Microwave Journal
www.microwavejournal.com/articles/3016-compatibility-issues-between-bluetooth-and-high-power-systems-in-the-ism-bands

Compatibility Issues Between Bluetooth and High Power Systems in the ISM Bands

July 1, 2000

COMPATIBILITY ISSUES BETWEEN BLUETOOTH AND HIGH POWER SYSTEMS IN THE ISM BANDS

Abstract:

Much discussion has taken place over the last several years about the tremendous growth potential for low power microwave chip devices for low power, unlicensed communications within the 2450 MHz band. Techniques for wireless communication consist of frequency hopping spread spectrum (FHSS) and direct sequence spread spectrum (DSSS). The generic name Bluetooth has recently been utilized to encompass all these types of devices even though it strictly applies only to FHSS. While the economic potential for such technology is undisputed, there is reason to believe that the wireless industry is not aware of the problems that are looming on the horizon in the form of interference from consumer and commercial microwave ovens and industrial microwave equipment installations. These types of non-communications equipment are operating legally in the industrial, scientific, and medical (ISM) bands and present spectra that are quite variable, and in many cases have quite high amplitudes. This article outlines the background and problems that exist and enjoins the two industries to work together to resolve the compatibility issues.

CHARLES R. BUFFLER AND PER O. RISMAN

Microwave Research Center

Marlborough, NH

The Federal Communications Commission (FCC) was established by the Communications Act of 1934. Of particular interest were the ISM bands which are contained in the Commissions rules Code of Federal Regulations, Title 27.1 These bands were established to explore and exploit applications of electromagnetic energy throughout the entire electromagnetic spectrum. One of the first major commercial uses of these bands came with the establishment of the industrial microwave heating industry, and began shortly after World War II as an outgrowth of the development of microwave power generating tubes used for radar. These applications took place primarily in the 915 and 2450 MHz ISM bands.

About 20 years ago, a ruling of the FCC allowed the unlicensed use of the ISM frequencies by any potential user as long as the new band inhabitant "accept interference that may be caused by the operation of authorized ISM equipment".2 ISM equipment, of course, included microwave ovens and industrial microwave installations. Taking note of this newly available windfall, the wireless communications community saw an enormous opportunity to proliferate worldwide, low power, low cost chip devices that could communicate over short distances. The ISM bands were the only frequencies where international systems could be built because the ISM frequencies were the same worldwide. Applications were immense, including wireless communications between computers, and computers and accessories, including modem access to the internet. Chips could be placed in credit cards, eliminating the swiping into point-of-sales terminals presently required for inputting information. Also myriad individual as well as multipoint/multisubscriber system applications could be imagined.

Fig. 1 A typical oven magnetron Rieke diagram.

The wireless community either did not realize the problems which might arise with microwave interference, or they chose to ignore it. Recent measurements of some of the available spread spectrum products show appreciable performance degradation when operated in the vicinity of microwave ovens. A number of confirmed interference problems have been reported with wireless local area networks (WLAN) when in the presence of industrial microwave processing installations. The interference has been so devastating as to make the WLANs completely nonfunctional. These situations have resulted in the costly removal of the WLANs and their replacement with hard-wired systems. Costs have run often into many tens of thousands of dollars, both in the US and in Europe. In most cases, the customer and dealer sales force were completely unaware of any potential interference problem. Sadder still, most problem cases seem to have been handled on an individual basis, with minimal information disseminated to colleagues in industry.

What may happen? The FCC regulations of Part 15 devices are abundantly clear; wireless may operate within the ISM bands. The wireless community has orders of magnitude more money, and thus more clout, than the microwave oven and industrial microwave industries combined. In a worst case scenario, the wireless community may pressure the FCC to such an extent that ISM spectrum limiting regulations will be adopted which are severely limiting. Under such regulations, microwave ovens may become considerably more expensive than they are now and industrial microwave installations may be shut down. On the other hand, perhaps the two industries can work cooperatively to effect a solution to reduce the incidence of interference. In order to begin a cooperative effort, communication links must be established so the microwave industry knows the spectral requirements of wireless systems and the wireless community understands the spectral emissions of industrial and consumer microwave equipment.

Unfortunately, no one in either the microwave industries or the communications community seems to care. The microwave industrial group is either oblivious to the threat or is of the mind that if they do nothing, the problem will go away. On the other hand, most microwave manufacturing operations are paralyzed with inactivity. They are caught between two conflicting operations within their own companies which have economic interest in both technologies. Thus, many are unwilling to participate in any discussions.

A TYPICAL EXAMPLE: THE MICROWAVE OVEN

Microwave ovens may legally emit significant levels of leakage in the ISM band, within the limits set by international safety standards. Any communication uses of the band must therefore be based on ad hoc communication industry agreements, since there cannot be any additional restrictions on microwave in-band emission from existing ovens and other industrial processing systems using microwave energy. Hence, it is important for the communication community to have technical data on the actual emission characteristics of ISM band ovens and equipment. Such information might make it possible to choose communication protocols, which minimize the interference from nearby ISM equipment.

The Starting Point: Spectral Emissions From a Simple Microwave Oven with a Large Food Load

The magnetron behavior in a typical microwave oven is normally quite stable if the oven has no stirrer. A stirrer is typically a rotating metal device, normally above a microwave-transparent protecting plate in the oven ceiling, intended to periodically change the balance between the different mode patterns in the cavity. This change provides a more even heating of the load if the load is nonrotating. The magnetron behavior under constant working conditions is well known. When generating more than 30 percent of its full power, it typically oscillates within a frequency band of less than 2 MHz. However, during start-up and shut-down (which occurs during approximately 1.5 milliseconds twice each cycle of the line frequency) it produces a spectrum of about 20 MHz, having 5 percent to 15 percent of the overall intensity of the full power operation spectrum.

If all microwave ovens always behaved as above, both the DSSS and FHSS communication protocols would probably work well with one or two microwave ovens in the vicinity, provided the protocols would utilize the fact that most power supplies used in microwave ovens have half wave rectifiers, resulting in power generation during only about 45 percent of the power line cycle time.

Frequency Fluctuations

A Rieke diagram, shown in Figure 1, shows the output power and frequency of the magnetron as a function of the phase and magnitude of the impedance mismatch. The operating point of the magnetron is typically adjusted for this load to be approximately along the 0.15l radius on the 800 W power contour. With an operating stirrer or a turntable with load, there are serious magnetron frequency fluctuations with fluctuations of the impedance mismatching. There are several causes, one of which is the gradual change of dielectric properties of the load during defrosting and heating. In particular, water without ions, such as in coffee, has a significant absorption change with temperature. About 100 grams of water in a small cup will typically change its absorption properties by a factor of 2 during typical heating for consumption. Also, drastic changes in load characteristics occur during popcorn popping. Figure 2 shows the sharpness of the spectrum from a turntable oven with a nonvarying load. In the same oven with a fixed bag of popping popcorn, the spectrum is widened somewhat, as shown in Figure 3. With a cup of water placed off center on the edge of the turntable, the spectrum has a very broad, time varying spectrum, between 2415 and 2475 MHz, as shown in Figure 4.

Fig. 2 Spectrum from empty turntable oven (courtesy of Ron Lentz, California Tube Laboratory).

Fig. 3 Spectrum during popcorn popping in a turntable oven (courtesy of Ron Lentz, California Tube Laboratory).

Fig. 4 Spectrum with a cup of water offset on the turntable (courtesy of Ron Lentz, California Tube Laboratory).

Another source of fluctuations is the use of a cavity stirrer. These fluctuations are quite rapid and may result in the very wide oscillation spectrum shown in Figure 5, even with a nonvarying load. Figure 6 shows the spectrum is equally broad in the stirrer oven with popcorn popping.

Efforts to Produce More Energy-efficient Ovens

Fig. 5 Spectrum from empty stirrer oven (courtesy of Ron Lentz, California Tube Laboratory).

Fig. 6 Spectrum from stirrer oven with corn popping (courtesy of Ron Lentz, California Tube Laboratory).

Magnetrons with power outputs from approximately 700 to 3000 W are used in 2450 MHz microwave ovens and in many industrial installations with up to hundreds of units per system. About seven years ago, several agencies, including the US Department of Energy, initiated efforts to influence the appliance industry to increase the energy efficiency of their products. Microwave ovens were no exception, and considerable preparations were made toward meeting the new requirements by both magnetron and oven manufacturers. As a consequence, both magnetrons and ovens were designed for operation closer to the sink region, the region on the Rieke diagram where all frequencies meet. This operation often resulted, with small food loads, in almost simultaneous emission in two frequency bands. One band was located around 2440 MHz, the other around 2475 MHz. It is estimated that at least one-third of high-end microwave ovens manufactured today have such spectral characteristics.

Multi-magnetron Ovens and Equipment

As previously mentioned, there are many other uses of 2450 MHz oven-type magnetrons than in domestic microwave ovens. Many foodservice (commercial) ovens have two or four magnetrons, which are then connected to their power supplies in counterphase to the mains so that only half of them oscillate simultaneously. These ovens are used with the same small food items as domestic ovens, and thus give the same type of spectrum emission. In addition, stirrers are more common in these ovens, which results in increased frequency fluctuations.

Multi-magnetron industrial installations, mainly in the food, ceramics and rubber industries and in foundries, may be designed for each individual magnetron to operate with better stability than in a domestic microwave oven; but the fact that there may be tens or even hundreds of simultaneously operating magnetrons in a single small building indeed creates an impossible situation for any communication system trying to operate in the same band. A more comprehensive discussion of microwave oven issues was previously published.3

A SECOND EXAMPLE: THE CORDLESS PHONE

As a practical example of the interference effect of microwave ovens upon a wireless device, a consumer-purchased 2450 MHz cordless portable phone was tested.4 The base unit sends a spread spectrum in the 904 to 925 MHz band and receives at 2406 to 2478 MHz. The microwave oven used was a Whirlpool model VIP27 with a fixed microwave feed and a glass turntable. Microwave leakage was 0.06 mW/cm2, less than 3 percent of that allowed.

Location of the Units

The microwave oven was operated in a kitchen on the first floor of a brick building. The base unit of the cordless phone was placed in two different locations for two series of experiments. The first location was on the second floor, approximately 3 m above and 6 m horizontally away from the oven. The second location was in a second building similar to the first one, located 25 m away from the first. The distance between the oven and the base unit was about 30 m. The mobile handset was used on the first floor and outdoors, at distances up to 25 m from the oven and base unit. In both locations, the handset was in a direction facing the oven front.

Oven Loading Conditions

A 1 liter fixed load is used for international standards measurements for out-of-band emissions. This type of load is not representative of actual oven operation and utilization of this load will not produce realistic results.

Thus, the test oven was operated with three different loads: a round, 1 liter water load located at the center of the turntable, approximately 150 milliliters of hot or boiling water in a ceramic mug and a bag of microwave popcorn. In addition, an insulated metal wire trapped between the door and oven frame was used with 150 milliliters of water to obtain a higher microwave leakage.

Results

The disturbances (audible noise) in the cordless unit caused by the microwave oven was noted in four subjective levels: from no noise to noise which makes communication impossible. The results have been published in detail.4 In summary, the following conclusions can be drawn from the results: The one liter load does not produce a strong noise. The directions in the user manual of the cordless phone are then reasonably correct ("If you are near a microwave oven which is being used, noise may be heard at the receiver. Move away from it and closer to the base unit.")

When there is a cup of boiling water in the oven, it may be difficult to use a handheld unit which has the same distance from the oven and base unit. Thus it may be difficult to use the unit at a location where a microwave oven is operating.

The noise problem is reduced if the oven and base unit are not close to each other and the handset is between them and closer to the base unit than to the oven.

If the oven leaks more than 1 mW/cm2 the situation becomes more severe, and communication may typically be difficult when the hand-held unit is closer to the oven than to the base station. (In actuality, an oven is allowed to leak up to 5 mW/cm2).

During the final stage of popcorn popping, a microwave oven 50 or more meters away from the handheld unit may make communication impossible on several channels. It may then be too inconvenient to try to find a new channel, and noise will also appear in a previously good channel due to the frequency variability of the oven.

The normal use and channel switching of the cordless phone system investigated here clearly indicates that still more severe problems will occur near dual magnetron food service ovens and ovens with a field stirrer.

The cordless field study is not intended to present conclusive scientific evidence of microwave vs. wireless communications problems. However, it does point to the need for further investigations before the general public is sold on a technology that may be unable to meet at least its implied warranty of serviceability.

The communications industry tends to minimize the potential interference implications stating simply that consumers may elect not to use their microwave ovens while utilizing wireless devices. While this may be true in a single household, interference at distances of 50 m will likely involve more than one household, a situation over which a consumer will have no control.

Also to be considered are the business applications. New wireless telephone systems using a dozen or more wireless handsets from one single or multiline transmitter are being sold to small businesses around the world. The International Microwave Power Institute's (IMPI) industry survey indicates that more than 70 percent of all businesses have microwave ovens for employee use. Thus, the possibility of interference to adjacent office suites is strong. In addition, it may be impossible to locate homes and offices near commercial restaurants and fast-food outlets where multiple microwave ovens are in almost continuous use.

Another point to consider in the preliminary field study is that the test oven is a current, state-of-the-art microwave oven engineered to today's high standards. However, the marketplace is populated with more than 200 million ovens, most of which are older and more likely to emit interference at levels higher than the field study. If the conditions reported here are the best to be expected, what might be the worst?

CONCLUSION

The allocated ISM band spectral emission characteristics of modern microwave ovens is far more complex in real life than that which is measured with simple water loads according to the standardized methods for immunity and emission testing outside the band. As has been shown, both the DSSS and FHSS protocols will suffer significantly from the emission characteristics of a majority of the world's microwave oven population if low power (1 to 10 mW transmitter power) is used, in configurations where the microwave oven is up to 10 times farther away from the communication units than their own distance. The situation will remain unfavorable for higher-powered communication units or systems if the microwave oven is closer than tens of meters to the nearest unit.

When communication equipment is operated within the ISM band, there will be phenomena, which have not yet been sufficiently analyzed by its developers. It is necessary to study these phenomena to understand and assess the compatibility problems and take measures in terms of systems design, specifications, user instructions and warnings, so that undue harm to their reputation and prospective business does not occur.

The electromagnetic environment in which these communication devices are intended to be used, cannot be changed for many years. With hundreds of millions of microwave ovens, as well as industrial installations already in use all over the world, any kind of national or global regulations can only be implemented for new and yet undeveloped microwave heating appliances and equipment. Knowledge and exchange of information between the ISM and communication communities can only be beneficial. The need for more wireless communication cannot be questioned, but the consumer who has or wants both a microwave oven or other devices and wireless communication equipment, wants all these products to work. Therefore, the ISM and communications communities must work together to find some practical solution to the compatibility problems.

References

1. Code of Federal Regulations Title 47, Part 18.

2. ibid., Part 15.5.

3. Per O. Risman and Ronald R. Lents, "Compatibility Issues Between Microwave Ovens and LAN Operations in the ISM Bands," Microwave World, 20(3):15.

4. Per O. Risman, "Consumer-Oriented Investigation of the Compatibility Between a Cordless 2400 MHz Phone and a Household Microwave Oven," Microwave World, 21(1):29.

Dr. Charles R. Buffler, vice president of the Microwave Research Center, is a well-known expert and author in the application of microwave technology. His specialization includes microwave physics, microwave oven design, industrial microwave processing, microwave food formulation and electromagnetic properties measurement. He has written extensively on these subjects. He is presently president of the International Microwave Power Institute.

Per O. Risman is the owner of Microtrans AB in Landvetter, Sweden. Through Microtrans, Mr. Risman devotes a substantial portion of his time to The Rubbright Group's Microwave Research Center and its developments in industrial and institutional microwave heating and drying. He also has an extensive background in consumer microwave oven design and development, as well as in IEC regulatory matters. He currently specializes in the design, development and testing of microwave heating devices with particular emphasis on the use of electromagnetic modeling. He also serves as secretary of the IMPI Board of Governors.