Temp Measurement and Cal (Fluke)

8/3/2019 Temp Measurement and Cal (Fluke)

http://slidepdf.com/reader/full/temp-measurement-and-cal-fluke 1/7

Application Note

Temperature measurementand calibration:

What every instrument technicianshould know

F r o m t h e F l u k e D i g i t a l L i b r a r y @ w w w . f l u k e . c o m / l i b r ar y

Introduction

Temperature may be the mostcommonly measured physicalparameter. Yet there have never

been so many ways to measure itas there are today. With so manyoptions it’s natural to have a fewquestions. How do I measuretemperature? How accurate is mymeasurement? What temperaturerange is required? What type of device best measures tempera-ture? Does my instrument requirecertification?

These are very common ques-tions when confronted with theneed to measure temperature. Avariety of measurement devices

may be used for temperature;liquid-in-glass thermometers(LIG), thermocouples (TCs),thermistors, resistance tempera-ture detectors (RTDs), platinumresistance thermometers (PRTs)and standard platinum resistancethermometers (SPRTs). This appli-cation note focuses on electronictemperature measurements, andhelps clarify the answers to someof these nagging questions. Moreinformation on these subjects isavailable at www.fluke.com.

How do I measuretemperature?

After inserting a temperaturesensor into the area to bemeasured, it takes time beforethe temperature reading isstabilized. For the thermom-eter to stabilize at the correcttemperature, the probe mustbe sufficiently immersed. Somethermometers require moreimmersion depth than others.Most precision thermometers

require four to six inches wheninserted into a liquid or snugfitting well, depending on thediameter of the probe. Best

results in terms of accuracy andstabilization time occur whenthe probe can be inserted intoa stirred liquid. Air pocketsbetween probes and solid sur-faces lead to longer stabilizationtimes and require more immer-sion than would be required ina liquid. Specialized thermom-eters are needed for measuringtemperatures on surfaces and forsituations where the probe cablewill be exposed to extremetemperatures.

Often devices that measureand display temperature needto be verified or calibratedagainst a reference thermom-eter. Accuracy is improved whenthe distance between the twothermometers is reduced. A bestpractice is to align the centersof the sensing elements of thereference thermometer and thedevice under test. Be aware thatthe location of the center of thesensor depends on the sensortype and model (i.e. PRT, ther-mocouple, bimetallic).

A common method of calibrat-ing temperature sensors is toremove them from where theyare installed and place them ina dry-well calibrator or a Micro-Bath. These calibrators provide astable temperature environmentover a range of temperatures tocompare the thermometer undertest to the calibrator display orto a reference thermometer formore accuracy. Alternatively,temperature sensors may becalibrated or verified without



2 Fluke Calibration Temperature measurement and calibration: What every instrument technician should know

removing them from theirinstalled location. Usually thisis done by inserting a referencethermometer into a thermowell,immersion well, or thermometer

pocket installed next to the ther-mometer to be tested. In othercases the sensing element of the reference thermometer mustbe placed inside of the freezer,oven, or environmental cham-ber being verified, calibrated oradjusted.

In these cases it is oftennecessary to record data overa period of time such as a fewhours to verify performance.Statistics such as average value,maximum and minimum, or

standard deviation are some-times recorded.Testing the energy perfor-

mance of steam systems, coolingtowers, heat exchangers, andrefrigeration systems, turbines,and internal and externalcombustion engines requiresmeasuring differences betweeninlet and outlet temperatures.Sometimes these measurementshave to be made from outsidethe pipe using thermocouples,thin-film sensors, or infrared

temperature measurements.However, the best accuracy willbe achieved when a thermowellhas been properly installed inboth the inlet and outlet pipes sothat a probe can be inserted andsufficiently immersed. Becausepipe diameters are sometimes alimiting factor for immersion, thebest location for a thermowell isat an elbow in the piping so thatthe probe can be inserted parallelto fluid flow with as much immer-sion depth as needed.

How much accuracy isneeded?

Decisions about accuracy shouldbe made carefully. Inaccuracyleads to mistakes and mistakescost money. Mistakes may leadto down time, excessive energycosts, high product defect rates,safety hazards, and public healththreats.

Thermometers are specified bydesign engineers for tempera-ture monitoring or control. Thesespecifications should includethe accuracy of the thermom-eters. A design engineer, qualityengineer or metrologist shouldalso specify the calibration

requirements. However, it is notuncommon for instrument tech-nicians to receive a calibration job and little or no informationabout calibration requirements.

A common calibration strategyis to reduce mistakes by keepingthe uncertainty of the calibrationstandards to a low percentageof the accuracy of the thermom-eter under test. This percentageis usually described as a TestUncertainty Ratio (TUR). Forexample, the 4:1 TUR used by

the military and other industrieskeeps the collective uncertaintyof the calibration standards to25 % of the thermometer undertest accuracy. For comparison,a TUR of 2:1 means that theuncertainty is 50 % of thethermometer accuracy, and if thereference thermometer has the

same accuracy as the thermom-eter under test, then the TURis 1:1. The latter TUR is neverrecommended for calibrationand would produce unreliableresults.

With a more accurate calibra-tion standard you can identify

more actual out-of-tolerancefield devices. Table 1 illus-trates the expected frequencyof making mistakes at variousTURs. Table 1 is based on ascenario where 950 of 1,000instruments are truly in toler-ance. For example, if all 1,000are calibrated with a 2:1 TURthen we expect that 926 will befound in tolerance (accepted), 12of which are truly out of toler-ance (false accept). Of the 74expected to be rejected, 41 are

expected to be truly in tolerance(false reject). The cost incurredfor each of those falsely rejectedinstruments could range from$50/each for a calibration houseto $10,000/each in down time inthe chemical process industry.

Temperature range Accuracy Cost

Noble-metal

thermocouples

(Special tolerances)

R, S: –50 °C to 1760 °C > ± 0.6 °C Med

Base-metal

thermocouples

(Special tolerances)

B: 0 °C to 1820 °C

E: –270°C to 1000 °C

J: –210 °C to 1200 °C

K: –270 °C to 1370 °C

N: –270 °C to 1300 °C

T: –270 °C to 400 °C

± 0.25 %

> ± 1 °C

> ± 1.1 °C

> ± 1.1 °C

> ± 1.1 °C

> ± 0.5 °C

Low

Low

Low

Low

Low

Low

PRTs and SPRTs Industrial: –80 °C to 480 °C

Reference: –200 °C to 660 °C

High temp: 0°C to 1000 °C

± 0.05 - 0.1 °C

±0.001 – 0.02 °C

±0.01 – 0.02 °C

Low - Med

Med - High

Med – High

Precision thermistors 0 °C to 100 °C ±0.002 °C Med

Table 2. Temperature sensor tradeoffs among temperature range, accuracy and cost. The most accurate sensors arethe most expensive. Often accuracy is sacrificed for wider temperature range.

TUR Accepted False Accept Rejected False Reject

1:1 843 17 157 128

2:1 925 12 75 41

3:1 941 9 59 224:1 947 8 53 15

Table 1. What-if table summarizing false accept and false reject risk for a hypothetical scenario of 1000 instrumentsthat are truly 95 % in tolerance. A normal distribution without guard-banding is assumed.




Thermometer probetypes

There have never been as manytemperature sensor (probe type)choices available for your mea-surements as there are today. With so many choices, the taskcan become time consuming anddifficult without some help. Themost important factors are tem-perature range, accuracy and cost.Table 2, on the previous page,illustrates the tradeoffs amongthese factors for several thermom-eter types.

Thermocouples (TCs)

Thermocouples are temperature

sensors that measure temperatureby generating a small voltagesignal proportional to the tem-perature difference between the junctions of two dissimilar metals.One junction (the measurement junction) is typically encased ina sensor probe at the point of measurement; the other junction(the reference junction) is typi-cally connected to the measuringinstrument. The measurementinstrument measures two things:the voltage signal and the refer-

ence junction temperature. Fromthose two things the instrumentcomputes the temperature atthe measuring end of the probe.It is important to note that thevoltage generated by the sensor

is not based on the absolutetemperature of the measurement junction, but rather a temperaturedifference between the measure-ment junction and the reference

junction.Thermocouple types are dis-tinguished by the metals usedin each leg of the thermocouple.Noble metal thermocouples allcontain platinum in one leg of the thermocouple and includeType S, Type R, Au/Pt, and Pt/Pd. Base metal thermocouplesinclude Type B, Type E, TypeJ, Type K, Type N, and Type T.These thermocouples come intwo accuracy classes: standardlimits of error and special limits

of error. The special limits of error thermocouples are the mostaccurate. Letter designated ther-mocouple tables are availablefrom NIST on the web or in NISTmonograph 175. A thermocouplevoltage and sensitivity calculatoris also available on the web atwww.fluke.com.

Reference junction com-pensation is one of the mostsignificant contributors to theaccuracy of a thermocouplemeasurement. Thermocouple

tables like those in NIST mono-graph 175 are based on areference junction temperatureof 0 °C. Although external refer-ence junctions can be used toachieve this with an ice bath,

thermocouple wire is usuallyconnected directly to the ther-mocouple readout binding postsat room temperature. Automaticreference junction compensation

is needed to compensate for thedeviation from 0 °C. A therm-istor bead is usually used tomeasure the temperature of the junction. The readout measuresthe resistance of the thermistorand calculates a correction forthe thermocouple temperature.In Figure 2, thermocouple wiremeets with copper wire at thebinding posts of the meter form-ing the reference junction (J).The temperature in the regionsurrounding the binding posts

(TJ ) is usually measured bya thermistor. Automatic refer-ence junction compensation isaccomplished by measuring thedifference from 0 °C at the bind-ing posts (TJ) and compensatingfor it digitally. The accuracy of this measurement has a signifi-cant impact on the accuracy of the overall temperature mea-surement.

Resistance based tem-perature measurement

An RTD is a temperature sensingelement that relates tempera-ture to its own resistance. Thereare several kinds of RTDs. RTDsensing elements include coils of

Figure 1. Model of a thermocouple circuit where A and B are dissimilar thermocouple wire, T1 represents the temperature at the measurement junction, and T2 repre-sents the temperature at the reference junction. The absolute temperature at T1 does not produce the voltage measured at V; rather, the temperature difference betweenT1 and T2 produces the measured voltage.

A

B C

C

T2

T1

V




+

-

+

-V ( T )TC TC

V (T )J1 J

V (T )J2 J

TJ

V =V (T )+[V (T )+V (T )]o TC TC J1 J J2 J

platinum wire (PRT), nickel wire,copper wire, thin films and more.

Another resistance based sensoris the thermistor which is made of semi-conducting material. Figure3 illustrates a simple 2-wiremeasurement circuit. The sens-ing element is labeled RT. Thelead-wires have finite resistanceslabeled RL1 and RL2.

When current passes throughthe sensor, the environmentis going to get a little warmerbecause of power dissipation.The more resistance or currentthere is, the more power gets

dissipated (P=I2R). The self-heating will be higher in airbecause the heat will not flowaway as efficiently as it wouldin a stirred fluid. Self-heatingerrors can be minimized byusing the same level of currentused during calibration. Usingthe correct current is particu-larly important in thermistorsbecause they can have verylarge resistances causing greaterself-heating.

Current reversal is a very

effective technique used inresistance measurements toeliminate errors associated withthermal EMFs. Thermal EMFsare unwanted voltages in aresistance measurement circuitcaused by the same principlethat produces a voltage in ther-mocouples. The measurement ismade with the current flowingin one direction and then againwith the current flowing in theother direction. ThermoelectricEMFs are removed by averaging

the results of both sets of measurements. This techniqueused by many modern instru-ments improves measurementstability and reduces significanterrors that are common in otherinstruments.

Platinum resistance

thermometersA platinum resistance thermom-eter (PRT) element contains coilsof highly-pure platinum wire. Theresistance of a PRT element variesmore linearly with temperaturethan any other temperaturesensor. A Standard PlatinumResistance Thermometer (SPRT)is the most accurate temperaturesensor available and is used inNational Standards Laboratoriesand in industry for traceabilityto the International Temperature

Scale of 1990 (ITS-90). The fulltext of the ITS-90 is available atwww.bipm.org.

Temperature measurementwith a PRT requires correlatingthe resistance of the sens-ing element with temperatureusing the correct equations andcoefficients. Fortunately, mostthermometer readout deviceshave support for these equa-tions, so the calculations arehandled automatically. Examplesinclude ITS-90 equations,

Figure 2. Reference junction compensation is one of the most significant contributors to the accuracy of a thermocouple measurement.Caution: some manufacturers may not advertize this important part of their accuracy.

RT

RL1

RL2

+

- Figure 3. Current is passed through the sensing element to producea voltage measured by a meter. Lead-wire resistance in two-wiremeasurements causes potentially large temperature measurement

errors. Other types of resistance measurements include three-wireand four-wire resistance measurements. Four-wire measurements arepreferred in temperature applications because they eliminate leadwire resistance from the measurement.

Figure 4. Current is required to measure resistance. Current thatpasses through a resistance dissipates power and generates heat,causing temperature errors.

C u r

r e n t




Callendar Van Dusen (CVD)equations, and polynomial equa-tions. Best performance withPRTs can usually be achievedwith the ITS-90 equations. Older

readouts and uncalibrated PRTsmay use CVD equations.The coefficients used with

these equations are customizedthrough calibration and can befound on individual calibra-tion reports along with a tablethat describes the resistancevs temperature relationship.Some inexpensive PRTs usedfor industrial measurement areintended to operate even with-out calibration. Out of the boxthey should conform to a stan-

dard such as IEC 60751 or ASTM1137. These standards specifyvalues for the CVD coefficientsand tolerances that dependon temperature. However, theaccuracy of these probes can besignificantly improved throughcalibration.

ThermistorsA thermistor element is made of semi-conducting material andhas an electrical resistance thatvaries non-linearly with tem-perature. Thermistors are widelyused because of their sensitiv-ity, small size, ruggedness andlow cost. Thermistor accuracydepends greatly on designand construction. Inexpensivethermistors are commonly usedin electronics applications whileprecision thermistors are cali-bration standards rivaling theaccuracy of SPRTs.

The resistance vs tempera-ture relationship in thermistorsis described by a few differ-ent polynomial equations. Oneform of the equations is usedtake resistance as an input tocalculate temperature T(R) andanother form of the equationtakes temperature as an inputto calculate resistance R(T). Thestandard version of these equa-tions take four coefficients, butthe Steinart-Hart version onlyrequires three.

Thermometer

accuracy, repeatability,and resolution

Two important components of accuracy are repeatability andresolution. They should be con-sidered along with other factorsaffecting accuracy. Repeatabil-ity refers to the consistency invalues of repeated measurements.Regular calibration is helpful forestablishing instrument repeat-ability. Another important methodof verifying repeatability in

thermometers is to measure peri-odically in an ice point or triplepoint of water cell and record theresults.

Resolution on a digital ther-mometer is often user selectable.A digital thermometer should bechosen with sufficient resolutionto achieve the desired accuracy;however, resolution is not thesame thing as accuracy, it ismerely a limiting factor in theaccuracy. In a liquid and glass ordial thermometer, the resolution

may be the most importantfactor affecting the accuracyaside from calibration.

Accuracy specifications can bestructured several ways. Speci-fications will usually be dividedinto ranges and may be givenin base units of temperature,resistance or voltage. Simplespecifications will either bea variable or fixed value, andcomplex specifications will bea combination of both. Whenvariable type specificationsare used, the allowed errorincreases when the magni-tude of the reading increases.

Calculations are required fora numeric result. Examplesinclude percent of reading orparts per million (PPM). On theother hand, fixed value specifi-cations remain constant over arange. Examples include percentof scale or span and numericconstants.

You can convert specifica-tions in base units of resistanceor voltage to temperature. Theconversion depends on thesensitivity of the temperature

sensor. For example, changetemperature 1 °C and it willresult in a 0.4 Ω change in resis-tance for a 100 Ω PRT and resultin a 0.1 Ω change for a 25 ΩSPRT, but it may cause a 1000 Ωchange in a thermistor. Thismeans a one degree change intemperature may have a verybig or very small impact on thechange in resistance. Conse-quently, a meter with ± 1 Ω

Table 3. What to look for in electronic thermometer readouts.

Readout device Requirements

Thermocouple readout Good accuracy from –10 mV to 100 mVLow noise floorVery low thermal EMFs

Good reference junction compensationPRT readout Excellent accuracy from 0 Ω to 400 Ω

Current reversalFour-wire resistance measurement1 mA excitation current

Thermistor readout Reasonable accuracy from 150 Ω and 500 kΩBetter accuracy required below 1000 Ω2 μA to 10 μA excitation current




accuracy will be most accuratefor those sensors with the high-est temperature sensitivity.

Each of the thermometerslisted in Table 2 requires a

digital readout. The best digitalreadout is going to be designedspecifically for temperaturemeasurement. Table 3 lists someof the requirements for goodelectronic thermometer readouts.

Reference junction com-pensation is one of the mostsignificant contributors to theaccuracy of a thermocouplemeasurement. Thermocoupletables are based on a reference junction temperature of 0 °C.Although external reference

junctions can be used to achievethis with an ice bath, thermo-couple wire is usually connecteddirectly to the thermocouplereadout binding posts at roomtemperature. Automatic reference junction compensation is neededto compensate for the deviationfrom 0 °C. A thermistor beadis usually used to measure thetemperature of the junction. Thereadout measures the resistanceof the thermistor and calculatesa correction for the thermocou-

ple temperature.NIST and calibration

NIST is a National MetrologyInstitute (NMI) responsible forUnited States national stan-dards. Metrology is the scienceof measurement and includesbasic research, activities thatensure adequate functioning of measurement instruments used inindustry, and also measurementsthat ensure economic transpar-ency, public health, and safety.

Traceability to NIST or anotherNMI is established through cali-bration. The three main reasonsfor having an instrument cali-brated are:1. To ensure readings from an

instrument are consistent withother measurements

2. To determine the accuracy of the instrument readings

3. To establish the reliability of the instrument

To ensure that temperaturemeasurements around the worldare consistent each country hasadopted the ITS-90, a tempera-ture scale defined by the BIPM

(International Bureau of Weightsand Measures). Instrument accu-racy is determined by comparisonto measurement standards of higher accuracy. These mea-surements are part of a chain of comparisons that extend from enduser measurements in the field tosecondary calibration laboratoriesand then all the way to national(i.e. NIST) and internationalstandards.

In addition to traceability,sometimes calibration customers

have an additional requirementfor accreditation. Accredita-tion provides assurance that anappropriate quality program isin place and that training andprocedures meet the technicalrequirements for the calibra-tion service provided. A logoappearing on accredited calibra-tion certificates distinguishesthem from other calibrationcertificates. There are manyaccreditation programs avail-able. NIST sponsors the National

Voluntary Laboratory Accredita-tion Program (NVLAP).

Maintain your standards

Maintenance of calibratedequipment is an important partof quality assurance. There isno guarantee that a calibratedthermometer will remain cali-brated over time. Changes in thetemperature relationship of usedthermometers over time needto be corrected by calibration atregular intervals. When thermom-

eters are found repeatedly out of tolerance remedial actions needto take place, such as shorteningthe calibration interval or replac-ing the thermometer. It is a goodidea to keep records and monitorcertified thermometers betweencalibrations using an appropri-ate method such as an ice bath,or triple point of water cell. Thiswill limit or prevent expensiveconsequences should a certifiedthermometer be found signifi-cantly out of tolerance.

What to look for in atemperature calibration

equipment solutionprovider

When choosing calibration equip-ment solution providers there aremany factors to be considered.Here are a few of them.

A complete solution: Make surethat the company you choose hasa complete range of products sothat your options can be based onyour needs and not just what theyhave available. If you’re consider-ing using automation software in

the future to improve productivityyou should plan to stick with onesupplier because the manufactur-er’s software will likely only workwith their own equipment.

Experience: You probably don’thave time to be an expert intemperature calibration alongwith everything else you do. It’simportant to get help from theexperts, so look for a companythat specializes in temperaturecalibration so you won’t have to.

Support: Expertise is great, butmake sure you have access to theexperts. You are in good hands if they answer their phones, treatyou respectfully and get answersto all your questions. You’ll alsowant to make sure they canresolve any equipment problemsfor you quickly. Ask about serviceturnaround times. If your equip-ment must be sent to a differentcontinent for repairs you could bein for some serious delays.




Accreditation: Even if you don’thave a specific requirement foraccreditation it makes sense towork with a company that has anaccredited laboratory. Here’s why.

An accredited laboratory has beeninspected very carefully to makesure they have the organizationalcapability to do what they saythey do. In the case of NVLAPaccreditation the laboratory haspassed a technical assessmentperformed by experts from NIST.

A guarantee: Calibration is allabout confidence. First, make sureaccuracy specifications are pro-vided. Second, make sure thosespecifications are guaranteed. The

last thing you need is to wind upwith a piece of equipment takingup space that doesn’t perform aspromised.

Fluke Calibration

Fluke Calibration provides thebroadest range available of calibrators, standards, software,service, training and supportsolutions. Our customers work inelectrical, temperature, pressureand flow calibration laboratoriesaround the globe, as well as inmanufacturing test, research anddevelopment, and service.

The Fluke Calibration tem-perature calibration divisionmakes everything needed forcalibrating temperature sensors,from constant temperature bathsand dry-well calibrators withunmatched stability to primarytemperature standards used innational metrology institutesaround the world. For precisionthermometry, Fluke Calibrationthermometer readouts andprobes are exceptionally accu-rate and easy to use, and ourtemperature and humidity datalogger eliminates the need forpaper chart recorders. In addi-tion Fluke Calibration also offerstemperature calibration train-ing courses from experiencedinstructors and expert servicefrom its NVLAP-accredited labo-

ratory (lab code 200348-0).

Fluke Calibration

PO Box 9090, Everett, WA 98206 U.S.A.Fluke Europe B.V.PO Box 1186, 5602 BDEindhoven, The Netherlands

For more information call:In the U.S.A. (877) 355-3225 orFax (425) 446-5116In Europe/M-East/Africa +31 (0) 40 2675 200 orFax +31 (0) 40 2675 222In Canada (800)-36-FLUKE orFax (905) 890-6866From other countries +1 (425) 446-5500 orFax +1 (425) 446-5116

Web access: http://www.fluke.com

©2011 Fluke Corporation.Specications subject to change without notice.Printed in U.S.A. 3/2011 3996588A A-EN-N

Modification of this document is not permittedwithout written permission from Fluke Corporation.

Fluke Calibration. Precision, performance, confidence.™

Documents

Temp Measurement and Cal (Fluke)