Understanding Pulse-Width Modulation (PWM) for Infrared Optical Communications, Lab Reports of Theatre

Download Understanding Pulse-Width Modulation (PWM) for Infrared Optical Communications and more Lab Reports Theatre in PDF only on Docsity! 1 © Bob York 2009 ECE 2C Laboratory Manual 4 Infrared PWM Transmitter In this lab you will construct a circuit to produce a 40kHz analog pulse-width-modulator (PWM), to be used later in our one-way analog audio communication link. Later we will combine this modulator with the microphone circuit from am earlier lab to form a complete IR audio transmitter. In the following lab session we will build the IR receiver circuit to complete the communication link. By studying this document and experimenting with the components and circuits in the lab, try to understand the following: ■ Basics elements of a communication link ■ Analog PWM modulation in particular ■ Some elements of optical signal propagation and detection ■ Limitations on data rate due to circuit components Table of Contents Pre-lab Preparation 2 Full Schematics for IR PWM Modulator 3 Parts List 3 Background: Analog Communication Links 4 Information and Bandwidth 4 Baseband Transmission vs. Modulated Carriers 5 Pulse-Width Modulation (PWM) for Optical Communications 7 In-Lab Procedure 8 4.1 Variable Duty-Cycle Circuit 8 4.2 Analog Pulse-Width Modulation 9 Other PWM Circuits 10 4.3 A Simple Infrared Link 10 Transmitter 10 IR Detectors 11 4.4 Hardwire the Transmitter Circuit 13 Lab 4 Record 14 2 Infrared PWM Transmitter © Bob York 2009 Pre-lab Preparation You have one week to complete this lab. Read through the lab experiment to familiarize yourself with the components and assembly sequence, and complete the calculations below. Before coming to the lab, each group should obtain a parts kit from the ECE Shop. Bring your solderless breadboard, tools, & wire jumpers. Required calculations: □ Determine the capacitance C1 in the 555 timer circuit of Fig 1 that is needed to provide pulse frequencies of 4kHz and 40kHz, respectively. 4kHz: C1=______________uF 40kHz: C1=______________uF □ Coupling capacitor C3 and the resistor chain R1-R2-R6 form a high-pass filter. For the component values shown in the schematic, find the maximum and minimum cutoff frequencies as potentiometer R6 is varied over its full range. minLf = ______________Hz maxLf = ______________Hz □ Determine an appropriate value for the IR LED biasing resistor R9 to provide a 50 mA current flow, assuming a power supply voltage of 12V. The forward voltage drop of the diode can be determined from graphs provided in the data sheet. R9 = ______________Ω Optional: Simulate the PWM circuit using MultiSim or Circuit Maker™ (circuit files available on the course web site). Background: Analog Communication Links 5 © Bob York 2009 5 with the information content must be concentrated in a finite range of the frequency spectrum. If the communication link can support the bandwidth associated with a given signal, this ensures that the information content of the signal will be accurately conveyed to receiver without distortion. The spectrum of a certain spoken vowel sound in the previous example is shown in Figure 4-2c. After analyzing the spectra of many different speakers in this way, and observing the influence of various electrical filters on the reproduction of their voices, researchers have concluded that a limited range of frequencies from 200-300 Hz to about 3-4 kHz is required to communicate intelligible speech. A somewhat larger bandwidth of 8 kHz or more is required for a faithful reproduction of a typical human voice. High quality music transmission may require an even wider range of frequencies, from 50Hz through 16-20 kHz, to accurately reproduce the more complex acoustical characteristics of musical instruments such as violins and horns, sounds that are rich in spectral content. So, depending on the kind of audio signal we wish to transmit, an audio communications system must support bandwidths of anywhere from 3 kHz to 20 kHz. A typical phone or intercom system usually has a bandwidth of about 3 kHz. We can generalize these ideas to all audio, video, and data communications: every signal has a characteristic band of frequencies, or bandwidth, representing the essential information content of the signal. Conversely, the bandwidth of a communications link will determine its maximum rate of information transfer, or data-rate. Thus in communications we will assume our information-bearing signals are band-limited; a baseband signal is one with a bandwidth concentrated at low frequencies, such as a voice or music signal as shown in Figure 4-3. Of course, real signals never perfectly fit within a finite bandwidth, and practical communication channels will never support an infinite bandwidth, so we must always expect some signal distortion in an analog communications. Baseband Transmission vs. Modulated Carriers Signals can be transported from place to place using acoustic or electromagnetic waves. When low-frequency band-limited signals can be transmitted directly, for example over wires, coaxial cables, or as a simple acoustic wave (human speech or music), this is referred to as a baseband communication link. This is the simplest and easiest system to implement. For long-distance or wireless communications, however, baseband communications are rarely possible and we must use a modulated carrier. Why? In order to efficiently radiate an acoustic or electromagnetic wave, any radiator (loudspeaker, ultrasonic transducer, antenna, etc.) must be comparable in size to the largest wavelength being transmitted. This is a fundamental property of all types of waves. The wavelength  is related to the frequency f by the fundamental relation /v f  , where v is the wave velocity. At audio frequencies the acoustic wavelength is reasonably small, typical of the audio speaker systems you are familiar with. But electromagnetic waves at audio frequencies have enormous wavelengths; at 3kHz, the wavelength is 100 km! So it is usually impractical to transmit electromagnetic waves directly at such low frequencies. On the other hand, electromagnetic waves are generally more desirable than acoustic waves for long-distance communication in air, since acoustic waves are more strongly attenuated than electromagnetic waves in air (the opposite is true for undersea communication). ( )X f fB0 Figure 4-3 – Fourier spectrum of a baseband signal x(t), concentrated in a bandwidth B. 6 Infrared PWM Transmitter © Bob York 2009 Thus, in order to take advantage of improved propagation characteristics of electromagnetic waves but keep the radiators to a reasonable size, we need to transmit and receive in a frequency range that is often much higher than the bandwidth of the information. AM radio stations in the U.S., for example, broadcast at frequencies in the range of 530 kHz to 1.7 MHz, but the information content of each radio station is restricted to a bandwidth of 10.2kHz or less. The high- frequency signal effectively “carries” the narrow bandwidth of information from the transmitter to the receiver,; this is the carrier frequency in the communication system. In this AM radio example, the information bandwidth represents less than 1% of the carrier frequency. Modulation is the process of shifting the information band onto a high-frequency carrier wave. The reverse process of extracting the information from the carrier after reception is called demodulation. There are many different ways to modulate a carrier. Two conceptually simple methods—Amplitude Modulation (AM) and Frequency Modulation (FM)—are illustrated in Figure 4-4. All modulation schemes have their own set of advantages and disadvantages. Some, like AM modulation, have the advantage of very simple hardware implementations but may be susceptible to noise and interference. Others, like FM, require more complex electronic implementations but have improved signal fidelity. In the frequency domain, modulating a high-frequency carrier with some baseband information-bearing signal has the effect of shifting the information bandwidth to a range of frequencies around the carrier. The information thus appears as “sidebands” around the carrier, as illustrated in Figure 4-5. In general there are two sidebands above and below the carrier that are mirror images of each other, and thus a bandwidth of 2B is required around the carrier for such “Double-SideBand” (DSB) transmission. AM radios in the U.S., for example, use DSB transmitters and have a channel bandwidth of 2 10.2 20.4kHz  . It is possible to make a “single sideband” (SSB) transmitter, but this requires more complex hardware. It is even possible sometimes to suppress or eliminate the carrier before transmission. So there are a wide variety of signal spectra that are encountered in communications. The key idea is that we can shift our information to a higher-range of frequencies through the modulation/demodulation process. This enables us to pack a lot of information into a certain range of frequencies. Again returning to our example of the U.S. AM radio system, there is a total FCC-approved bandwidth of 1700-530=1170kHz, so in any given geographical area, approximately 58 different AM stations could be broadcasting simultaneously (1170 / 20 58 ). Figure 4-4 – Two simple modulation schemes. ( )X f f cf Baseband signal spectrum Carrier “sidebands” representing baseband signal cf Bcf B Figure 4-5 – Effect of modulating a sinusoidal high- frequency carrier with a baseband signal. Background: Analog Communication Links 7 © Bob York 2009 7 Moving to higher frequencies also makes it somewhat easier to increase the information bandwidth for a communication link. All transmitting and receiving systems will have a finite bandwidth, but a bandwidth of 5-10% of the carrier frequency is often possible with simple circuits and antennas. So, for example, a point-to-point wireless link operating at 24 GHz using antennas with a 5% bandwidth corresponds to an available information bandwidth of 1.4GHz, enough to support a rather large number of simultaneous voice, video, and data channels. Now you can appreciate the potential for fiber-optic communications; the carrier frequency is on the order of 1410 Hz, so the intrinsic capacity of the optical fiber as an information conduit is enormous. To summarize: it proves advantageous to modulate for either or both of the following reasons: (1) to take advantage of the superior propagation or radiation characteristics of high frequency acoustic or electromagnetic waves, or (2) to transmit multiple information-bearing signals simultaneously from one point to another. As time goes on, and the demand for ever- faster data rates increases, wireless communication devices will therefore continue to evolve towards higher carrier frequencies; count on it! Pulse-Width Modulation (PWM) for Optical Communications Unfortunately it is beyond the scope of this course to probe the subject of analog an ddigital modulation techniques more deeply; that is firmly in the realm of more advanced junior- and senior-level coursework in communications theory. In this lab we will explore one modulation format in particular: so-called pulse-width modulation (PWM), which for various reasons is well-suited to optical communications and is relatively easy to implement. The basic idea of PWM is illustrated in Figure 4-6. The technique requires a pulse train at some frequency 1/T with a variable pulse width or duty cycle. The modulating signal is used to control the duty cycle; that is, the instantaneous pulse width at any given time is proportional to the amplitude of the modulation signal. As the signal increases the pulse width and duty-cycle increase; as the signal decreases, the pulse width decreases. In optical communications the light source is usually an LED or Laser diode. It is relatively easy to construct a circuit to switch the device on and off quickly, so PWM is well- suited to diode-based emitters. All we need is a variable duty-cycle circuit to control the time the emitter is on or off, and you may recall that we already discussed how to make those last quarter in ECE 2B when we built a variable-duty-cycle fan speed controller. De- modulating the signal is relatively easy too; we will discuss this in more detail in the next lab, but you can see from the figure above that the average value of the PWM waveform over each carrier period is proportional to the pulse width, which is turn is proportional to the modulating signal. This all we need to do at the receiver end is low-pass filter the PWM waveform to eliminate the carrier and recover the modulating signal! Signal PWM T Figure 4-6 – Pulse-width modulation. 10 Infrared PWM Transmitter © Bob York 2009 of the comparator is then the sum of the DC value set by the R1-R6-R2 divider, plus a time- varying modulation signal. Remember that anytime we add a capacitor in this fashion we are potentially adding a pole in the AC frequency response. In this case, C4 and the resistor divider create a high-pass response for the modulation signal. The cutoff frequency will change a bit depending on the setting of R6, so we must be sure to choose a large enough capacitance to insure that the cutoff frequency is always well below our lowest desired frequency for all possible states of R6. In the prelab calculations you should have estimated the range of cutoff frequencies for C4=1μF as shown in Figure 4-9. □ Add the AC coupling capacitor C4 as shown in Figure 4-9. □ Set up your function generator to supply a low- frequency (1 Hz) sinusoidal modulation signal with a peak amplitude that is approximately what you recorded in the last step of section 4.1 for a 30% duty cycle. Apply this signal to the modulation terminal of your circuit and record your observations. Do you understand what is happening? Vary the amplitude of the function generator a little to enhance the effect. Note that you may need to adjust the oscilloscope for negative-edge triggering for clarity, since the falling edge of the waveform always coincides with the falling-edge of the sawtooth generator, and that provides a more stable reference for triggering the oscilloscope. The timing of the leading edge will vary with the modulation signal. □ Increase the frequency of the function generator to 100 Hz and record the waveform. Other PWM Circuits Anyone who has bothered to read the 555 data sheet will note that there is a very simple PWM circuit that can be made using ONLY the 555 and a handful of passives1. You may wonder, then, why we bothered with the more complicated circuit in Figure 4-8. In fact we could probably get away with the simpler version too, but for reasons which will become clear later, we wanted a duty cycle that could be varied above and below 50%, whereas the simpler 555-only circuit is constrained to duty-cycles <50%. 4.3 A Simple Infrared Link Transmitter Now we have all the circuit components we will need to assemble our transceiver system. The last ingredient is the infrared transducers: the devices that will generate or receive the infrared signal. You may recall that we used an infrared emitter and phototransistor pair in ECE2B to as part of a “chopper” circuit to generate a pulse train proportional to the speed of a fan blade. 1 Fig. 8 on p.8 of the National 555 data sheet. LM393N Figure 4-9 – AC-coupled modulation signal controls the duty cycle. A Simple Infrared Link 11 © Bob York 2009 11 The transmitter portion of the link is simple, we just use the output of the PWM circuit to drive a transistor switch as shown in Figure 4-10. Here we have chosen a high-efficiency IR emitter, designed to operate at somewhat higher current levels than we normally encounter in LEDs in order to produce a high- intensity output. □ Using the value for biasing resistor R9 that you calculated in the prelab, add the IR transmitter circuit in Figure 4-10 to your PWM circuit. □ Observe the voltage across the LED or R9 to confirm that it is being driven by a pulse train. IR Detectors Now add a simple phototransistor detector circuit as shown in Figure 4-11. This is again similar to the IR detector circuit that we used in 2B: when the base region of the phototransistor is illuminated, charge carriers are created which bias the transistor “on” and drive a current through the load resistor R11: □ Add the phototransistor-based detector as shown. In this phase, the emitter and detector should be spaced a few inches apart on your breadboard. To start with, use a 3.3k resistor for R11 □ With the PWM transmitter operating at 4 kHz, observe the output waveform on the detector, noting the peak amplitude of the signal and the pulse shape. Can you estimate a rise-time and fall-time for this waveform? This simple test illustrates one of the challenges that always comes up when transmitting pulse trains. In this case, the internal capacitances of the phototransistor, along with R11 and re, create an RC low-pass response that distorts the signal. We can improve things a bit by using a smaller load resistor, which should reduce the RC time constant in the circuit: □ Replace R11 with a 1k resistor and record the output waveform again, noting the rise- and fall-times as before. We could continue reducing R11 for further improvement, but note that each time we do, we also reduce the amplitude of the detected signal! Even so, you might think that what we have so far is “okay”: the detected signal is a reasonable approximation to a square wave, and we Figure 4-10 – Infrared emitter circuit. emitter flat Figure 4-11 – Infrared emitter circuit using a phototransistor. 12 Infrared PWM Transmitter © Bob York 2009 could always amplify the signal if we have to. However, there is still a major problem: we’re still at 4 kHz! □ Replace timing capacitor C1 with the value that gives a 40kHz output. You may have to re-adjust R6 at this stage to obtain a 50% duty-cycle. Record the waveform across Q2 or R9 to confirm that the emitter is still being driven appropriately. □ Now observe and record the output waveform on our photodetector circuit, using a 1k resistor for R11. Yikes! Ultimately the problem here is the phototransistor capacitance. As a circuit designer there are some tricks we can use to mitigate the effects of the capacitance; for example, using a cascode arrangement where the phototransistor drives a fast CB or CG gain stage, or similarly using a trans-impedance (current-to-voltage) amplifier with a very low input impedance. But there are other issues too: the phototransistor itself is an amplifying device; it has a built-in current gain, and thus amplifies all incoming light indiscriminately. This is a disadvantage in situations like ours where the incoming signal could be relatively weak in comparison to ambient light and other potential sources of interference. Also, phototransistors tend to have rather large “dark” currents in comparison to other detectors, that is, the leakage current that flows even when no light is incident on the device. This shows up as additional background noise after amplification. A better solution is to use an intrinsically faster detection device, one with smaller internal capacitance and (ideally) a greater sensitivity to low-level signals, and then build in some frequency-selective amplification to extract only the modulated carrier that we are interested in, rejecting other interfering signals. Enter the PIN photodiode. This is a diode that is designed to be operated under a large reverse bias. The leakage current that flows in the device is proportional to the incident light. Such devices are quite fast and very sensitive. The one we use also has a built-in IR filter to help reduce interference from ambient visible light, and optimized for use with the TSAL6100 emitter that we are using. It has a cutoff-frequency that is similar to the gain-bandwidth product of our LF353 op-amps (4 MHz), corresponding to an expected pulse rise-time of ~80 ns, which should be adequate for our 40kHz pulse train ( 25 sT  ). □ Using the BPV23F PIN photodiode in your kid, construct the detector circuit shown in Figure 4-12. Note that the supply connections for the op-amp and decoupling capacitors are not shown. The decoupling capacitors are IMPORTANT!. Also, Note that the anode is connected to Vss, which is a NEGATIVE supply voltage, i.e. ss ddV V  . That this detector circuit is a current-to-voltage amplifier, otherwise known as a trans- impedance amp: the input impedance seen by the photodiode is very small, and the Iphoto Figure 4-12 – Photodiode and detector circuit.