evalKits.com Blog

Bidirectional Optocoupler Isolates Communication Signals

2011-01-20T08:01:14Z

Fairchild Semiconductor has a new logic-gate optocoupler IC, the FOD8012, that offers a pair of isolated channels that work with 3.3- and 5-V CMOS devices. Although the announcement and the datasheet tout the IC as "bidirectional," they mean you can set up one channel for transmit and the other for receive. Neither of the channels is truly bidirectional. But that does not diminish the value of the FOD8012. You can use it in I2C, SPI, SMBUS, and other digital communication circuits when you need electrical isolation of several thousand volts.

If you wish, you can use both optocoupler elements in the small-outline SMT package for unidirectional signals and to perform logic-level "translations." Fairchild specifies the following timing information for this device: a 15 Mbit/sec data rate (non return to zero data), and a 60 nsec maxium propagation delay. Find a complete data sheet at: www.fairchildsemi.com/ds/FO/FOD8012.pdf.

I recently created a logic-level "converter" circuit that used an SN74LVC4245 octal bus transceiver, but I needed only two or three level conversions. It didn't occur to me to try an optocoupler as a way to convert 3.3-V signals to 5-V logic signals and vice versa. I'll ask Fairchild for a few samples to see how they work in this type of application. Price for 1000 ICs: $US 3.08 each. Fairchild quotes delivery of ICs in 8 to 10 weeks, but the company has samples available now. --Jon Titus

Does Everything Need a Wireless Link?

2011-01-06T10:01:13Z

During the break between Christmas and New Year's day, I started to experiment with several XBee IEEE 802.15.4 transceiver modules from Digi International. These modules can operate in a network or provide basic point-to-point communications.

As I worked with the wireless modules I thought that this approach to wireless communications, where engineers and designers get a ready-to-go transceiver, makes a lot of sense. Products can now take advantage of a variety of wireless devices, from individual chips through easy-to-use modules that have FCC approval.

On the other hand, the ease of using wireless might make for a crowded spectrum in the license-free Instrumentation, Scientific, and Medical frequency bands. So, do we really need more wireless devices? Years ago I attended a presentation at the MIT Media Lab by then-head Nicholas Negroponte who talked about types of electronic communications.

Negroponte made the point that the form of communication should fit the need. Thus, if you have two fixed locations and can run a wire, then that's the way to communicate. If you have two fixed locations, but one exists in a dangerous area, across a wide canyon, and so on, then wireless might provide the best option.

Negroponte also stressed that wireless communications don't always need radio signals. In many cases, infrared will work just as well. I have seen announcements recently about RF remote controls for home-entertainment equipment. That strikes me as odd, because if you're watching a TV program obviously you have a line-of-sight path to the TV--and so would an infra-red (IR) hand-held control. So, except in unusual situations, I question the need for RF controls for home-entertainment equipment. So even though companies have made RF devices easy to integrate into products, that doesn't mean every product needs a RF link. --Jon Titus

Hey Kids! Enter the "Student LED Design Challenge"

2010-12-20T03:12:13Z

Attention, middle-school and high-school students and teachers. Electronic Engineering Times (EE Times) has started an "Innovation Generation" competition that will involve teams of students that designs interesting projects that use light-emitting diodes (LEDs). During the first phase, teachers will submit a short essay of 500 (or fewer) words that describe how they would use the "LED Challenge" as a teaching tool. The top 50 submissions will each receive a grant that includes $100 worth of LEDs and components, and a $100 educational grant to attend an online seminar that describes the kit and how to use it in a classroom. The first phase ends on January 15th, 2011, so get your essay on its way ASAP.

Each kit includes individual LEDs, 7-segment numeric displays, bar-graph LEDs, photocells, buzzers, resistors, capacitors, switches, pushbuttons, transistors, and a plug-in power supply.

You can find a grant application and more information such as Challenge Rules (legal stuff), Challenge Guidelines, and Project Submission Guidelines at: igen.eetimes.com/student-led-design-challenge. You'll also find a tutorial about LEDs that explains how to control them, how to use transistors to turn LEDs on or off, and so on. A short video shows how to control many LEDs with only a few lines, and a link to the "What's a Microcontroller?" web site points you to more information if your team wants to use a microcontroller in its project. Microcontroller use is optional.

Teachers who receive a grant and a parts kit will form a student team to work on an LED project, due on March 15th, 2011. That's the second phase of the competition--your team builds something interesting or useful with the components in the kit of parts. An LED team project might include a game, an aid for people with disabilities, a colorful and arty display, a score keeper for school teams, an alarm circuit, emergency lighting, lighting effects of model railroads or radio-controlled airplanes, and so on. Teams can use their imagination. (You don't submit the completed project, just a description and photos that show how it works.)

Awards for Phase-2 projects include a grand prize that takes the winning team's teacher to the San Jose, CA 2011 "Embedded Systems Conference" and honors the teacher and the project at the Ace Awards Ceremony (Tuesday, May 3, 2011).

The second-place team and honorable-mention teams will receive awards and grants for science, technology, engineering, and math equipment and materials.

If you know a student, middle-school or high-school teacher, let them know about this exciting kids-only competition. Have fun! --Jon Titus

STMicro Extends ARM MCU Family

2010-12-17T08:12:27Z

ST Microelectronics has extended its line of STM32 MCUs based on the ARM Cortex-M3 core and announced new ARM Cortex-M0 and the Cortex-M4 MCUs for mid-2011.

New STM32 F-2 Series

STMicro now has 30 new members of its popular STM32 MCU family all based on the ARM Cortex-M3 processor core. The new devices, in what ST calls the F-2 series, complement those in the earlier STM32 F-1 and L-1 series. The F-2 devices offer pin-to-pin and software compatibility with the other family members and the four variations in the new F-2 series let engineers choose memory size, package type, and peripheral options. You can find more technical details, specifications, and a block diagram of the STM32F217--the most-capable member of the family--in the press release at: www.st.com/internet/com/press_release/p3029.jsp.

The ST Adaptive Real-Time Accelerator, or ART Accelerator, operates as a prefetch cache in the F-2-series MCUs and provides 128 bits simultaneously to the Cortex-M3 core. The combination of the accelerator and a cache of the 64 most-recently-used branches provides a 0-wait state execution from flash memory at a CPU clock frequency as high as 120 MHz. Unfortunately, STMicro hasn't provided details about how the ART Accelerator works and how engineers can use it to improve code performance. Perhaps it all happens without any programming intervention. I have asked STMicro for more info about the ART Accelerator, but have yet to receive a reply.

A Cortex-M4 and -M0 roadmap

STMicro expects to have MCUs based on the Cortex-M4 and -M0 processor cores available in mid 2011. The Cortex-M4 (ARMv7-ME architecture) includes digital-signal processing (DSP) instructions and a floating-point [math] unit (FPU) and it provides a code-compatible path upward from the company's MCUs that use the Cortex-M3 core. For a lower-cost ARM-based MCU, look at the Cortex-M0 MCU (ARMv6-M architecture)that gives you the flexibility to move upward to a Cortex-M3 or -M4 device if necessary.

For more information about the ARMv7-ME and ARMv6-M architectures, download and read the "ARMv6-M Architecture Reference Manual" available from ARM, after site registration. www.arm.com.

If you need an ARM "score card," find more information about the various ARM families at: en.wikipedia.org/wiki/ARM_architecture. --Jon Titus

Silicon Labs ICs Bridge USB-to-Serial Port Gap

2010-12-08T09:12:40Z

Engineers who plan to include a touch screen in a human-interface device (HID) might lack the expertise needed to create a USB port for an HID-class, as defined by the USB Implementers Forum (USB-IF). This group has developed USB "classes" for many device types, from printers and digital cameras to test-and-measurement instruments and HID equipment. Each class provides specific information about how to communicate with a class device.

Silicon Laboratories has a new device in its CP21xx USB-bridge family that gives engineers and product designers a way to quickly add a USB connection to an HID-type product. Because the CP21xx ICs require no USB expertise, engineers can focus on their hardware and software rather than try to master USB code, stacks, and ports. Engineers who already have such a UART-based product can use the bridge chip to create a USB port that connects to the internal universal asynchronous receiver/transmitter (UART). For more information, data sheets, and to order chips, samples, or evaluation kits, visit: www.silabs.com/pr/USBbridge. You also can download a copy of the 3-page white paper, "USB Simplified: Adding USB Connectivity to Applications with Legacy Serial Connections," after you supply your email address.

The CP21xx USB-bridge family includes four new devices: the USB-to-UART bridge (CP2104), the USB-to-dual-UART bridge (CP2105), the HID-USB to UART bridge (CP2110), the HID-USB to SMBus and I2C bridge (CP2112)

According to Silicon Labs, the new CP2110 and CP2112 bridge ICs comply with the USB-HID class specification natively supported by most operating systems, thus removing the need to install drivers. This compliance eliminates concerns about incompatible and driver updates. Customers no longer need companion CDs and companies no longer bear the cost to furnish driver updates.

The CP2110, for example, provides as many as 10 general-purpose I/O (GPIO) lines if you do not require the UART RTS and CTS flow-control signals. The chip also includes an RS485 interface shared with four of the GPIO lines. Windows applications communicate with the CP2110 through a DLL provided by Silicon Labs. The company also supplies the interface specification for the CP2110 to let engineers develop an application programming interface (API) for any operating system that supports the HID-USB class.

Evaluation kits ($US 29.00 to $US 39.00 each) let engineers try the four bridge devices:

CP2104EK = a simple USB-to-UART bridge.
CP2105EK = USB-to-dual-UART bridge.
CP2110EK = evaluation and customization of an HID-USB-to-UART bridge.
CP2112EK = HID-USB-to-SMBus/I2C bridge connections (see image below).

The CP21xx USB-bridge ICs are available now in a 4-by-4 mm 24QFN package. Price: $US 1.06 (10K). --Jon Titus

Beagle Tracks and Analyzes USB 3.0 Traffic

2010-11-29T10:11:38Z

I have worked with serial communications from slow 110 bits/second current loops up to CAN bus communications and thus appreciate how much time a protocol analyzer can save. Working at even higher speeds must tax engineers who rely solely on a scope or logic analyzer without protocol-analysis tools.

Anyone who develops a product that will use a USB 3.0 connection might need to monitor, track, and analyze USB communications to ensure they comply with the latest USB standards. Recently, Total Phase introduced the Beagle USB 5000 SuperSpeed Protocol Analyzer (TP320910) that provides real-time interactive capture and analysis of USB 3.0 and USB 2.0 (as high as 5 Gbits/second) bus operations. The company's free Data Center protocol-analysis software runs under the Windows, Linux, and Mac OS X operating systems. (People often refer to the USB 3.0 specifications as "SuperSpeed USB.")

The software lets product designers and engineers observe class-level decoding that converts protocol-level information (basic bytes) into more useful USB class-level commands and instructions such as Request Sense, Read Capacity, Data Transport, and so on. Individual device classes cover products such as printers, audio devices, cameras, mass-storage equipment, human-interface devices, and others. So, if you have a new USB 3.0-based communication device, for example, you can view the commands and traffic in a format that make sense to you and your device. The Beagle USB 5000 offers both a protocol-level and a class-level view of USB information.

Although the Beagle 5000 specifications note a memory capacity of 2 Gbytes, the instrument streams data to a host PC, so it can capture much more bus information. Because USB traffic generally occurs in bursts, the host PC can "catch up" to data stored locally in the instrument. If you will use other analysis instruments, you can use the Beagle 5000 triggers and digital I/O lines to synchronize data capture with specific bus transactions or external-device trigger signals.

Learn more about the Beagle USB 5000 at: www.totalphase.com/products/beagle_usb5000. A standard analyzer will cost $US 5000 and customers can purchase two upgrade options: Option A for simultaneous USB 3.0/2.0 monitoring and

advanced triggers (software upgrade only), and Option B, which increases the buffer to 4 Gbytes and includes current/voltage measurements (hardware upgrade, available early 2011).

Find the USB Implementers Forum at: www.usb.org/about. This group establishes the USB standards, runs compliance workshops, and develops USB tests. --Jon Titus

Vinculo Board Expands Arduino I/O Capabilities, Adds USB 2.0

2010-11-19T10:11:12Z

Engineers who like the popular Arduino Duemilanove / Uno microcontroller boards now can take advantage of the wide variety of I/O boards designed and supported by the Arduino open-source community. The Vinculo board from Future Devices Technology International (FTDI) combines hardware and software needed for rapid development of innovative, economical embedded systems. The new board includes a Vinculum II VNC2-64 dual-channel USB Host/Slave controller that can connect to Arduino-compatible I/O application boards, called "shields."

The on-board VNC2 chip combines a 16-bit, 48-MHz CPU, 256 Kbytes of flash memory and 16 Kbytes of SRAM. The chip also gives designers UART, FIFO, PWM, GPIO and SPI Slave/Master interfaces. And the Vinculo module integrates an 8-channel, 10-bit analog-to-digital converter (ADC) and USB A and mini B (host and device) connectors.

The onboard ADC simplifies connection to sensors and the PWM interface can control many external devices such as motors and FETs. Arduino-compatible I/O boards connect to the Vinculum II VNC2 via the GPIO signals. A separate Proto board lets designers to create their own I/O or special-function boards. FTDI makes the schematics and PCB layout of the Vinculo board available free of charge to third-party designers who choose to replicate or enhance the original design.

For a data sheet, visit: www.ftdichip.com/Support/Documents/DataSheets/Modules/DS_Vinculo.pdf.

Designers can create firmware with the Vinculum II software-development tools that provide a C compiler, assembler, linker and debugger tools. FTDI also provides device-class driver libraries that comprise USB Mass Storage, HID (human-interface devices), Printer Class, Image Class, Communication Class, and Instrumentation Class code. Code generated by the Vinculo tool chain and FTDI libraries is royalty free for both commercial and non-commercial use.

The FTDI Vinculo base module (VNCLO-MB1A) costs $US 29.95 and the Proto board kit (VNCLO-SHLD-1A) sells for $US 9.75 for single units. The modules measures 55.48 mm x 68.6 mm.

For Arduino information, visit: www.arduino.cc/en/Main/ArduinoBoardDuemilanove.

Have you used an Arduino processor or I/O board in a project? Share your experiences with us with a comment below. --Jon Titus

Microchip Licenses MIPS M14K Cores for New PIC32 MCUs

2010-11-11T10:11:05Z

You might not know the PIC32 family of microcontrollers (MCUs) from Microchip Technology uses a processor core--the 32-bit M4K--from MIPS Technologies. Why roll your own processor if someone will license one already tested and ready to go?

Microchip has now licensed the MIPS32 M14K and will implement it in new PIC32 MCUs. This core lets programmers reduce the amount of memory their software requires, which in turn lets engineers create circuits with smaller, less-expensive MCU chips. The new MCUs will maintain code compatibility with the earlier PIC32 MCUs that use the M4K core. Microchip has not yet announced a timetable for release of the new PIC32 M14K-based MCUs.

The program-memory saving comes about from the microMIPS code-compression technology that combines optimized 16- and 32-bit instructions in a single, unified Instruction Set Architecture (ISA). The ISA creates a high code "density" by combining variable-length re-encoding of the MIPS instruction set and additional code-size-optimized 16- and 32-bit instructions. The smaller amount of code leads to better cache use and a lower fetch "bandwidth" that helps improve performance and reduce power use.

Microchip's press announcement noted, "Executing the microMIPS ISA results in at least a 30-percent code-size reduction with little or no compromise in performance. Additional features of the M14K cores that will be beneficial in the next generation of PIC32 microcontrollers include interrupt-latency improvements and low power consumption."

For information about the Microchip PIC32 family, visit: www.microchip.com/en_US/family/pic32/index.html.

For more information on the microMIPS architecture, visit: www.mips.com/products/architectures/micromips/.

For more information about the M14K processor architecture, visit: www.mips.com/products/cores/32-64-bit-cores/mips32-m14k/.

For my review of a PIC32 Starter Kit, visit: www.designnews.com/article/509516-Microchip_s_PIC32_Development_Kit_Kick_Starts_Ethernet_Designs.php.

--Jon Titus

Are Some Kits Too Complicated to Document Well?

2010-11-04T09:11:40Z

I started to work with a small, inexpensive demonstration kit that provided an ARM Cortex-M3 microcontroller (MCU). Sadly, I quickly discovered myself in a morass of code. The kit manufacturer supplied sample code and needed .c and .h files for various general-purpose I/O operations, but I found them difficult to understand. The listings included few comments about how the code worked and users like me faced blocks of code similar to:

/** * @brief Initializes the GPIOx peripheral according to the specified * parameters in the GPIO_InitStruct. * @param GPIOx: where x can be (A..G) to select the GPIO peripheral. * @param GPIO_InitStruct: pointer to a GPIO_InitTypeDef structure that * contains the configuration information for the specified GPIO peripheral. * @retval None */ void GPIO_Init(GPIO_TypeDef* GPIOx, GPIO_InitTypeDef* GPIO_InitStruct) { uint32_t currentmode = 0x00, currentpin = 0x00, pinpos = 0x00, pos = 0x00; uint32_t tmpreg = 0x00, pinmask = 0x00; /* Check the parameters */ assert_param(IS_GPIO_ALL_PERIPH(GPIOx)); assert_param(IS_GPIO_MODE(GPIO_InitStruct-&gt;GPIO_Mode)); assert_param(IS_GPIO_PIN(GPIO_InitStruct-&gt;GPIO_Pin));

I posted a call for help on the MCU vendor's Web site and got a suggestion to read a help file that described GPIO (and other) library functions. But the descriptions seemed incomplete and example code duplicated material I had examined earlier.

In this particular MCU, the manufacturer provides I/O ports that are so flexible that they complicate the set up of simple operations. And the lack of good explanations about how to set up and I/O port and do something useful make we wonder if some products are just too complicated to properly document. There are just too many other ARM MCUs out there to bother with one so poorly supported. --Jon Titus

Test Coil-Cell Capacity in Wireless Devices

2010-10-28T10:10:41Z

Texas Instruments has published an interesting new white paper, "Coin cells and peak current draw," that compares battery-energy capacity with current drawn for a wireless circuit, in this case, a Bluetooth device. Although a test model mimics the load of a Bluetooth device, the report notes the model applies to many other wireless devices that use ZigBee, RF4CE, 6LoPAN, and other protocols.

The report concludes that, "...minimizing average current is the key to achieving long battery life with CR2032 [coin cells]." That might seem like common sense, but the author also explains that adding a capacitor in parallel with a CR2032 coin cell provides the most effective choice a designer can make to maximize battery capacity utilization in low-power RF applications. The white paper includes information about the test circuit, the load profile, and includes charts of test results. An appendix explains how to calculate the capacitor value needed across a battery to produce the best use of power. Tests used CR2032 coin cells from six vendors--four from name-brand manufacturers and two from "no-name" vendors.

You can find the 13-page paper (TI document number SWRA 349), written by Mathias Jensen at: focus.ti.com/lit/wp/swra349/swra349.pdf. I found the paper worth reading. --Jon Titus

TI's MSP430 MCU Runs on 0.9 Volts

2010-10-21T12:10:12Z

The Texas Instruments MSP430 microcontroller family now has a 0.9-volt sibling, the MSP430L09x, aimed mainly at high-volume products that can benefit from code in factory-programmed read-only memory (ROM). When you supply external memory, you can economically use an MSP430L092 MCU in applications that require fewer than 100K devices.

The MSP430L092 comes with "loader code" already programmed in its ROM. That code, which performs several functions, lets engineers create applications without the need for a custom-mask ROM. Such an application comprises an MSP430L092 and a serial SPI EEPROM such as an M95512 or a 25AA040. The MSP430L092 can hold as many as 1984 bytes of code in internal RAM, which leaves 64 bytes free for variables. (You could trade fewer bytes of code for more RAM for variables.)

According to TI, "The major use cases for an application with a loader device and external SPI memory for native 0.9-V supply voltage are late development, prototyping, and small series production. Table 8-1 and Figure 8-1 [in the MSP430x09x Family User's Guide] list various debugging scenarios possible for ultra-low supply voltage. A loader approach is the only choice for an autonomous application with MSP430L092, as no nonvolatile memories are available on the market for native ultra-low supply voltages." For the MSP430L09x Family User's Guide, visit: http://focus.ti.com/lit/ug/slau321/slau321.pdf.

Thus you can have a fully functioning MCU with a single AA battery or even a small coin cell. An MSP430L092 consumes as little as 45 ?A when active and it can "awaken" in 5 ?sec. Prices for the new MSP430L092 MCUs start at $US 0.85 (10K units), and TI offers two low-cost development tools.

1. The MSP-TS430L092 ($US 99) offers a target board you can use to program and debug the MSP430 via a JTAG interface or Spy Bi-Wire interface. The development board supports the MSP430L092 device in a 14-pin TSSOP package and a zero insertion-force (ZIF) socket on the target board lets you drop in a device. Because the MSP430L092 MCU requires a low voltage, a second board translates 3-volt power and JTAG signals from an MSP-FET430UIF down to 1.5 volts for the MSP-TS430L092 board. The two-board set provides the means to load code into and debug code on a 0.9-volt MSP430L092 with TI's many software-development tools.

2. If you need an MSP-FET430UIF USB pod, purchase the MSP-FET430U092 ($US 149) kit that includes it and the two boards mentioned above. You can order software tools and hardware directly from TI at: www.ti.com/l092-pr-es.

One of the appeals of the MSP430L092 is what TI calls its analog-function pool, or "A-Pool" that provides a series of functions that programmers can configure as an 8-bit digital-to-analog converter (DAC), an 8-bit multichannel analog-to-digital converter (ADC), supply-voltage supervisor (SVS), and comparator. Input voltage dividers and an internal reference source allow engineers to create a wide range of combined analog functions. For a complete data sheet, visit: http://focus.ti.com/lit/ds/symlink/msp430l092.pdf. --Jon Titus

Jump Into the Renesas MCU Design Competition

2010-10-20T04:10:41Z

Renesas Electronics America has started the RX Design Contest that will give away a top prize of $US 5000, a $US 3000 second-place prize and a $US 1000 third-place prize. Honorable-mention awards of $US 500 each will go to five other designs. Renesas gives embedded-system engineers an opportunity to showcase their skills in any type of design that uses an RX-type MCU.

The new Renesas RX family of 32-bit microcontrollers (MCUs) and a "software-support" environment, provided by Renesas and its Alliance Partner community, which includes CMX Systems, IAR Systems, Micrium, Micron Technology, SEGGER, Rowebots and TotalPhase let engineers and programmers employ a variety of tools. Participating Alliance Partners also will present their own prize awards, based on criteria for their respective prizes. At a special ceremony during the Embedded Systems Conference Silicon Valley 2011, Renesas will announce the contest winners.

The contest, which ends on March 4, 2011, includes 1000 free RX62N Renesas Demonstration Kits (RDKs) available for contestants who register at http://www.renesasrulz.com/rx-contest. Be ready with a 600-character (not word) description of the project you plan to create. Renesas will evaluate project ideas and award free kits accordingly. Although Renesas encourages contest entrants to use the RX62N RDK, you can use your own board and RX62N MCU. Renesas noted it can ship free RX62N RDK kits only to North America and South America addresses. (The RX62N RDK has a retail value of approximately $US 99.)

For information about the RX62N kit, visit: am.renesas.com/products/tools/introductory_evaluation_tools/renesas_demo_kits/yrdkrx62n/yrdkrx62n.jsp. You'll find documentation, software, diagrams, and other information on this Web site.

According to information on the ResesasRulz Web site, the kit includes, a 3-axis digital accelerometer, real-time IEEE-1588 Ethernet PHY, 96 x 64 mono graphic LCD, FOC [field-oriented-control] motor-control simulation, digital microphone, speaker, and Micro-SD card slot. The RX62N-series MCU has on-chip flash memory and enhanced communication functions that include an Ethernet controller, USB 2.0, and CAN bus.

Have fun. --Jon Titus

New Wireless Kits from Texas Instruments

2010-10-07T08:10:52Z

Texas Instruments has three new wireless evaluation kits that give engineers opportunities to check out the 16- and 32-bit Stellaris microcontrollers (MCUs) in RFID, low-power RF, and ZigBee configurations. The Stellaris MCUs provided in these kits give engineers an ARM Cortex-M3 CPU to work with.

The kits extend TI’s Stellaris DK-LM3S9B96 development kit and give engineers complete ready-to-use software and reference designs. For more information, go to www.ti.com/wk-pr-lp and for a video overview of the kits, go to www.ti.com/wk-pr-v. For information about the Stellaris DK-LM3S9B96 kit, visit: http://focus.ti.com/docs/toolsw/folders/print/dk-lm3s9b96.html.

According to TI's recent announcement, when coupled with the DK-LM3S9B96 development board, each kit supplies the hardware and software needed to jumpstart a design, and the "quick-start" application included in each kit lets developers set up and start to evaluate networks in 10 minutes or less.

Here's a brief description of each kit:

DK-EM2-7960R: A 13.56-MHz RFID wireless system that includes a TI TRF7960TB HF RFID Reader Module target board, two ISO/IEC 14443A (MIFARE®-1K) contactless smart cards, additional TRF7960-supported tags/inlays, and Stellaris DK-LM3S9B96-EM2 Expansion Board. Price: $US 99. Visit: www.ti.com/wk-7960r-pr-es

DK-EM2-2500S: A 2.4-GHz SimpliciTI-protocol-based kit aimed at small RF networks and designed for easy implementation and deployment out-of-the-box. The kit providesmultiple heterogeneous network nodes accessed via TI’s eZ430-RF2500 kit, one CC2500 evaluation module, and one Stellaris DK-LM3S9B96-EM2 Expansion Board. Price: $US 125. Visit: www.ti.com/wk-2500s-pr-es

DK-EM2-2520Z: Stellaris ZigBee Networking Kit with compliant Z-Stack 2.4 software and example coordinator application software that helps you take advantage of a growing portfolio of IEEE 802.15.4 compatible ZigBee products. The kit supplies multiple heterogeneous network nodes with one CC2520 evaluation module, two battery operated sensor modules, and two CC2530 evaluation modules pre-programmed with a temperature sensor application, along with one Stellaris DK-LM3S9B96-EM2 Expansion Board. Price: $US 249. Visit: www.ti.com/wk-2520z-pr-es

You can also purchase a Stellaris EM2 Expansion Board ($US 39 US) www.ti.com/wk-em2-pr-es and the DK-LM3S9B96 ($US 425) www.ti.com/wk- lm3s9b96-pr-es. Development kits include software, reference designs and other technicakl information. Designers and engineers will also find an active community of peers on TI's E2E online network. --Jon Titus

How the Semiconductor Industry Got Off the Ground

2010-10-03T01:10:49Z

When many non-engineers think about the US semiconductor industry company names such as Intel and IBM probably come up right away. But at the time of the semiconductor revolution in the 1950's and 1960's, Intel didn't exist, and IBM had no means to produce its own semiconductors. Instead, IBM bought transistors for $150 each(!) from a small company called Fairchild Semiconductor, probably the company whose people had a greater impact on semiconductor development than any others at the time.

To give people insight into the creation of Fairchild and the engineering and scientific developments at the company, Christoph Lécuyer and David C. Brock have written, "Makers of the Microchip: A Documentary History of Fairchild Semiconductor." Instead of writing a dry corporate history, the authors have put together information about the technical advances as well as the initial funding, office and lab preparation, and "management" of the company in its early days. But that information takes only the first 44 pages of the book. The remainder, called "Facsimiles and Interpretive Essays" looks at letters and lab notebooks from the company's founders and first employees. Only by reading this information can you understand the ways personal connections, college and business friendships, and concentration of technical details shaped the start of the semiconductor industry.

These interpretive essays deal with the basis for the planar transistor structure as well as with the details of furnishing a new office and getting the floors waxed. Looking at photos of lab notebook pages might sound rather dull until you realize that people such as Jay Last, Gordon Moore, Bob Noyce, and Jean Hoerni put the words, charts, and diagrams on paper. If you don't know who these people are or were, you should read this book. Then you will understand how they took crude lab demonstrations and turned them into profitable mass-produced products.

Frankly, I could not put the book down. And although it might seem to appeal mainly to tech-history afictionado, I recommend it to anyone interested in how the semiconductor industry came into its own. The book is well documented with notes and an appendix covers semiconductor technology in the late 1950's and early 1960's.

Lécuyer, Christoph, and David C. Brock, "Makers of the Microchip: A Documentary History of Fairchild Semiconductor," MIT Press, Cambridge, MA. 2010. $24.95. ISBN: 978-0-262-01424-3, 368 pages. www.mitpress.mit.edu. --Jon Titus

A Case for Basic But Expandable Eval Kits

2010-09-22T08:09:40Z

I received news about Actel's SmartFusion Development Kit (A2F500-DEV-KIT) that gives engineers a SmartFusion FPGA with an ARM Cortex-M3 processor core built in. I can understand the appeal of such an FPGA and why engineers might like to use it in a design. The board, though, seemed overly cluttered with "stuff" that includes:

1 RJ-45 connector for 10/100 Ethernet
2 RJ-45 connectors for EtherCAT ports
1 Potentiometer for current measurements
1 Mixed-signal header
1 DAC and ADV I/O header
1 OLED display
1 SPI header
1 I2C header
1 USB connector for UART0
1 RS-485 serial port
1 FlashPro header
1 RealView header
2 CAN bus ports

That seems like a lot to stuff onto one board, and everything uses power. I single out the Actel board only because I just received information about it. Other chip manufacturers and third parties take the same approach and aim to get as many I?O devices and connectors as possible on a board.

Whoa. Not everyone needs that much electronics on one board. In my opinion, companies should offer an alternate route for engineers who want to learn about a new FPGA, microcontroller, analog subsystem, and so on. How about a basic "starter" board that gives people an opportunity to evaluate hardware and get used to software tools? Then, people can move up to boards with additional hardware a step at a time. I could start with a few pushbuttons and LEDs and perhaps a temperature or light sensor.

Microchip Technology makes a series of PICtail daughter boards that connect via a small header to many of the company's PICDEM demonstration boards. The Ethernet PICtail board, for example, lets people add a 10-Base-T Ethernet controller and they can download software examples, too. This approach lets engineers take a look at hardware and software and make incremental steps with hardware as they learn more.

I understand some engineers will jump in with boards and software that provide everything but a TV tuner and Dr. Emmett Brown's flux capacitor. Others, though might benefit from a slower approach that doesn't overwhelm them with a lot of "stuff" they'll never use.That "stuff" includes a lot of software drivers, header and include files that can clutter basic software demonstrations. -- Jon Titus