Heat Lights the Way
Infrared sensing technology has broad application, ranging from motion detection in security systems to proximity switches in consumer devices. In this article, George looks at the science, technology and circuitry of infrared sensors. He also discusses the various types of infrared sensing technologies and how to use them.
Invisible infrared (IR) radiation, whose wavelength is longer than that of visible red light (for example greater than 780 nm), was discovered by William Herschel in 1800. The IR range most interesting for our topic here lies between 780 nm and 14 µm and is called Near Infrared. Its spectrum with respect to other radiation is shown in Figure 1.
Physicists Planck, Stefan, Boltzmann, Wien, Kirchhoff and others, working in the late nineteenth and early twentieth centuries, defined IR’s electromagnetic spectrum and the physical characteristics of its energy. To explain IR behavior, they created a theoretical Blackbody model, which absorbs all incoming radiation while reflecting or transmitting none.
Temperature is an objective measure of hot or cold and is determined with a thermometer. Some thermometers are based on the bulk behavior of thermometric materials such as mercury. Others may rely on measurement of thermal radiation, or assessment of particles’ kinetic energy. Every physical body with temperature above “absolute zero” (equal to 0° Kelvin, -273.15ºC or -459.666ºF) emits electromagnetic radiation proportional to its surface temperature. This radiation can be detected by IR sensors.
One application of IR radiation is for detection of movement of objects whose temperature differs from the background. The first extensive use of such infrared technology was by the military for personnel detection and in missile guidance systems. After declassification, the technology was commercialized by, among others, the security industry, for the design of passive infrared intrusion detectors (PIR). PIR has now become the most popular type of an intrusion detector, because it is inexpensive to manufacture, versatile, and—unlike ultrasonic, microwave or light beam interruption devices—emits no energy, thus making its presence undetectable. IR sensors have become an inexpensive commodity that can be found in many consumer applications, such as proximity switches, security lights, toys and various gadgets.
TWO MAJOR TECHNOLOGIES
The two major infrared-sensing technologies are thermopiles and pyroelectric sensors. Thermopiles comprise a stack of thermocouples—typically of bismuth and antimony—arranged on a chip around a heat-absorbing black material, integrated with a preamplifier to boost the output signal. Pyroelectric sensors are constructed of crystals such as lithium tantalate (LiTaO3) or deuterated-triglycine sulfate (DTGS), whose exposure to IR radiation causes surface loading. This, due to the pyroelectric effect, causes a voltage to be generated across the crystal. Because of the extremely high output resistance of the pyroelectric crystal (on the order of 1012 Ω), an FET source follower is integrated into the sensor to provide low-output resistance.
Most modern pyroelectric sensors for motion detection contain two pyroelectric crystals, X1 and X2, in series or parallel-opposite configuration, as shown in Figure 2. They attenuate the sensor’s response to common mode stimuli caused by ambient temperature variations, RF irradiation, vibration and other factors. This arrangement keeps the sensor output fairly constant when no moving object is within its field of view.
For reliable operation, the packaging of a PIR detector is crucial. In addition to holding the lens to properly focus the IR beam, the enclosure must prevent free air movement across the sensor. PIR detectors need optics to achieve good sensitivity, although a small aperture may suffice for short-distance detection. In the past, faceted mirrors were the mainstay, but eventually, Fresnel lenses replaced them, due to their low cost, low weight and easy manufacturing. You can still find mirrors in long-range, perimeter-protection devices. In my experience, the enclosure should not be hermetically sealed. In fact, it should be allowed to “breathe” to maintain the same environment inside and out.
Figure 3 illustrates detection of a moving object by a PIR. The sensor is located in the focal plane of an array of Fresnel lenses, often comprising many lens segments to achieve a desired pattern of detection field . Today, there are many commercially available, inexpensive lens arrays . They are usually molded from polyethylene, which, unlike glass, is transparent to near-IR radiation. IR radiation from an object depicted by the green rectangle and moving across the field of view from left to right strikes one sensor element (red rectangles) first, then the other. This causes the sensor to output two pulses. It also explains why PIR detectors are rather insensitive to the axial movement toward or away from the sensor. Both elements get irradiated at the same time, and the resulting common-mode signal is attenuated. The dual output pulse results only when the object moves across the field of view, even at a small angle.
Many commercially available Fresnel lenses provide both horizontal and vertical segments, whereas some are designed with fewer segments for a narrow field of view or for long-distance detection. Some detectors claim up to 140-degree detection view, but at 70 degrees off the center axis, the IR beam arrives at the sensor at too obtuse an angle to provide good sensitivity. A wider view is commonly achieved by arranging two sensors at 90 degrees to each other. This renders a reliable 180-degree angle of view. I purchased several devices for external security lighting claiming a 270-degree view, but none of them actually achieved that.
Figure 4 shows the inside of a PIR intrusion detector I designed some time ago. Nowadays, ICs containing all the needed circuitry are available from several vendors  . Nevertheless, to explain the PIRs’ inner workings I’ll describe the discrete circuits I utilized at the time. Before I discuss the circuitry, consider the close-fitting, albeit not hermetically sealed enclosure. Mounting holes and openings for wires allow it to breathe. Also notice the plastic ring, called the thermal ring, surrounding the sensor. It isolates the sensor from internal air movement.
THE SENSOR CIRCUITRY
The circuitry is uncomplicated and can be built with a single quad operational amplifier, such as the old faithful Texas Instruments LM324 or most modern operational amplifiers. Figure 5 is a simplified schematic diagram in which, for clarity, I omitted some frequency-band-limiting filter components, pulse discrimination and relays with their drivers. The pulses generated by the pyroelectric sensor in response to the movement of humans range from approximately 0.1 Hz to 10 Hz. This is not critical, but suppressing frequencies outside this range avoids false triggering.
As mentioned previously, the pyroelectric sensor contains an integral FET, operating as a source follower to provide low impedance output. U1A amplifies the signal and sends it to a window comparator comprising U1B and U1C. The comparator outputs are OR-ed and buffered by U1D for further processing. U2 is a three-terminal regulator, such as the ubiquitous 780x. To avoid false alarms, it is vital to keep the power supply stable, especially within the detector’s operating frequency band of 0.1 Hz to 10 Hz.
Modification of this basic PIR design for battery operation was, at the time, a challenge. The circuit with LM324 drew around 8 mA (relays not counted), and would rapidly drain the battery. A typical 9 V alkaline battery has roughly 500 mA-hours capacity. I tested this, as no battery manufacturer published their battery capacity below about 10 mA current drain. Still, at least 1-year battery life was the design requirement.
To reduce the current, I replaced the LM324 with a programmable quad op amp LM146 and modified the voltage regulator as shown in Figure 6a. It relies on a JFET Q1 as a series regulator, together with one of the LM146 amplifiers. A stable, low-power voltage reference presented the biggest challenge. I tested several devices and established that a specific red LED provided a sufficiently stable reference voltage at merely 2.5 μA current. (Being driven from stable regulator output helped.) Several storage capacitors throughout the circuit ensured the internal voltage stability during the RF transmitter operation.
Having given up one op amp for the power supply, I redesigned the window comparator as shown in Figure 6b. The circuit is essentially what’s known as a bit slicer. C5 averages the amplified signal of the PIR sensor, while the surrounding resistors and diodes determine the window size. The average current draw of the end product was approximately 14 μA, including regular 477 MHz transmission bursts reporting status every 4 minutes. The 9 V alkaline battery lasted more than 1 year.
As one would expect, present day integrated circuits do all the processing transparently to the designer. Their current draw is higher, but they operate at a lower voltage. This, combined with huge capacity batteries such as LiPo batteries (not available at the time of my design), make new designs quick and efficient. Nevertheless, I am sure my old discrete circuits illustrate to you how the PIR detectors work.
For detailed article references and additional resources go to:
References  through  as marked in the article can be found there.
Texas Instruments | www.ti.com
PUBLISHED IN CIRCUIT CELLAR MAGAZINE • FEBRUARY 2019 #343 – Get a PDF of the issue