Bluetooth Challenges Testers
RF-capable ATE and dedicated Bluetooth test systems will team up to gauge product quality.
Rick Nelson Senior Technical Editor -- Test & Measurement World, 9/1/2001 2:00:00 AM
| Bluetooth represents a wild card in the wireless market. The technology has been touted as everything from a printer-cable replacement to a full-fledged wireless local area network (LAN). Whether it will excel in all of these areas is doubtful, but even if it exceeds in only a few, it promises to insinuate itself within a large volume of products, and you’re likely to find yourself testing Bluetooth-enabled DUTs at the chip, board, or system level.
Indeed, the availability of effective test strategies could be the key to whether or not Bluetooth succeeds, as Dan Strassberg of EDN magazine has argued persuasively (Ref. 1). The target price of production quantities of Bluetooth chips is $5 per node—a target that when reached will leave very little cash for test. The per-device test costs of Bluetooth chips or chip sets will measure in the single digits, says one ATE vendor, and that’s pennies, not dollars. Of course, OEMs will expect to charge a premium over that $5 for their Bluetooth-enabled products. But for a capability that, at its most elemental level, replaces a computer printer cable, OEMs, too, will face severe cost pressures and will be unlikely to devote much in-circuit or functional test time to verifying Bluetooth operation. Consequently, vendors of Bluetooth chips as well as Bluetooth-enabled consumer products must adopt test equipment and strategies that accurately measure Bluetooth performance parameters at high speeds and low costs. Bluetooth basics Bluetooth (Ref. 2) is intended to support data transmission at 432.6-kbps rates over a 10-m distance in the 2.4-GHz unlicensed ISM (industrial, scientific, and medical) band. Bluetooth transmissions occupy an 83.5-MHz slice of this band, with each product in turn implementing a frequency-hopping spread-spectrum technique that switches 1600 times/s among 79 carrier frequencies within the 83.5-MHz slice. The technique’s Gaussian frequency-shift-keyed (GFSK) operation generates pseudorandom hop sequences that can operate for nearly 24 hrs without repeating. The pseudorandom nature of the frequency hopping helps to minimize collisions among products competing for available bandwidth. Bluetooth’s symmetrical data rates are 432.6 kbps; asymmetrical rates are 721 kbps in one direction and 57.6 kbps in the other. An optional 20-dB power amplifier would extend the transmission distance to 100 m. In addition to these distance and data-rate specs, the Bluetooth standard defines sets of protocols that products must meet to earn the right to carry the Bluetooth trademark. Standards define the necessary protocols for various classes of Bluetooth products, ranging from those that handle simple unidirectional data downloads between two nodes to ones that mimic peer-to-peer networking in a so-called piconet—a miniature wireless LAN in which as many as seven devices can interoperate simultaneously. Ref. 1 provides more information on piconet operation, and Ref. 3 details the procedures for obtaining Bluetooth qualification for a product. The likely structure of your Bluetooth tests will depend on the application your product serves and on the means you choose to differentiate your product from the competition. You’ll certainly be testing the RF and baseband performance of the Bluetooth interface. But the test parameters and the way in which the functions are distributed throughout a product will determine the exact test process you employ. All Bluetooth products will require RF and baseband tests as well as protocol analysis. Dedicated Bluetooth test instruments are emerging to perform each test type. An article in Test & Measurement Europe (Ref. 4) describes the instruments. Versions available in the Americas include Agilent Technologies’ Model E1852A RF/baseband test set, Anritsu’s MT8850A test set, Tektronix’s BPA100 Bluetooth protocol analyzer, and Yokogawa’s BX1000 protocol analyzer. Protocol analysis Protocol analyzers are critical for determining whether your products will communicate properly with other Bluetooth products in the same class. Protocols define data formats at various stages—or layers—between the raw bit stream of an unmodulated Bluetooth RF carrier and an application that displays the data in a format useful to the user of a Bluetooth product. Together, these layers make up what’s called a protocol stack; the specific layers used will vary from application to application. Ref. 5 provides details on Bluetooth-related protocols; they include • Bluetooth Baseband protocol, which performs such functions as synchronizing the hopping frequencies among communicating Bluetooth products; • LMP (Link Manager Protocol), which establishes baseband packet sizes, provides power control, and handles authentication and encryption; • L2CAP (Logical Link Control and Adaptation Protocol), which interfaces the Baseband protocol with higher-level protocols; • SDP (Service Discovery Protocol), which queries remote products about available services; • RFCOMM (an abbreviation for “RF communications”), a protocol providing wireless emulation of RS-232-like serial communication functions; and • TCS (Telephony Control Protocol), which provides functions for use with mobile phones. Bluetooth also makes use of protocols not specific to Bluetooth itself, including • the Internet’s PPP (Point-to-Point Protocol) and UDP/TCP/IP (User Datagram Protocol/Transmission-Control Protocol/Internet Protocol); • the OBEX (Object Exchange) protocol, which was originally developed for infrared communications; it is a lightweight version of the Hypertext Transfer Protocol; and • vCard and vCal, open specifications for electronic business cards and calendars, respectively. A typical Bluetooth protocol stack might include vCard, OBEX, RFCOMM, L2CAP, LMP, and Baseband protocols. To determine whether your product implements such protocols properly, you can employ an instrument such as Yokogawa’s BX1000, which connects to a host PC through a USB port to provide protocol analysis of a Bluetooth product under test. The BX1000 can serve as a piconet node, monitoring the status and performance of products under test residing elsewhere on the same piconet. It can also generate and receive host-controller-interface (HCI) commands, displaying results on the host PC. The BX1000 handles protocols including L2CAP, RFCOMM, and SDP.
RF test sets Beyond performing the protocol analysis required of any Bluetooth product, a protocol analyzer can also determine whether a product under test’s radio works—after all, if the radio doesn’t, then the analyzer won’t receive any bit streams to analyze. For low-end applications—a Bluetooth-enabled remote-controlled toy, for example—such a basic functional test of a radio might be sufficient. For mission-critical applications, though, you’ll want to augment protocol tests with quantifiable details on RF performance—details that only an RF test set can provide. In fact, RF performance might be your product’s claim to fame. Beyond meeting the basic requirements of Bluetooth qualification, you might target your product at robust operation despite interference from other denizens of the 2.4-GHz ISM band—ranging from IEEE 802.11 wireless LAN systems to microwave ovens—or to ensure that your product doesn’t interfere with other ISM-band inhabitants. To that end, you might have to test for specific levels of parameters such as transmitter power and receiver sensitivity. Specific tests you’ll want to make include transmitter output power, including the average power of bursts as well as peak power and power density (power within a 100-kHz bandwidth). If your Bluetooth product supports battery-saving power-control functions, you’ll need to ensure that transmission power responds properly to power-control commands. In addition, you’ll want to test the output spectrum to ensure that out-of-channel emissions remain within spec. You also will want to make modulation measurements, including measurements of carrier-frequency deviations and drift and of modulation quality, to look for excessive levels of intermodulation distortion. On the receiver side, you’ll want to test receiver operation by measuring bit error rates under various conditions—with an impaired input signal, for example, or in the presence of an interfering signal, which could be an adjacent-channel carrier or the product of intermodulation distortion. Finally, you’ll want to test the entire Bluetooth transceiver for spurious conducted emissions (through a cable connection to a host PC, for instance) and radiated emissions (RFI leakage though the product’s enclosure). You can make such measurements using general-purpose signal generators, time-domain signal analyzers, oscilloscopes, and spectrum analyzers (Ref. 6). To make some of these measurements more conveniently, you now can employ dedicated Bluetooth RF test sets such as Agilent Technologies’ Model E1852A and Anritsu’s MT8850. Such instruments are particularly useful for conveniently exercising a Bluetooth product’s frequency-hopping operation, because the appropriate frequency-hopping signals are difficult to generate and measure using standard instruments.
Testing Bluetooth chips Semiconductor test will also be an important component of a successful Bluetooth product. The nature of that test will depend on how Bluetooth functionality is ultimately implemented in silicon—it will depend on the number of chips required for the Bluetooth implementation. Ideally, full Bluetooth functionality would be implemented within one chip that combines RF and digital baseband functions. Among various radio-transceiver applications, Bluetooth, with its relatively low power, is the most amenable to a single-chip implementation (Ref. 7). Thus far, though, implementations tend to require multiple chips, as vendors find isolating baseband and RF circuitry can optimize price/performance ratios. Texas Instruments, for example, offers a two-chip implementation: the TRF6001 Bluetooth transceiver, which contains the functions shown within the solid blue line of Figure 3, coupled with the BSN6040 Bluetooth baseband processor, which contains the functions contained within the green line of Figure 3.
Traditionally, semiconductor ATE systems have been tailored to test either RF or digital devices. With the convergence of RF and digital functionality on single chips, though, ATE manufacturers are gearing up to provide one-pass processes for testing Bluetooth chips. Advantest, for example, targets its T7610 8-GHz RF test system at a variety of semiconductor devices aimed at wireless applications. The T7610 includes 32 20-MHz digital channels for testing baseband logic circuits integrated on RF chips.
Similarly, Agilent Technologies is targeting its 93000 Series SOC ATE systems (Figure 4 ) at Bluetooth chip tests. In fact, Silicon Wave has chosen the 93000 for design verification and characterization of its Bluetooth devices. The 93000 includes 12 RF ports that handle signals to 8 GHz. Several features make it appropriate for Bluetooth testing: its ability to generate a modulated stimulus to 3 GHz, its frequency-hopping capability, and its bit-error-rate measurement capability. Not all ATE vendors are convinced that RF-plus-digital SOC will dominate the market for Bluetooth chips. Credence Systems, for example, touts its RFx RFIC tester (Figure 5) for testing Bluetooth RF components. Echoing concerns expressed at this year’s IEEE Microwave Theory and Techniques Symposium (Ref. 7), Credence spokesmen contend that it will always be more expensive to implement relatively low-performance digital circuits in RF-ready silicon processes, and they expect a significant market for RF-only test systems for Bluetooth components. In this scenario, RF testers and digital testers would ensure the production of known-good RF and digital dice, respectively, which could be assembled into multichip system-in-package configurations. However it all shakes out, it’s unlikely that one test strategy will fit all applications. T&MW References 1. Strassberg, Dan, “Test may provide salvation for a technology on the cusp,” EDN, June 7, 2001. p. 58.www.e-insite.net/ednmag/contents/images/85634.pdf. 2. You can find an overview of technology plus FAQs and other information at The Official Bluetooth Website, www.bluetooth.com. 3. Bluetooth Qualification Program Website, qualweb.opengroup.org/Template2.cfm. 4. Kerridge, Brian, “Product survey: Checking and verifying Bluetooth traffic,” Test & Measurement Europe, August/September 2001. p. 14. 5. Mettala, Riku, “Bluetooth Protocol Architecture,” the Bluetooth Special Interest Group, 1999. Available for download at www.bluetooth.com/developer/whitepaper/whitepaper.asp. 6. “Performing Bluetooth RF Measurements Today,” Application Note 1333, Agilent Technologies. Available for download, along with other Bluetooth-related papers, at www.get.agilent.com/bluetooth. 7. Nelson, Rick, “International Microwave Symposium Commentary: What is a single-chip radio and is it a good thing?” www.tmworld.com/news/2001_0525_microwave.htm. 8. Woo, Howard, “Bluetooth-IC testing meets chip design,” EDN, May 3, 2001. p. 77. www.e-insite.net/ednmag/contents/images/82757.pdf. 9. Lee, Nelson T.K., and Yogan Senthilkumar, “Essential testing on Bluetooth transceiver ICs,” Test & Measurement Europe, April/May 2001. p. 14. Rick Nelsonreceived a BSEE degree from Penn State University. He has six years experience designing electronic industrial-control systems. A member of the IEEE, he has served as the managing editor of EDN, and he became a senior technical editor at T&MW in 1998. E-mail: rnelson@tmworld.com. |
Talkback
No related content found.
- 0 rated items found.
Datasheets.com Electronic Parts & Inventory Search
185 million searchable parts
- Part Number
- Description
- Inventory
- Products
- Manufacturers
Sponsored Links
View All Webcasts
