Product Characterization Through Data Logging April 3 2002

Product Characterization Through Data Logging April 3, 2002 presented by: Bill Griffith 1

DAC Back to Basics Product characterization through data logging Click to edit Master subtitle style • Why perform product characterization • Data logging system overview • • Sensors And, how to make better Signal Conditioning measurements! Multiplexing / Switching Analog to Digital Converter PC Connectivity An example of characterizing a power supply Questions and answers 2

What is Product Characterization The use of measurements on Click to edit Master subtitle style a product or design to verify properation • Does it meet design goals? • Does it meet standards (UL, CSA, • • GS/TUV, CE, etc. )? Is it reliable? Can it be manufactured ? Is it energy efficient? Is the product safe? 3 Automated test can collect more data and will have more repeatable results

Product Characterization Some examples Click to edit Master subtitle style • • PC board / Components • Life / Durability tests • Operation vs. temp and humidity • Noise, accuracy, gain checks • Incoming value screening Temperature is Battery Tests the most common measurement • Charge / Discharge rates typically changes • Life test slowly • Temperature Rise 4

Product Characterization Some more examples Click to edit Master subtitle style • Mechanical Enclosures • • • Heat rise, hot spots Cooling effectiveness Air flow • Environmental tests (extreme temp and humidity) Standards (UL, CE, etc. ) Specification development through margin analysis Reliability - Vibration, Stress Tests • Whole Product • • • 5

Why Product Characterization Temperature can affect reliability Click to edit Master subtitle style • • • Reliability • Changes in temperature create mechanical stress • Strain will lead to failure • Possible safety issues Temperature changes can occur during power up or thermal gradients caused by hot spots Hot spots can be caused by: • Components • Dead air spots 6

Why Product Characterization Temperature can affect performance Click to edit Master subtitle style • Performance • A change in temperature can cause frequency drift • In addition, voltage can change with temperature • Temperature can affect energy efficiency • Solutions - reduce differences in temperature • Add vents • Add fan - optimize design (size, cost, quiet operation, power consumption, magnetic flux) 7 Hitachi has developed what it says is the first water cooled laptop. The machine has no fan to cool its CPU.

Why Product Characterization An example of product characterization Click to edit Master subtitle style • Power Supply - unique challenges • • Smaller transformer often more desirable • drawbacks include less efficiency and more heat • Transformers create more heat at lower frequencies Measurement challenges • Magnetic flux and RF energy can interfere with temperature measurements • Parts maybe grounded creating potential ground loops with instruments used to measure temperature 8

Why Product Characterization Product characterization is the process of using electrical and physical measurements to gain insight and improve a design Click to edit Master subtitle style • Most common measurements include: • • • Temp, DCV, Current, ACV, Frequency, and Events Measurement characteristics include: Multiple inputs: typically < 20 channels Relatively slow reading rates: ~ 1 channel/sec Standalone or PC-connected: Most analysis done in a PC Data storage: Electronic memory or documented on paper 9

Data Logger System Architecture Overview Click to edit Master subtitle style Signal Transducer Multiplexer Conditioning Display Analysis & Reporting Digital Multimeter (DMM) A/D SCAN MON VIEW 28. 32 C Consider PC connectivity - Cables - Adapters - SW 10

Data Logger System - Transducers Physical parameter to electrical Click to edit Master subtitle style signal Physical Parameters • • Temperature Flow Pressure Strain Position Weight Speed Electrical Signals Transducer • • • dc Volts ac Volts dc current ac current Resistance Frequency Choose the correct sensor Mount it correctly Position it correctly 11

Data Logger System - Transducers Temperature background Click to edit Master subtitle style • • How is heat transferred • Glass is a poor conductor • Gap reduces conduction • Metallization reflects radiation • Vacuum reduces convection Thermal mass • Don’t let the measuring device change the temperature of what you are measuring • Response time is a function of the mass of the sensor and the mass of object being measured 12 Dewar Sensor

Data Logger System - Transducers Making temperature measurements Click to edit Master subtitle style • • • What is your normal temperature? 97. 6 98. 6 99. 6 36. 5 37 37. 5 Thermometer resolution, accuracy Contact time Thermal mass of tongue, thermometer Human error in reading 13

Data Logger System - Transducers Thermocouples Click to edit Master subtitle style • Need accurate, sensitive measurements-- • J-type thermocouple at room temp = < 1 m. V • A 1 C change at 0 C = 50. 38 V To see a 0. 1 C change in a J-type TC at 0 C, your instrument must be able to resolve down to 5 V. • 14 Metal A Metal B + VAB - • Good junction • Thermal shunting • Noise and leakage current • Thermocouple specs • Calibration of TC

Data Logger System - Transducers Thermocouples subtitle style Click to edit Master - Gradient Theory • • The wire is the sensor, not the junction Heating one end of a wire creates a voltage Vax The voltage is a non-linear function dependent on wire type and the temperature difference from one end of the wire to the other The Seebeck coefficient (e) is used to quantify the voltage that is created 15 TX VAX + Metal A - T A Metal B TB Tx VAB = e d. T + A Ta + VAB - Tb e. Bd. T Tx

Data Logger System - Transducers Thermocouples subtitle style Click to edit Master - Gradient Theory • If a “thermocouple junction” is made with two wires of the Tx same metal - what voltage would the TC produced (VAB)? a) 2*VAX - twice the gradient voltage of metal A b) No voltage c) Can’t determined without knowing the temperature difference d) Can’t determine without knowing the metal type 16 + VAX Metal A Ta + VAB Metal A T b

Data Logger System - Transducers A modern data logger will make it easy to Click to edit Master subtitle style make measurements • • We are taking a deeper look at how measurements are made - a lot of the details are handled by the data logger The architecture that we have presented makes it possible to connect any type of transducer to any channel. • Simply setup the data logger with the type of transducer and channel information • The data logger will make the necessary measurements, conversions and display the correct results in scientific units 17

Data Logger System - Transducers More edit Master subtitle style Click toinformation is available • For more details on making temperature measurements - see application note 290. • Having a good understanding of transducers will allow us to make better measurements • In order to make good measurements - care has to be taken 18

Data Logger System - Transducers Thermocouples subtitle style Click to edit Master - Gradient Theory • • • + If a thermocouple was made of a single metal Metal A the TC would produce zero volts Two different metals are combined to create a TC Tx VAX Metal A Ta + VAB Metal B T b How do you measure a TC? Tx Metal A Metal B 19 Metal C + V Metal C

Data Logger System - Transducers Thermocouples subtitle style Click to edit Master - How to measure a TC • Create a reference • • Metal A junction Tx V X Measure Tref using Metal B thermistor Lookup Vref for TC at Tref reference junction Compute Vx=V+Vref Compute Vx = V+Vref Solve for Tx using Vx Metal C + V Metal C Vx Vref V o 0 Tref Tx

Data Logger System - Transducers Thermocouples subtitle style Click to edit Master - Common Types m. V E 60 Platinum T/Cs Base Metal T/Cs K J N 40 20 T 0 500 RS 1000 21 deg C

Data Logger System - Transducers Thermistor Click to edit Master subtitle style • Better accuracy and more expensive • • Absolute temperature A common thermistor type has 5000 at 25 C with 4%/ C Rlead + Rtemp Rmeas Rlead - Often physically small mass • Won’t cause thermal 1 C = 200 loading 10 of lead resistance • A large measurement would cause a. 05 C current will cause self error heating 22

Data Logger System - Transducers RTD - edit Master subtitle style Click to Resistance Temperature Detector • + Vlead RTDs • Absolute measurement Rlead RTD • Accurate measurement Rtemp • A common RTD has 100 Rlead at 0 C with. 385 / C • 10 of lead resistance + Vlead would cause a 26 C error • Avoid self heating, a 5 m. A current source would create 2. 5 m. W of power in our RTD. • At 1 m. W/ºC, that is an error of 2. 5/ºC 23 DMM i + V - R=V/i

Data Logger System - Transducers RTD - edit Master subtitle style Click to Four wire ohm measurement DMM Rlead I=0 RTD Rlead i Rtemp Rlead I=0 Rlead 24 + V=Vtemp=i*Rtemp V Rtemp=V/i • For more R=V/i information on 4 wire ohm measurements see application note 1389 -2

Data Logger System - Transducers IC Temperature sensor Click-to edit Master subtitle style • Linear change in voltage or current with change in temperature • • 10 m. V/K (voltage IC) Room temp approx. 3 V with a 10 m. V/°C change OR • • • 1 A/K (current IC) Need an external power source Absolute measurement 25 + 10 mv/K V temp v - • Have limited temperature range < 150 C • Fairly large mass

Data Logger System - Transducers Temperature sensors Click to edit Master subtitle style IC Thermocouple + Most accurate + High output + Wide variety + Most Stable + Fast + Most linear + Cheap + Fairly linear + 2 -wire meas. + Inexpensive + Wide T. range - Expensive - Very nonlinear - Limited variety + No self-heating - Slow - Limited range RTD Thermistor - Hard to measure - Needs I source - Needs V source - Relative temperature only -Self heating - Self-heating - Nonlinear -4 -wire meas. - Fragile - Special Absolute temperature sensors connectors 26

Data Logger System - Transducers Polling Question Click to edit Master subtitle style • Your designing an electronic device and have included an IC temperature sensor in your design and want to verify the accuracy of the IC sensor - what type of sensor would you use with your data logger to make the measurement? A) Thermocouple B) RTD C) Thermistor D) IC Sensor 27

Data logger System Architecture Overview Click to edit Master subtitle style Signal Transducer Multiplexer Conditioning A/D V V Wiring errors 28 Display Analysis & Reporting y Digital Multimeter (DMM) SCAN MON VIEW 28. 32 C

Data Logger System - Wiring Errors Thermocouple extension wires Click to edit Master subtitle style Small diameter measurement wires • Long rugged extension wires Possible problem Extension wires are less expensive, more rugged, but have slightly different temperature curves than thermocouple 29 • Keep the junction of the extension wire and thermocouple wire near room temperature

Data Logger System - Wiring Errors Common Master subtitle Click to edit Mode Errors style Metal A Metal B H L Icom ZH DMM ZL • Common mode errors create a • Make Z as large as L current into both the high and the low measurement paths • Typically caused by a difference in grounds 30 possible • Avoid connections to grounds

Data Logger System - Wiring Errors Normal Mode Errors Click to edit Master subtitle style magnetic coupling V Inorm DMM Reduce the size of the • Normal mode noise measurement loop introduces a current that is - twisted pair wire in the same direction as the - move measurement hardware measurement current closer to the source • Magnetic flux or RF energy - run measurement wires perpendicular to high current are common sources of wires normal mode errors 31

Data Logger System - Wiring Errors Normal Mode Errors Click to edit Master subtitle style Electrostatic noise Inorm DMM ZL • Large ZL is important to avoid capacitance coupling through tip of thermocouple Shielding will provide an alternative path for electrostatic energy - Only ground one end 32

Data logger System Architecture Overview Click to edit Master subtitle style Signal Transducer Multiplexer Conditioning A/D V V 33 Display Analysis & Reporting y Digital Multimeter (DMM) SCAN MON VIEW 28. 32 C

Data Logger - Signal Conditioning - Examples Click to edit Master subtitle style • Convert signal to be compatible with system • Thermocouple reference junction • • • Metal A Filter Strain gauge bridge completion circuit Shunt resistor for indirect current meas. Amplifier / Attenuator AC converter 34 Metal C + V Metal C Metal B Isothermal Reference Junctions + V - I

Data Logger - Signal Conditioning Filter Click to edit Master subtitle style magnetic coupling V • • Inoise Reduce noise that has been coupled into the signal “normal mode error” Allow for filter settling time 35 I < Inoise Include lead and switch resistance, plus lead and switch capacitance when calculating settling time

Data logger System Architecture Overview Click to edit Master subtitle style Signal Transducer Multiplexer Conditioning DMM V V 36 Display Analysis & Reporting y Digital Multimeter (DMM) SCAN MON VIEW 28. 32 C

Data Logger System - Multiplexing - Examples Click to edit Master subtitle style H H H L L L 1 -wire, 4: 1 MUX 2 -wire, 4: 1 MUX 37 4 -wire, 2: 1 MUX

Data Logger System - Multiplexing Switch characteristics Click to edit Master subtitle style Typical MAX Voltage Offset Speed Life Armature 300 V 1 A 3 µV 1 60 ch/s 10 M Reed 300 V 50 m. A 16 V 1 m. A 300 V 1 A 6 µV 1 25 µV 1 k 3 µV 3 FET Solid State 38 500 ch/s 10 M 100 k ch/s Infinite 200 ch/s Infinite

Data logger System Architecture Overview Click to edit Master subtitle style Signal Transducer Multiplexer Conditioning Digital Multimeter (DMM) DMM V V 39 Display Analysis & Reporting y SCAN MON VIEW 28. 32 C

Data Logger System - DMM - edit Master subtitle style Click to Digital Multimeter H Amplifier, attenuator L and shunt resistors for current AC RMS Current Source for resistance A/D Converter ZL Control and reading memory • Reduce the complexity of signal condition • • May only need one RMS converter May not need additional attenuation and amplifiers in signal conditioning 40

Data Logger System - DMM Normal Mode Errors Click to edit Master subtitle style magnetic coupling Inorm A/D Converter When using an integrating A/D the integration period will be introduces a current that is set to line frequency in the same direction as the - Use at least 1 PLC to measurement current reject noise - Higher frequency noise • An integrating A/D can be will be integrated and reduced used to reduce normal along with power line noise • Normal mode noise 41

Data Logger System - DMM Integrating A/D subtitle style Click to edit Master. Converter • • V Normal mode noise is Vnoise typically generated by large currents that have the same VTemp 1 PLC frequency as line frequency 2 PLC integrating over one Power NPLC Line Cycle will make Vnoise=0 Can trade off reading rate (speed) for resolution T • 1 line cycle = 60 rdgs/s • 10 line cycles = 6 rdgs/s 42

Data Logger System - DMM Measurement Characteristics Click to edit Master subtitle style • Resolution - How many bits or digits the ADC/system produces • Repeatability - How consistent results are reading-to-reading • Accuracy Relative accuracy statistical determination of error (deviation from norm) Absolute accuracy statistical determination of error from a traceable source 43

Data logger System Architecture Overview Click to edit Master subtitle style Signal Transducer Multiplexer Conditioning Digital Multimeter (DMM) SCAN DMM MON VIEW Display Analysis & Reporting 28. 32 C System error is a sum of the error from each component - avoid noise sources! 44

Data Logger System - Reporting Tasks Common Master subtitle style Click to edit reporting tasks • • • Quickly setup data logger • Self guiding menus Make sure that you are collecting valid data • Verify setup / possibly monitor a channel • Set limits and alarms Upload results to PC • Create final reports in Microsoft applications Use away from bench • Display viewing angle • Bright in multiple lighting conditions Save channel setups and easily modify 45

Data Logger System- Reporting Tasks PC to edit Master Click. Connectivity subtitle style • Connecting to data logger • • RS-232 USB LAN GPIB • • • Easy to use Create documentation Compatible with Microsoft Office Apps • Software 46

Data Logger System - Reporting Tasks Creating reports Click to edit Master subtitle style • • • Typically report results in Microsoft Office applications; Excel, Word, or Power. Point Need a way to cut and paste graphical data into these office applications easily Could create a program or use an off-the-shelf application designed for the task 47

Data Logger System - Reporting Tasks Using edit Master subtitle style Click to. RS-232 / USB • • • Although a common interface RS-232 has a couple of drawbacks • RS-232 can be slow • Can be difficult to find the correct cable USB has become prevalent and addresses the common problems of RS-232 USB to GPIB converters have made it possible to take advantage of USB and create an opportunity to take advantage of portable computers away from the bench 48

Data Logger System - Reporting Tasks Using edit connectivity Click to. LANMaster subtitle style • • LAN connectivity has several advantages • Fast - although not tuned for measurement data like GPIB • Can be used over long distances • Allows resources to be shared LAN to GPIB converters can take advantage of LAN capability 49

Data Logger System - Reporting Tasks PC to edit Master subtitle Click. Software Summary style Software Create Report Data logger SW Make 1 st Flexibility Reading Cost Low product specific Free to $300 Textual Time language intensive Great $300 to $1, 000 T&M graphical Medium language Great $600 to $1, 400 Quick Medium 50

Data Logger System - Reporting Tasks PC to edit Master Summary Click. Connectivity subtitle style Interface Common Speed Distance Cost to PC to Instr. Short $0 to Low Yes Need an Medium Short adapter $500 Yes Need an Fast adapter Long $1100 Fast Short $400 to $600 RS-232 Yes USB LAN GPIB Need an adapter Yes Low Yes 51

Data Logger System Demo OEM Power Supply Characterization Click to edit Master subtitle style • Specified an OEM power supply to speed product development. Requirements include: • 100 Vac - 240 Vac +/- 10% line voltage • Characterize temperature measurements in critical areas • Voltage stability under full load • Voltage regulation under low line voltage 52

Data Logger System Demo OEM Power Supply Characterization Click to edit Master subtitle style • Test setup • • Data acquisition / switch unit connected to: • Four temperature measurements using T -type thermocouples • Four dc voltage measurements A PC with USB/GPIB converter Data logger software Create a report, Temperatures vs. Time, Voltage vs. Time 53

Data Logger System Demo OEM Power Supply Characterization Click to edit Master subtitle style • Getting Started • • Use front panel of DAC system to quickly verify voltage and temperature setup The PC and SW is used to automate the measurements over time Varied line voltage manually The data logger screen images were then cut and pasted into a Microsoft Word document for the final report 54

Data Logger System Demo Power edit Master Click to Supply Test subtitle style A/D Vab Vcd Vef 55 SCAN MON VIEW 28. 32 C

Summary - System Accuracy Click to edit Master subtitle style Analog-to. Signal Transducer errors Conditioning Thermal gradient Multiplexer Thermal EMF / Offsets Additional cable errors Display digital converter DMM - offset, noise, linearity, resolution, accuracy A/D SCAN MON VIEW 28. 32 C Data logger temperature conversion algorithm Ref transducer accuracy System accuracy is the sum of the measurement errors for each component 56

Summary Modern Data logger - Can style Click to edit Master subtitlehelp • • • Data logger may include PC cables and software Data logger will make thermocouple measurements easy - built-in conversion routines, reference junction, reading storage, monitor data from display Data logger may specify system accuracy - to determine complete error add transducer error • For our example the data logger specifies the system error for a T-type TC as 1 o. C. The TC has an error of 1 o. C or. 75% which ever is greater. So the total error is 2 o. C Work to minimize external sources of error 57