Loading...
HomeMy WebLinkAboutOn the Effects From High Powered Microwave Illumination 1993ABSTRACT ON THE EFFECTS FROM HIGH-POWERED MICROWAVE ILLUMINATION February 1993 C. D. Taylor and N. H. Younan Mississippi State University Mississippi State, MS ON THE EFFECTS FROM HIGH-POWER MICROWAVE ILLUMINATION C.D. Taylor and N.H. Younan Mississippi State University Mississippi State, MS ABSTRACT The effects of wireless power transmission using high-power microwaves are considered for biological materials and electrical/electronic systems. Quantification of electrical stress from microwave illumination is discussed. And conditions for deleterious effects are identified and discussed. INTRODUCTION Wireless power transmission is the theme of this conference. The eventuality of this concept depends upon a number of factors, some technical, some economical, and some environmental. The technical and economic problems can be, and to a great extent have been, solved. This paper addresses the environmental concerns for microwave power transmission. The effects from microwave illumination can be separated into three major categories: Enthalpic effects, biological effects, and electrical effects. Enthalpic effects include increases in temperature from dielectric hysteresis and ohmic heating of electrical conductors. Biological effects include the variation in life processes within living organisms resulting from microwave illumination. Electrical effects include the development of consequential currents and voltages within electrical and electronic systems. Conditions for deleterious effects are identified and discussed. Microwave heating of dielectric materials is well known. It varies with the material and the frequency of the illumination. The analysis is accomplished by the introduction of complex permittivity. Electric conductors are heated by ohmic power absorption in the region very near the conductor surface. These processes are examined in detail. Most of the data from biological effects from microwave radiation are explained by thermal energy conversion, almost exclusively as enthalpic energy (heating) phenomena. Results from recent studies are used to determine power density levels that are known to cause deleterious effects in humans and laboratory animals. These results are compared with standard safe exposure levels set forth by the U.S. government. Electrical effects are studied as an electromagnetic interaction process. Data are presented for specific effects on electronics under microwave illumination. The general quantification of electrical stress from microwave illumination is also discussed. ENTHALPIC EFFECTS Microwave heating of dielectric materials is well known. the process is known as dielectric hysteresis and is analogous to magnetic hysteresis in magnetic materials. Dielectric heaters are in widespread use in various industries to seal, emboss, dry, and mold materials. These devices generate from hundreds to thousands of watts RF power at 20 to 70 MHz. Moreover, their exposure fields exceed 200 V/m or 0.5 A/m. This corresponds to a power density of 11 mW/cm?. The analysis of microwave bulk heating can be accomplished by the introduction of complex permittivity, e=-¢ - je’ (1) where e’ = etand - o/o (2) Here, tan6 is called the loss tangent and o is the conductivity of the material. The power absorbed into heat per unit volume is P= (we/tand) E? + oF? (3) where E is the electric field strength of the microwave. Typically, e’ and e” change with frequency. Over very wide frequency ranges, materials may exhibit resonance absorption and associated permittivity changes due to various molecular vibrational modes. The water molecule is an effective absorber of microwave radiation. Electrical properties of human tissue at microwave frequencies have been studied extensively. According to the Kronig-Kramer relation, the complex relative dielectric constant, €, = €/€, for dielectric materials is ' & - En __ (4) 1+ jot €,(m) = €, + where €, and €, are the high and low frequency limits respectively and t is the dielectric relaxation time. these may vary depending upon the type of biological tissue. Similarly, the equivalent conductivity is o,(@) =o + wee, (wt)? (5) 1+ (wt)? Wt 0, + (0, - O) where ©, and ©, are the high and low frequency limits. For skin tissue, appropriate parameter estimates are: €, = 42.0, €, = 4.0, t = 6.9 ps, and 6, = 1.4 U/m. Therefore, at 1 GHz, €, = 41.9-j1.64 and o, = 1.4 U/m ”. Corresponding measured data agrees quite well. Other types of tissue at 1 GHz are listed in Table I. Table I. Tissue Parameters at 1 GHz? 1.27 to 13.3 0.83 to 1.49 0.43 to 1 BIOLOGICAL EFFECTS It is generally accepted that biological effects from microwave radiation occur as a result of power absorption, conversion to thermal energy, and the consequential increase in temperature. For animals and humans, this process is complicated by nonuniform power absorption and the internal thermal regulation processes. Predicting distributed temperature increase within irradiated bodies becomes largely an intractable problem. Moreover, a specific biological effect on the functioning of a given organ may be the result of a stimulus from the power absorbed by an organ some distance away. Under microwave illumination, that produces increases in temperature from diverse etiologies and pathological effects which have been observed include chromosomal alterations, mutagenesis, virus activation and inactivation, as well as behavioral effects 7°. Table II provides exposure levels for common biological effects from microwave 3 illumination. Table I. Biological Effects and Exposure Levels *? OBSERVATION EXPOSURE LEVEL No effects observed in < 10 mW/cm? Rodents Mutagenesis < 10 mW/cm’ Fetal Development No effects observed Central Nervous System Transient function < 10 mW/cm” changes occur 5 to 20 mW/cm” Behavioral Effects Effects on Rodents occur Initial stimulation of < 10 mW/cm’ leukocyte production seen in Rodents Immunity System Audible clicks perceived > 1 Wem? under modulated illumination Auditory Response Ocular Effects Epithelial and Stromal 10 mW/cm’ @ 35-107 GHz injuries to human eyes expected Skin Sensitivity Warning sensation occurs 27 mW/cm? @ 2.45 GHz 1 mW/cm? @ 35-107 GHz | The non-uniform pattern of microwave absorption, with differing rates of temperature rise at absorption sites, results in a pattern of heating which can not be replicated with radiant, convected, or conducted heat. Therefore, experimental control groups for temperature increases to discriminate between direct and indirect effects from microwave illumination, or between effects due to temperature rise and effects independent of temperature rise, may have limited value. MICROWAVE HEATING EFFECTS Microwave radiation can produce appreciable heating effects in animal and human tissues. For frequencies up to 10 GHz, a single exposure fluence of about 1 MJ/m? (or 100 J/em?)can 4 produce significant heating, where fluence is the product of the average incident power density and the time of exposure. At higher frequencies, the microwave absorption occurs near the body surface which results in significant heating at lower fluences. For example, at tens of Gigahertz, a fluence of only 20 J/cm’ can burn the skin. The brain is particularly sensitive to heating effects. Microwave induced temperature increases of only a few degrees have caused convulsions, unconsciousness, and amnesia in laboratory rats. Corresponding effects in humans would be expected at power densities of 10 to 50 mW/cm? under continuous exposure for frequencies below 10 GHz. The mammalian blood-brain barrier is a layer of cells around the small capillaries in the brain that is believed to help regulate the fluid environment of the brain. It has been observed that the blood-brain barrier is changed by stresses of physical, chemical, and physiological forms. Under relatively low fluence levels, as low as 100 mJ/cm” , single microwave pulses have depressed electroencephalograms of rodents. Overall, an increase in the total body temperature of 1°C from microwave radiation is considered to be harmful, and prolonged exposure to a great increase in temperature from microwave radiation can be fatal. At frequencies below 400 MHz and above 3 GHz, less than one-half of the incident energy on a human is absorbed. However, between 1 GHz and 3 GHz, the amount of incident energy absorbed approaches 100%, depending upon the skin thickness and the subcutaneous layers of fat. It has been found that pulsed power can induce biological effects not seen with CW power of the same average value. The maximum CW power density considered to be safe by the U.S. government is 1 mW/cm2 under unlimited exposure and 10 mW/cm? for exposure times less than one hour”. Some common exposure levels from radio frequency (RF) sources, including microwave, is shown in Table III. Table III. Common Exposure Levels from RF Sources 7! SOURCE EXPOSURE LEVEL FREQUENCY Dielectric Furnace > 10 mW/cm? 20 to 70 MHz Microwave Oven < 5 mW/cm? 2.45 GHz Leakage Microwave Diathermy < 10 mW/cm? 2.45 GHz Leakage Short Wave Diathermy > 10 mW/cm? 27.12 MHz Leakage Hand-Held Transceivers > 10 mW/cm? 27.12 MHz > 10 mW/cm? 150 MHz MICROWAVE-INDUCED AUDITORY EFFECT When human beings are exposed to modulated microwave radiation, an audible sound can be perceived 4. At times, the sound has been described as a click, buzz, or chirp, depending on the pulse width and repetition rate. Pulses of 1 to 32 ys pulse width at a frequency of 2.45 GHz and fluences as low as 40 J/cm? have been heard as distinct clicks that appear to originate from within or immediately behind the head. The pulsed microwave energy is believed to induce a thermoelastic wave of pressure in brain tissue that drives the inner ear receptors via bone conduction. Unfortunately, the presently available physiological data are not sufficient to permit a complete evaluation of the health risk to human beings. ELECTRICAL EFFECTS Under microwave illumination, metal surfaces and wiring will have induced surface currents. Wire currents, in turn, will lead to termination voltages. In this section, the coupling of microwave energy into electrical systems is considered in detail. Both theoretical and experimental data are presented. RECEIVING CROSS SECTION When characterizing the electromagnetic coupling from the exterior of an electrical system to a specific interior point, a convenient parameter is the receiving cross section defined by P, Pp inc (6) Or= where P, is the power delivered to a load, Z,, connected across the specified terminals and p™ is the incident power density. As shown by Park and Tai *, the receiving cross section can be expressed by the general relationship 2 Orp= A = 4 7 aq NPT D(8, >) (7) Here p is the polarization matching factor and q is the impedance matching factor, 4R,R,, = 8 qg Tz, * Zi + Zin (8) where R, and R,, are the resistive components of the impedances Z, and Z,, respectively. The impedance Z,, is the impedance seen when the subject terminals are driven forming a radiating antenna with ohmic efficiency varying from zero to unity and with directivity gain D(@,o). Note that D(0,) = 422° p(r,6,) (9) Prag where p(r,0,0) is the power density radiated in the (0,0) direction and P,,, is the total radiated power. The ohmic efficiency is Prad zi Pin n= (10) where P,, is the power delivered when the subject terminals are driven interior to the electrical system. The quantity p appearing in equation (7) represents the polarization matching factor with a value varying from zero to unity. For linear polarization incident on the system and linear polarization transmitted from the driven interior terminals, the polarization factor is simply the magnitude of the cosine of the angle between the illuminating and the radiated fields. The receiving cross section expression of equation (7) is the most general formula characterizing the transfer of power from an incident field to a specific set of terminals. Note that for an aperture antenna, the receiving cross section is the same as the effective aperture area. Using equation (7) in equation (6) with Bi pi? = ape = = EQlHie/? (11) 0 relating the incident power density to the incident electric and magnetic field strengths enables the determination of the received power in terms of the incident field strength. Table IV presents the maximum receiving cross sections for certain canonical configurations *’. Using the maximum receiving cross section provides an upper bound on the power coupled to the interior terminals. From Table IV, it appears that for a set of interior terminals, the maximum power that can be coupled decreases with an increase in frequency. This is somewhat misleading since the efficient extraction of power from an electrically small antenna is very difficult because of the very small resistive part of the antenna impedance, the radiation resistance. For example, the radiation resistance of a 10 cm dipole at 300 MHz (1m wavelength) TABLE IV. Receiving Cross Sections (Effective Areas) and Radiation Resistances for a Few Wire Configurations *’. CONFIGURATION ILLUSTRATION MAXIMUM og (2: = Zent) RADIATION RESISTANCE (Re{Zani}) t in Ohms h 2 Center-Loaded 3 2, he<a 802? (:) » <<a Linear Zt 2h 8x ; Dipole ; ant | a 2, h-= nd forodd n 30C;,(2mm), 2h=n 2 forodd n = Cin(2am) = 2.44 n=1 Zr = 3.11 n=3 = 3.33 n=5 3 mca 3205 (2)', 2a<<a —— 4 Sa , z) ° << pata 0.33844 , a>A 12023 ($) , ark 3 a\? A =P, m<<a 40n27 (2), 2n<<a Monopole 4x a Linear h 60 z h Antenna ZL —— % , 2h =n foroddn 15C;,(20h) , 2h =n= forodd n 732 2 2 2, </> 3p teca 6Hont ME teed Transmission 4n > Line h A Antenna hf Zr 2 2, ten x forodd n 24022 (:) » f= n> forodd n is only 2 Q (see Table IV for formula). Whereas the corresponding radiation resistance at 4.5 GHz is about 93 Q (again see Table IV for formula). FIELD-TO-WIRE COUPLING The computation of currents induced on wire structures at microwave frequencies becomes more difficult as the electrical length increases. Adams and colleagues * performed numerical studies of field-to-wire coupling for a few antenna configurations illuminated at microwave frequencies. They found that the wire currents were developed in a pattern consisting of standing waves and traveling waves where the peak currents tended to remain relatively constant as the frequency increased. Their results were obtained for cases where the direction of the incident microwave beam and the polarization were chosen to yield the greatest peak currents. When they considered transmission line antenna configurations, their results indicated that the higher order (waveguide) mode excitation contributed little to the induced wire currents, even when the transmission line separation was several wavelengths. Consequently, the subsequent analysis of microwave coupling to transmission line configurations will be directed toward obtaining only the TEM mode response. Note that the results in Table I include the maximum receiving cross section and radiation resistance for a simple transmission antenna configuration where h<<A is assumed. EFFECTS ON ELECTRONICS Once microwave energy reaches the illuminated object, a sequence of penetration and propagation processes will take place from the object’s outer surface into its interior and ultimately arriving at its electronics. The energy of arrival can be affected via either front-door or back-door paths. A front-door path is referred to as an intended path for microwave transmission and reception, an example of which is an antenna is connected to a coaxial cable terminating at an electron box. A back-door path is an inadvertent point of entry for energy penetration, such as windows, doors, cracks, seams, connectors, cable shields, or non-electrical lines. Rectification. For electronics under microwave illumination, rectification is usually the principal mode by which the microwave energy is coupled into the system electronics. it is the cause for computer malfunction under radar illumination, for erroneous output from an EKG when a physician’s paging system is activated, or for a stereo system to receive a citizen’s band (CB) signal. It is also because of rectification that warnings are posted for heart pacemaker interference from microwave ovens and airport radar transmitters. Generally, microwave energy coupling via rectification occurs by the signal entering a victim amplifier, digital circuit, etc., through interconnecting signal and/or power cables, and perhaps is enhanced by parasitic resonances. A typical response is then characterized by the microwave signal being propagated to a non-linear device, such as the detector of an AM radio or the bipolar junction transistor input to a digital gate. The resulting non-linear response produces a "video pulse" or a wideband signal, which propagates throughout the electronic system, perhaps as a "legitimate" signal upsetting the normal data transmission and storage. In some cases, there may be overstresses occurring that damage system components. The resulting effects may be temporary, i.e., ceases as soon as the source is removed, or they may be permanent. Intermodulation. Intermodulation is a phenomenon associated with electromagnetic interference (EMI). It arises from the simultaneous operation of two or more sources with non- linear elements located either in the victim’s response circuits, the transmitters, or the propagation path °. For example, suppose f, is the frequency of a signal normal to the system operation and f, is the frequency of an incident microwave radiation. The superposition of these two signals combined with non-linearities produces intermod product signals with frequencies f, = Mf, + Nf, where M and N are integers, either zero, positive, or negative. The order of the modulation product is O= |M| + |NI An example of severe intermodulation interference problems have been aboard ships e Here, the interference results from electromagnetic scattering from metallic structures such as ladders, life-raft rangers, guard rails, antenna guying wires, booms, anchor chains, etc., where oxidized metal-to-metal joints develop junction diodes. As a result, significant intermod product signals were generated. Latchup. A general definition for latchup is the condition where a semiconductor device no longer responds to an input. In some cases, latchup may lead to the destruction of the device as occurs in CMOS circuits. There are a number of preventative measures that can make devices less sensitive. Generally, the latchup state arises from a parasitic transistor condition that occurs in an IC fabrication. An inadvertent multijunction SCR-type switch is formed, either a pnpn or a npnp. When the switch is triggered, the device is disabled. Possible sources of latchup are: (1) minority carriers injected into the substrate by a transient forward bias on parasitic pn junctions, (2) Photoelectric generation from ionizing radiation, and (3) impact generation from thermal heating. A common problem for CMOS circuits is latchup. The operation of CMOS circuits incorporates a series combination of Complementary Metal-Oxide-Semiconductor transistors where one transistor is "ON", i.e., conducting, and the other is "OFF", i.e., not conducting. When latchup is triggered, both transistors go to the "ON" state. With both transistors conducting, there 10 will be a large current through the devices to ground. Generally, this condition leads to burnout. Thermal Damage Process. A major failure mechanism for semiconductor devices under a short pulse stress is thermal second breakdown. This failure mechanism is illustrated in Figure 1, where the voltage-current curve for a pn junction is shown. The first breakdown , which occurs in the reverse bias configuration, is an avalanche breakdown that is associated with the Zener diode. As the current is increased in the reverse bias configuration, the second breakdown, a destructive-irreversible process, occurs. Thermal second breakdown is a result of heating in the junction region. For a reverse bias voltage applied to the junction, the voltage drop appears across a narrow region about the junction, i.e., known as the depletion region. Consequently, the heating is localized to the junction region. Second breakdown is thought to be a filamentation process that occurs in three stages: nucleation of the filament, the growth of a relatively broad filament across the depletion region, and the growth of a second filament of molten material within the first filament. The filamentation growth begins in a region of high current density, and more than one filament may occur. With the formation of the filament of molten material, the device is irreversibly damaged. The occurrence of thermal second breakdown for a video pulse depends upon the stress pulse width and average power. For pulse widths in the range of 100 ns to 100 Us, the damage threshold power, Pp, required for thermal second breakdown is " Po 2 y-3/2 (12) To where A, is the junction area, t is the pulse width, and K, the proportionality constant, is called the K-factor. The K-factor is characteristic of the device. Since breakdown may occur over only a fraction of the junction area, one tenth of the junction area is often used for A, in equation (12) for bounding purposes. The device K-factor can be computed from product data or it can be measured ". Moreover, significant device-to-device variation in the K-factor may occur because of the nature of a breakdown process. Typical ranges of damage threshold levels for devices are shown in Figure 2. Here, the damage threshold power is expressed in units Kilowatts for 1 [1s rectangular pulse stresses . According to these data and equation (12), the damage threshold power may be as low as 1W for microwave diodes or as large as 160 KW for high power transistors under a 100 ns rectangular pulse stress. The foregoing damage threshold data are strictly applicable for rectangular pulses. In order to apply the data to a general video pulse, it is recommended that an equivalent rectangular pulse be identified by equating the pulse energy (time integral) and duration of the video pulse to that of the rectangular pulse. 11 Saturation Current Breakdown Voltage 7 7 Secondary Breakdown Reverse Bias Figure 1. Voltage-Current characteristic of an ideal p-n diode illustrating both the forward and reverse bias behavior 12 HIGH POWER TRANSISTORS SILICON CONTROLLED RECTIFIERS GERMANIUM TRANSISTORS SWITCHING TRANSISTORS LOW POWER TRANSISTORS RECTIFIER DIODES REFERENCE DIODES SWITCH DIODES POINT CONTACT DIODES MICROWAVE DIODES INTEGRATED CIRCUITS 0.01 0.1 1 10 100 (kW) 0.001 0.01 0.1 1 10 100 (kW) Figure 2. Damage threshold power range of representative transistors, SCR’s, diodes, and integrated circuits for 1 [1s rectangular pulse stresses |. 13 For a very short stress pulse, < 100 ns, the thermal process causing device damage is adiabatic. In this case, damage occurs before the absorbed heat is conducted away from the junction region. Consequently, the energy available to cause damage is proportional to tPp. And for long stress pulses, > 100 ps, the thermal stress producing damage approaches the constant power rating that is provided in the manufacturer’s specifications for the device. Combining all three of the foregoing failure mechanisms into a single relationship yields P u =? ~ K'YA;t + Krw2 4 KO 40 (13) A; VAs where the proportionality constants K’ and K” are characteristic of the specific device ". In this section, only discrete devices have been considered up to this point. integrated circuits (IC’s) consist of a number of devices with a large number of pn junctions. However, after testing a number of IC’s, it appears that the damage threshold powers for stress pulse widths of 100 ns to 100 ps appear to follow equation (12) dependence on pulse width. Table V provides damage threshold powers measured for a few IC’s. These data are normalized for a 100 ns pulse width. Measurement of the damage threshold power for various devices under microwave pulse excitation were performed by Antinone and Ng ". Their results indicate that power levels, required for device damage, are typically near 100 W for 1 p1s pulses at frequencies of a few GHz. Using equation (12) to estimate the dependence on pulse width would indicate threshold power levels of 320 W are needed for 100 ns pulses. This result is consistent with the data presented in Table V. Punch-through. In the fabrication of integrated circuits, multiple pn junctions are normal occurrences. When the pn junction is reversed biased, the depletion region expands with increase in voltage. If the depletion region expands to another junction, then the resulting current may be sufficiently large to damage the junction producing what is called punch-through. Digital Circuit Upset. Advanced VLSI/VHSIC digital devices using either bipolar or CMOS technology have shown to be highly susceptible to upset (change in stored or transmitted information) as well as damage when HPM illumination is directed upon the chip or multi-chip package '*. These devices utilize sub-micron dimensions and operate at clock rates up to 100 MHz with logic levels of 3.3 V. As device dimensions and logic levels correspondingly decrease, even more sensitivity will occur. Test results from HPM testing of various analog and digital devices show that the microwave power that is required to cause upset in a digital device is more than that required to cause disturbance in an analog device '. Moreover, the test results do not vary significantly 14 Table V. Threshold Power for Damage Measured for ICs Failure Power (W) Device Tye elif Fairchild 9930 Dual 4-input Gate 730 290 660 Signetics SE 8481 Quad 2-input Nand Gate 230 149 ©1230 T1946 Quad 2-input Nand Gate 50 60 870 Sylvania SG140 Quad 2-input Nand Gate 170 210 660 Motorola MC301G 5-input Gate 2020 950 4400 Radiation Inc. 709R Operational Amplifier 50 57 206 Motorola MC1539G Operational Amplifier 890 15000 5400 TI 709L Operational Amplifier 1600 11000 8400 Radiation Inc. RD211 Dual Quad-Diode Gate Expander 63 63 _ Radiation Inc. RD220 Hex Inverter 110 430 1080 Radiation Inc. RD221 Dual Binary Gate 850 570 2180 Radiation Inc. RA239 Amplifier — 160 210 Philbrick Q25AH Hybrid Amplifier 630 50 1000 Philbrick Q25AM Hybrid Amplifier 320 6300 3200 Fairchild MA709 Operational Amplifier 35 95 _ between devices. The power levels required for burnout/damage generally are several order of magnitude higher than the levels required for causing degradation in the performance of analog devices. Damage threshold levels for digital devices are only slightly less than that required for analog devices. The extent of computing equipment susceptibility to microwave radiation has been studied by Everett and Everett '*. They irradiated both microprocessors and small computers in a mode- tuned chamber and looked for occurrences of digital circuit upset. A simple program to read into and fetch from RAM memory was executed as the device was being continuously illuminated. With each memory fetch, there was a check to see if the recalled number had been affected by the radiation. A changed number indicated that an upset had occurred. In some cases, the effect was dramatic in exhibiting a wild-running display. The maximum field strength of the test was 200 V/m (rms) corresponding to an incident average power density of 106 W/m?. 15 Irradiation of several unshielded microprocessors and small computers, including the TRS- 80 and the ZX81, indicated that upsets for microwave frequencies in the range of 1 GHz to 10 GHz were rare. With the advent of faster microprocessors and small computers operating with clock rates in the tens of MHz, more upsets are expected for frequencies above 1 GHz. For the KIM-1 microprocessor, the most susceptible in the test, there were upsets observed for field strengths as low as 2 V/m (rms) and incident average power density of 106 pW/cm?. STRESS QUANTITATION It is not clear which attribute of the induced microwave signal is most important in producing stress on electronics. Castillo and Marin '* suggest a set of 6 scalar quantitators to be considered. For a given induced transients, x(t), representing either the induced voltage across a set of terminals or the induced terminal current, they suggest the important quantifiers may be: sup 1. Time -domain peak, |x(t) | 0<t<00 2. Total signal energy, i |x(t) |? dt sup 3. Peak signal power, |x(t) |? 0<t<00 sup 4. Peak time rate of change, Lee x( 0)| 0<t<00 sup 5. Peak time integral of the pulse, f * x(t) dt 0<t<00 Dielectric breakdown and arc-over effects are directly related to the peak induced voltage. For a resistive termination, the voltage is proportional to the current. Hence quantifier #1 is important. For a capacitive termination, the voltage is proportional to the time integral of the current. This indicates that quantifier #5 may be important. For an inductive load, the voltage is proportional to the time derivative of the current, which indicates that quantifier #4 may be important. Heating effects, including burn-out, are directly related to the energy delivered by induced signal. This, of course, is quantifier #2 when the induced current or voltage is used and the load 16 has a resistive component. Note that the quantifiers have been normalized for simplicity and do not include certain constant factor multipliers, such as resistance. Some electronics effects, such as thermal second breakdown, occur as a result of the stress from the peak signal power. This is quantifier #3 when either the induced current or voltage is used and the load has a resistive component. The foregoing list of quantifiers may not be sufficiently complete to characterize microwave effects on electrical systems. Also, the list may not include the most important quantifiers, e.g., a quantifier for rectification effects is not included. Baum 7 addresses this problem by taking a more mathematically rigorous approach. He suggests that norms should be considered for the quantifiers of microwave effects. The use of norms for stress quantifiers offers several benefits. First, it simplifies the comparison of strength and stresses for electronic systems, subsystems, parts, and components. Second, it provides a means of comparing transient stresses from different waveforms, e.g., video pulses, chirp pulses, pulse-modulated sinusoids, damped sinusoids, etc. For a full discussion of the applications of norms to electronic system protection from various electromagnetic environments, see the paper by Baker et al. '*. For a time-domain waveform, the commonly used p-norm is defined as beep = { f° |xce) pach”, 1spse (13) where the oe-norm is defined as sup Ix. = |x(t)| 0<t<00 A comparison of the p-norm with the Castillo and Marin quantifiers reveals that quantifier #1 is the oo-norm and quantifier #2 is the 2-norm. It is also noted that the 1-norm is the time integral of the rectified signal. Consequently, it should provide a quantifier for microwave rectification effects. Although the other quantifiers suggested by Castillo and Marin are not p- norm, they are norms in that they satisfy the appropriate mathematical properties. An advantage of using norms for quantifiers is that they provide opportunities for developing rigorous bounds on transfer functions and hence the interaction process '°. Baum '?”° has developed bounds on scattering matrices, general linear time-invariant system response, black box response, and nonlinear system response. 17 CONCLUSION Environmental concerns from wireless power transmission utilizing microwave may never be completely eliminated. Wire guided power transmission that is currently used provides distinct environmental hazards. EMF, or electromagnetic field, concerns not withstanding, wire- guided power transmission present electrical shock hazards that are a constant threat to electrical workers and occasionally a threat to the public, attracts lightning, generates ozone, and provides a constant source of electrical interference. Now, biological effects from EMF are generating concerns. What is the difference between wire guided power transmission and wireless power transmission? And is wire guided power transmission inherently more safe than wireless power transmission? These questions are not really that difficult to answer. Both modes of power transmission utilize EM fields to deliver power with exclusion regions about the paths of the power delivery required for safety purposes. The greatest difference is the oscillation frequency of the fields. Consequently, wire guided power is not inherently more safe than wireless power transmission. Studies have shown that deleterious effects can occur from relatively low levels of microwave power densities, both biological and electrical effects. Biological effects are difficult to analyze and are not very predictable. Electrical effects are amenable to analysis. However, for large systems, the analysis may not be tractable. The general problem in determining the stress from microwave illumination is ascertaining the attribute of the illumination that is most important in the interaction process. Signal norms have been suggested as a quantifier. However, more work is needed in this area. REFERENCES {1] H.P. Schwan and K.R. Foster, "RF-Field Interactions With Biological Systems: Electrical Properties and Biophysical Mechanisms," Proceedings of the IEEE, Vol. 68, No. 1, pp. 228-113, January 1980. [2] O.P. Gandhi and A. Riazi, "Absorption of Millimeter Waves by Human Beings and Its Biological Implications, "IEEE Trans. Microwave Theory Tech., Vol. MTT-34, No. 2, pp. 228-235, February 1986. [3] S.M. Michaelson, "Microwave Biological Effects: An Overview," Proceedings of the IEEE, Vol. 68, No. 1, pp. 40-49, January 1980. [4] S.Y. Liao, Microwave Devices and Circuits (2nd Ed.), Prentice Hall, Englewood Cliffs, NJ, 1985, Chapter 15. [5] R.K. Park and C.T. Tai, "Receiving Antennas," Antenna Handbook, edited by Y.T. Lo and S.W. Lee, Van Nostrand Reinhold, New York, 1988, Chapter 7. 18 [6] [7] [8] (9] {10} {11} {12} {13] (14] {15} [16] {17] [18] E.A. Wolff, Antenna Analysis, John Wiley & Sons, Inc., New York, 1966, Sections 3.3 and 3.7. R.W. King, Transmission Line Theory, Dover Publications, New York, Chapter 4. A.J. Adams, J.Perini, M. Miyabayashi, D.H.U. Shau, and K. Heidary, "Electromagnetic Field-to-Wire Coupling in the SHF Frequency Range and Beyond," IEEE Transactions on Electromagnetic Compatibility, Vol. EMC-29, No. 2, pp. 126-131, May 1987. J. Gavan, "Main effect of Mutual Interference in radio Communication Systems Using Broad-Band Transmitters," IEEE Trans. on Electromag. Compat., Vol. EMC-28, No. 4, pp. 211-219, November 1986. W.M. Chase, J.W. Rockway, and G.C. Salisbury, "A Method of Detecting Significant Sources of Intermodulation Interference," IEEE Trans. on Electromag. Compat., Vol. EMC-17, No. 2, pp. 47-50, May 1975. G.C. Messenger and M.S. Ash, The Effects of Radiation on Electronic Systems, Van Nonstrand Reinhold Co., New York, 1986, Section 8.9. I.N. Mendel, "DNA EMP Awareness Course Notes (3rd Ed.)," DNA Report no. DNA2772T, Defense Nuclear Agency, Washington, DC, 20305, October 1977. R. Antinone and W.G. Ng., "HPM Testing of Electronic Components," Technical Report UCID-21687, Lawrence Livermore National Laboratory, Livermore, CA. Program Solicitation 90.1 Small Business Innovative Research Program, U.S. Department of Defense, SBIR Program Office, Washington, DC 20301, October 1989. W.W. Everett, III and W.W. Everett, Jr., "Microprocessor Susceptibility to RF Signals- Experimental Results," Proceedings of the 1984 Southeastcon, pp. 512-516, April 1984. J.P. Castillo and L. Marin, "Measurements of System Responses to Transient Excitations," Time-Domain Measurements in Electromagnetics, E.K. Miller, editor, Wan Nostrand Reinhold Co., New York, 1986, Chapter 8. C.E. Baum, "Black Box Bounds," Interaction Note 429, Weapons Laboratory, Kirtland AFB, NM, May 1983. G. Baker, J.P. Castillo, and E.F. Vance, "Potential for a Unified Topological Approach to Electromagnetic Effects Protection," IEEE Trans. on Electromagn. Compat., Vol. 34, No. 3, pp. 267-274, August 1992. 19 [19] CE. Baum, "Norms of Time-Domain Functions and Convolution Operators," Recent Advances in Electromagnetic Theory, edited by H.N. Kritikos and D.L. Jaggard, Springer- Verlag, Berlin, 1990, Chapter 2. [20] CE. Baum, "The Theory of Electromagnetic Interference Control," Modern Radio Science 1990, edited by J.B. Anderson, Oxford University Press, London, 1990, pp. 87-101. [21] HI. Bassen, "RF/Microwave Radiation Measurement, Risk Assessment and Control Activities at the Bureau of Radiological Health," IEEE Antennas and Propagation Society Newsletter, pp. 5-9, April 1981. 20