Humidity Resistance (Bias) Carbon Composition Aging (Powered) Life (Unpowered) Resistance (Bias & Random) Resistance (Bias & Random) Vacuum Resistance (Bias) Mechanical Resistance (Bias & Random) Capacitors Temperature Capacitance (Bias and/or Random) ESR (Bias), DF (Bias, Non Linear) Aging ESR (Bias), Capacitance (Bias and/or Random) Mechanical Capacitance Electrical Voltage Coefficient Vacuum Capacitance (Bias, Non Hermetic) Humidity Capacitance (Bias) Linear ICs Radiation Voltage, current offset (Random), AOL (Bias) Temperature Voltage, current offset (Bias & Random), Random AOL (Bias) Digital ICs Temperature Rise/Fall Time Propagation Delay Radiation Propagation Delay (Bias) Magnetics (strongly dependent on materials) Temperature Saturation flux density (Bias) Permeability (Bias) Core Loss (Bias, Nonlinear, Nonmonotonic) Aging Saturation flux density (Bias, very small) Mechanical Permeability (Bias) Saturation flux density (Bias) Relays Temperature Pull in/Drop out current/volts (Bias) Contact resistance (Bias, Secondary effect) Mechanical Contact Resistance (Bias) HFE gain VF forward voltage ESR equivalent series resistance ICBO collector-base output current Ts storage time DF dissipation factor RDSON on-drain source resistance IR reverse current AOL open loop gain VTH threshold voltage Vz zener voltage VCE collector-emitter voltage Zz zener impedance DIABLO MOUNTAIN RESEARCH, LLC Walnut Creek, CA 94597 925.788.4807 DIABLO MOUNTAIN RESEARCH, LLC Walnut Creek, CA 94597 925.788.4807 Worst Case Circuit Analysis (WCCA) (Cont'd) Table 5. Circuit Parameters for WCCA Digital Circuit Parameters Circuit Logic Pulse Widths Compatibility Circuit Timing Analog Circuit Parameters Current Draw Comparator- Threshold Precision- Hysteresis- Switching Speed/Time Constant- Offset Stability Oscillator- Frequency, Accuracy, Stability- Output Power Level Stability- Output Impedance- Phase Stability- Noise & Spurious Output Filter- Insertion Loss- Frequency Response- Phase Response- Linearity- Input/Output Impedance - VSWR- Spurious or Out-of-Band Feedthrough Modulator- Frequency Response- Phase Response- Input/Output Impedance - Linearity- Insertion Loss Detector- Bias Voltage- Frequency Range- VSWR- Load Impedance- Input Impedance- Output Impedance- Input RF Switch (Solid State/Mechanical)- Drive Requirements- Insertion Loss- Power Dissipation- Output Level- Deviation- VSWR- Power Handling- Switching Speed- Frequency Response- Video Feedthrough- Switch Duty Cycles- Input/Output Impedance - VSWR Multiplier- Output Power- Frequency Response- Input/Output Impedance Mixer (Converter)- Noise Figure- Frequency- Output Spectrum- Intercept Points- Input Drive- Group Delay- Power Dissipation- Conversion Loss Coupler, Circulator- Insertion Loss- Frequency Response- Power Handling- Input/Output Impedance- Magnetic Leakage- VSWR- Directivity Stripline, Waveguide, Cavity- Mode Suppression- Insertion Loss- Adjustment Range, Resolution- Dimensional Stability- VSWR- Driver Require-- Compression Points- Terminating Impedance ments- Input/Output Impedance- Material Stability Copyright  2020 DIABLO MOUNTAIN REAASEARCH, LLC. All rights reserved. Source: • RAC Publication, CPE, Reliability Toolkit: Commercial Practices Edition . For More Information: • RAC Publication, CRTAWCCA, Worst Case Circuit Analysis Application Guidelines . 6