EEWeb Pulse - Volume 117

PERFORMANCEAll Sensors:

Dale DauenhauerVP of Engineering, All Sensors

Dennis DauenhauerPresident, All Sensors

Rocket Thruster TestingSmart Grid Diagnostics

under

PRESSURE

concepts to reality Bringing your

is as easy as...

Copyright ©2013 Aspen Labs LLC.

Visit: digikey.com/schemeit • partsim.com • pcbweb.com

1.

Create schematics, technical diagrams, and flowcharts using your browser.

• 600+ Symbol Library• Share Schematics Online• Export High Quality Images

digikey.com/schemeit

2.

Free and easy-to-use circuit simulator that runs in your browser.

• SPICE Simulator• AC/DC/Transient Sims• Waveform Viewer

partsim.com

3.

Full featured online CAD application for designing and manufacturing electronics hardware.

• Schematic Capture• PCB Layout• BOM Integration

pcbweb.com

http://bit.ly/SJ9ugm

http://bit.ly/13zBBCn

http://bit.ly/XVO5Wt

applications. Their latest memory architecture is a system-level solution that combines innovative software and hardware with non-volatile memory on the Memory Channel of the system processors. This latest approach to storage and memory promises to revolutionize servers that drive big data, cloud computing, databases, virtualization, and many other critical enterprise workloads.

Each day, society produces and captures more data than ever before—over 2.5 quintillion (that’s a 1 followed by 18 zeros!) bytes daily. Humans have created 90% of the world’s data in the last two years. In order to effectively manage

this explosion of information, enterprise IT managers needed an infrastructure that linked data directly to their applications. Since 2003, Diablo Technologies has been pushing the frontier of high-speed enterprise memory subsystems for Datacenter

Takes Storage Architecture to the Next Frontier

3Visit: eeweb.com

CONTENTS

10Diablo Technologies Takes Storage to the Next Frontier

Diablo’s latest memory architecture is a system-level solution that will revolutionize the

storage industry.

Dennis & Dale DauenhauerALL SENSORS CORPORATION

How this unique sensor company implements complex MEMS technology into its

pressure sensors.

2216Approach to Rocket

Thruster TestingA look at how to monitor rocket thrusters with

various sensors using the LabVIEW software.

28The Need for Smart

Diagnostics in FuturePower Grids

Measuring and controlling current flows in smart grids has never been more important.

22

10

28 4Featured ProductsThis week’s latest products from EEWeb.

RTZReturn to Zero Comic 42

PULSE

34

Product OverviewFreescale K20D50M

Freedom Development board

34

4 Visit: eeweb.com

PULSE

Dual LIN 2.2A/SAE TransceiverThe TJA1022 is a dual LIN transceiver that provides the interface between a Local Interconnect Network (LIN) master/slave protocol controller and the physical bus in a LIN network. It is primarily intended for in-vehicle subnet-works using baud rates up to 20 kBd and is compliant with LIN 2.0, LIN 2.1, LIN 2.2, LIN 2.2A and SAE J2602. The TJA1022 is pin compatible with the TJA1020, TJA1021 and TJA1027. The TJA1022 and TJA1027 are also software compatible. The transmit data streams generated by the protocol con-troller are converted by the TJA1022 into optimized bus signals shaped to minimize ElectroMagnetic Emissions (EME)...Read More

High-Performance Automotive SoCAn R-Car M2 software development board is also now under development and will be available in December 2013 to further simplify and speed the development cycle for designers working with the R-Car M2 SoC. With the newest member of its R-Car series, Renesas aims to expand and reinforce the ecosystem led by the R-Car Consortium, the partner program of oper-ating system (OS) and software providers. The new R-Car M2 SoC exceeds the previous high-end R-Car H1, with more than three times the enhanced CPU and approximately 6 times the graphics capacity and three times the memory bus function compared to the previous R-Car M1...Read More

Leadframe-Based SMT ChipLEDAvago’s ultra-thin ASMT-Rx45 ChipLEDs were developed based on the industrial standard ChipLED 0603 platform which requires less board space. These ChipLEDs provide a wide viewing angle of 130 degrees to improve visibility in bright sunlight. In addition to the high-brightness and compact size, Avago’s ASMT-Rx45 ChipLEDs provide two significant advantages in the production environment: They can be easily soldered using IR solder reflow process, and the package is qualified to a Joint Electronic Device Engineering Council moisture sensitive level (MSL) rating of 2a...Read More

Low Power Integrated Video AmpThe THS7376 is a low-power, 3-V to 5-V single-supply, four-channel, inte-grated video amplifier. The device incorporates one standard definition (SD) filter channel for CVBS video and three high-definition (HD) filter chan-nels. The CVBS filter features a sixth-order filter and the HD channels feature eighth-order filters. These filters are useful as digital-to-analog converter (DAC) reconstruction filters or as analog-to-digital converter (ADC) an-tialiasing filters. The HD filters can be bypassed to support 1080p60 video or up to super extended graphics array (SXGA) RGB video...Read More

http://www.eeweb.com/company-news/nxp/dual-lin-2.2asae-transceiver


http://www.eeweb.com/company-news/renesas/high-performance-r-car-m2-automotive-soc


http://www.eeweb.com/company-news/avago_technologies/leadframe-based-smt-chipled/


http://www.eeweb.com/company-news/texas_instruments/low-power-integrated-video-amplifier/






5Visit: eeweb.com

FEATURED PRODUCTS

First Integrated Audio SuppressorThe MAX9890 provides click-and-pop suppression for devices such as CODECs with integrated headphone amplifiers that lack a clickless/pop-less startup/power-up or shutdown/power-down. The device controls the ramping of the DC bias voltage on the output-coupling capacitors and the application of the audio signal to ensure that no audible transients are present at the headphones. The MAX9890A features a 200ms startup time for use with up to 100µF coupling capacitors. The MAX9890B features a 330ms startup time for use with greater than 100µF coupling capacitors. The MAX9890 consumes 14µA of supply current and 0.001µA in shutdown, while contributing less than 0.003% THD+N into a 32Ω load...Read More

PCB-Mounted Line Sense RelayThe M-949-11 Line Sense Relay is a small, PCB-mounted loop current detec-tor with the safety and reliability features required for UL and British Stand-ard (BSI) regulated telephone applications. The M-949-11 is designed for both North American and international use, and offers superior protection against voltage surges such as lightning strikes. When connected to the voice pair (tip and ring) of an ordinary telephone line, the M-949-11 pro-vides a 1-Form-A relay closure in response to current flowing through the wires. This closure can be used with control circuitry for on-hook/off-hook monitoring, switch hook flash detection...Read More

Optimized for 2D Rendering GDCMB87P2020 “Jasmine” is an enhanced version of MB87J2120 “Lavender” which adds 8Mbit embedded SDRAM, and some more improved features. The device is fully compatible to the Lavender chip and uses the same internal architecture, the same layer and interface concept. No external memory devices are required for graphic memory because of the 8Mbit embedded SDRAM. As a consequence, the package could be reduced to QFP208. Therefore, Jasmine is optimized for compact automotive or con-sumer applications where a high integration of functions is required. Like Lavender, the Jasmine is optimized to work as a companion chip for the Fujitsu 32-Bit RISC devices MB91F36x...Read More

Integrated Analog Video DecoderIntersil Corporation introduced the TW9990, a highly integrated analog video decoder that eliminates the need for external op amps while ena-bling performance advantages through programmable short-to-battery and short-to-ground detection. The TW9990 video decoder is ideal for au-tomotive rear camera displays and other camera systems. Automotive rear camera displays have become a popular driver safety feature. According to Edmunds market research, 70 percent of 2012 vehicles were outfitted with rear cameras...Read More

http://www.eeweb.com/company-news/maxim/first-integrated-audio-click-and-pop-suppressor/


http://www.eeweb.com/company-news/ixys/pcb-mounted-line-sense-relay/


http://www.eeweb.com/company-news/fujitsu_semiconductor/optimized-for-2d-rendering-gdc/



http://www.eeweb.com/company-news/intersil/highly-integrated-analog-video-decoder/





6 Visit: eeweb.com

PULSE

Plug-and-Play DriverThe 1SP0335 single-channel SCALE-2 Plug-and-Play drivers are designed to safely and reliably drive 130 × 140 mm and 190 × 140 mm IGBT modules with an isolation voltage of 10.2 kV and blocking voltages from 3.3 kV to 6.5 kV. They are optimised for high-reliability applications in the rail sector. The driver concept uses a master/slave structure that allows the safe operation of parallel-connected IGBT modules. The master (1SP0335V or 1SP0335S) can be used as a stand-alone driver without slave to drive a single IGBT module or it can be used with up to three slaves (1SP0335D) to drive up to four parallel-connected IGBT modules...Read More

Clock Buffer for Wireless SystemsThe IDT8V79S690I is a JESD204B Clock Fanout Buffer with Configurable Phase Delay. The device has been designed for clock signal conditioning and frequency/phase management of wireless base station radio equip-ment boards. The device supports the division and clock distribution of high-frequency clocks and low-frequency system reference signals. This is a very flexible device that allows clock amplitude and phase to be config-ured, and it is optimized to deliver excellent phase noise performance. The IDT8V79S690I uses SiGe technology for high output clock frequency and low phase noise performance, combined with high power supply noise rejection...Read More

Wi-Fi Module Development KitMicrochip Technology Inc. introduced a simple Cloud Development Plat-form that is available on the Amazon Web Services Marketplace and ena-bles embedded engineers to quickly learn cloud based communication. Microchip’s platform provides designers with the ability to easily create a working demo that connects an embedded application with the Amazon Elastic Compute Cloud service. At the heart of this platform is Microchip’s Wi-Fi® Client Module Development Kit (Part # DM182020), which offers developers a simple way to bridge the embedded world and the cloud, to create applications encompassing the Internet of Things...Read More

Reliable Compact Deutsch ConnectorTE Connectivity Deutsch 369 Connectors are compact, high reliability connectors suitable for a wide variety of applications and industries. The Deutsch 369 Connectors are based on the Deutsch ARINC 809/EN4165 interface and constructed using lightweight, high performance composite materials that conform to low smoke, toxicity, and flammability require-ments of the aerospace industry. These connectors are sealed to IP68 for use in areas with high levels of moisture. The 369 Series eliminates the need for separate part numbers by allowing the housing types to use either pin or socket contacts...Read More

http://www.eeweb.com/company-news/power_integrations/plug-and-play-driver-with-masterslave-operation/


http://www.eeweb.com/company-news/idt/clock-buffer-for-wireless-infrastructure-systems/


http://www.eeweb.com/company-news/microchip/wi-fi-client-module-development-kit/


http://www.eeweb.com/company-news/mouser/highly-reliable-compact-deutsch-369-connectors/






FEATURED PRODUCTS

Automotive Smart Sensor InterfaceThe MLX90320 covers the most typical resistive type of Wheatstone bridge applications for use in an automotive environment. It is a monolithic silicon analog sensor interface that converts small changes in resistors, configured in a full Wheatstone bridge on a sensing element, to large output voltage variations. The signal conditioning includes gain adjustment, offset control and second order temperature compensation in order to accommodate variations of the different resistive sensing elements. Compensation values are stored in EEPROM and can be reprogrammed with an interface circuit and provided software...Read More

Lead-Free SlimLine Slo-Blo FuseThe 468 Series Time-Lag (Slo-Blo®) SMF is a small (1206 size) thin-film device designed for secondary protection of circuits used in space constrained applications such as hand-held portable electronic devices.This series is 100% lead-free and meets the requirements of the RoHS directive. New Halogen-Free 468 Series fuses are available–to order use the “HF” suffix. See PartNumbering section for additional information. The device complies with electronic industry environmental standards for lead reduction. The product is also compatible with lead-free solders and higher temperature profiles...Read More

Low Power 32-bit Stereo CODECThe AK4954A is a low power consumption 32-bit stereo CODEC with a microphone, a headphone and a speaker amplifiers. The input circuits include a microphone amplifier and an ALC (Automatic Level Control) circuit, and the output circuits include a capless headphone amplifier and a speaker amplifier. It is suitable for portable application with recording/playback function. The integrated charge pump circuit generates a nega-tive voltage and removes the output AC coupling capacitors. The speaker amplifier has a wide operating voltage range, which is from 0.9V to 5.5V, enabling a direct drive to batteries...Read More

High Voltage Step Down RegulatorThe A4447 device is a 2 A, high efficiency general-purpose buck regulator designed for a wide variety of applications. The output voltage is adjust-able from 0.8 to 24 V based on a resistor divider and 0.8 V ±2% reference. External components include an external clamping diode, inductor and filter capacitor. The off-time is determined by an external resistor to ground. It operates in both continuous and discontinuous modes to maintain light load regulation. An internal blanking circuit is used to filter out transients due to the reverse recovery of the external clamp diode. Typical blanking time is 200 ns...Read More

http://www.eeweb.com/company-news/melexis/automotive-smart-sensor-interface/


http://www.eeweb.com/company-news/littelfuse/lead-free-slimline-slo-blo-fuse/


http://www.eeweb.com/company-news/akm_semiconductor/low-power-32-bit-stereo-codec/


http://www.eeweb.com/company-news/allegro_microsystems/high-voltage-step-down-regulator/






eeweb.com/register

PULSE

8 Visit: eeweb.com

FEATURED PRODUCTS

AC/DC Driver ICs with Built-In MOSFETThe BM2P Series was designed for use in AC adapters, home appliances, office equipment, and the like. A low ON resistance super junction MOSFET is built in for significantly improved efficiency, ensuring compliance with the latest Energy Star* standard. Delivers an optimized power supply system for sets of all types. For 25W adapters, simply replacing the IC itself will boost efficiency to over 87%. This can be increased to over 90% by optimizing the circuit. As a result, the ICs easily pass the new EnergyStar 6 requirements for efficiency...Read More

2.8W Monno Class-D Audio AmplifierThe CS35L00 is a 2.8 W high efficiency Hybrid Class-D audio amplifier with low idle current consumption and a selectable gain. It features an advanced closed-loop ar-chitecture to provide 0.02% THD+N at 1 W and -88 dB PSRR at 217 Hz. A flexible Hybrid Class-D output stage offers four modes of operation: Standard Class-D (SD) mode offers full audio bandwidth and high audio performance; Hybrid Class-D (HD) mode offers a substantial reduction in idle power consumption with an integrated ClassH controller...Read More

DC-DC Converter for Telecom ApplicationsMurata announced the availability of the RBQ series of isolated 12Vout, 400 Watt DC-DC converters. The RBQ-12/33-D48 model provides what are believed to be the industry’s highest efficiencies of up to 96% from a standard DOSA-compliant quarter-brick package. The RBQ-12/33-D48 is designed to operate in most applications with convection cooling. The RBQ-12/33-D48 can operate from a standard telephone network voltage (TNV) network supply of 36 to 75 VDC around a nominal 48 VDC. The RBQ series is ideal for use in a host of telecommunications...Read More

POP IP Technology for Mali-T600 GPUsARM introduced the first POP IP solution for ARM Mali-T600 series graphics processor units (GPUs). This latest offering of POP IP — core-hardening acceleration technology that produces the best implementations of ARM processors in the fastest time-to-market — is optimized for the Mali-T628 and Mali-T678 on TSMC 28nm HPM process technology. Developed in synergistic collaboration by ARM’s Media Processing and Physical IP divisions, the optimized POP IP technology has been created to produce the most efficient GPU implementations at 28nm...Read More

75 Watt DC-DC Converter Astrodyne’s extensive line of DC/DC converters provide both a high performance and low cost power conversion solutions for telecommunications, industrial, data acquisition, embedded systems, or any application where efficient distributed power is required. Our 0.75 watt to 350 watt power offering gives you the flexibility of high density, wide input range converters in fully encapsulated SIP, DIP, Standard 1” x 2” and ½ Brick modules as well as open frame with integral heatsink and cage-enclosed styles. The modules and open frames are PCB mounted while the enclosed modules offer screw terminals for Input/Output connect ability...Read More

http://www.eeweb.com/company-news/rohm/acdc-driver-ics-with-built-in-mosfet/


http://www.eeweb.com/company-news/cirrus_logic/2.8-w-mono-class-d-audio-amplifier/


http://www.eeweb.com/company-news/murata_power_solutions/dc-dc-converter-for-telecom-applications/


http://www.eeweb.com/company-news/arm/pop-ip-technology-for-mali-t600-series-gpus/


http://www.eeweb.com/company-news/astrodyne/75-watt-dc-dc-converter/







eeweb.com/register

Join Today

Portability & Power All In One...Debug digital designs on an iPad, iPhone, iPod.

Can a logic analyzer be sexy? Watch the video and weigh in...

Logiscope transforms an iPhone, iPad or iPod into a 100MHz, 16 channel logic analyzer.

Not only is it the most intuitive logic analyzer available, the triggering is so powerful you’ll be able to count the hair on your bug.

See why our innovation keeps getting recognized.

http://bit.ly/XYAU15

10 Visit: eeweb.com

TITLE OF SECTIONLOREM IPSUM DOLOR SIT AMET, CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MATTIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR SIT AMET, CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MATTIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR SIT AMET,

CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MATTIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR SIT AMET, CONSEC-TETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MAT-TIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR SIT AMET, CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MATTIS TURPIS.

TITLE OF SECTIONLOREM IPSUM DOLOR SIT AMET, CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MATTIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR SIT AMET, CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MATTIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR SIT AMET, CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MATTIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR S

IT AMET, CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MATTIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR SIT AMET, CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MATTIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR SIT AMET, CONSEC-

TITLE OF SECTIONLOREM IPSUM DOLOR SIT AMET, CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MATTIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR SIT AMET, CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MATTIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR SIT AMET, CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBOR-TIS IPSUM. NUNC VEL MATTIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR SIT AMET, CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MATTIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR SIT AMET, CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MATTIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR SIT AMET, CONSEC-TETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MAT-

TITLE OF SECTIONLOREM IPSUM DOLOR SIT AMET, CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MATTIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR SIT AMET, CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MATTIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR SIT AMET, CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBOR-TIS IPSUM. NUNC VEL MATTIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR SIT AMET, CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MATTIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR SIT AMET, CONSECTETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MATTIS TURPIS. FUSCE SED SEM LOREM IPSUM DOLOR SIT AMET, CONSEC-TETUR ADIPISCING ELIT. NAM SIT AMET LOBORTIS IPSUM. NUNC VEL MAT-

PULSE





11Visit: eeweb.com

TECH ARTICLE





12 Visit: eeweb.com

The Secret Behind a Start-up's Success

The fusion of talent between current CEO Riccardo Badalone, VP for Business Operations Michael Parziale, and VP of Strategic Customer Engineering Franco Forlini, began at Concordia University. Their unified vision and technical expertise led to the creation of Diablo Technologies, an IP company, which focused on building serialized/deserialized or SerDes products for communication applications. Diablo quickly forged a foothold in the communications field and was pivotal in delivering cutting edge SerDes technologies for switching and framer applications. Diablo’s early success enabled the founders to raise the necessary funding to establish a business model, develop a core team, and expand their reach. Paramount to Diablo’s success was the addition of Maher Amer, chip architect and current Chief Technology Officer. Maher took over as CTO in 2005, when Diablo was developing the Advanced Memory Buffer (AMB), an “advanced ASIC that communicates with the system memory interface,” Maher told us. With over a decade of experience focused solely on the system memory interface, Diablo has leveraged their collective understanding and knowledge to create a portfolio of products designed to enhance the performance and capabilities of memory system designs. “Since 2005, we’ve designed the Advanced Memory Buffer, then we built a BOB buffer board, then we built a load reduction chipset,” indicated Maher. Introducing Memory Channel Storage

Diablo’s most recent development and novel ap-proach to storage and memory architecture is sure to disrupt current attempts. Leveraging their deep expertise with the memory bus protocol and transac-tions, they recently introduced a technology platform called MCS™, or Memory Channel Storage™. “You can think of it as a bus protocol that is running directly on the CPUs memory interface,” Maher said. Diablo began exploring the concept several years ago, back when Flash was first introduced in solid state drives. “Companies began placing the solid state drives behind the storage interfaces, such as SATA and SAS. It was a very, very easy integration exercise because all they had to do was swap the spinning disc with Flash devices.” But, despite the advantages of Flash – “IOPS went from a few hundred, to a couple thousand, to a few thousand - it was captive behind the data interface,” explained Maher. In an attempt to lower latency and increase IOPS, developers began moving Flash from behind the stor-age interface to the PCI Express bus, creating what is known today as PCIe-based Flash storage. Although these approaches significantly reduced latency and

increased IOPS, they still faced major limitations. “Data was still routed through the PCIe interface, or the IO controller on the CPU itself. Even though your media now [had] a bigger pipe, it wasn’t leveraging the full potential of Flash,” said Maher. While repositioning Flash on the PCIe interface improved performance, it was far from optimal. The PCI Express bus is a fast but general purpose bus with inherent limitations. First, competing PCI Express devices create bottlenecks that hinder application performance as data transfers to and from various devices collide. Second, active CPU intervention is required while interrupts are used to take control of the bus by the storage, as well as other devices. Moreover, latency in-creases dramatically as performance attempts to scale, resulting in high and extremely variable response times. With Memory Channel Storage, Diablo is placing Flash memory in the most obvious location. By attaching Flash directly onto the CPU’s memory system, MCS is “Moving Flash from behind the PCIe bus to be part of the memory system, where it be-longs.” Through protocol and software interfaces, MCS combines DRAM and Flash modules on the same bus, allowing applications to leverage the Flash associated with all the memory channels of the CPU with nearly equal latency. “This distrib-utes the flash, controllers and FTL across all of the memory channels in the system in a very parallel and highly leveragable fashion. This provides very high bandwidth and IO’s, while keeping write latencies extremely low”, says Maher; “Achieving both simultaneously is just not possible with PCIe based flash.” With this architecture, the memory space is no longer limited to volatile DRAM and also includes non-volatile persistent memory, which can handle NAND Flash or any other type of future media. “In the current iteration, our architecture uses NAND, but we’ve designed this architecture to be future-proofed, allowing for any type of non-volatile memory to be easily supported in the future,” indicated Maher. MCS can also co-exist with other DRAM, “It literally fits in the system like an RDIMM and co-exists with DRAM,” added Maher. Since MCS modules have the same pin-out, electrical, and thermal characteristics of RDIMM’s, DRAM can be replaced or complemented with MCS based modules without restrictions. “It can be placed on its own channels on multiple CPU’s, or they can be attached to the same CPU. Through our IP and deep understanding of how memory controllers work, we were able to overcome [the mixed latency] issue. We allow heterogeneous latency

PULSE

Click to view Memory Channel Storage

Video Overview

13Visit: eeweb.com

devices to exist in the memory interfaces that were designed to only work with homogene-ous latency." “What’s really exciting from a workload per-spective, is that this technology is so flexible, that it addresses problems that are faced by such a broad range of applications”, Maher said. While there have been other, even re-cent, attempts to mix DRAM and NAND flash on the memory channel, those products are meant for a very small, niche set of applica-tions. As such, they are more expensive in terms of dollars per gigabyte than Memory Channel Storage. While those other attempts are meant for applications that require a very small data footprint and nearly one hundred percent writes, MCS ad dresses a broad ar-ray of applications. “We have been working closely with multiple da tabase, virtualization, financial service, and other software applica-tion providers to prove the great benefits of this architecture”, Maher says; “The results have been amazing!”

Finding Memory Channel Storage

“Diablo Technologies has worked closely with multiple server OEMs to ensure that MCS technology works with their most popular servers – only minor modifications in the serv-ers UEFI BIOS are necessary. In the past few

years, Diablo has fostered an ecosystem of partners to ensure the endurance, reliability, compatibility, and distribution of their product. IBM, for example, has been working closely with Diablo Technolo-gies on the MCS architecture. And most recently Diablo has teamed with Sandisk™ to create ULL-traDIMM™, a new tier of low-latency storage, which is the first product to utilize the MCS architecture. The ULLtraDIMM currently comes in capacities of 200GB and 400GB and achieves an astounding five microsecond write latency. The ULLtraDIMM is fully interoperable with RDIMMS and can fit into existing memory slots on the motherboard. This allows for its deployment across a full spectrum of server form factors, even blade sys tems. Several other OEM’s have since begun working to integrate ULLtraDIMM’s into their server offerings.” With the launch of MCS, Diablo has introduced a tech-nology that disrupts the land-scape of both memory and Solid State Storage designs. As Kevin Wagner, Diablo’s Vice President of Market-ing ex pressed, “What we’ve developed is a very unique and flexible memory and storage architecture that is extremely versatile and solves a wide array of enter-prise application problems.” ■

Diablo's MemoryChannel Storage™

FEATURED PRODUCTS

http://www.youtube.com/watch?v=rVDkwsxf6Ec&list=PLxBGCQ7vEcuqrjf4_wOmKiVtkYcCt6EgQ&index=3

http://bit.ly/ZUVpwJ

http://bit.ly/1aazU4s

http://bit.ly/1fYq34h

http://bit.ly/ZUVpwJ

http://bit.ly/1aIrz39

http://bit.ly/1afhnjT

16 Visit: eeweb.com

PULSE

Rocket Thruster Testing Control Design

Recently, the space industry has been becoming more and more commercialized. New space

companies with great new ideas and projects are arising. At the core of many space-related projects

are rocket thrusters, which are used to ascend, descend, or to control the attitude of a craft. In

order to use these rocket thrusters they must first be tested to understand how they work—various

characteristics such as operating temperatures and pressures, and other limitations that might not

be foreseen. Creating a robust control program using LabVIEW and various National Instruments (NI)

hardware may seem daunting at first, but given the right direction turns out to be straightforward. In

this article, I will be explaining the approach that I used to design and implement a control program

to monitor the thruster with various sensors in order to control it with solenoid valves. A brief note:

This article focuses strictly on the requirements for the LabVIEW program control designer. Parts

such as power subsystem requirements and propulsion details are left out of this article. Such

information would go more into supporting power and signal conditioning circuitry and how the

thrusters themselves work.

Approach to

with LabVIEW

Kenrick Dacumos

Rocket Thruster Testing Control Design

Recently, the space industry has been becoming more and more commercialized. New space

companies with great new ideas and projects are arising. At the core of many space-related projects

are rocket thrusters, which are used to ascend, descend, or to control the attitude of a craft. In

order to use these rocket thrusters they must first be tested to understand how they work—various

characteristics such as operating temperatures and pressures, and other limitations that might not

be foreseen. Creating a robust control program using LabVIEW and various National Instruments (NI)

hardware may seem daunting at first, but given the right direction turns out to be straightforward. In

this article, I will be explaining the approach that I used to design and implement a control program

to monitor the thruster with various sensors in order to control it with solenoid valves. A brief note:

This article focuses strictly on the requirements for the LabVIEW program control designer. Parts

such as power subsystem requirements and propulsion details are left out of this article. Such

information would go more into supporting power and signal conditioning circuitry and how the

thrusters themselves work.

Approach to

with LabVIEW

Kenrick Dacumos

17Visit: eeweb.com

TECH ARTICLE

18 Visit: eeweb.com

PULSE

There are many components of a rocket thruster test control program that must be addressed. These components then form into control design requirements for the subsystem in charge of making the LabVIEW control program. The first is data acquisition. Data acquisition, in this case, is the use of LabVIEW software and NI Hardware to acquire, monitor for hazards, and save data from various sensors. This data is crucial in characterizing your particular rocket thrusters and for protecting your system and users from dangerous hazards. Your types of sensors may vary based on your specific test requirements, but most often include load cells, thermocouples, pressure transducers, etc. After the sensors are selected, their datasheets must be examined for a few specifications. The control program designers will need to worry about the type of data signal that the LabVIEW program will be acquiring, such as digital or analog, and any offsets and calibrations that must be included in the program in order to convert the data signal into a usable measurement.

For example, if you have a pressure transducer that outputs a voltage between 0 and 5V, you will need to convert that using a linear mapping from voltage to PSIG, which can be found from the sensors datasheet. Particularly if a sensor is linear within its operating range, all you need to do is plot the Pressure vs. Voltage calibration data found on the datasheet and find the linear trendline equation. This equation can then be used to convert from voltage to pressure. For redundancy and verification, it is a good idea to confirm the datasheet’s sensor calibration data with your own testing of the sensor to ensure you have the most accurate data. This will also ensure that your sensor is functioning properly. Most rocket thruster tests are time sensitive. If you find yourself with a broken sensor it may take weeks to replace it, so the sooner you know it is broken, the better. The last requirement is the required data sampling rate. Different tests require different data sampling rates. Changing sampling rates within LabVIEW is very simple.

The second component of the LabVIEW program is device control. Once again these devices vary by test, but are likely to include control solenoid valves and a spark plug. Other devices might be necessary such as a PWM generator. The first requirement you must focus on for controlling valves is the type of signal you must send. This may be an analog signal, a digital signal, or even a mixture of both. Based on your rocket thruster test requirements the speed of your valves may even be an issue to address. Normally, this can be solved through hardware changes on the power subsystem side by selecting quicker response solenoid valves.

The last control design requirement includes specific testing information. This is heavily based on your test, but may include the following: Test duration, Valve control functionality (automatic/manual control), testing procedure, test abort conditions (e.g. maximum temperature that will trigger the test to stop), soft kill procedures (the steps your test

LabVIEWLabVIEW is short for Laboratory Virtual Instrument Engineering Workbench. Developed by National Instruments in 1986, LabVIEW is a system design platform and development environment for visual programming languages.

19Visit: eeweb.com

TECH ARTICLE

should take when an abort condition is reached), startup procedures, etc. This testing information will affect your design for your LabVIEW program. Here is a scenario. Say you are testing a rocket thruster you designed. In your test you would like to acquire data at 1 Hz while you set up your system, and then after pressing a START button you want to acquire data at 300 Hz. After 30 seconds, you want to stop acquiring data. This procedure gives you two choices to approaching how you design your program. First you can use a flat-sequence structure. Below is a picture of an example flat sequence structure, which can be found in Programming >> Structures >> Flat Sequence Structure.

In each different frame of the flat sequence structure are different pieces of code that correspond to different steps of the test. One downside to using this structure is that it must go in order from left to right and can become very hard to maintain as your program becomes larger and more complex. The second way to go about designing your LabVIEW program is to use a state machine. This is a much better system to use for large, complex programs. However for the scenario above you do not need to use a state machine.

After identifying your control requirements you should have a good idea of how you want to approach designing your control program. Your next step is to choose your hardware that you would want to use. The LabVIEW software enables you to use not only National Instruments data acquisition hardware, but also third party hardware. National Instruments provides many drivers to support such third party hardware. In the off chance they do not have a driver for your hardware, you can make your own using the LabVIEW software. The choice of your sensors, control devices, and your budget will determine what hardware you need. For example, if know you are using the NI SCXI 1001 Chassis, you have various modules to select from to connect your sensors and control devices. For example, if you are using a thermocouple sensors you can use the SCXI 1102 module, which includes some signal conditioning and a built in thermistor to measure the cold junction temperature, which allows you to obtain an absolute measurement from your thermocouple. Continue to do this for the rest of your sensors and control devices ensuring that the modules provide adequate signal conditioning. You will also need to make sure that your hardware can handle the power draw from your sensors and devices. One quick

Figure 1: Flat Sequence Structure

20 Visit: eeweb.com

PULSE

note: If you need a real time system, then you may want to look into hardware that has a reprogrammable FPGA in it.

After choosing your hardware you will need to choose compatible software. You can do most things with the base LabVIEW software. However, depending on your hardware and control requirements you may need particular “modules” or add-ons. For example, if you are using NI real-time hardware that includes an FPGA you will need a LabVIEW software module called the FPGA Module in order to program the FPGA with the LabVIEW software. You can visit NI’s site to learn more about the various LabVIEW modules that they offer.

In the beginning we discussed control design requirements and how these requirements will drive your LabVIEW program design. I will now go over a few tips and guidelines to use when designing your program.

The goal of your program is to be as simple and robust as possible, while also being easy to debug and maintain. While designing your code, you will want to leave comments, describing the function of certain parts of your program. This allows other users to understand how your program works in case they will need to edit, debug, or maintain it. Designing your program to be easy to maintain may be a difficult task especially if you will have different phases of testing where your testing procedures might change. You would want to design your program so that between each phase you are changing the minimal amount. Re-designing your entire program will be a waste of a time when you only need to implement only a few changes in the testing procedure.

When setting up data acquisition in your program, I recommend using the low level DAQmx blocks shown in Figure 1.

Figure 2: Low Level DAQmx

21Visit: eeweb.com

TECH ARTICLE

The reason for using these blocks instead of the NI-DAQmx Express VI’s is that these blocks give you more control in your program. More control means more ways you can choose to design your program in terms of how your program runs as well as timing. Another tip is to initialize all buttons and indicators in their appropriate default states to ensure that they are sending the correct signal at the start of your test. If a valve is open when it should be closed at the start of your test, many disastrous things can happen.

As you make your program you will find that there are many wires being drawn across your program and it is causing your program to look very messy, and thus, difficult to debug and maintain. In order to fix this problem you should utilize clusters. Clusters allow you to bundle different signals (or wires) together so that instead of sending five different signals to a SubVI you can send those same signals in a single cluster wire. I haven’t talked much about SubVI’s previously. SubVI’s are sub-programs. In these SubVI’s’ can be portions of your main program that may take up a lot of space or make your main program cluttered and messy. To help make a cluttered program cleaner, you can utilize SubVI’s.

One example of a good way to use a SubVI is to use one which functions as a calibration for your pressure transducers. So within this particular SubVI, will be an algorithm which will calibrate your raw pressure transducer measurements. As mentioned before, you will want to be able to debug your program easily. This ability comes in very handy when you make small changes in a complex program and cannot figure out what it is not functioning properly. In order to make your program easier to debug, you should include error messages and warnings at all major parts of your program. Such parts include such functions as saving a file, starting a data acquisition read, etc. If an error goes wrong here, your test may become a failure. These can be avoided by including error signals which activate a dialog box and stops your program when an error is hit. You can also program a message into different error so that you can see exactly why your program may be malfunctioning.

To wrap things up, I will include a short and simple 6 step process which summarizes the process you should go through when programming your LabVIEW program for rocket thruster testing.

1. Create calibration and offset SubVI’s’ for each sensor type containing the algorithms to go from the input signal, likely a voltage, to a usable unit of measurement such as degrees Celsius.

2. Create the rough frame of your main program with comments as to how certain parts of it function.

3. Create a user dialog box SubVI that will pop up at the beginning of a test so the user can input all appropriate test information and parameters, such as the file path to save your files, date, time, location, etc.

4. Create your Graphical User Interface with all required buttons and indicators with clear, appropriate labels.

5. Add supporting control code and algorithms to your main program.

6. Clean up your code to ensure it is easy to debug and maintain. ■

to comment on the article.

» CLICK HERE

http://www.eeweb.com/blog/kenrick_dacumos/approach-to-rocket-thruster-testing-control-design-with-labview

22 Visit: eeweb.com

PULSE

Freescale K20D50M Freedom Development Board

The FRDM-K20D50M development board from Freescale is part of the Freedom Development Platform array of boards. The Freedom Platform is a small, low-power, cost effective series of development boards featuring the

Freescale Kinetis family of MCUs. The Kinetis family is based on a variety of ARM Cortex-M series cores and processors.

Overview of the

23Visit: eeweb.com

PRODUCT OVERVIEW

Freescale K20D50M Freedom Development Board

The FRDM-K20D50M development board from Freescale is part of the Freedom Development Platform array of boards. The Freedom Platform is a small, low-power, cost effective series of development boards featuring the

Freescale Kinetis family of MCUs. The Kinetis family is based on a variety of ARM Cortex-M series cores and processors.

Overview of the

24 Visit: eeweb.com

PULSE

Included Hardware Setup & UseThe K20D50M has two USB ports—one is labeled SDA and one is labeled K20. The K20

USB port is for the K20 to act as a USB host and the SDA is what you want to access the

device for programming and debugging. With each of the default applications, the LED light

will cycle through various colors and when it is done, it goes to white. From there, you can

tilt it around and see that it controls the color of the LED based on the accelerometer data

that it is receiving. The capacitive touch slider can be used to control brightness.

One of the main features of the open SDA interface is that it will populate a mass storage

device on your computer. Once you open the device, you can use the files on it in case the

serial port does not populate properly. Open the device manager and verify the ports you

have open such as SDA, CDC serial port, and the web address. To use the serial port, you

will need to find what the port number actually is. To find it, look under “Other,” and there

will be another category labeled “Other,” and the device will be shown. Right click on the

device and select “Update Driver Software.” From there, browse for the device and point it

to the SER CDC 89 file that is on the Freedom Development Platform Board.

To obtain a quick start pack, go to Freescale’s website and download the pack that contains

open SDA applications as well as precompiled examples. It will also give you the pin outs

and a quick start guide, which you can use to go through some of these same steps. If you

want to load another application, load the .srec file on to your board and copy it onto the

mass storage device, which will populate it in Windows under the K20D50 name. Once it

receives the information, you will be able to turn the LED on the board on and off.

ConclusionThe Freescale Freedom Development Platform is a fantastic platform, and because it’s

ARM based, you can use your favorite ARM tool chains. Also, because it is compatible

with the Arduino R3 shield, you have the ability to access to a ton of hardware for

developing your project.

To purchase the FRDM-MK20D50M, visit Mouser’s website.

Watch Video

Capacitive Touch SliderFRDM-K20D50M

Freedom Development Platform

J19 I/O Header

J2 I/O Header

K20D50M

AccelerometerMMA8451Q

J10

J9

OpenSDAK20D50M USB

25Visit: eeweb.com

PRODUCT OVERVIEW

Included Hardware Setup & UseThe K20D50M has two USB ports—one is labeled SDA and one is labeled K20. The K20

USB port is for the K20 to act as a USB host and the SDA is what you want to access the

device for programming and debugging. With each of the default applications, the LED light

will cycle through various colors and when it is done, it goes to white. From there, you can

tilt it around and see that it controls the color of the LED based on the accelerometer data

that it is receiving. The capacitive touch slider can be used to control brightness.

One of the main features of the open SDA interface is that it will populate a mass storage

device on your computer. Once you open the device, you can use the files on it in case the

serial port does not populate properly. Open the device manager and verify the ports you

have open such as SDA, CDC serial port, and the web address. To use the serial port, you

will need to find what the port number actually is. To find it, look under “Other,” and there

will be another category labeled “Other,” and the device will be shown. Right click on the

device and select “Update Driver Software.” From there, browse for the device and point it

to the SER CDC 89 file that is on the Freedom Development Platform Board.

To obtain a quick start pack, go to Freescale’s website and download the pack that contains

open SDA applications as well as precompiled examples. It will also give you the pin outs

and a quick start guide, which you can use to go through some of these same steps. If you

want to load another application, load the .srec file on to your board and copy it onto the

mass storage device, which will populate it in Windows under the K20D50 name. Once it

receives the information, you will be able to turn the LED on the board on and off.

ConclusionThe Freescale Freedom Development Platform is a fantastic platform, and because it’s

ARM based, you can use your favorite ARM tool chains. Also, because it is compatible

with the Arduino R3 shield, you have the ability to access to a ton of hardware for

developing your project.

To purchase the FRDM-MK20D50M, visit Mouser’s website.

Watch Video

Capacitive Touch SliderFRDM-K20D50M

Freedom Development Platform

J19 I/O Header

J2 I/O Header

K20D50M

AccelerometerMMA8451Q

J10

J9

OpenSDAK20D50M USB

To learn more about the Freedom Development Platform and purchase the FRDM Board, visit Mouser’s website.

http://www.youtube.com/watch?v=nr184SZQibY&feature=c4-overview-vl&list=PLxBGCQ7vEcupTLZjnZA4Isf_1UBEZQdne

http://www.mouser.com/ProductDetail/Freescale-Semiconductor/FRDM-K20D50M/?qs=sGAEpiMZZMvFPGEOwQcrYzDymgRYjreQUwMSQLOprtk%3d

http://www.mouser.com/new/Freescale-Semiconductor/freescale-freedom-boards/

Wide VIN 500mA Synchronous Buck Regulator ISL85415The ISL85415 is a 500mA Synchronous buck regulator with an input range of 3V to 36V. It provides an easy to use, high efficiency low BOM count solution for a variety of applications.

The ISL85415 integrates both high-side and low-side NMOS FET's and features a PFM mode for improved efficiency at light loads. This feature can be disabled if forced PWM mode is desired. The part switches at a default frequency of 500kHz but may also be programmed using an external resistor from 300kHz to 2MHz. The ISL85415 has the ability to utilize internal or external compensation. By integrating both NMOS devices and providing internal configuration options, minimal external components are required, reducing BOM count and complexity of design.

With the wide VIN range and reduced BOM the part provides an easy to implement design solution for a variety of applications while giving superior performance. It will provide a very robust design for high voltage Industrial applications as well as an efficient solution for battery powered applications.

The part is available in a small Pb free 4mmx3mm DFN plastic package with an operation temperature range of -40°C to +125°C

Related Literature• See AN1859, “ISL85415EVAL1Z Wide VIN 500mA

Synchronous Buck Regulator”

Features• Wide input voltage range 3V to 36V

• Synchronous Operation for high efficiency

• No compensation required

• Integrated High-side and Low-side NMOS devices

• Selectable PFM or forced PWM mode at light loads

• Internal fixed (500kHz) or adjustable Switching frequency 300kHz to 2MHz

• Continuous output current up to 500mA

• Internal or external Soft-start

• Minimal external components required

• Power-good and enable functions available.

Applications• Industrial control

• Medical devices

• Portable instrumentation

• Distributed Power supplies

• Cloud Infrastructure

TABLE 1. KEY DIFFERENCES BETWEEN PARTS

PART NUMBER

TEMP RANGE (°C) EN THRESHOLD PG DELAY

ISL85415 -40°C to +125°C Standard TTL input

10% of soft-start time

ISL85415A -40°C to +85°C High accuracy 2ms

NOTE: See Electrical specifications for more details on the ISL85415.

FIGURE 1. TYPICAL APPLICATION FIGURE 2. EFFICIENCY vs LOAD, PFM, VOUT = 3.3V

GNDCBOOT100nF

CFB

R3

R2

PHASE

SS

SYNC

BOOT

VIN

PGND

FS

COMP

FB

VCC

PG

EN

CVCC1µF

CVIN10µF

L122µH

COUT10µF

1

2

3

4

5

6

9

10

11

12

INTERNAL DEFAULT PARAMETER SELECTION

COUT10µF

VOUT

50

55

60

65

70

75

80

85

90

95

100

0 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 OUTPUT LOAD (A)

EFFI

CIE

NC

Y (%

)

VIN = 5V VIN = 15V

VIN = 24V VIN = 33V

VIN = 12V

September 5, 2013FN8373.1

Intersil (and design) is a trademark owned by Intersil Americas LLC. Copyright Intersil Americas LLC 2013.All Rights Reserved. All other trademarks mentioned are the property of their respective owners.

Get the Datasheet and Order Samples

http://www.intersil.com

http://bit.ly/nHxRhB

“Like” EEWeb’s New Facebook Page

One of

Freescale’s New

Freedom Boards!

and Get a Chance to

Win

http://on.fb.me/1dJtAT4

28 Visit: eeweb.com

PULSE

All Sensors Corporation is a leading manufacturer of high accuracy pressure sensors and transducers. Started by company president Dennis Dauenhauer in 1999, All Sensors was the culmination of his decades of sensor technology experience. The company specializes in low and medium-pressure sensors using complex MEMS technology, yielding highly accuracy and quality sensor products.

We spoke with Dennis Dauenhauer and his brother Dale, Vice President of Engineering, about the challenges in implementing MEMS technology, what drives the company to succeed, and some of the surprising new applications of their products.

Performanceunder

PRESSURE

Dennis & Dale Dauenhauer

All Sensors

29Visit: eeweb.com

INTERVIEW

All Sensors Corporation is a leading manufacturer of high accuracy pressure sensors and transducers. Started by company president Dennis Dauenhauer in 1999, All Sensors was the culmination of his decades of sensor technology experience. The company specializes in low and medium-pressure sensors using complex MEMS technology, yielding highly accuracy and quality sensor products.

We spoke with Dennis Dauenhauer and his brother Dale, Vice President of Engineering, about the challenges in implementing MEMS technology, what drives the company to succeed, and some of the surprising new applications of their products.

Performanceunder

PRESSURE

Dennis & Dale Dauenhauer

All Sensors

30 Visit: eeweb.com

PULSE

What was your personal motivation for getting into sensor technology?

Dennis: Well it’s the thing I know best—it’s the business I’ve been in for pretty much most of my working career. We’re always developing new products, we’re always developing new processes, we’re always establishing new relationships, so it’s fun to do.

Could you give an overview of the types of sensors you make?

Dennis: It’s predominately low-pressure sensors, although we periodically look at other types of sensors. We really focus on low pressure where there are still not as many competitors and

technology, so it’s still a developing market. We don’t do acceleration, we don’t do force, and we don’t do chemical, whether it’s humidity or anything else. If technology comes along where we could provide some value added that is not existent in the marketplace, we’ll take a look at those.

Right now, if you look at some of those other sensor type sensors. If you take high pressure, the market is pretty well served with both the technology and companies that are out there. If you look at acceleration, such as Inventis, a guy that used to work for me started the company. They do most of all the accelerometers used in all handheld devices. The accelerometers in automotive applications are fairly well served. If you look at chemical sensors, some of that is evolving, but a lot of that is still on the university level.

Your technology is based on MEMS, which is pretty complex technology. Can you give us an overview of what makes the technology used in your sensors unique?

Dennis: In the chip itself, the pressure sensing element has a combination of very old technology that was incorporated in thin film devices forty years ago using strictly Poisson equations for transfer of pressure into an electrical signal. In combination with

MEMS Microelectromechanical systems is the technology of extremely

small devices. Despite the miniature sizes, technological advancements

have allowed MEMS devices and MEMS microsensors to reach significantly higher

speeds and sensitivities.

“ Being a small company, we have a lot of customers that come to us for specific solutions, so a lot of our products aren’t necessarily offered to everyone. We do a lot of custom projects.”

31Visit: eeweb.com

INTERVIEWearly piezoresistive technology that used longitudinal and transverse strain and the mobility of the silicon to get a large gauge factor over what’s done in thin film. In the combination of those two, we started on that chip development about six years ago and that’s still an ongoing development. Not so much in the topography of the chip, but more in the process to enhance the performance of the topography. That’s what really gives us a technical advantage. What we then do is combine the technical advantage incorporated in the chip combined with the manufacturing that we’ve set up all over the world. The chip itself is done in Europe in two different locations.

What are the biggest challenges in manufacturing these sensors?

Dale: One of the biggest challenges is finding contract manufacturers outside of our facilities that are capable of processing our technology. It’s not a standard process like an IC as far as the sensor attached. Most contract manufacturers aren’t used to having to be so careful in attaching a die. With ours, you can’t necessarily use a standard piece of equipment the way it’s designed to process the die because it is so sensitive. A lot of the equipment the way it is—even the wafer processing equipment in a MEMS fab—has to be modified to process our die because it’s so sensitive.

Standard saws usually jet water to keep the wafer clean and free of particulates. If it was used with our die, the way the system was designed, it would break out all the diaphragm. We have to modify standard die patch equipment to ensure the sensors aren’t damaged when they’re processed.

Back to the products in your bigger picture: what types of customers are using your products?

Dennis: Probably the biggest market we service is medical—anything for respiratory breathing and wherever else they need it. They tend to use it as a flow sensor, reading differential pressure that they turn to a flow sensor for breathing applications of any type, such as infant monitors and for inhaling and exhaling for sleep apnea. The next biggest market is industrial automation equipment used in everything from semiconductor

equipment to leak detection. Following that market would be UAVs, unmanned aerial vehicle autopilot controls. These customers use pressure sensors in that application for airspeed and altimeters. Especially in the UAVs in predator drones where they want really lightweight, relatively high accuracy, with an extended temperature range for performance characteristics.

The MLV Series Low Voltage Pressure Sensorsbased on All Sensors’ CoBeam2 TM Technology.

“A lot of the equipment the way it is—even the wafer processing equipment in

a MEMS fab—has to be modified to process our die

because it’s so sensitive.”

32 Visit: eeweb.com

PULSE

Outside of pressure sensors, what are your core products?

Dennis: Probably flow—that would be one of the next ones we’re taking a look at.

Dale: Our same medical customer base uses pressure and flow in the same applications so it kind of lends itself to a good fit for us.

Dennis: It looks like there are some emerging chip developments that are being done that we could borrow and apply to the markets that we’re serving. They’re doing the development aimed at some high volume consumer applications but the same sensing elements could be used in our kind of applications. For instance, this particular company wants to put a flow sensor in all the electronic cigarettes.

Would you say the culture at All Sensors is driven from bottom-up innovations or top-down visions? How would describe the way your company is moving, and how it works internally?

Dennis: I think it’s drive from the market, and then we’ve got three US sales guys that cover geographic area in the US. Then we have a guy who we’ve worked with for a while that

“ Our products have been involved in two very big aerodynamics projects for the new Tesla Type S and Hendricks Motorsports racing R&D program.”

covers Europe. If it’s driven from anywhere it’s outside in, rather than internal. Then everyone sits around and listens to what is said. We have one particular technology guy from the outside who has been retired for quite a while, so use him when we have technology questions that are related to the process. Periodically we have meetings to discuss things. I don’t think there is one direction where it’s coming from

Dale: Being a small company, we have a lot of customers that come to us for specific solutions, so a lot of our products aren’t necessarily offered to everyone. We do a lot of custom projects.

What are some of the most interesting applications you haveseen for your sensors?

Dale: Our products have been involved in two very big aerodynamics projects for the new Tesla Type S and Hendricks motorsports racing R&D program. These are both very high profile applications that we’re involved in. In the new Tesla Type S sedan. Our sensors are used in aerodynamics testing for that project. With Hendricks Motorsports, which is a huge racing company with many champions, our sensors are used in those applications as well. Obviously more aerodynamic vehicles on an electric car adds to the efficiency, which is the big game play there. It allows the vehicle to get the more miles out of charge. Anything that helps boost energy efficiency is in high demand right now. ■

Prem Magnetics, Inc. Johnsburg, IL 60051 USA

Call Us: 815-385-2700

Our transformers are 100% final tested for quality.

ZERO DEFECTS GUARANTEED

DIRECT LINKS Power

Switchmode

Inductors

Audio & Telephony

High Speed Data

Economy Audio

Line Matching

Instrument & Current

CRT Display Magnetics

Factory Direct Transformers 18 Hour Express Delivery ¥ Buy Online

www.partsim.com

PartSim includes a full SPICE simulation engine, web-based schematic capture tool, and a graphical waveform viewer.

Some Features include:• Simulate in a standard Web Browser• AC/DC and Transient Simulations• Schematic Editor• WaveForm Viewer• WaveForm Viewer• Easily Share Simulations

Online Circuit Simulator

Try-it Now!

http://bit.ly/13zBBCn

http://www.premmagnetics.com/Power/Category

http://www.premmagnetics.com/Switchmode

http://www.premmagnetics.com/Inductors

http://www.premmagnetics.com/AudioTelephony

http://www.premmagnetics.com/HighSpeedData

http://www.premmagnetics.com/EconomyAudio

http://www.premmagnetics.com/LineMatching

http://www.premmagnetics.com/Instrument

http://www.premmagnetics.com/CRT

34 Visit: eeweb.com

PULSE

SMARTThe Need for

A Practical Example for MV Cables

Peter Van Der Wielen & Edwin Maurer DNV KEMA

The majority of discussions around Smart Grids is about smart metering and application of smart control systems into the grid in order to e.g. increase the flexibility and optimal use of the grid. Measuring and controlling current flows and voltages in an intelligent way is an important part of the smart grid of the future. However, although the energy flows may not be unidirectional and from point to point anymore in the modern and future grids, the core purpose of a power grid remains transporting and distributing energy. One of the methods to help ensuring this primary function is by applying diagnostics to assess the condition of the various components of the grid. In future smart grids, possibly also smarter diagnostics will be needed.

Diagnostics

35Visit: eeweb.com

TECH ARTICLE

SMARTThe Need for

A Practical Example for MV Cables

Peter Van Der Wielen & Edwin Maurer DNV KEMA

The majority of discussions around Smart Grids is about smart metering and application of smart control systems into the grid in order to e.g. increase the flexibility and optimal use of the grid. Measuring and controlling current flows and voltages in an intelligent way is an important part of the smart grid of the future. However, although the energy flows may not be unidirectional and from point to point anymore in the modern and future grids, the core purpose of a power grid remains transporting and distributing energy. One of the methods to help ensuring this primary function is by applying diagnostics to assess the condition of the various components of the grid. In future smart grids, possibly also smarter diagnostics will be needed.

Diagnostics

36 Visit: eeweb.com

PULSE

DEFINITION OF SMART DIAGNOSTICS

In order to be able to discuss smart diagnostics and their need in the future grid, a clear understanding of the definition of a “smart diagnostic” is needed.

The word “diagnostic” is a well-known term in the power engineering world. This word is about determining the condition of equipment in order to assess whether it will fail during some future period and how well it will perform. An important part of the definition of diagnostic is also related to the difference with “test.” A diagnostic is intended to be non-destructive and is focused on measuring certain properties of the equipment in order to estimate the condition and remaining life. A test can be destructive and is intended to subject the equipment to specific conditions in order to determine whether the equipment’s properties change during the test.

The less trivial part in the term “smart diagnostics” is the word “smart.” The question is: what makes a diagnostic smart? Obviously, this answer is related to its application in the future smart grid: it should be able to be implemented in the grid itself. Furthermore, a smart diagnostic is usually using various IT related capabilities in order to perform its diagnosis. Many people will think in the direction of something like a smart solution to a property that is difficult to measure, or by using smart and new technologies. All is true, but the most important aspects according to the authors’ opinions are more related to the diagnostic results, how they are obtained and how they are made available. These can be summarized as follows:

First of all, a smart diagnostic should be specific. This has to do with two elements. (1) It should be specific in its results in a sense that it should be clear whether the measured properties are properties of the component that is measured and not some adjacent component in the grid. (2) A smart diagnostic is optimally also specific in the location of deviating properties of the measured component.

The second characteristic of a smart diagnostic is that the outcome should be meaningful. Applying a diagnostic is measurement of a certain measurable property. This property, however, is usually not

the required information (i.e. the condition) itself, but some property that can be directly or indirectly an indication of this condition. In other words, usually, the measured property still needs some interpretation. A smart diagnostic should have this translation as part of the diagnostic, either implemented in the diagnostic equipment itself or as a service associated with the diagnostic. The resulting outcome should usually be something like condition, failure risk or remaining life.

Another property the outcome of a smart diagnostic should have is: being actual. This means that the moment the results of the diagnostic are available, they should still be valid, i.e. not the properties of a week or month ago. This property gets more important in a future grid, where the loads can potentially be very unpredictable and also fluctuating a lot (e.g. decentralized generation). The measured property that acts as a diagnostic indicator can e.g. fluctuate as a function of load (and heat). Furthermore, the conditions and reliability can become a momentary input for (automated or not) network switching decisions, as the result is related to risk of operation. This also points towards another meaning of the word actual: the diagnostic should be able to provide the condition of the component at any moment in time and its associated data should therefore be available at any time. This points towards monitoring: applying the diagnostic (preferably automatically) continuously or in a repetition rate related to the speed in which a condition can change.

One of the most important aspects of a smart diagnostic is that its outcome is reliable. Having a diagnostic of which the outcome is not reliable might be worse than having no diagnostic at all. This reliability is related to two elements. (1) The measurement itself should be reliable. As diagnostic measurements often depend on subtle changes of small values, the measurements can often be considered as difficult to perform correctly and therefore specialized. Measurements in reality are not always performed correctly or by well-educated staff. (2) The second element of a diagnostic is the interpretation. Even if the measurement was performed perfectly, the interpretation is often also something that needs experience, the right background information and the right experts or expert

37Visit: eeweb.com

TECH ARTICLErules. If either of the two mentioned elements goes wrong, the outcome of the diagnostic should be considered as unreliable. Therefore, a smart diagnostic should have a system implemented in which the chance of this to happen is eliminated or minimized.

Finally, a smart diagnostic should provide its result on time, i.e. if the condition is changing, this should be communicated immediately, in order to enable timely measures or adjustments. Especially degradation mechanisms that evalu-ate fast or that can only be detected close before failure are important to be detected immediately and the associated diagnostic results should be communicated almost instantly in order to enable countermeasures. Otherwise, the degradation mechanism is missed, or the detected change in condition is only realized after it is too late. This also points towards con-tinuous monitoring of the diagnostic property.

Summarizing the above properties of a smart diagnostics results in the word S.M.A.R.T.:

• Specific,• Meaningful,• Actual,• Reliable and• Timely.

THE NEED FOR SMART DIAGNOSTICS

As mentioned in the introduction section of this paper, it is well understood that the core function of a power grid is and remains transporting and distributing energy, despite of all kinds of extra features the future smart grid has or may get. Furthermore, the primary features around this core function are, and will always be, doing this in a safe and reliable manner. All other features may be useful, needed and efficient, the core function and its primary features are still the base. Without a reliable base, all the secondary features are useless.

Looking at it this way, a grid would not be very “smart” if it would not be able to monitor its own core function and accompanying primary features. Often, one of the first features any intelligent system or product gets (or should get) is some kind of self-check in order to determine at least its capability to perform its most important function. This is why intelligent systems to monitor the capability of a grid to perform its core function, transport and distribute energy, cannot be excluded from any grid, especially from a smart grid.

This function becomes more important when transforming to the future smart grid. There are three elements related to the current load that strengthen this:

1. in the future smart grid, the current loads can be more fluctuating (e.g. wind farms and decentralized generation), causing more stress variations to several components in the grid,

2. the current loads and variations will be less manually controlled / more automated and therefore less predictable in a conventional way,

3. due to the high controllability, the future smart grid will enable postponement of expansion investments, because the current loads may be distributed such, that the equipment is used more optimally or longer. This will increase the dependency on this equipment and therefore the importance on its reliability.

Furthermore, if the condition of equipment is known continuously, this is valuable information for the grid to use in switching actions. Switching actions have impact on the risk state of the grid, so the conditions of the equipment in operation are the input parameters leading to possible new risk states and therefore altered switching decisions (either automated of manual).

Besides these arguments, the more conventional arguments for condition monitoring obviously remain without strength. Having (preferably continuous) insight in the condition and risk on failure of the power system’s components is inevitable for correct implementation of reliable asset management systems. It will at least help to:

• prevent failures and unplanned outages,• determine the optimal time for replacement

(too late means failures and too soon means too early investments).

SMART DIAGNOSTIC WITH SMART CABLE GUARD FOR MV CABLE CONNECTIONS

Background

The behavior of the medium voltage cable feeders has a major influence on the overall reliability figures of many grids. So when smartening a grid, it seems logical to start with putting intelligence in that part of the network.

38 Visit: eeweb.com

PULSEFurthermore, it is known that load variations can have a strong relation with certain degradation mechanisms in especially MV cable accessories. So to smarten the cable connection, the health of a cable feeder should be monitored continuously.

Detection of Partial Discharges (PDs), if applied correctly, is a well-known and proven way to detect weak spots in the insulation material of high-voltage equipment. PD measurements for medium-voltage cable connections have therefore become popular since the early 1990-s. These are mainly off-line measurements that can both measure and locate PDs, based on one sensor at one cable end and PD pulse reflection. This implies that the cable has to be disconnected from the grid and energized with a separate power supply in order to generate the PDs. Depending on both details of the measurement itself and the quality of the result interpretation, these measurements have shown very good results and many weak spots in cable connections have been discovered before leading to an outage. Also due to these good results, many grid owners indicated the strong need for a similar method that would even go beyond this: a diagnostic measurement method that is able to assess the cable systems’ condition on-line, i.e. while the cable connection remains in-service. This would also be a requirement for making this diagnostic a smart diagnostic.

Another requirement for such a system was that it should be able to be applied on old (current) cable connections, as those are often the ones with increased chances on problems.

After a couple of years of research activities a prototype of a smart diagnostic, able to measure and locate PDs in an on-line MV power cable connection became available in 2005. Since then, energy was spent in realizing practically applicable and commercial equipment, which has become available recently (2007). The system is called Smart Cable Guard (SCG) and was during the development phase known as PD-OL (Partial Discharge monitoring On-line with Location).

Approximately 225 SCG systems have been installed meanwhile and put into operation since that time.

Basic Measurement Technology

As the system needs to be able to measure continuously, it consists not only of measure-ment units, but also of a centralized data center, which collects all data and interprets the results.

In order to measure a complete cable feeder, two measurement units are needed. Each of these is to be installed at one of the cable circuit ends (e.g. substation or RMU, Ring Main Unit). Each measurement unit consists of a Sensor/Injector Unit (SIU) and a Control Unit (CU). See Fig. 1 for an illustration.

The SIU contains both a sensor, to measure pulses from the cable, and an injection de-vice, to inject pulses into the cable. The SIU is placed around the cable or earth lead (not both) and is of the inductive type. The SIU is separable and thus can even be mounted without disconnecting the cable termina-tion. Depending on the local safety rules, it can in this way be mounted while the cable remains in-service. The installed sensor will not have any galvanic contact with cable and will therefore never introduce any reliability threat itself.

The SIU also injects pulses into the cable sys-tem, for both synchronization and calibration. Synchronization is needed for accurate time-base alignment of both measurement units in order to eliminate any pulses originating from outside the monitored cable circuit (i.e.

Figure 1: SCG measurement setup. At each cable end there is a control unit (CU) for signal processing, data handling and com-munication and a sensor/injector unit for the actual measurement and pulse injection.

39Visit: eeweb.com

TECH ARTICLEto eliminate noise and disturbances), and to locate the weak spots in the complete cable circuit by means of the detected PDs. The synchronization is done by injecting a pulse on one site of the cable circuit which will travel in exactly the cable propagation time to the other cable end where it is meas-ured and recognized. As both measurement units will incorporate the measured injection pulse in their recorded signals, these injection pulses can be used as extremely accurate time beacons (by applying appropriate signal processing and corrections for propagation time of the cable).

The SIU also offers the possibility of calibra-tion, because the injected pulse of the SIU is measured by the sensor part of the SIU and its transfer impedance is depending on the local impedances, i.e. those of the RMU including the power cable(s) at hand. Thanks to the calibration it is possible to calculate the actual PD charge from the measured PD pulse shape instead of only having an indication in the mV scale. This is important for the data inter-pretation later on. Calibration also offers the possibility to calculate and implement digital noise suppression if needed. This digital noise suppression is a task of the Control Unit (CU).

The CU is connected to the SIU by means of an optical fiber. This eliminates possible EMC interference, which can easily occur in the environments in which the units are applied. The CU controls the measurement sequence, the data collection, the signal processing and the data communication towards the Control Center.

Signal processing is crucial, since practical (and especially on-line) PD measurements are inherently impeded by noise and interfer-ence to a large degree. Disturbance by radio stations can be suppressed using adaptive notch filters or cancellers, which hardly af-fect PD signals. However, broadband noise poses a fundamental limit on PD detection. Matched filtering is the technique for detec-tion of pulses in the presence of noise that is optimal in the sense that the average signal to noise ratio (SNR) at the filter output is (by definition) maximized.

The CU also has communication facilities on board (LAN or mobile cellular phone card) in order to upload the resulting data via the internet to the Control Centre at DNV KEMA for further interpretation after which the infor-

mation will be projected on the secured SCG Customer Web Interface where each SCG user can access their own data only.

Centralized Control and Surveillance

After the data is uploaded and combined, various statistical parameters are calculated in order to be able to see the behavior of groups of PDs from e.g. one specific period of time or from a specific location along the cable. After this, the results are interpreted by means of knowledge rules and experts in this area. The data is also evaluated on trends. Having this centralized Control Center at one location instead of having many people interpreting small parts of the full set of diagnosed cables has important advantages, among which:

• the interpretation is done with uniform quality,

• since the interpretation is centralized, the growth of knowledge rules is much faster than would be otherwise,

• for this kind of expertise, there is within a rea-sonable period (or maybe never) enough data to be able to construct reliable rules purely from the statistics. This means also much theoretical, fundamental and practi-cal knowledge from experts has to be in-corporated,

• network owners do not need to invest in obtaining and maintaining specialized, technology specific knowledge as the measurements are translated into risk lev-els, which can be used directly into asset management systems.

Furthermore, in this way, the SCG units can be reached via the internet for diagnostic purposes and updates (e.g. adjust measure-ment repetition rate and update soft- and firmware), so no physical access to the units is necessary once installed.

This all results in a large set of diagnostic equip-ment, currently installed on in-service MV ca-ble feeders, from which interesting results can be learned. A few will be presented in the next section.

SCG EXAMPLES FROM THE FIELD

Since the time practical equipment became ready (2007), various weak spots have been allocated and in this article some of the new results are presented.

40 Visit: eeweb.com

PULSEIn Fig. 2 the 3D graph shows the continuous risk determined from measured PDs from an XLPE based circuit of 5561m in length. The cold shrink joint at 3526m clearly showed significant PD activity during 4 days. A warning was given to the network owner, who replaced this joint after which the PDs disappeared, as can be seen. The defect causing these PDs was a loose conductor connector, further shown and described in Fig. 3.

The second example is an example from SCG units that were installed on two cable feeders after each other, with in between a RMU (Ring Main Unit, customer substation, transformer house). One cable feeder was 260m, the other 220m and the RMU in between. All cable and accessories are extruded material.

Clearly some PD activities were visible from the location of the RMU in between. Severity is however hard to determine from these PD graphs. In the KEMA Control Center, data and statistical analysis is performed. This, together with applied knowledge rules and experience, leads to risk plots, of which one is shown in Fig. 4. In this plot, the dangerous location is clearly identified; therefore a high risk level situation was communicated to the network owner. As a result, the suspicious RMU was inspected and clear traces of electrical sparking in one of the insulating booths of a termination were discovered (see Fig. 5).

CONCLUSION

Smart diagnostics are essential for moving the grids further towards smart grids. In order to make a diagnostic SMART, it should incorporate the following properties:

• Specific (specify component and lo-cation)• Meaningful (result in terms of condition

or risk level)• Actual (continuous monitoring and

result availability)• Reliable (reliability of measurement

and interpretation)• Timely (early detection and communi-

cation of condition deviations)

Figure 2: 3D graph continuous risk level (vertical) as a function of location (from left to front) and time (from front to right). A period of a few weeks is shown, indicating the PD activity with high risk during the first 4 days.

Figure 3: The cable conductor connector has generated a lot of heat as it became a loose contact with high resistivity; due to this heat development the sleeve became brittle and this caused the measured PDs.

41Visit: eeweb.com

TECH ARTICLEThe required properties for making a diagnostic SMART are met using an SCG system. The SCG system locates the obtained deviations in condition, it determines risks on failure, it monitors continuously and interpretation is done by specialists in this area, together with automated knowledge rules.

The obtained effectiveness of the SCG system to detect problems in cable connections before they lead to an actual outage is over 80%. The SCG system can help network owners to:

1. prevent outages and thereby increasing the reliability of the (smart) grid,

2. have constant insight in the risk on failure of each cable connection and thereby insight in the derived operational risks,

3. have better insight in the remaining life of the cable connections and thereby determine optimal time for replacement and/or maintenance.

Furthermore, the system can be applied to newly installed cable connections as a method of performing a prolonged field acceptance test after installation in order to capture the faults that occur during the first period of the lifespan where faults happen more frequently according to the well-known ‘bath tub’ curve (e.g. due to non-ideal installation).

This all makes the SCG system a good example of a smart diagnostic which cannot be excluded from future smart grids.

Acknowledgments

This work was supported by the energy consulting and testing firm DNV KEMA, the automation company Locamation B.V. and the Dutch utilities Alliander N.V. and Enexis B.V. ■

Figure 4: 3D graph of continuous risk plot of circuit A, with horizontally location, vertically continuous risk level and into the paper date/time.

Figure 5: Traces of sparking and degradation on the insulating booth

to comment on the article.

» CLICK HERE

http://www.eeweb.com/blog/nicholas_abisamra/the-need-for-smart-diagnostics-in-future-smart-grids

eeweb.com/register

Join Today

Martian Salary

Good, Fast, Cheap

Glyph Notes

http://bit.ly/jd6Wcw

Date post:	23-Mar-2016
Category:	Documents
Upload:	eeweb-magazines
View:	223 times
Download:	1 times

EEWeb Pulse - Volume 117

Documents