Use Infrared: It’s Already There

Pyroelectric detectors make mid-range infrared affordable. You usewhat is already there — 100%natural and harmless.

You can use the invisible glow ofobjects and people to detect, count,monitor, locate, activate, conserve,protect, or warn. It is passive technol-ogy.

Beyond Photodiodes

Visible light goes from 0.4 to 0.7micrometers on the wavelengthspectrum. Beyond that is infrared.Photodiodes are inexpensive andpractical even to 1 micrometer. But,1 micrometer corresponds to the

"wavelength of maximum energy" ofa blackbody at about 2,900 Kelvin(4,700oF) which is the temperature ofan incandescent bulb’s white-hot fila-ment.

To use the infrared emitted fromourselves or objects that we cantouch, wavelengths well beyond 1micrometer and especially thosearound 10 micrometers must bedetected.

Pyroelectrics Are Practical

Detection of mid-range infrared isnot new. Thermistors and ther-mopiles (thin-film thermocouples)have long been available. Althoughthese components are relatively inex-pensive, the circuitry required to

make them work is not. Moreover,both thermistors and thermopiles aregenerally found wanting in terms ofs igna l st rength and speed o fresponse.

Pyroelectrics are today’s practicalchoice for broad-band IR detection.Pyroelectrics offer technical ad-vantages in signal strength, speed ofresponse and in minimizing intercon-necting circuitry. And, as has hap-pened with other components, use ofmore sophisticated production techni-ques pioneered at ELTEC INSTRU-MENTS, INC., has increased theavailabi l ity of l i thium tantalatepyroelectric detectors while loweringcost.

ELTECdata # 100

Introduction ToInfrared PyroelectricDetectors

ELTEC INSTRUMENTS, INC.

LO

G10

(SP

EC

TR

AL

RA

DIA

NT

EX

ITA

NC

E-W

AT

T-C

M–2

WAVELENGTH–MICROMETERS0.1 1 10 1000100

-6

INFRARED

1

-3

2

-4

3

VISIBLE

-5

0

-1

-2

3000K

2897λMAX = (µm)T(oK)

Relative Intensities andWavelength Distributions of

Tungsten Filament and People

300K

M

ICR

OM

ET

ER

–1)

The Pyroelectric Effect:

The Material

If a material has an internal electri-cal symmetry, it’s neutral. If it’s un-symmetrical — like water — it has apermanent electric dipole. Most un-symmetrical materials in bulk have azero dipole effect because of a ran-dom or self-cancelling arrangement.

There are some unsymmetricalmaterials which maintain a net dipoleorientation even in bulk. Heatingsuch a material (within limits) doesn’trandomize the dipoles, but rotatesthem in unison and thus maintains apolarization. Since this occurs in theabsence of an external electric field,it is called spontaneous polarization.

Dipoles will act in unison to anupper temperature point called theCurie point. Lithium tantalate is apractical pyroelectric material be-cause it has a Curie point of 610o C.Also, lithium tantalate is a veryresponsive synthetic crystal with anestablished, long-term stability.

The Pyroelectric Effect: Simplified

The Greeks discovered thepyroelectric effect 23 centuries ago.They observed that when tourmalinewas placed in hot ashes, it first at-tracted and then repelled them(charge generation ... attraction byinduction ... contact/reversal ... repul-

sion of like charge). Hence "pyro", forfire, plus electric !

Pyroelectric isn’t thermo-electric.In a thermoelectric device, like a ther-mocouple, a steady voltage isproduced when two junctions of dis-similar metals are held at steady butd if ferent temperatures. In apyroelectric device, a change intemperature creates a change inpolarization. "Electrical polarization"is just another way of saying "electri-cal charge". The charge is collectedby electrodes on the crystal. So the"open circuit" Voltage = (Q, charge) /(C, crystal capacitance).

In THEORY, if the crystal werelevitated in a vacuum in a perfectlyreflective Dewar, at infinite im-pedance (and some other condi-tions), and a thermal step functionapplied, the voltage would follow thestep function.

In REALITY, the crystal has a ther-mal time constant, so it will quicklythermalize to its environment (returnto ambient) after a step input. Thisreleases the strain on the crystal lat-tice and the crystal "reabsorbs" theelectrons as the lattice returns to itsneutral state. Thus, "step function in,""voltage pulse out."

The nature of the pyroelectricdetector makes it both fast and useful.Since every object is emitting infraredlight, every object is a transmitter.And since the infrared detectorresponds to infrared, it is a receiver.An intruder entering a room is like aninvisible light being turned on; thedetector responds to the change ininfrared light, generating a useful sig-nal. A glass object (transparent in thevisible and near-infrared) may passright through a light beam undetected,but its infrared emissions will identifyit every time. In short, whereverthere’s a change in infrared light,there’s a potential pyroelectric ap-plication.

Heated tourmaline developselectric charges

H

O

105o

+ + +

– – –

H

Infrared input to detector (step)

Response of pyroelectric detector

Water is unsymmetrical

Dipoles acting in unison

Light

Crystal

SignalElectrodes

PyroelectricDetector Capacitor Generator

Page 100-2

The Pyroelectric Detector

A thin wafer of lithium tantalate haselectrodes deposited on both faces.The electrodes gather the chargewhich is unable to leak through be-cause the material is such a gooddielectric (insulator). In its simplestform, the pyroelectric detector is botha capacitor and a charge generator (inresponse to infrared light striking aface, being absorbed as heat, creat-ing change in polarization). And allthis at room temperature, without theneed for cooling or electrical biasing.

Electrical Considerations

Think of a pyroelectric detector as atiny flat-plate Active Capacitor. Typi-ca l capaci tance is about 30picofarads. Insulation resistance is5x1012 Ohms. So, except in laserapplications, the extreme source im-pedance makes use of the crystal byitself impractical. PRACTICALpyroelectrics contain either a JFETsource follower (Voltage Mode) or atransimpedance amplifier (CurrentMode).

For a rough idea of the signal youmight get using a pyroelectric detec-tor, use the following formula:

V responsivity = (current respon-sivity)(effective impedance)

I • (R/√1+(2πfRC)2)

for I, use 0.5 to 1 microamp per watt

for R, use either your load resistor valueor feedback resistor value

for C, use detector capacitance for Volt-age Mode (typically 30 pF) or usestray feedback capacitance forCurrent Mode (typically 0.03 pF)

The formula is very useful to get anapproximation of voltage responsivityat different frequencies (modulationrates).

Both o f these ampl i f ica tionschemes have positive and negativefeatures. The voltage mode circuitwill generally yield the best signal tonoise ratio and it can operate at a verysmall supply voltage and current.However, it does not have a largeoutput responsivity and outputresponse will be frequency depend-ent (unless a low value RL is used).The current mode offers a substantial

increase in output signal. It can havea "flatter" frequency response andthat response can be set independentof the crystal. Unfortunately, thenoise characteristics of the operation-al amplifier limit the signal-to-noiseratio and the operating voltage andcurrent requirements are greater.

NOTE: Although the voltage (FieldEffect Transistor) or current (Op Amp)circuits can be added externally to thebasic detector package, it is ac-complished with the addition of straycapacitance, susceptibility to EMI,testing problems, expense and pos-sibly a compromise in reliability. Tocircumvent these problems, detectorsare offered with the FET and ap-propriate load resistor or op amp andappropriate feedback resistor in thedetector package.

The voltage follower is basically aFET connected as a source follower.

In this configuration, the voltageoutput will be:

RV = Ri • Z eff • Aowhere RV = voltage response in V/W

Ri = current responsivityZ eff = lumped impedance of

crystal, RL, and straycapacitance at the input

Ao = follower gain (approx. 0.8)

The current to voltage convertercan be an operational amplifier con-nected as shown.

In this configuration the voltage out-put will be:

RV = Ri • ZFwhere ZF = lumped impedance of

feedback loop includingRF and CS, stray feedbackloop resistance and capa-citance

RL CS

Detector connected with sourcefollower

RF

CS

–

+

Detector connected with a currentto voltage converter

Page 100-3

Laser Applications

In high speed or fast pulse applica-tions with a great deal of incidentpower, the detector can be operatedwithout an impedance converter. Ifpulse resolution is required, thedetector can be loaded down with aresistor – the value of which is deter-mined by the speed of the event to bemonitored. In this case, responsivityis RV. The detector can also be usedas an energy monitor by loading theoutput with a capacitor. In this case,responsivity is RE.

Demystifying D-Star

The ultimate sensitivity of an in-frared detector is determined by thesignal-to-noise ratio. No matter howprecise or noise-free the amplificationscheme, there is a point where theoutput signal cannot be distinguishedfrom background noise. This point,when related back to the responsivityof the detector gives the minimumdetectable power level. In IR jargon,this is called "Noise EquivalentPower" or NEP. It is defined as:

Noise WattsNEP = =

Responsivity √ Hz

Note that the NEP for any givendetector is dependent on wavelength,operating frequency, noise center fre-quency, noise bandwidth (usually 1Hz), and temperature. For example:

WNEP = 2.2 x 10-10

√ Hz(500oK, 20Hz, 1 Hz, RT)

The source temperature is 500oKand implies the optical bandwidth; 20Hz is the operating frequency andnoise center frequency; 1 Hz is thenoise bandwidth; RT (25oC) is thetemperature of the sensor.

Since minimum noise power isdesired, the smaller the NEP the bet-ter. Unfortunately, sensor manufac-turers want to complicate sensorperformance factors even further byalso specifying a parameter called D-Star (D*). This parameter normalizesthe NEPs to a given constant detectorarea. This permits all detectors to becompared on an equivalent basis.

√ AdD* = NEP

where Ad = area of detector in cm2.

The larger the D* the better.

As with NEP, D* must be specifiedfor wavelength, frequency, noisebandwidth and temperature. For ex-ample:

D* = 2.8 x 108 cm√ Hz Watt

(10.6 µm, 10 Hz, 1 Hz BW, RT)

Even though the concepts of NEPand D* were created to facilitate ap-ples-to-apples comparisons, practicalpyroelectric detector performance isarea-dependent rather than square-root-of-the-area-dependent (seeELTECdata #103). Consequently,direct comparison of different detec-tors with different sizes is still difficult.Also, the temperature of the sensor isoften not specified (and some othertypes of IR sensors are very tempera-ture dependent) and also somedevices have outputs which are notlinear with input power.

Ri CD RL

RV= R i • RL

Circuit for pulse resolution inhigh power applications

Ri CD CL

RE =RiCL

Circuit for energy monitor ofhigh power pulses

Page 100-4

.320 INCH∅(8.13 MM)

.190 INCH(4.83 MM)

METAL CASE

FILTER

LITHIUMTANTALATE

CRYSTAL

ELTECTHICK FILM

LOADRESISTOR

ELECTRODES(lollipop top;full bottom)

FET

HEADER

SUBSTRATE

PINS

PATENTED ELTEC LOOP SUSPENSION OF CRYSTAL

(3 loops)

PYROELECTRIC DETECTOR WITH INTEGRATED ELECTRONICS

Page 100-5

The top abscissa on the curveabove shows an object’s temperaturewhile the bottom abscissa shows thewavelength of the maximum energyfor an object at the correspondingtemperature. Note that there is al-ways a distribution of energy over allwavelengths for any object with about

25% of all energy in wavelengthsshorter than the wavelength of maxi-mum energy and 75% in the longerwavelengths.

The curve within the coordinatesystem relates the temperature/max-imum wavelength to the total emitted

energy given in watts per square cen-timeter of surface on the ordinateaxis. The value given are for a trueblackbody and the value for any realobject will be a percentage repre-senting the ratio of the actual radiationemitted to the energy emitted by ablackbody at the same temperature.

NOTICE: The information provided herein is believed to be reliable. However, ELTEC Instruments, Inc. assumes no responsiblity for inaccuraciesor omissions. ELTEC Instruments, Inc. assumes no responsibilities for the use of the information, and all use of such information shall be entirelyat the User’s own risk. Publication of this information does not imply any authority or license for its use nor for any infringement of patents or rightsof others which may result from its use.

ELTEC Instruments, Inc. P.O. Box 9610 Daytona Beach, Florida 32120-9610 U.S.A.Tel (USA and Canada): (800) 874-7780 Tel (Outside USA): (386) 252-0411 Fax: (386) 258-3791

Web: www.eltecinstruments.com E-Mail: Sales@eltecinstruments.com

©ELTEC INSTRUMENTS, INC. 03/2006 Printed in U.S.A. Form DN100 (E-07/2009)

2000

1000

500

200

100

50

+25 0

–25 5035

00

5000

8000

100

100K

10K

1K

1

10

100

0.1

0.01

0.00110.1 10 15 20

Em

itte

d B

lack

Bo

dy

Rad

iati

on

W/c

m2

Wavelength (µm)

OBJECT TEMPERATURE VS. WAVELENGTHTemperature (oC)

5

15

0.110.04 0.03

Page 100-6