Archive for July, 2010
Too funny: Spencer Webb, antenna designer and president of AntennaSys, blogs about his own tests for the iPhone 4 and the challenges of getting quantitative results:
“What about the Consumer Reports “Duct Tape” fix? Yep, it will help. Any insulator over the “gap” area of the antenna is going to help in direct proportion to its thickness. I think Consumer Reports was going for style points in the selection of Duct Tape. Nice move – they sure dominated the news cycle. But, hey, Consumer Reports guys: you don’t do radiated tests in a shield room. That’s like measuring the light output of a desk lamp in a house of mirrors.”
To measure the effect of holding the iPhone different ways Spencer crafted a unique gripping device:
The GripOfDeathInator is built from one-inch thick foam sheet. It’s glued together with a foam-friendly adhesive from Liquid Nails; thanks to the smart Home Depot guy for that one. The foam is pretty much RF-invisible, but it is strong enough to securely clamp our test devices. I’d love to tell you all the details of the extensive mechanical engineering that went into the design, but really, we winged it. It works as good as it looks. We also found all-plastic saw-horses to create our work table, and a wooden stool for our bag of salty water…um… Steve.
Check out Spencer’s blog post for the full test results. Basically the bumper works.
Apple, give everyone a bumper. And give me my twenty nine bucks back for the one I bought. Or, just buy me sushi and we’ll call it even.
Guest blog post by Ross Sabolcik, Director of Marketing, Wireless Products, Silicon Labs
In the early days of Silicon Labs, we were developing a device called a Silicon DAA that replaced a bunch of discrete components in PC modems. Ultimately, we replaced about 70 discrete components with a pair of ICs, a couple of capacitors, and three transistors. We thought the Silicon DAA was a really cool technology advancement – simple to use, software reconfigurable to support multiple worldwide standards, and much smaller than any alternative solution. But a funny thing happened when we launched the product. Several really big modem customers balked at the whole concept of integrating everything into one device. “I can’t customize it,” some would say. “I can do it better than you guys,” others would boast. But the comment that really made me stop and think was this: “If you integrate everything and sell it to me and my competitors, how will I differentiate myself?”
At that time, that question was a real head scratcher for me. “Why would someone buy one product over another when the hardware on the board is exactly the same?” Fast forward a few years, and let’s move into the world of GSM transceivers. Silicon Labs started off with a really neat RF synthesizer, and we decided we needed to integrate more stuff on the chip to remain competitive in our largest market – the GSM cell phone radio. So we did. Lots of stuff! Basically everything except the baseband and power amplifier of a GSM cell phone wound up on our chip. We thought we had made a pretty neat chip — high performance, simple to use, software reconfigurable to support multiple standards, and much smaller. Sound familiar? But like Bill Murray in “Ground Hog Day,” a funny thing happened while showing several large customers what we had accomplished. “I can get better performance if I go with a discrete solution” or “I can’t customize the blocking performance by tweaking the LNA current” or “We need to build our own radio in house to differentiate ourselves from the competition.”
Today virtually no one makes discrete DAA solutions, and multi-band standard cell phone chips of unimaginable integration and performance (compared to a few years back) are the norm. Everyone has access to the same radio solutions so how do you differentiate? In both cases I think the guys who survived and flourished in the market looked at these new solutions as opportunities to focus more of their resources on what their end customers cared about. The engineering time formerly spent on tweaking a radio design could be spent on passing compliance testing sooner to speed up product development, or crafting clever antenna structures that were smaller, or better unit performance testing to ensure that the overall product performed better.
This was the secret I learned about integration and differentiation: Use the integration to spend more time on what makes your overall solution better and more differentiated. It might be the software, the user interface, interoperability or customer support. It might be that killer antenna or clever modulation scheme that doubles the battery life. It might be a killer PA and low-noise amplifier design to put in front of an off-the-shelf radio to get the high performance you need. Whatever the secret sauce may be, you have to know that this is where you should be spending your time and focusing your integration resources.
In this tradeoff between integration and differentiation, you must achieve a balance to be sure. Some markets are so large and (mostly) homogenous that numerous suppliers are going to develop killer turnkey solutions. Other markets are so specialized that the engineering team has no choice but to build it themselves as no one has (or likely ever will) built it for them.
To survive in today’s market, it’s crucial to recognize where along that integration/differentiation continuum the product lies. Determining where to cut and partition the system to maximize the solution effectiveness is the key for anyone in the radio business, whether you’re an RF IC, module or end product provider. Choose wisely and suddenly you’re spending your time on the things that really matter to your customers. Make the wrong choice and you’re in a quagmire of activity that isn’t really adding value to your product, your company and ultimately your end customers.
Sometimes as RF engineers, we can get caught up in the idea that the black magic that is RF breaks this rule of balancing integration and differentiation. “RF is different,” we tell ourselves. “I can build a better radio because I understand my application better.” Sometimes it’s true, and getting that extra 100 feet of range is the difference between a killer product and an elegantly designed flop. Sometimes we spend so much time on those issues that we miss the bigger picture of what our customers are really going to pay us for.
This week the Bluetooth SIG formally adopted Bluetooth Core Specification Version 4.0, which adds Bluetooth Low-Energy (LE) to what’s now being called Classic Bluetooth or Basic Rate (BR), namely what you have in your Bluetooth headset. Designed to work from a coin cell for up to a year, Bluetooth LE is targeting healthcare, sports and fitness, security and home entertainment applications.
Not Your Dad’s Bluetooth
The Basic Rate system includes optional Enhanced Data Rate (EDR) Alternate Media Access Control (MAC) and Physical (PHY) layer extensions. Classic Bluetooth offers synchronous and asynchronous connections with data rates of 721.2 kbps for Basic Rate, 2.1 Mbps for Enhanced Data Rate and high speed operation up to 24 Mbps with 802.11 alternate MAC/PHY (AMP) controllers.
The LE system includes features designed to enable products that require lower current consumption, lower complexity and lower cost than BR/EDR. The LE system is designed for use cases and applications with lower data rates and shorter duty cycles.
Bluetooth LE chips will feature a lightweight Link Layer providing ultra-low power idle mode operation; simple device discovery; and reliable point-to-multipoint data transfer with advanced power saving and secure encrypted connections. Bluetooth LE technology supports very short data packets (8 octet minimum up to 27 octets maximum) that are transferred at 1 Mbps. LE packets are far shorter than Classic Bluetooth ones, which include a 68-72 bit access code, a 54-bit header and a payload of up to 2745 bits—for a maximum packet size of 2971 bits. In addition all connections use advanced sniff-subrating to achieve ultra low duty cycles.
It Pays to Advertise
Bluetooth’s handshaking scheme, while secure, isn’t exactly dynamic. Setting up a trusted link is a time-consuming process, which doesn’t lend itself to ad hoc mobile networks, where nodes can join or drop out as they move into or out of range.
Like the BR/EDR radio, the LE radio operates in the unlicensed 2.4 GHz ISM band. According to the 4.0 spec, LE employs two multiple access schemes: Frequency division multiple access (FDMA) and time division multiple access (TDMA). Forty physical channels, separated by 2 MHz, are used in the FDMA scheme. Three are used as advertising channels and 37 are used as data channels. The physical channel is sub-divided into time units known as events. Data is transmitted between LE devices in packets that are positioned in these events. There are two types of events: Advertising and Connection events.
Devices that need to form a connection to another device listen for connectable advertising packets. The advertising event is ended and connection events begin if the advertiser receives and accepts the request for a connection be initiated. Once a connection is established, the initiator becomes the master (host) device in the piconet and the advertising device becomes the slave device.
The slave device then synchronizes with the host’s clock, mirrors its frequency hopping algorithm and goes to sleep, waking up only periodically as dictated by the application to pass data to the host. Nodes in low-power sensor networks might only go into active mode for a few microseconds every several seconds to pass along a short data string, leading to long battery life. This is a far simpler scheme with less overhead than that used in classic Bluetooth.
Bluetooth LE ≠ ZigBee
At first glance it would seem that Bluetooth LE would be going after low-power ZigBee applications. There would be some irony in that, since the ZigBee camp broke off from Bluetooth several years ago in order to develop a smaller, low-power stack for energy-efficient applications.
But Bluetooth is a very different animal than ZigBee and suited to a different range of applications. ZigBee was designed for mesh networking, using asynchronous communication requiring dedicated routers that must always be powered up since nodes can wake up at any time. It’s basically a low-power wireless LAN protocol with flexible routing to deal with nodes that may not respond. ZigBee is suited to fixed-location networks, with always-on routers to track the network status.
Bluetooth LE was designed to accommodate portable devices in a star network. Nodes can dynamically connect with and drop off the network as they move into or out of range. Bluetooth LE uses a synchronous protocol, allowing both master and slave to wake up simultaneously. Since both can be powered down most of the time, this keeps energy consumption low.
Bluetooth LE is suited to dynamic mobile networks where energy efficiency is paramount. It’s also a leading candidate for ultra-low power wireless sensor networks, such as those used to monitor vibrations on aircraft wings and industrial motors as well as cracks in bridges and buildings. According to analysts at Research and Markets, “Bluetooth Low Energy will be a significant contributor to the overall Wireless Sensor Network market, representing nearly half of all shipments in 2015.”
Despite its early designs on medical and other ultra-low power applications, ZigBee seems to have homed in on the one application where it’s found commercial success, namely smart metering. Expect it to expand later out the power line to other ‘smart grid’ applications. Meanwhile Bluetooth LE could have a great run in low-power medical applications, where energy efficiency is paramount and cost is apparently no object.
Despite the lack of apparent overlap with ZigBee, the Bluetooth SIG is also promoting LE applications in home automation and smart energy, so in these areas the horse race may just be getting started.
Are We Really Compatible?
With any technology advance there’s always the issue of backward compatibility. Here the answer is “yes, but.” According to the Bluetooth SIG:
“This enhancement to the Bluetooth Core Specification allows two types of implementation, dual-mode and single-mode. In a dual-mode implementation, Bluetooth low energy functionality is integrated into an existing Classic Bluetooth controller. The resulting architecture shares much of Classic Bluetooth technology’s existing radio and functionality resulting in a minimal cost increase compared to Classic Bluetooth technology. Additionally, manufacturers can use current Classic Bluetooth technology (Bluetooth V2.1 + EDR or Bluetooth V3.0 + HS) chips with the new low energy stack, enhancing the development of Classic Bluetooth enabled devices with new capabilities.”
In short, your current Bluetooth phone can’t communicate with LE devices but future ones will be able to since they’ll include Bluetooth BR and LE (dual mode) in the same chip. When a Bluetooth LE peripheral wants to come onto a BR network, the host will downshift to LE mode and be able to communicate with it. This compatibility will have to wait for the next generation of consumer devices that incorporate dual mode Bluetooth chips.
First Silicon is Already Here
On the same day that the Bluetooth SIG released the new spec, Texas Instruments announced it had “achieved complete Bluetooth v4.0 controller qualification on the CC2540 low-power, single-mode system-on-chip (SoC) running both protocol stack and application software. TI has now certified the full set of components, including ATT, GATT, GAP and SMP that encompass Bluetooth protocol implementation.”
In addition to TI, Broadcom, CSR, Atheros, Wicentric and Nordic Semiconductor have all qualified 4.0 products (silicon) since qualification opened on Tuesday. With more than 13,000 members of the Bluetooth SIG, you can be sure that plenty of Bluetooth LE chips are in the pipeline, with products starting to appear early next year.
“The finalization of Bluetooth low energy wireless technology within the Core Specification is a monumental achievement,” said Michael Foley, Ph.D., executive director, Bluetooth SIG. Amen to that. The SIG and its members deserve a lot of credit for a disciplined development process that didn’t give rise to a lot of “pre-LE” products as happened with 802.11n, where “pre-n” silicon appeared months before the spec was finalized. With a stable spec and the SIG’s rigorous certification process, Bluetooth LE should see a fast ramp, leveraging the huge base of Bluetooth applications.
In low-power sensor networks, medical, security and home automation applications, Bluetooth LE looks set to be a major player in the low-power wireless world for years to come.
Apple has been taking a lot of flack recently for reception problems with the iPhone 4—much of it for denying a problem that was immediately obvious to iPhone users. Steve Jobs’ reaction to being told about dropping calls when you hold the phone a certain way was reportedly, “Just avoid holding it in that way.”
“Doctor, my arm hurts when I move it like this.” “Well, don’t move it like that.”
Yesterday Apple finally acknowledged the problem and attributed it to a longstanding software error:
“We have discovered the cause of this dramatic drop in bars, and it is both simple and surprising. Upon investigation, we were stunned to find that the formula we use to calculate how many bars of signal strength to display is totally wrong… Their big drop in bars is because their high bars were never real in the first place. To fix this, we are adopting AT&T’s recently recommended formula for calculating how many bars to display for a given signal strength… We will issue a free software update within a few weeks that incorporates the corrected formula.”
That explanation is complete b.s., and for reasons we’ll explain in a moment, Apple apparently knows it. Judging from the angry reaction in the blogosphere, tech savvy iPhone users who’ve experienced the problem aren’t buying it either.
Hire More Engineers, Not PR People!
The problem is the result of external antennas that wrap around the case—one for Bluetooth/Wi-Fi/GPS that goes down the left-hand side of the case and another for UMTS/GSM that wraps around the rest. The stainless steel antenna band lacks any insulation, which makes touching it a very bad idea.
The dramatic drop is signal levels occurs when you cup the phone in your hand in what’s now referred to as the Death Grip. Note in the photo above that the user’s hand is touching the lower left side of the iPhone, which where the two antennas terminate. Antennas are conductive devices, and so are people, especially people with sweaty palms. When you wrap your hand around the phone you connect both antennas, completely detuning them and providing a very effective ground path, your being essentially—with all due respect—a big bag of water.
It’s a miracle that any signal gets through, but some still does. Both Gizmodo, who inadvertently got a look at the first iPhone 4, and Anandtech have done excellent analyses of the iPhone’s reception problem, and there’s no doubt it’s the result of an uninsulated external antenna. Their results (table) show clearly that reception varies directly and dramatically depending on how you hold the phone. The higher penalty you pay compared to the iPhone 3GS, with its internal antenna, is equally dramatic.
Anandtech’s review is detailed, entertaining and highly recommended reading for anyone interested in antenna design—or who just enjoys a good detective story.
Meanwhile, while denying that it has an antenna problem, Apple has posted a job listing for “Antenna Engineer – iPad/iPhone”. Welcome to cognitive dissonance.
Until Apple can hire that engineer to fix their antenna problem, there’s a cheaper solution than the $29 rubber case they’re pushing as the solution: Livestrong wristbands that sell for $1 apiece.
Livestrong bands are available from Livestrong.org, a nonprofit foundation started by Lance Armstrong in 1997 to help people fight cancer. Proceeds from the sale of the yellow bands are used to fund its cancer survivor and support programs.
Wrap a snappy yellow Livestrong band around your iPhone 4 and you simultaneously solve your reception problem, make a fashion statement and help fight cancer. It doesn’t get better than that.