R16.50 (incl VAT)

25 July 2012

FEATURING: Test & Measurement, Analysis, Calibration & Compliance • Audio & Video Electronics

Dataweek, 25 July 2012www.dataweek.co.za

News 125 July 2012 Vol. 35 No. 14

News

Events

Literature showcase

Systems, components, design

General

Hot chips

QuickFind index

Test & measurement, analysis, calibration & compliance

Audio & video electronics

regulars

All rights reserved. No part of this publication may be reproduced, adapted, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying,

recording or otherwise, without the prior written permission of Technews Publishing (PTY) Ltd. Reg No. 2005/034598/07 ISSN 0256-8934

features

Published by:Technews Publishing (Pty) Ltd

1st Floor Stabilitas,

265 Kent Avenue, Randburg

Tel: (011) 543 5800

Fax: (011) 787 8052

dataweek@technews.co.za

Printed by: Intrepid Printers (Pty) Ltd

9 Grix Road, Willowton

Pietermaritzburg

GautengTracy Wolter

Tel: 011 543 5800

Fax: 011 787 8052

tracy@technews.co.za

KwaZulu-NatalJane van der Spuy

Tel: 031 764 0593

Fax: 031 764 0386

jane@technews.co.za

AdvertisingPrint and OnlineSales Manager: Malckey Tehini

malckey@technews.co.za

EditorBrett van den Bosch

B Eng (Electrical & Electronic)

brett@technews.co.za

Subscription servicesFor address changes,

subscriptions, renewal status

or missing issues call 011 543

5800 or fax 011 787 8052 or

subs@technews.co.za or

WRITE TO:Technews Publishing (Pty) Ltd,

Box 385, Pinegowrie 2123

Subscribe onlinewww.technews.co.za

Western CapeContact Durban or

Jhb numbers for details

2

3

4

6

28

32

33

contents

10

18

highlights . . .

www.dataweek.co.za

On the cover

Microchip’s article on page 18 details

how designers of high-quality audio

systems can benefi t from the digital audio

features built into the PIC32 family of

microcontrollers.

Fluke designs tools for profes-

sionals who take pride in their

work. The process begins with

observing and listening to its

customers, with the result that,

in this year alone, through its

distributor, Comtest, Fluke

has launched six new instru-

ments, with more to come. The

companies bring expertise in

power and indoor air quality

monitoring, process and predic-

tive maintenance with thermal

imagers, infrared thermometers

and vibration screening, help-

ing customers to accurately

troubleshoot problems, predict

failure and prevent costly

repairs.

For more information contact

Comtest, +27 (0)11 608 8520.

OnOnOnOOO ttttthhhhhe cover

With the buzz surrounding wireless charging

and the push to get the technology into

mainstream use, IDT has developed a high-

power transmitter/receiver combination

(page 8).

Dataweek, 25 July 2012 www.dataweek.co.za

News2 Electronics news digest

DisclaimerEditorial contributions relevant to the focus of this publication are welcome but the publishers cannot accept responsibility for unsolicited material. While every eff ort has been made to ensure the accuracy of the information contained herein, the publisher and its agents cannot be held responsible for any errors contained, or any loss incurred as a result. Articles published do not necessarily refl ect the views of the publishers. The editor reserves the right to alter or cut copy. Articles submitted are deemed to have been cleared for publication.

Advertisements, inserts and company contact details are printed as provided by the advertiser. Technews Publishing (Pty) Ltd can not be held responsible for the accuracy or veracity of supplied material.

• Cirtech has been awarded its ISO 9001: 2008

accreditation after passing the fi nal audit in

March this year. The set of standards dictates

specifi c levels that need to be maintained for

quality, environmental friendliness, safety, reli-

ability, effi ciency and interchangeability. This

news follows the recent awarding to Cirtech of a

BBBEE level one contributor certifi cate.

• Neil Kayton, former Chief Financial Offi cer

of Power Technologies (Powertech) has been

appointed Chief Executive Offi cer (CEO) of the

company eff ective from 1 July 2012. Kayton

succeeds Norbert Claussen, who recently

stepped down from the position after 16 years

with the group.

• Mantech has bolstered its stockholding by

purchasing Avnet Kopp’s locally warehoused

stock. This deal came about as a result of Avnet’s

centralisation of its warehouse in Europe. The

stock in question includes passives, integrated

circuits, microprocessors, enclosures, transform-

ers, switches and more. Mantech expects the

deal to increase its already wide product off er-

ing which now stands at close to half a million

product lines. The products will be available

through its three South African distribution cen-

tres, and could benefi t existing users of these

components in industry where off -the-shelf

stock is required.

• The long-standing distribution partnership

between TTI and TRX has been made offi cial,

with the companies having signed an agree-

ment that gives TRX access to TTI’s entire range

south africa

overseas

of products, including passive components,

discretes, connectors, relays and switches.

Business

• XP Power announced the opening of its new

manufacturing facility in Ho Chi Minh City,

Vietnam. The factory, believed to be the world’s

most environmentally advanced power supply

manufacturing facility, incorporates the use of

sustainable power and energy effi ciency; the

building’s highly insulated construction includes

the use of a solar PV array, eco-friendly glass

to prevent solar heat gain, low-energy lighting

and ultra-effi cient air conditioning. In addition,

rainwater is captured for use in the building and

for irrigation systems. All the environmental fea-

tures are managed by a fully integrated building

management system.

• Xilinx announced its intention to invest

$50 million in an expansion of its electron-

ics engineering operations at the company’s

EMEA headquarters in Dublin and Cork, Ireland.

The company will recruit 45 senior silicon and

electronics engineering staff for its regional

headquarters in Dublin and for its engineering

centre in Cork. In addition, a further 15 staff

will be hired across a broad range of disciplines

supporting the continued growth of Xilinx in

Ireland.

Companies

• Aeroflex / Metelics announced a strategic

alliance conveying a licence for proprietary

information with Picosecond Pulse Labs for

the manufacture and design of a family of

non-linear-transmission-line based comb gen-

erators. Under the terms of the agreement,

Aeroflex / Metelics will produce the prod-

ucts in its Sunnyvale, California facility

and supply parts both to Picosecond for its

internal use, and to serve the test and measure-

ment industry.

• NXP Semiconductors announced the

appointment of Sean Hunkler as executive vice

president and co-general manager of opera-

tions. The company also announced that

René Penning de Vries has stepped down as

member of the NXP management team and

as CTO. Hans Rijns, senior vice president of

research and development, will take over as

interim CTO.

• Sierra Wireless has entered into an exclusivity

agreement for the acquisition of Sagemcom’s

machine-to-machine (M2M) business. Sierra

Wireless has submitted a binding off er to

acquire substantially all of the assets of the busi-

ness for 44,9 million Euros in cash consideration

plus assumed liabilities.

• u-blox announced the acquisition of Cognovo,

a company specialising in software defi ned

modem chip development technology. The

deal gives u-blox access to 4G chip technol-

ogy and enables it to address new markets for

high-speed LTE modems. u-blox acquired 100%

of the shares of Cognovo at a price of $16,5

million.

• Ramtron’s board of directors has rejected

Cypress Semiconductor’s off er to acquire all

outstanding stock of Ramtron at a price of $2,68

per share in cash, stating that, in consultation

with its fi nancial and legal advisors, the board

“unanimously concluded that the tender off er is

inadequate, does not refl ect the intrinsic value

of the company, and is not in the best interest of

Ramtron and its stockholders.”

• Cadence Design Systems has acquired

Sigrity, a signal and power integrity technology

provider. Sigrity provides gigabit signal and

power network analysis technologies, including

a unique power-aware signal integrity analysis

capability for system, printed circuit board (PCB)

and IC package designs. The acquisition was

made at a purchase price of approximately $80

million.

• Micron Technology and Elpida Memory have

signed a defi nitive sponsor agreement for

Micron to acquire and support Elpida. Under

the agreement, approximately $2,5 billion, less

certain reorganisation proceeding expenses,

will be used to satisfy the reorganisation claims

of Elpida’s secured and unsecured creditors.

Micron will acquire 100% of the equity of Elpida

for roughly $750 million.

Industry

• The Semiconductor Industry Association (SIA)

announced that worldwide sales of semicon-

ductors reached $24,4 billion for the month

of May 2012, a 1,4% increase over the prior

Dataweek, 25 July 2012www.dataweek.co.za

News 3

EVENTSmonth’s fi gure of $24,1 billion. Combined global

sales for the March-April-May period increased

over December-January-February by the high-

est rate (6,4%) since June 2010. Additionally,

May marked the third consecutive month that

sales have grown over the previous month – the

longest streak of sequential monthly growth

since September 2010. However, sales from May

2012 were 3,4% lower than the May 2011 total

of $25,2 billion, and 2012 year-to-date sales

were lower across all regions than at this time

last year.

• Following a spectacular recovery in 2010, the

power semiconductor discrete and module

market grew by a more modest 9% in 2011,

according to a new report from IMS Research.

(NYSE: IHS). Total power semiconductor reve-

nues grew to just under $18 billion in 2011. The

market for power modules grew by 32%, much

The DM330014 kit enables

designers to quickly leverage

the capabilities and perform-

ances of the dsPIC33F GS

series of digital signal control-

lers (DSCs) to develop LED

lighting products.

The DSC and kit allow

developers to create a 100%

digitally controlled ballast

function, while including advanced features

such as dimming and colour hue control.

The kit includes an LED baseboard with an

onboard dsPIc33FJ16GS504, a DC-DC buck

daughter board and a DC-DC boost daughter

board.

LED lighting applications supported by

the development kit include dimmable LCD

backlighting, signage, LED replacement of

LED lighting development kit up for grabs

PIC24 hands-on workshop

3 August 2012Military Museum, Johannesburg

A hands-on training course based on

Microchip’s 16-bit PIC24 platform; suitable for

students and professionals alike.

Email arnold.perumal@avnet.eu to register

Microchip MASTERS 2012 Conference

6 – 11 August 2012Arizona, USA

Technical training for embedded systems

design engineers, with numerous topics

taught by Microchip product developers and

technical experts.

Register at www.microchip.com

NEPCON South China

28 – 30 August 2012Shenzhen, China

Exhibition showcasing electronics manu-

facturing technologies; co-located with ATE

Register at www.nepconsouthchina.com

Electronica India & Productronica India

11 – 13 September 2012Bangalore, India

The leading trade fair for the electronics

industry in India and South Asia, covering

components, manufacturing technologies

and much more.

Register at www.electronica-india.com

European Microwave Week

28 October – 2 November 2012Amsterdam, The Netherlands

Incorporates three major conferences cover-

ing microwave ICs and technology, wireless

technology and radar. Includes workshops

and seminars.

Register at www.eumweek.com

faster than that for discrete power semiconduc-

tors, which grew only 3%. Infi neon remained

the leading supplier of power discretes and

modules, while other manufacturers whose

share of the market grew included power

module specialists Mitsubishi Electric, Fuji

Electric and Semikron.

Technology

• Audi is using Sierra Wireless AirPrime embed-

ded wireless modules to help develop and test

an LTE-connected infotainment system. Still in

the development and testing stage, the system

has been showcased at several industry events

this year. Sierra Wireless AirPrime modules are

available with varying combinations of LTE

frequency bands and fallback technologies, and

were the fi rst LTE modules available, allowing

Audi to work on the leading edge of wireless

technology as LTE networks roll out.

Dataweek is off ering

readers the chance to

win a Microchip LED

lighting development kit.

fl uorescent tubes and incandescent bulbs,

architectural lighting and automotive lighting

applications. Automotive lighting prod-

ucts include exterior applications, such as

headlights, daytime running lights and signal

lights.

For more information visit

www.microchip-comps.com/dw-led

Dataweek, 25 July 2012 www.dataweek.co.za

News4

LITERATURE SHOWCASE

Contact Technews to place your literature

in the Literature Showcase

IP&E Component Guide

Avnet Kopp has released its 2012 IP&E components

guide containing an overview of passive, electro-

mechanical, interconnect and enclosure products

available from the company’s suppliers. The guide

includes colour photos and performance sum-

maries of all products, as well as line cards for

the various product types. An electronic ver-

sion can be downloaded at www.avnet.co.za/

Download/2012_IPE_Components_Guide.pdf.

Avnet Kopp, +27 (0)11 809 6100

u-blox has released its Product Catalogue 13,

containing the latest updates to all its products

and services, including the latest generation

of low-power GNSS receiver products based

on the new u-blox 6 platform, plus compact,

cutting-edge wireless communication

modules and reference designs implement-

ing 2G and 3G mobile communication

technologies.

RF Design +27 (0)21 555 8400

u-blox Product Catalogue

ts

Interconnect Solutions Catalogue

Samtec’s new 2012 Interconnect Solutions

Catalogue includes a number of new products,

new high-speed test data and updated specifi ca-

tions on existing products, as well as an overview

of Samtec’s ever-expanding RF product line. In

addition, the catalogue highlights expanded

product features such as lower-cost plating

options, new rugged screw-down options for

board-to-board and discrete wire systems,

and latching options for ultrafi ne-pitch

connectors.

Otto Marketing, +27 (0)11 791 1033

Skiff y Catalogue No. 65

The latest version of Skiff y’s catalogue includes

information on the wide range of plastic and

metal parts supplied by the company, such as

nuts, washers, spacers, fasteners, caps and plugs.

Among the new products featured are PCB spac-

ers, knobs and handles, shrink tubes, grommets,

cable glands and clamps.

Skiff y, +27 (0)11 314 8750

Samte

Catalo

new

tion

of S

add

pr

o

b

a

O

Skiff

The l

info

me

nu

Am

e

c

IP&E Co

Avnet K

guide

mech

avail

incl

ma

th

s

D

Avnet Kopp, +

N 65

Schroff Innovations Brochure

This brochure details some of Schroff ’s new

developments in the areas of subracks,

card locks, systems and cabinets. Colour

photographs and product specifi cations

are provided.

Actum Electronics, +27 (0)11 608 3001

Small-scale renewable energy solutions, produced from zero- or low-carbon

source technologies and derived from naturally regenerative or practically

inexhaustible resources, are the latest addition to the list of energy-effi cient

technologies funded through Eskom’s Standard Off er Programme.

Introduced in consultation with NERSA as a pilot project with total capacity

achieved during this phase limited to 10 MW, Eskom views small-scale renew-

able energy solutions as a way to further broaden the scope of optimal energy

usage in South Africa through its range of energy-effi ciency rebate programmes.

For this purpose, renewable energy sources include biomass waste, wind

energy, and geothermal, solar, thermal gradient and ground source heat.

Solar thermal liquid heat is excluded from the programme but municipal solid

waste will be considered a source. Small-scale renewable energy solutions will

be limited to an installed capacity range of between 10 kW and 1 MW.

Customers or project developers wishing to apply for funding must

submit electrical and structural designs certifi ed by qualifi ed engineers as

part of the application, and a letter from the electricity supplier (Eskom or a

municipality) confi rming knowledge and acceptance of the proposed

network connection, and registration of the project as per NERSA requirements.

To qualify for funding, equipment should be installed on the energy user’s

side of the meter; systems should be new and in compliance with perform-

ance and safety standards applicable to all components of the installation.

In addition, standalone systems not tied to the grid must replace existing

national electricity connections to substitute energy that would have been

drawn from the grid. Grid-tied systems must comply with all regulatory

and embedded generation interconnection requirements; a letter from the

relevant local authority or electricity utility will be required as part of the

application to confi rm the applicant’s knowledge and acceptance of the

proposed connection onto the network.

Initially, funding will only be available for systems that do not feed elec-

tricity onto the grid at a rate of R1,20 per kWh, and be paid as per the usual

Standard Off er Programme milestone payments based on the agreed rate,

and the total savings delivered over the contract period of three years.

For more information visit www.eskomidm.co.za/industrial

Eskom promotes small-scale renewable energy

Ember is the latest manufacturer to board the ZigBee Light Link band-

wagon. The company’s ZigBee platform – chips, software and tools – has

achieved Golden Unit certifi cation from the ZigBee Alliance for the newly

ratifi ed Light Link standard.

ZigBee Light Link is a wireless interoperability standard for consumer

LED lighting and control products. It allows consumers to wirelessly gain

control over their LED fi xtures, light bulbs, timers, remotes and switches.

The standard ensures that ZigBee certifi ed products from diff erent vendors

will easily connect into a single network, without the need for additional

devices to coordinate the network.

The standard will also give consumers the ability to control lighting

remotely over the Internet, providing yet another avenue for managing

energy use and making their homes greener. In addition, the standard

enables lighting manufacturers to target the broader ZigBee home automation

market through the capability to integrate with other ZigBee standards.

As one of the fi rst fi ve Golden Units certifi ed by the ZigBee Alliance,

Ember’s ZigBee platform, including EM300 series wireless chips, EmberZNet

PRO networking software and development tools, will serve as a preferred

development platform to build and test future ZigBee Light Link products.

For more information contact Michelle Parreira, CST Electronics,

+27 (0)11 608 0070.

More ZigBee Light Link ‘golden units’

Dataweek, 25 July 2012www.dataweek.co.za

News 5

Local electronic research and development house

Keystone Electronic Solutions has been awarded

the contract to provide its Remote Site Monitor

solution for all DFA monitoring facilities in order

to provide proactive monitoring of the cable on

DFA’s new generation networks (NGN).

Jacques van Loggerenberg, DFA technical execu-

tive, comments: “DFA has a build, own and maintain

model whereby it is responsible for deploying the

fi bre infrastructure, for our clients to light up. The

initial build was focused around metro build in

Johannesburg, Pretoria, Cape Town and Durban.

“We then completed two long haul links

from Mtunzini to Pretoria and Cape Town to

Yzerfontein. We now cover all three international

cable landing stations through Mtunzini (EASSy,

Seacom and SAFE), Melkbos Strand (SAT2, SAT3

and SAFE) and Yzerfontein (WACS).

“Last year we initiated an NGN project where

we evaluated all the existing building blocks we

were deploying on the network. The project was

part of the DFA Total Quality Management Pro-

gramme (TQM) whereby products get evaluated

on a continuous basis to ensure that the best prod-

uct is used for immediate and future requirements.

“One of the new building blocks identifi ed

in this process was the need for a Primary Cross

Connect (PCC) facility. The purpose of the primary

cross connect is to partition and provide fl exible

connectivity between the metro core network

and the long haul fi bre network.”

He continues: “These PCC nodes will be housed

in containers or building locations, and will cater

for up to 4320 fi bre connections at a single node.

These facilities will also be used to house the

remote testing units (RTUs) which are used for

proactive monitoring of the DFA fi bre cables.

“Because monitoring and security are of the

utmost importance to DFA it was decided that we

required a remote site monitoring solution for the

PCC sites. Although we had used monitoring equip-

ment before, we needed to integrate the best prod-

uct for our specifi c requirements on this project.

“We developed a list of requirements, and

together with an external consultant for objectiv-

DFA chooses Keystone’s remote site monitor

ity, we eventually shortlisted three products, one

of which was the Remote Site Monitor developed

locally by Keystone. We installed the products

in a test environment in one of the facilities, and

then began stringent testing,” he explains.

One of the criteria was fast and effi cient

local support, and so the consultant requested

changes and additions to the functionality on a

regular basis, in order to assess how the compa-

nies would handle rapid and diverse changes.

Keystone was able to turn around most of the

changes within 6-8 hours.

The fi ndings of the testing were documented

and evaluated according to the predetermined

criteria. Keystone scored the highest and was

selected to provide its monitoring solution for the

DFA PCC monitoring facilities rollout. In fact, the

Keystone test site was able to go live onto the net-

work immediately as the fi rst monitored facility.

John Eigelaar, Keystone technical director,

explains the solution: “The Remote Site Monitor

consists of modular rack-mounted hardware as

well as a server-based management platform that

can be accessed remotely through a web-based

interface. The geo-referenced interface plots each

facility’s location and status on a map of South

Africa, and immediately refl ects any critical change

in the status of the facility on the icon on the map.”

For the DFA installations, the RSM monitors

access, power and environmental factors such

as air temperature, movement and humidity.

This enables DFA to have full monitoring capa-

bility to all its PCC facilities.

Says van Loggerenberg: “One of the advan-

tages of the Keystone solution was that it easily

integrated with our existing network software.

We had used Keystone previously to provide

other network management solutions, so it was

really just a case of expanding the functionality

of the existing network software.

“Another bonus was that unlike other prod-

ucts, Keystone was able to provide us with only

those modules that we required. Other products

came as a bundled package, so would have cost

a lot more, and come with a range of features

we didn’t require. And additionally, we have

confi dence that Keystone can develop modules

for any future requirements that we may have,

making it totally scalable,” he concludes.

For more information contact Keystone

Electronic Solutions, +27 (0)12 460 4135.

Jacques van Loggerenberg (DFA), John Eigelaar (Keystone) and Derek Koekemoer (DFA).

Dark Fibre Africa (DFA), founded in 2007,

is an open access, carrier neutral dark

fi bre provider to all licensed entities. It

currently has more than 6000 km of duct

route installed across South Africa.

Dataweek, 25 July 2012 www.dataweek.co.za

News6 regular Systems, components, design

Premier Farnell has extended its relation-

ship with ARM through the launch of

the latest series of ARM Cortex-M3 and

Cortex–M4 processor-based STM32F2xx

and STM32F4xx development kits.

The deal will see the kits shipped with

Keil MDK-Lite development tools and the

ULINK-ME debug adaptor. Developers will

benefi t from support by Premier Farnell’s

ARM trained engineers, the element14

ARM development tools group and rich

technical content, video support materi-

als, and design examples on the Knode

on element14.

STM32 development kits

The MCBSTM32F200 board features

an STM32F207IG ARM Cortex-M3 based

microcontroller which off ers 1 MB Flash

and 128 KB RAM on-chip memory, and

the MCBSTM32F400 features an STM-

32F407IG Cortex-M4 based microcon-

troller with 1 MB Flash and 192 KB RAM

on-chip memory.

In addition, both evaluation boards

are populated with 8 MB NOR Flash,

512 MB NAND Flash, 2 MB SRAM, and

8 KB I2C EEPROM with an NFC interface.

They provide a colour QVGA LCD and

interfaces for SD card, CAN, 10/100 Ether-

net and Full/High-speed USB. The boards

also feature a 3-axis digital accelerometer,

a 3-axis digital gyroscope, a digital micro-

phone and a digital VGA camera.

Also featured are a dedicated audio PLL

and two full-duplex I2S. The kits have up

to 15 communication interfaces (including

six USARTs running up to 10,5 Mbps,

three SPI running up to 42 Mbps, three

I2C, two CAN, SDIO). Analog functions

include two 12-bit DACs, three 12-bit

ADCs reaching 2,4 MSps or 7,2 MSps in

interleaved mode. The boards have up to

17 timers, 16-bit and 32-bit running up to

120/168 MHz.

For more information contact

Electrocomp Express, 0860 10 20 20.

It off ers high linearity with

25,9 dBm IIP3 (input third-order

intercept), high conversion gain

of 7,4 dB owing to its integrated

IF amplifi er, and a noise fi gure of

11,3 dB.

These performance character-

istics enable excellent dynamic

range performance for both main

receivers and digital pre-distortion

receivers. Moreover, the IC’s RF

input can withstand strong interfer-

ence signals without signifi cantly

degrading its noise fi gure, thus

preserving receiver sensitivity and

enhancing robustness.

The LTC5544’s 6 GHz reach is

ideally suited for wireless systems

such as point-to-point broadband

microwave links, 5 GHz licence-free band

WiMAX radios, satellite receivers, radar systems,

avionics, public safety radios and RF test

systems.

The device incorporates an on-chip, high-

gain IF amplifi er capable of supporting IF

frequencies up to 1 GHz. Also integrated are an

LO buff er along with two RF balun transformers.

They enable the RF and LO inputs to operate

single-ended and with 50 Ω matching. The LO

Down-converting RF mixerLinear Technology announced the LTC5544, a 4 GHz to 6 GHz, high dynamic

range down-converting mixer.

input requires only +2 dBm drive level.

The chip operates on a single 3,3 V supply,

drawing 194 mA of current. It has a power-down

feature with 0,6 μs turn-on and turn-off time,

supporting Burst Mode operation or TDD (time

division duplexed) systems. When disabled, the

device consumes a maximum of 500 μA.

For more information contact Arrow Altech

Distribution, +27 (0)11 923 9600.

Texas Instruments has

expanded its award-

winning WEBENCH suite

of tools by introducing

System Power Architect,

an online tool to speed

design of multi-output,

high-performance

DC-DC power supply

systems.

The tool enables the

user to build and opti-

mise a complete power

system from the edge of

the card to the smallest

point-of-load. In addi-

tion to the design and

simulation of individual

point-of-load converters,

it adds the ability to accommodate multiple

intermediate bus architectures and safety isola-

tion in conjunction with TI’s hot-swap control-

lers and isolated power products.

With the System Power Architect, TI claims

designers can create isolated power systems

with hot swap capability in just about 60

seconds. When driving multiple loads, the

designer can customise output voltages and

Power supply design tool

load currents, and also select FPGAs and pro-

cessors, which are preconfi gured with critical

requirements, load grouping and sequencing.

Other options to select include ripple fi lter,

low dropout regulators (LDO), external sync or

separate supply.

For more information visit www.ti.com/

webenchspa-pr-eu

on element14.

Dataweek, 25 July 2012www.dataweek.co.za

News 7Systems, components, design

Nordic Semiconductor announced the fi rst members of its new nRF51

series of ultra-low power (ULP) RF ICs that feature a new multi-protocol

2,4 GHz radio and a 32-bit ARM Cortex-M0 based processor.

Nordic’s novel software architecture for Bluetooth low energy and

ANT system-on-chip (SoC) solutions frees designers from the integra-

tion eff ort, complexities and restrictions of chip vendor-supplied

software frameworks and instead allows customers to develop their

designs quickly and easily using the highly popular ARM Cortex pro-

gramming environment.

This is achieved by using a new software architecture featuring

a unique and powerful separation between protocol stack and user

application code. This separation provides developers a clean bound-

ary between application and protocol stack, and removes the need

to struggle with integration of application code as part of a vendor-

imposed application development framework.

Multi-protocol 2,4 GHz SoCs

These GaAs pHEMT MAPS-010145 and MAPS-010165 are 4- and 6-bit

packaged devices for use in C-band applications from 3,5 to 6,0 GHz.

They maintain low phase error and minimal attenuation variation over

the full 360 degrees and frequency range.

The MAPS-010145 has a built-in CMOS driver that allows for serial

or parallel control and low power consumption. The 360 degree phase

shift is broken down to steps of 22,5 degrees. It off ers insertion loss of

5,0 dB with a ±1,0 dB attenuation variation and high phase accuracy

over the operating band.

The MAPS-010165 also has a built-in CMOS driver and is capable of

either serial or parallel operation, while the 360 degrees of phase shift

is divided into 5,6 degree step sizes. It has 5,5 dB of insertion loss with

a low ±0,8 dB attenuation variation and high phase accuracy over the

entire band.

The phase shifters operate using both positive and negative voltage

supply. They are packaged in industry standard 4 mm, 24-lead PQFN

style packages. They are intended for use in commercial weather radar

and military phased array radar applications.

For more information contact M/A-COM Technology Solutions,

+27 (0)11 908 6679.

Digital phase shiftersM/A-COM Technology Solutions has added two new

members to its digital phase shifter product family.

The fi rst two Flash-based ICs to début in the new series are the

nRF51822 multi-protocol Bluetooth low energy / 2,4 GHz proprietary RF

SoC, and the nRF51422 ANT SoC.

The new chips provide -92,5 dB receive sensitivity in Bluetooth low

energy mode, and up to +4 dBm output power in all modes. They

feature sub-10 mA peak currents running off a 3 V coin cell battery.

For more information contact Andrew Hutton, RF Design, +27 (0)21 555 8400.

Dataweek, 25 July 2012 www.dataweek.co.za

News8 regular Systems, components, design

Hittite has expanded its clocks and

timing product portfolio to include the

HMC1031MS8E, a low-power integer-N PLL

clock generator. This new device is ideal as

a building block for jitter attenuation, OCXO

and TCXO frequency multiplication and low-

power VCXO-based clock generators driving

data converters (ADCs/DACs), FPGAs and

SERDES applications.

The chip includes a phase/frequency de-

tector with a charge pump current output

and a low-power integer-N divider resulting

in reference frequency multiplication ratios

of x1, x5 and x10. Together with an external

loop fi lter and a VCXO, VC-TCXO or OCXO,

the HMC1031MS8E forms a complete clock

generator solution targeted for jitter clean-

up and reference clock multiplication /

generation applications.

The IC may be used with a low phase

noise VCXO to multiply 10 MHz TCXO or

OCXO sources to 50 MHz or 100 MHz with

excellent long-term stability. Such a topol-

ogy reduces system cost and improves

control over the design because high-

frequency OCXOs and TCXOs may be more

expensive, have higher ageing rates and

degraded phase noise compared to their

low-frequency 10 MHz counterparts.

PLL clock generator

Additional applications of the HMC-

1031MS8E may include low-frequency clock

jitter attenuation and clock frequency trans-

lation or synchronisation for 25 MHz and

125 MHz Ethernet clocks and 27,000 MHz,

74,250 MHz and 74,1758 MHz video clocks.

The integrated phase detector and

charge pump are capable of operating up

to 140 MHz which means the designer can

eliminate noise-generating divider circuitry,

further reducing cost; additionally, VCXOs as

high as 500 MHz can be used.

The low-power architecture of the de-

vice consumes less than 2 mA under typical

conditions, making the device well suited

for power-sensitive battery powered and

portable applications. Additional features

include integrated lock detect indicator on

a dedicated hardware pin, and a built-in

power-down mode for further power

saving.

For more information contact Andrew

Hutton, RF Design, +27 (0)21 555 8400.

The ADN469xE series

includes eight transceivers,

each of which is capable of

connecting 32 data/clock nodes using a single

diff erential cable pair at data rates of 100 Mbps

or 200 Mbps. Traditional LVDS communication

links, by comparison, must use 32 separate

point-to-point nodes, which signifi cantly

increases power, connector size, cabling costs

and total board space.

The devices deliver 8 kV IEC ESD protection

to improve the reliability of insertable/

removable boards and cards used in wireless

base station and networking infrastructure, data

acquisition, automated test equipment and

other high-speed, highly networked backplane

and cable applications.

Available in half- and full-duplex versions,

the transceivers are fully compliant with the

Multipoint LVDS transceivers

TIA/EIA-899 M-LVDS standard. The ADN4691E/

ADN4693E are Type 1 receivers with 200 Mbps

data rates that include symmetrical input

threshold voltage and are designed to improve

clock timing performance.

The ADN4696E/ADN4697E Type 2 receivers

with 200 Mbps data rates are designed for

data applications and feature a receiver-input

threshold voltage off set of -100 mV. This ensures

a known receiver output state when the bus is

idle (bus-idle failsafe) or the bus inputs are open

(open-circuit fail-safe).

For more information contact

Arrow Altech Distribution,

+27 (0)11 923 9600.

The IDTP9030 and IDTP9020 comprise a

wireless power transmitter and receiver solu-

tion designed to meet the Wireless Power

Consortium (WPC) Qi standard, which ensures

interoperability with any other device meeting

the standard.

Both transmitter and receiver are capable

of multi-mode operation, supporting both the

Qi standard as well as proprietary formats for

added features, improved safety and increased

power output capability. Built-in protocol

detection enables dynamic switching between

Qi and proprietary modes, creating seamless

transitions and a trouble-free user experience.

These devices can be used in a myriad of mobile

applications for convenient battery charging.

The IDTP9020 is a high-effi ciency wireless

power receiver delivering up to 5 W in Qi mode.

When used with the IDTP9030 transmitter

in a proprietary confi guration, it can deliver

up to 7,5 W. The devices also feature propri-

etary multi-layered foreign object detection

Wireless charging chipsetIDT has developed a single-chip wire-

less power transmitter accompanied

by a high-output-power single-chip

receiver solution.

(FOD), utilising sophisticated multi-parameter

algorithms to assure a high level of safety while

avoiding FOD false alarms.

The solution features over-temperature,

over-voltage and over-current protection. In

addition, the receiver includes support for

charging the battery via USB cable when a wire-

less charging station is not available, eliminating

the need for external USB adaptor switches in

mobile devices.

For more information contact Marian

Ledgerwood, Future Electronics,

+27 (0)21 421 8292.

Recently introduced

by Analog Devices is

a series of multipoint,

low-voltage, diff erential

signalling (M-LVDS)

transceivers with

the highest ESD

(electrostatic discharge)

protection of any

multipoint LVDS

transceivers.

Dataweek, 25 July 2012www.dataweek.co.za

News 9regularSystems, components & design

The MCUs are system-on-chip devices based on

the Cortex-M4 core with FPU, and are optimised for

effi cient handling and processing of mixed signals in

circuits such as three-phase motor controls, biometrics

and industrial sensor outputs or audio fi lters. They are

intended to simplify design, cut power consumption

and reduce PCB size in consumer, medical, portable fi t-

ness, system monitoring and metering applications.

The new F3 series extends the scope of ST’s STM32

family by allowing designers to tackle mixed-signal

control applications. Featuring versatile analog

circuitry, the series provides low to medium memory

sizes. The Cortex-M4 with FPU core augments the

proven Cortex-M3 CPU with additional digital signal

processing (DSP) capability, optimised single-cycle and

saturating arithmetic instructions, and fl oating-point

unit (FPU), to boost performance over ST’s STM32 F1

Cortex-M3 devices.

Its features position the new series between the

STM32 F1 and the STM32 F4, which features the

Cortex-M4 with FPU core and large memory capacity

combined with high-speed 168 MHz maximum CPU

frequency for complex applications.

With the new F3 microcontrollers, ST’s STM32

family now off ers more than 350 variants addressing

applications from entry-level, price-sensitive designs

through to those demanding high performance and

rich on-chip features. The F3 family comprises four

product lines, diff erentiated according to on-chip

memory density and integrated peripherals.

STM32 F30x devices have seven fast comparators

with 50 ns response time, four programmable gain

amplifi ers supporting four diff erent gain settings with

1% accuracy, two 12-bit DACs, and four 12-bit, 5 MSps

ADCs. ADC performance can extend up to 18 MSps in

interleaved mode.

The devices also have two advanced motor-control

timers running at up to 144 MHz. These allow one

microcontroller to manage two motors plus power

factor correction (PFC) via its on-chip analog peripher-

32-bit ARM MCUsSTMicroelectronics is now delivering the fi rst samples of its new STM32 F3 series of micro-

controllers, enabling early evaluation of this new line in ST’s product family featuring the ARM

Cortex-M core embedded.

als, in equipment such as home appliances. Among

other capabilities of the advanced timers, timing

resolution better than 7 ns allows applications such as

digital power supplies in telecom infrastructure or data

servers, micro-inverters in solar installation and LED

lighting.

To help developers optimise code execution time,

the F30x devices have 8 KB of SRAM mapped to the

instruction bus as Core Coupled Memory (CCM-SRAM).

Critical routines loaded in the 8 KB CCM at startup

can be completed at full speed with zero wait states,

achieving 94 Dhrystone MIPS at 72 MHz. This compares

with 62 DMIPS when executing from Flash or SRAM;

equivalent to a 52% performance increase for critical

routines.

The F30x series also has up to 40 KB of SRAM

mapped on the data bus. The 8 KB CCM-SRAM can

also be used for data storage with no performance

trade-off . Devices are pin-compatible with the STM32

F1 series, and also share the same API, which simplifi es

migration to higher-performing devices.

The F37x devices deliver a diff erent blend of

peripheral functions. They are the fi rst microcontrollers

from ST to integrate a 16-bit Sigma-Delta ADC on-chip,

allowing the STM32 family to solve a wider range of

precision-sensing applications and providing a single-

chip solution capable of replacing a discrete general-

purpose processor and separate ADC.

Up to three 16-bit Sigma-Delta ADCs are integrated

on-chip, with separate analog supply from 2,2 V to

3,6 V, up to 21 single or 11 diff erential channels, and

seven programmable gains per channel.

Other STM32 F37x analog peripherals include two

fast comparators, three 12-bit DACs and one 12-bit,

1 MSps ADC. There is also an enhanced CEC (Consumer

Electronics Control) unit, making these devices relevant

for consumer multimedia products.

For more information contact Avnet Kopp,

+27 (0)11 809 6100.

Dataweek, 25 July 2012 www.dataweek.co.za

News10 feature Test & measurement, analysis, calibration & compliance

Using a particle counter is relatively simple;

however, understanding the features

that distinguish counters can sometimes

be a challenge. The following terms are

commonly used to describe the accuracy,

effi ciency and other attributes of an air-

borne particle counter.

Count mode: The count mode defi nes

how the particle counter displays data to

the user. Concentration and raw count are

two typical sampling modes, and the Fluke

985 adds an audio mode. Concentration

mode samples a small volume of air then

calculates the value based upon the volume

setting (cm3, ft3 or litre) in the counter while

raw count mode allows the user to collect

actual particle counts as they accumulate

per sample time.

It can be set to cumulative (the total

number of particles > each channel size)

or diff erential (the number of particles fall

between the channel sizes). Audio mode is

useful when searching for areas that exceed

predefi ned particle levels. Once exceeded,

the counter notifi es the user in audio alarm.

The count mode defi nes how the particle

counter displays data to the user.

Zero count: Zero count is a measure of

the particle counter’s accuracy, and should

be taken prior to use and periodically there-

after, or when sampling error is suspected.

The zero count fi lter is attached to the par-

ticle counter per the manufacturer’s instruc-

tions, then the counter is run for 15 minutes.

The counter should not have detected more

than one particle greater than 0,3 μm in a

fi ve-minute period.

Coincidence loss: Coincidence loss occurs

when two particles cross the counter’s light

beam simultaneously, creating a single

pulse and resulting in a single particle

count. This type of error occurs more fre-

quently as the concentration of particles

increases within the sample. Per 21501-4,

coincidence loss must be ≤10% by the maxi-

mum particle number concentration. Fluke

985 is 10% at 4 000 000 particles per ft3.

Counting effi ciency: The probability that

the counter will sense and count a par-

ticle passing through the sample volume.

Counting effi ciency is a function of size up

to a minimum sensitivity threshold, above

which all particles are sensed and counted.

A counting effi ciency of 50% at the most

sensitive threshold is typically considered

optimal, and facilitates consistent compari-

sons between counts from OPCs and those

of higher-resolution instruments.

Sensitivity: A device’s ability to detect

small particle sizes at a certain counting

effi ciency. The Fluke 985 detects 0,3 μm at

50% counting effi ciency.

Calibration: A set of operations or ac-

tions taken to establish the relationship

between the measured values obtained

through a device and the values of the

corresponding parameters as defi ned in a

standard. The Fluke 985 is calibrated using

PSL (polystyrene latex) spheres, widely

used due to their uniform size and light-re-

fraction properties and meets ISO 21501-4:

Light scattering airborne particle counter

for clean spaces.

NIST traceable: Traceability is a charac-

teristic of a measurement or standard and

its relationship to stated references, which

are often national or international stan-

dards. The PSL spheres used in the Fluke 985

calibration process can be traced to NIST (US

National Institute of Standards and Technol-

ogy) standards.

Making sense of particle counter features

Leaving the system off until occupants arrive

in the morning might save money, but allows

pollutants to build up in the air. If an HVAC

system is not properly maintained and the air

fi lter checked, it could lead to IAQ problems. The

eff ects of IAQ have been highlighted in numer-

ous environmental risk reports and joint agency

eff orts by the US Environmental Protection

Agency (EPA). Further studies indicate that

indoor air in some commercial buildings is up to

fi ve times more polluted than outdoor air.

Airborne particulates come in a variety of

forms ranging from animal dander, plant pollen

and airborne bacteria, to fi bre-glass, asbestos

and combustion particles that are generated

from part of the production equipment or

process. To properly identify and troubleshoot

IAQ problems, the technician needs a tool that

not only reads particle concentrations for spot

check, but also provides continuous monitoring

for process control.

Why particle counts matter

Diff erent locations have varying levels of accept-

able particulate concentrations. In a residential

and commercial environment (i.e. homes, offi ces,

Tips for monitoring indoor air qualityInformation from Fluke.

Energy savings are top of facilities owners’ agendas, and HVAC systems tend to be the largest contributor to the energy bill. Tax

rebates are often based on energy savings per square metre, but savings on HVAC energy costs can cause signifi cant indoor air

quality (IAQ) issues due to improper ventilation, temperature and humidity control.

hotels), health and comfort concerns and fear of

litigation often drive IAQ investigations.

In industrial and institutional environment

(hospitals, food and beverage plant, electronic

and precision manufacturing), energy cost,

contamination control and production yields

are the primary concern. Excessive levels can

result in medical conditions such as Sick Building

Syndrome, lower productivity, contaminated

product or all of the above.

Maintaining acceptable air quality levels not

only lowers the costs associated with downtime,

but also reduces or removes costs associated

with expensive fi xes in the future. The fi rst step

in establishing an IAQ maintenance programme

is to determine if a problem currently exists.

IAQ investigation

An IAQ investigation is the fi rst step in an ongo-

ing maintenance programme or in responding

to complaints potentially associated with air

quality. No matter if it is a commercial/residen-

tial or industrial/ institutional environment, the

methodology is similar:

1. Conduct a survey of credible staff at the facil-

ity. Who has fi led complaints, and what are

the symptoms? Are those who complained

centrally located, or dispersed throughout

the facility? The purpose is to gauge the level

of toxicity or concentration as evidenced by

allergic reactions or irritation in human com-

fort or increased contamination in product.

2. Research the building’s history. When was the

building built and/or remodelled? Have there

been any instances of excessive damage,

and how were the repairs carried out? What

Dataweek, 25 July 2012www.dataweek.co.za

News 11Test & measurement, analysis, calibration & compliance feature

are the maintenance practices

within the facility? For exam-

ple, roof or plumbing leaks

may have been repaired, but

the underlying water damage

may not have been addressed.

3. Perform a physical inspec-

tion. Technicians need to be

aware of their test environ-

ments by considering harmful

particulate sources. Within a

given location, areas contain-

ing exhaust vents, furnaces,

cleaning supplies, as well as

areas with fresh paint and/or

carpeting must be accounted

for, especially if present within

a complaint area. Are there any

odours or visible sources (i.e.

mould)?

4. Take air quality measurements.

When conducting a complete

IAQ investigation, tempera-

ture, humidity, CO and CO2

readings should also be taken

to identify problems related to

inadequate and/or contami-

nated ventilation, potentially

creating a particulate problem.

For example, temperature and humidity

readings play a key role in identifying mould

and bacteria. A location with high relative hu-

midity and higher concentrations of particles

3,0 μm or larger may indicate the presence

of mould spores, which should be remedied

once identified. High concentration of 0,3 μm

to 10 μm particles may imply the presence of

bacteria which may endanger patients in a

surgery room or special infirmary.

The most efficient procedure

for assessing indoor air quality is to

obtain several outdoor air read-

ings as baselines, noting where the

readings are taken in relation to the

facility. At least one of the readings

should be from near the building’s

fresh air intake vent.

Note, however, the location of

the intake vent to ensure baseline

readings are not skewed by pollut-

ant sources, for example, locations

near a loading dock. An indoor

air particulate ‘target’ is then cal-

culated by modifying the baseline

readings by the efficiency of indoor

filtering.

In the case of cleanroom applica-

tions, ‘as built’, ‘at rest’ or ‘opera-

tional’ (the three cleanroom stages)

can be used as baseline. But once

the baseline is established, data col-

lected should always be compared

against the same baseline.

Particles tend to diffuse very

quickly into the surrounding air,

making source identification a

challenging task. One method is to

take multiple indoor readings, starting with

the complaint area first, then move outwards.

Get a map of the HVAC systems, as installed

and use that to create an inspection route.

Within each zone, take a measurement in the

Dataweek, 25 July 2012 www.dataweek.co.za

News12 feature Test & measurement, analysis, calibration & compliance

middle of the space as well as near the air

intakes, outtakes and any other HVAC system

elements.

Make particular note to measure both

upstream and downstream at any HEPA

filters. As data is collected, take note of any

unusual increases in particulate quantity

and size. One can make use of the Fluke 985

location naming and data storage capability

to distinguish particle concentration in dif-

ferent locations for comparison.

Compare the particle readings against the

baseline to get a feel for the relative severity

of the particulate concentration, and identify

hotspots and pathways that may lead to the

particulate source. Continue to follow the

path of higher concentrations until

the source is identified. Once the source

is remedied, the area is reassessed to

ensure the corrective action addressed the

problem.

Combined with the Fluke 971 tempera-

ture and humidity meter, the technician can

also gauge the temperature and humidity

readings against the accepted parameters

(see ASHRAE Standards 55 and 62) for a basic

IAQ investigation, and take appropriate steps

to treat the problem.

Interpreting the data

A correct interpretation of the data requires

an understanding of the test area. Is the

area residential or commercial? Is it a clean-

room or precision manufacturing area? Is

the cleanroom at rest or operational? Is the

location exposed to tobacco smoke or com-

bustible particles? Is there construction at

or near the location? A proper assessment of

the environment can help narrow down the

list of problem particulates.

Concentration limits vary widely accord-

ing to the size and type of facility, among

other variables. However, a high-level assess-

ment can provide direction on whether or

not a problem exists. The outside air read-

ings in Figure 1 provide a high-level point of

reference for the technician.

Scenario 1: The particulate levels dis-

played in Figure 2 are from a new residence

(<5 years), and do not indicate any concen-

trations outside of the norm. In a residential

setting, particle levels are sometimes higher

than outside readings due to more potential

particle sources (i.e. pet dander), smaller

diffusion area, and often less sophisticated

filtration.

Scenario 2: The particulate levels dis-

played in Figure 3 are representative of

an average office workspace, and do not

indicate any concentrations outside of the

norm. In a commercial setting, particle levels

should be significantly less than outside

readings due to better filtration and better

dilution with outside air.

Scenario 3: The particulate levels in Figure

4 are from an older residential location with

visible mould. The readings are significantly

higher, and steps should be taken to remedy

the mould and address the root cause of the

problem.

Scenario 4: If the particle source in Scen-

ario 3 is not visible, use particle size tables such

as Figure 5 to identify possible sources. Obtain

a sample of the particles and submit to a lab for

further analysis.

A cleanroom exercise (certification

and monitoring)

Cleanrooms are an excellent application for a

particle counter. Certification for cleanrooms

is usually conducted at ‘as built’ stage. For

illustration purposes, let’s put the Fluke 985

to the test in evaluating an ISO Class 5 (per

ISO 14644) cleanroom. To qualify as a Class

5 cleanroom, levels cannot exceed the limits

for the class in each particle size stated in

Table 1 (Page 14).

Our test is concerned with the concen-

tration of 0,3 μm particulates in the room.

Several 2 litre samples are taken from six

different locations over six different times

inside the cleanroom, with the results as

shown in Table 2 (Page 14).

The individual readings are well within

the limitations for the cleanroom; however,

we can take the following steps to determine

the statistical validity of the readings:

SD = ( (AC1 – M)2 + ... + (AC

6 – M)2)/(L –1)

116

= ( (674–995)2 + (1154–995)2 + (1097–995)2 + (841–995)2 + (828–995)2 + (1376–995)2)/(6–1)

SE = SD/ L

47,36 = 116/ 6

Step 1: Calculate the mean average particulate concentration.

M = (AC1 + AC

2 + AC

3+ AC

4+ AC

5+ AC

6) / L

995 = (674 + 1154 + 1097 + 841 + 828 + 1376) / 6

Step 2: Calculate the standard deviation of the averages.

Step 3: Calculate the standard error of the mean of the averages.

Step 4: Establish the upper confi dence limit (UCL) – see Table 3 (Page 14).

Dataweek, 25 July 2012www.dataweek.co.za

News 13Test & measurement, analysis, calibration & compliance feature

Figure 1. Reference reading.

Figure 2. Typical particulate levels in a new residence.

Figure 3. Typical particulate levels in an offi ce space.

Figure 4. Typical particulate levels in an older

residence with visible mould.

Figure 5. Typical particle sizes in μm.

The resulting mean count for all locations is within the

requirements of a Class 5 cleanroom.

The Fluke 985 provides particulate data over six chan-

nels on a single display, allowing the technician to view

all readings at a glance. Though the cleanroom exercise

focused on 0,3 μm particulates, the single display would

immediately alert the technician to anomalies in other

particle size concentrations.

With the Fluke 985, the technician can confi gure the

number of channels displayed for diff erent particle sizes

and also set an alarm alert via audio or highlighted text for

count over-concentration for the particular particle size of

concern.

Once the cleanroom is certifi ed, the technician can

also use the trend graph function of the instrument to

monitor particle concentration over specifi c time periods

they can defi ne. This can be used at the operational stage

of the cleanroom. This allows the technician to monitor

the manufacturing process if there is any unusual spike

of particle concentration and track down the event that

introduces contaminates to the process.

This also allows the technician to view the trend imme-

diately on the unit before data is exported, saving time on

data download. With the Fluke 985 USB/ Ethernet cradle,

the technician can quickly transfer the data to computer

via USB or over the network for immediate data analysis so

that corrective actions can be quickly deployed.

The delay function of the Fluke 985 is useful for

determining ‘at rest’ cleanroom particle concentrations

after facility occupants have exited the cleanroom and

the room’s ventilation system has fi ltered the air and the

environment has stabilised. Continued on page 14

Dataweek, 25 July 2012 www.dataweek.co.za

News14 feature Test & measurement, analysis, calibration & compliance

EXFO announced the addition of its iOLM

technology to its FTB series OTDR (opti-

cal time domain reflectometer) modules.

Originally developed for the challenging

FTTH/PON fibre test environment, iOLM

technology now extends its realm of fibre/

cable testing applications to more common

architectures, such as access, point-to-point,

FTTA/DAS, metro, CWDM and long-haul

networks.

Having iOLM on these OTDRs will enable

network operators to eliminate complex

OTDR trace interpretation and human errors

from their network testing cycles. Using an

automated multi-pulse acquisition approach

and advanced algorithms, the iOLM is an

OTDR-based software application that deliv-

ers detailed information on every element on

the link, in a single button operation.

Powered by Link-Aware technology, the

EXFO proprietary iOLM self-manages the

setting of all test parameters, requiring

virtually no training for technicians of any

skill level. Now available on new FTB-7300

and FTB-7400 OTDRs, the application can be

combined with various advanced test func-

tionalities – CD, PMD, WDM-Aware, Ether-

SAM, SONET/SDH, Ethernet, etc. – inside

EXFO’s FTB Ecosystem platforms, including

the popular FTB-200 Compact platform.

Since its first release about a year ago,

EXFO’s iOLM has been well received by

the market, as it is already being used in

60 countries around the world and cur-

rently undergoing numerous field trials with

major operators. EXFO intends to develop

additional targeted features for the iOLM to

address the needs of even more customers in

multiple applications.

For more information contact Chris Nel,

Lambda Test Equipment, +27 (0)12 349 1341.

EXFO enhances OTDR modules

Continued from page 13

Concentrations (C1) AVE. Concentration

Location (L) 1 2 3 4 5 6 (AC1)

A 750 560 655 730 674

B 1575 1250 750 950 1100 1300 1154

C 1300 850 980 1125 1350 975 1097

D 1150 775 450 825 845 1000 841

E 825 855 730 940 695 925 828

F 1700 1585 1135 900 1725 1210 1376

Upper control limit (UCL) factor for 95% confi dence

# Locations 1 2 3 4 5 6 7 8 8+

95% UCL factor 6,31 2,92 2,35 2,13 2,02 1,94 1,9 1,86 NA

UCL=M + (UCL Factor * SE)

1,087 = 995 + (1,94 * 47,36)

Table 1. ISO particulate limits.

Table 2. Example concentration levels.

Table 3. Calculating the upper confi dence limit.

ISO Class particulate limits

Classifi cation 0,1 μm 0 0,2 μm 0,3 μm 0,5 μm 1,0 μm 5,0 μm

m3 m3 m3 m3 m3 m3

1 10 2

2 100 24 10 4

3 1000 237 102 35 8

4 10000 2370 1020 352 83

5 100000 23700 10200 3520 832 29

6 1000000 237000 102000 35200 8320 293

7 352000 83200 2930

8 3520000 832000 29300

9 35200000 8320000 293000

Particle counting in perspective

The key to a successful IAQ investigation is

to be aware of the environment as a whole.

Location, building history, complaints, process

records, and measurable factors such as tem-

perature and humidity, pressure and particle

concentration all play a role in uncovering IAQ

problems.

When using a particle counter, be aware

that a particle’s source may only be a symp-

tom of a much larger issue looming under the

surface. Remediation of the source may not

address core problems of poor filtration, ven-

tilation or excessive moisture. Left unchecked,

these conditions will cause the same symp-

toms, or worse, to reoccur with certainty.

Particle counting as part of ongoing

preventive maintenance is critical to ensure

a healthy environment and good production

yield. The Fluke 985 is a powerful, rugged,

and easy to use tool to assist the techni-

cian in identifying particulate problems and

authenticating the efforts to address their

root causes.

For more information contact Comtest,

+27 (0)11 608 8520.

Dataweek, 25 July 2012www.dataweek.co.za

News 15Test & measurement, analysis, calibration & compliance feature

Newsteo recognises that some customers

are likely to have continuing and evolving

needs for product capability and adaptation

requirements. Newsteo has strong experi-

ence in radio transmission and sensors

interfacing, and is willing to supply exper-

tise as needed.

The Newsteo Logger LGS allows users

to connect any potentiometric sensor and

record the measurement or transmit it by

radio to the supervision centre. The installa-

tion and maintenance of the sensor are thus

simplified. The logger is available in two

configurations, one with a potentiometric

Potentiometric sensor loggersensor input and the second with three

inputs.

The units are made versatile thanks to

the capability to read a number of sensors.

These sensors include crack meters, pressure

sensors, photoresistance sensors and poten-

tiometer for angular measurement, to name

but a few. In areas where the supervision

centre is out of reach of the radio transmit-

ter, Newsteo also has a repeater with an

external antenna.

For more information contact Conical

Technologies, +27 (0)12 345 1545.

Unlike standard cor-

rugated test leads

that experience rapid

failures due to kinking

and connector/cable

interface breakage, the

SilverLine-LP is steel-

armoured and has a

massive, solid metal

backshell to protect

the cable-to-connec-

tor interface against

virtually all possibili-

ties for damage. The

robust design results

in much longer prod-

uct life and reduces

the occurrence of faulty test results.

SilverLine PIM is ideal for use with portable

PIM testers in fi eld test applications and for use

with benchtop PIM analysers in a lab or factory

environment.

In the fi eld this test cable represents a de-

pendable test lead, eliminating the need to re-

build or re-terminate a test lead on site or cancel

a test entirely. In the factory it saves labour by

providing more accurate and consistent results

over a far longer product life, thus reducing the

PIM testing cableWebb Industries has added Times Microwave’s SilverLine-LP, a test cable specifi cally

designed for fi eld and production passive intermodulation (PIM) testing, to its

product off ering.

possibility of rejects caused by faulty test leads.

Other features and benefi ts include: better

than -117 dBm (-160 dBc) performance; a set

of low PIM adaptors with connector saver (op-

tional); and a retractable, OneTurn connector for

fast-on, fast-off mating. The SilverLine-LP test

cable is easier to handle than corrugated cable,

is water-resistant and RoHS compliant.

For more information contact Lodie Potgieter,

Webb Industries, +27 (0)11 719 0000.

Model 7131A from Measurement Specialties is

a miniature triaxial IEPE accelerometer specifi -

cally designed for NVH, engine vibration, road

response, modal analysis, structural testing, ride

handling and fl ight testing.

Available in ±50 G to ±2000 G dynamic

ranges, the rugged device features a hermeti-

cally sealed titanium construction with a 4-pin

Microtech connector interface. It incorporates

stable piezoceramic crystals in annular shear

mode which provide a fl at frequency response

up to >10 kHz.

The accelerometer features stable operating

performance over an operating range of -55°C

through +125°C with linearity of ±1% (FSO).

Operating from 18 to 30 V d.c., it provides a

±5 V d.c. output with an 8 – 12 V d.c. bias volt-

age. Excitation current is 2-10 mA. The 7131A

is case grounded and adhesively mounted for

ease of installation.

Each unit is shipped with a NIST traceable

amplitude calibration certifi cate. A 4-pin 1/4-28

to 3x BNC cable assembly (315) is off ered as an

accessory, as is the Model 161A 4-channel PE &

IEPE signal conditioner.

For more information contact Accutronics,

+27 (0)11 781 2645.

Triaxial accelerometer

+27 (0)11 781 2645.

Dataweek, 25 July 2012 www.dataweek.co.za

feature16 feature Test & measurement, analysis, calibration & compliance

The target of cost optimisation means that

standard commodity silicon processes with no

special manufacturing steps must be used. This

leads to a paradox: how to use an ASIC made

entirely of elements whose critical parameters

vary by several hundred ppm /°C to make mea-

surements with a precision of 1% to 2% over all

operating conditions.

The target is made even more challenging

when the need for a very fast response time is

added. This article describes a new

LEM ASIC which achieves this objec-

tive while at the same time adding

extra functionality.

Technology and quality

evolution

In many current measurement

applications, it is advantageous that

the measuring circuits are isolated

from the measured current, and in

these cases the magnetic fi eld due to

the current can be concentrated at a

sensor to which it is sensitive. In an

ASIC, a Hall cell may be used to trans-

form the magnetic fi eld into a voltage.

A Hall cell, however, is not only a particularly

unstable element, but additionally its properties

are usually not guaranteed in an ASIC manufac-

turing process.

Current transducer architectures have been

developed which are insensitive to the stabil-

ity and linearity of the magnetic sensor, for

example the closed-loop architecture shown in

Figure 1a.

However, such an architecture is still sensi-

tive to off sets in the sensor, and in any case

the ultimate goal for cost and reliability is to

develop an open-loop transducer containing

Current transducers benefi t from ASIC advances David Jobling, LEM.

Optimisation of cost and reduction of size are driving the need to integrate the components in current transducers using

application-specifi c integrated circuits (ASICs).

Figure 1a. Closed-loop current transducer architecture. Figure 1b. Open-loop current transducer architecture.

no electrical components except the ASIC – see

Figure 1b – while having the performance of a

closed-loop architecture.

Fully integrated ASICs for open-loop current

transducers such as LEM’s HXS series have been

developed in the past. The performance is sat-

isfactory, with typical specifi cation parameters

as follows:

• Gain drift: 350 ppm/°C;

• Input related off set drift: 2 μT/°C;

• Response time: 5 μs;

• Output noise density: 7 μV/rtHz with a

sensitivity of 40 V/T in a fi xed bandwidth;

• Current consumption: about 15 mA.

These ‘fi rst generation’ ASICs have a rigid

architecture where only gain (sensitivity) is

adjustable and is stored in a memory which can

only be written once, which limits the fl exibility

of their use.

Advances in silicon manufacturing tech-

nology and also in quality standards give a

serendipitous combination of process changes

that permit the design of a ‘second generation’

family of ASICs. These advances are:

• Analog elements which are fast and dense;

• Digital elements which are dense enough to

add useful functionality within a limited cost

budget;

• EEPROM memory which can be rewritten as

many times as needed;

• The three above features are available in

processes such as the 180 nm and 350 nm

nodes which are somewhat behind

the state of the art, and so are at a

reasonable cost;

• There is also an apparently unre

lated development, which is that

quality standards such as the auto

motive AEC-Q100 require testing

at 2 or 3 temperatures.

Of course these factors are

accompanied by some disadvan-

tages. The need for both high speed

and low noise increases power

consumption. The best transistors

are 3,3 V devices, but the market

requires operation in a 5 V environ-

ment.

However, putting these develop-

ments together has allowed LEM to design an

ASIC in which the company targeted at least

a twofold improvement in the critical param-

eters of the fi rst generation listed above. At

the same time the company has been able to

add features such as an over-current detection

(OCD) output, and the fi nal circuit confi guration

may be chosen at any time in the process, even

by the end user of the current transducer using

the ASIC.

The result is a cost-eff ective Hall-cell based

ASIC with fast response time and low tempera-

Dataweek, 25 July 2012www.dataweek.co.za

feature 17featureTest & measurement, analysis, calibration & compliance

Figure 3a. Gain error validation measurements. Figure 3b. Input related off set validation measurements.

Figure 4. ASIC die showing eight Hall cells

in the centre.

Figure 2. Block diagram of LEM’s new ASIC.

ture drifts. Figure 2 shows a block diagram of

the ASIC. To minimise the eff ect of off sets in the

Hall cells, conventional spinning techniques are

used with an output at 1,5 MHz and the eight

cells are placed in an optimised symmetrical

arrangement .

Using eight cells improves the

signal-to-noise ratio of the signal at

the input to the electrical chain to a

level similar to that of dedicated Hall

cells made with III-V compounds

such as Gallium Arsenide, though

the low signal level at this point

means that the following amplifi er

has to have a low noise level. It is

physically large and consumes a

high current: about 8 mA for the Hall

cell and amplifi er combination.

The signal is amplifi ed with high

gain-bandwidth amplifi ers and is

fi ltered before demodulation to

limit the noise which is folded back

into baseband. The fi lter uses the

same elements as the oscillator which sets

the spinning frequency, and so can have quite

a narrow bandwidth despite the process

variations.

The OCD block input is taken before the

demodulator and fi nal output amplifi er / fi lter

which has two advantages: it accommodates

signals outside the linear operating range of

the output, and it has a fast response which is

independent of the output fi lter bandwidth.

The latter may be chosen for the best

compromise between the response time of the

signal channel and the noise level at the output.

A timer block at the OCD output allows a hold

time of up to 1 ms, if needed, to ensure that

a short overload may still be detected by an

external microcontroller.

In order to meet a quality commitment at

the ppm level, testing of the packaged ASIC is

done at cold and/or hot before the fi nal test at

ambient temperature. This allows drifts of gain

and off set to be measured and an opposite cor-

rection to be stored in EEPROM.

These data are combined with the output of

an on-chip thermometer and digital-to-analog

converters to generate analog corrections of

the unavoidable gain and off set drifts in a stan-

dard silicon process. Indeed, several corrections

may be stored for diff erent ASIC confi gurations

and correction automatically applied for the

confi guration chosen.

If the Vref pin of the ASIC is forced to the

supply voltage, the output pin becomes the I/O

port of a single-wire bus interface. This means

that the fi nal ASIC confi guration may be de-

cided at any time, even in the fi nal application

of the LEM current transducer.

For example, two users may both want to

measure currents in the same range. For one a

fast response time may be the most important

parameter while the other prefers low output

noise. Each may choose the output fi lter band-

width by confi guring the ASIC according to their

preference.

The magnetic sensitivity may be

chosen between 5 and 200 V/T with

0,5% precision by sharing the 12 bits of

data between exponential steps of 15%

in the coarse gain adjustment and 32

steps of 0,5% in the linear adjustment.

The real challenge in design is to

improve all the critical parameters,

since generally they impose opposite

constraints. The measured specifi ca-

tion parameters of the new LEM ASIC

are given below:

• Gain drift: typical 100, maximum

200 ppm/°C;

• Input related off set drift: typical 0,3,

maximum 1,0 uT/°C;

• Response time: < 2 μs;

• Output noise density: 6 μV/rtHz with a sensi-

tivity of 40 V/T in a selectable bandwidth;

• Current consumption: 19 mA;

• Die area: slightly smaller than the fi rst-genera-

tion equivalent.

Measured parameters from LEM’s standard

validation setup which checks 9 circuits at a

time are shown in Figures 3a and 3b. A die

photograph is shown in Figure 4 – the eight Hall

cells can be seen at the centre of the die.

Conclusion.

The new LEM ASIC described meets the target

of improved precision for open-loop, Hall-eff ect

based current transducers. Looking forward,

the company believes that more improvements

are possible, both by adding more features and

by using new design techniques to achieve

the same level of precision while reducing the

calibrations needed.

For more information contact

Denver Technical Products,

+27 (0)11 626 2023

Dataweek, 25 July 2012 www.dataweek.co.za

feature18 feature Audio & video electronics

High-quality audio applications using the PIC32 Jayanth Murthy Madapura, Microchip Technology.

In a typical docking station and device acces-

sory, a digital audio source that plugs into the

unit sends a serial stereo audio data stream

over the dock’s data transfer interface, such as

USB. The dock captures the data stream while

performing other crucial tasks, and routes the

stream to an audio codec or digital-to-analog

converter (DAC) at a specifi c sample rate, as

shown in Figure 1.

The captured stereo audio stream then

fl ows through a serial interface to the codec

or DAC. Since there

are many possible

sources of digitised

audio, and not all of

the sources use the

same sampling rate,

this serial interface

typically adapts the

sampling frequency to

the source, or converts

the sampled data

stream into a common

data rate.

Therefore, one of

the challenges in the

design of the docking

system or device acces-

sory is to perform the

sample-rate conversion

without degrading the

audio quality, and at

the lowest cost pos-

sible.

To deal with these

challenges, designers

have typically used a

dedicated sample-rate

conversion circuit and/

or a high-end audio

DAC that incorporates

Sample

rate

(kHz) Codec master clock/Time base with diff erent oversampling (MHz) Bit clock (MHz)

fs 128 fs 192 fs 256 fs 384 fs 512 fs 768 fs 1152 fs 32 fs 64 fs

32,0 — — 8,1920 12,2880 16,3840 24,5760 36,8640 1,024 2,0480

44,1 — — 11,2896 16,9344 22,5792 33,8688 — 1,4112 2,8224

48,0 — — 12,2880 18,4320 24,5760 36,8640 — 1,536 3,0720

88,2 11,2896 16,9344 22,5792 33,8688 — — — 2,8224 5,6448

96,0 12,2880 18,4320 24,5760 36,8640 — — — 3,072 6,1440

176,4 22,5792 33,8688 — — — — — 5,6448 11,2896

192,0 24,5760 36,8640 — — — — — 6,144 12,2880

Table 1. Typical master clocks and bit clocks required by the codec/DAC/ADC for diff erent audio sample rates.

When designing audio docking stations and accessories for portable digital audio devices, and other digital audio

sources, designers are constrained by cost, while trying to deliver the highest-quality audio playback.

Figure 1. Typical audio application.

Figure 2. I2S format.

complex phase-locked loops (PLLs) to ensure

fl exible sample rates for stable communication

of the sampled audio data.

The USB interface is a convenient interface

for the transfer of audio data. However, to

meet the requirements of professional audio,

the subtle loss of quality due to USB clock and

codec clock mismatch must be addressed.

In this article, the available audio-specifi c

features of the PIC32 MCU are explored to

address these needs. The SPI module supports

diff erent standard audio communication modes

and off ers high bit resolution for high-quality

audio applications.

The fl exible reference clock output feature

of the PIC32 can be used to provide the master

clock to the analog front-end to generate the

diff erent sample rates. The reference clock

output also eliminates the need for an external

crystal/oscillator by a codec.

It also eliminates the need for a PLL on

the codec. The reference clock output can be

tuned to prevent buff er

under-run and over-run

that arise because of

clock mismatches. The

MCU also off ers the USB

Host and Device module

with fl exible PLL clocking

schemes at low power.

Digital audio data

basics

When analog audio is

converted into a discrete

digital format, the analog

signal is sampled at a

frequency of at least twice

the highest frequency

component in the analog

signal, or the Nyquist rate.

Therefore, an audio signal

that spans 0 to 20 kHz

can be sampled at a data

rate of 44,1 kHz, which

in this case is the suit-

able Nyquist rate, so that

the signal can be recon-

structed without aliasing

when converted back to

the analog domain.

In addition to the

Dataweek, 25 July 2012www.dataweek.co.za

feature 19featureAudio & video electronics

Sample Bit clock (MHz) SPIxBRG MCLK rate 32 fs 64 fs 32 fs 64 fs (MHz) (kHz) (16-bit) (24-bit) (16-bit) (24-bit)

11,2896 44,1 1,4112 2,8224 4 2

12,2880 48 1,536 3,072 4 2

32 1,024 2,048 6 3

8,1920 32 1,024 2,048 4 2

Table 2. Baud rate values when the PIC32 SPI

is confi gured as a master.

Table 3. Comparison of master clock

generated using the codec PLL and reference

clock output on a PIC32 device.

8,192 8,192 8,192

11,2896 11,2889025 11,28893

12,288 12,288 12,288

Expectedmaster

clock (MHz)

Average masterclock using

external crystal

(MHz)

Average masterclock using

REFCLKO (MHz)

Table 4. Divisor and trimming values for diff erent master clocks.

8,1920 5,859375 5 440 8,1920 0

11,2896 4,251701 4 128 11,28893 0,005937

12,2880 3,90625 3 464 12,2880 0

16,3840 2,929688 2 476 16,3840 0

16,9344 2,834467 2 427 16,93728 0,01703

18,4320 2,604167 2 309 18,43661 0,02501

22,5792 2,12585 2 64 22,58824 0,04002

24,5760 1,953125 1 488 24,5760 0

33,8688 1,417234 1 213 33,85124 0,051848

36,8640 1,302083 1 154 36,84558 0,049975

MCLK (MHz)

Divisor with

USB-PLL of

96 MHz as clock

source

RODIV

(N)

ROTRIM

(M/512)

Observed MCLK

(MHz) % Deviation

Figure 3. Left-justifi ed format.

Figure 4. Right-justifi ed format.

Figure 5. DSP/PCM format.

sampling rate, data bit resolution can be 16-

bit or 24-bit stereo audio data. For compact

disc (CD) quality audio, the standard is 16-bit

resolution with a 44,1 kHz sample rate. How-

ever, there are higher-performance CD music

options. One such standard encodes the data

with a 24-bit resolution and increases the

sampling rate to 96 kHz.

For professional audio, the audio fi les

are encoded with a resolution of 24 bits per

sample, which provides headroom when the

audio is mixed and manipulated. Also, the

resolution choice allows for the trade-off

of sound quality versus fi le size, even with

compression.

The USB interface can readily handle

the streaming of high-quality audio over

isochronous transfers. Its ability to deliver

high-quality audio is quite evident, as it is

popular among many audio

users. With its universal

ease-of-use, USB audio can

transfer high-resolution and

high-sample-rate audio with

negligible jitter, when pack-

aged with a fl exible audio

interface.

Isochronous data transfer,

amongst its various other

uses, is utilised to stream

audio data to and from a

source at a constant rate in

real-time. Stereo audio data

packets, with size governed

by the sample rate of the

audio stream, are transferred

as part of USB frames every

1 ms on the USB full-speed

link. USB audio also provides

controls for common features

such as volume, tone, gain

control and equalisers,

among many control and

processing units.

The diff erences in bit rates

and sample rates require the

hardware in the playback

system, or dock, to be able

to handle the diff ering rate

data streams. To do that,

the system must either use a

more complex DAC that is expensive and can

phase-lock to each sample rate and adjust

itself to each playback option, or use an

external sample-rate converter IC with the

low-cost DAC, or convert all the streams into

a standard sample rate and bit rate using an

algorithm running on a microcontroller that

a simple low-cost DAC can handle.

The PIC32 MCU off ers a fl exible reference

clock output and audio mode to address

these requirements to achieve high-quality

audio while maintain a low design cost. The

serial interface with the Audio mode and

fl exible reference clock output module are

explored in the following sections.

PIC32 serial peripheral interface

module with Audio mode

Most codecs off er serial communication over

a 4-line serial interface. The transmission and

reception of the stereo audio samples between

the MCU and codec occur over this serial inter-

face. Typical serial interfaces have the following

signals:

• Serial Data Output (SDO) to transmit stereo

audio data to the codec.

• Serial Data Input (SDI) to receive stereo audio

data from the codec.

• Serial Bit Clock (SCK/BCLK) is the required bit

clock provided by the Master.

• Left/Right Clock (LRCK) is the phase clock

provided by the Master for stereo data.

PIC32 devices have a serial peripheral inter-

face (SPI) module with Audio mode.

Audio mode off ers various

interface formats, bit resolu-

tions and Master/Slave con-

fi gurations. The communica-

tion modes supported include

the following:

I2S format

I2S mode is where the most

signifi cant byte (MSB) is avail-

able on the second rising edge

of the BCLK following an LRCK

transition (Figure 2).

Left-justifi ed format

Left-justifi ed mode is where

the MSB is available on the fi rst

rising edge of BCLK following

an LRCK transition (Figure 3).

Right-justifi ed format

Right-justifi ed mode is where

the least signifi cant byte (LSB)

is available on the rising edge

of the BCLK preceding an LRCK

transition and the MSB is still

transmitted fi rst (Figure 4).

DSP/PCM format

In DSP/PCM mode, the left-

channel MSB is available on

Dataweek, 25 July 2012 www.dataweek.co.za

feature20 feature Audio & video electronics

Table 5. Observed THD and SNR using codec PLL and PIC32 reference clock output – sample rate of 32 kHz, expected master clock of 8,192 MHz.

fs = 32 kHz Using external crystal as time base Using reference clock output as time base

Average Average

observed observed

Expected frequency frequency

frequency (Hz) (Hz) THD (%) SINAD (dB) (Hz) THD (%) SINAD (dB)

500 499,9649 0,002 88,62 499,9668 0,00224 88,476

1000 999,9318 0,00224 88,752 999,9342 0,002 88,752

1500 1499,895 0,00208 88,308 1499,901 0,00192 88,308

10000 9999,318 0,002 82,212 9999,335 0,002 82,212

the fi rst rising edge of the BCLK following a

rising edge of the LRC (Figure 5). Right-channel

data immediately follows left-channel data.

Depending on word length, the BCLK frequency

and sample rate, there may be unused BCLK

cycles between the LSB of the right channel

data and the next sample.

PIC32 Audio mode supports 16-, 24- and

32-bit stereo audio data. It also provides

advanced error handling with receive over-

fl ow and transmit under-run fl ags

and control bits to disable them as

required. Also, the clocks gener-

ated by the BCLK and the LRC when

the PIC32 MCU is a serial Master are

free-running, which is essential for

uninterrupted audio data transfer in

streaming applications.

The audio codecs and DACs allow

either or both of these confi gura-

tions: Master or Slave. In the PIC32

SPI Audio mode Master confi guration,

the MCU provides the BCLK and the

LRCK to the codec. This confi guration

is shown in Figure 6. In the PIC32 SPI

Audio mode Slave confi guration, the

codec provides the bit clock (BCLK)

and stereo phase clock (LRCK) to the

MCU. This confi guration is shown in

Figure 7.

PIC32 devices also provide a

control interface channel over the

Inter-Integrated Circuit (I2C) peripheral

module. The control interface is used

to confi gure the codec for a specifi c

mode of operation by confi guring the

control registers on the codecs.

Flexible reference clock output

The PIC32 devices have a fl exible

reference clock output. The reference

clock output module (REFCLKO) can

be used to generate the fractional

clock that can be used by audio

codec/DACs to accommodate vari-

ous sample rates. Typical examples of these

sample rates and associated master clocks are

shown in Table 1.

The REFCLKO module can be used to gener-

ate these audio-specifi c Master clocks and is

not limited to just these. The reference clock

Figure 6. PIC32 SPI-codec confi guration with PIC32 SPI as master

and the codec as slave.

Figure 7. PIC32 SPI-codec confi guration with PIC32 SPI as slave

and the codec as master.

Figure 8. SPI and codec confi guration showing the reference

clock output signal.

can be mapped to any port pin on the device

if the Peripheral Pin Select (PPS) functional-

ity permits on the specifi c device. REFCLKO

can be confi gured to the source clock for the

SPI module on the PIC32 MCU instead of the

peripheral bus clock.

This ensures synchronisation of the USB

clock with the SPI channel clock to minimise

eff ects of clock jitter when the USB PLL clock is

selected as the source clock for the REFCLKO.

The stereo word select signal and the bit clock

will be synchronised with the reference clock

output since all of them are sourced off the

reference clock output.

The clock source for the reference clock

output module has various options. It can be

selected from USB-PLL, Primary PLL, POSC,

FRC, SOSC, LPRC, PBCLK, SYSCLK clocks.

Based on the selected clock source, a frac-

tional divisor needs to be confi gured for the

REFCLKO to generate the needed MCLK based

on the oversampling rate as noted in Table 2.

The fractional divisor needed is confi g-

ured as an integer part and fractional part

determined by RODIV and REFOTRIM bits in

the REFOCON register, respectively. Figure 9

shows the PIC32 oscillator module

with the reference clock output indi-

cating the diff erent clock sources for

the reference clock output and the

divider with the trimming scheme.

When the PIC32 SPI is the Master,

the BCLK and the LRCK are gener-

ated by the device. With REFCLKO

as the clock source for the SPI

module, the BCLK can be confi gured

by loading the appropriate values in

the SPI baud rate generator register,

SPIxBRG. If needed, the REFCLKO

can provide the master clock input

to the codec.

As shown in Table 2, the SPIxBRG

register values and the required

REFCLKO frequencies for the audio

sample rates are 32/44,1/48 kHz for

stereo audio data with both 16-bit

and 24-bit resolution. The bit clock,

BCLK, can be 32 fs or 64 fs. When

the PIC32 SPI is the Slave, the BCLK

and the LRCK are generated by the

codec. The reference clock output

can provide the master clock for the

codec/DAC.

Example implementation

Let us explore an example of

generating a REFCLKO signal. If

the required MCLK is 12,288 MHz

and if the USB-PLL clock of 96 MHz

is used as source for the reference

clock output, the required divisor is

7,8125. The divisor needs to be programmed

for a half period of 3,90625, which is

(7,8125/2). The RODIV bits will be confi gured

with a value of 3 and ROTRIM confi gured for

464/512 = 0,90625.

Table 4 provides the required RODIV and

Dataweek, 25 July 2012www.dataweek.co.za

feature 21featureAudio & video electronics

Table 6. Observed THD and SNR using codec PLL and PIC32 reference clock output – sample rate of 44,1 kHz, expected master clock

of 11,2896 MHz.

fs = 44,1 kHz Using external crystal as time base Using reference clock output as time base

Average Average

observed observed

Expected frequency frequency

frequency (Hz) (Hz) THD (%) SINAD (dB) (Hz) THD (%) SINAD (dB)

500 499,9649 0,00208 88,632 499,9668 0,00168 88,62

1000 999,9318 0,00184 88,332 999,9342 0,002 88,548

1500 1499,895 0,00168 88,38 1499,901 0,00176 88,212

10000 9999,318 0,00192 83,148 9999,335 0,00208 82,968

Example 1. RODIV and ROTRIM values for REFCLKO of

12,288 MHz.

Figure 9. Block diagram of the PIC32 oscillator showing the reference

clock output.

Figure 10. Clock jitter DAC underrun or overrun manifesting as subtle

audible clicks.

ROTRIM values needed for the

standard audio MCLKs previously

listed in Table 1. The clock source

selected here is the USB-PLL

clock. However, any other clock

source can be selected and the

RODIV and ROTRIM values can be

calculated in a similar manner.

The reference clock output can

be changed with a fi ne precision.

The unit resolution is between

approximately 0,02% and 0,05%

based on the generated clock

frequency. This ensures that the

sample rate deviation is well

below the limits required by the

codec when the reference clock

output is only being used as the

SPI clock source and not as the

master clock to the codec. If the

reference clock output is used as

the master clock for the codec, the

deviation is only limited by the

specifi cation of the codec.

Table 3 shows an example of

an audio master clock generated

by an expensive codec with fl ex-

ible internal PLL using a standard

external crystal of 12 MHz and the

master clock using reference clock

output. The data indicate that the

master clock generated by the

PIC32 reference clock output is

very close to the required master

clock.

Also, the reference clock out-

put performs as well as or better

than the master clock generated

using the internal PLL on an ex-

pensive codec. This would elimi-

nate the need for a PLL on the

codec or DAC used as an analog

front end and eliminate the need

for an external crystal used as time

base for the codec. This will result

in cost savings in the audio design

while providing the same or better

quality than an expensive codec.

Table 5, Table 6 and Table 7 show examples

of the signal quality metrics, SNR and THD, for

a set of (pure) audio tones generated using

external crystal with internal PLL on the codec

and using the reference clock output as master

clock on the same codec.

The SNR metric or signal-to-noise ratio

measures the power of the noise

induced compared to the signal.

The higher the SNR metric the

better is the quality of the signal

produced.

The THD metric or the total

harmonic distortion measures

the strength of harmonic distor-

tion caused by the harmonics

of the fundamental frequency.

The lower the THD metric the

better is the quality of the signal

produced.

The SNR and THD data indi-

cates that the quality of audio

signal generated by an AFE with

reference clock output as the

time base performs exception-

ally well compared to that of

a very expensive codec with

internal PLL.

Tuning reference clock

output

The reference clock output has

the ability to be tuned on-the-

fl y, and can be tuned in steps

between a specifi ed range. The

range should be such that it

ensures a swing of the sample

rate, typically about ±0,2%,

which is well below a range

that might introduce audible

artefacts. For example, a swing

of ±0,2% of the data stream

with a sample rate of 48 kHz

requires a tuning of the refer-

ence clock output between the

ranges of 12 263 424 Hz and

12 312 576 Hz.

Table 8 lists the required

RODIV and ROTRIM values for

tuning the sample rate within

±200 Hz of the sample rate. The

required master clock column

indicates that the sample rate

needs to be tuned between the

range of 47,9 kHz and 48,1 kHz, ROTRIM needs

to be tuned between 459 and 468 and the

required RODIV is 3. Similar tuning ranges for

RODIV and ROTRIM to tune the reference clock

Dataweek, 25 July 2012 www.dataweek.co.za

feature22

Table 8. REFCLKO tuning example with RODIV and ROTRIM values for 48 kHz audio stream.

47800 12236800 3,922594 3 472 48010 12290560 3,905436 3 463

47810 12239360 3,921774 3 471 48020 12293120 3,904623 3 463

47820 12241920 3,920954 3 471 48030 12295680 3,90381 3 462

47830 12244480 3,920134 3 471 48040 12298240 3,902998 3 462

47840 12247040 3,919314 3 470 48050 12300800 3,902185 3 461

47850 12249600 3,918495 3 470 48060 12303360 3,901373 3 461

47860 12252160 3,917677 3 469 48070 12305920 3,900562 3 461

47870 12254720 3,916858 3 469 48080 12308480 3,89975 3 460

47880 12257280 3,91604 3 469 48090 12311040 3,898939 3 460

47890 12259840 3,915222 3 468 48100 12313600 3,898129 3 459

47900 12262400 3,914405 3 468 48110 12316160 3,897319 3 459

47910 12264960 3,913588 3 467 48120 12318720 3,896509 3 459

47920 12267520 3.912771 3 467 48130 12321280 3,895699 3 458

47930 12270080 3.911955 3 466 48140 12323840 3,89489 3 458

47940 12272640 3,911139 3 466 48150 12326400 3,894081 3 457

47950 12275200 3,910323 3 466 48160 12328960 3,893272 3 457

47960 12277760 3,909508 3 465 48170 12331520 3,892464 3 456

47970 12280320 3,908693 3 465 48180 12334080 3,891656 3 456

47980 12282880 3,907878 3 464 48190 12336640 3,890849 3 456

47990 12285440 3.907064 3 464 48200 12339200 3,890041 3 455

48000 12288000 3,90625 3 464

Legend: Shaded cells indicate Sample Rate Tuning Range values

LRCK

(Hz)

LRCK

(Hz)

MCLK

(Hz)

MCLK

(Hz) Divisor DivisorRODIV RODIVROTRIM ROTRIM

feature Audio & video electronics

fs = 48 kHz Using external crystal as time base Using reference clock output as time base

Average Average

observed observed

Expected frequency frequency

Frequency (Hz) (Hz) THD (%) SINAD (dB) (Hz) THD (%) SINAD (dB)

500 499,9959 0,002 88,62 499,9965 0,00224 88,476

1000 999,9938 0,00224 88,368 999,9932 0,002 88,752

1500 1499,989 0,00208 88,332 1499,997 0,00192 88,308

10000 9999,915 0,002 82,164 9999,934 0,002 82,212

Table 7. Observed THD and SNR using codec PLL and PIC32 reference clock output – sample rate of 48 kHz, expected master clock of 12,288 MHz.

output can be determined for other standard

audio sample rates.

The tuning capability of REFCLKO prevents

buff er under-run and over-run and alleviates the

audible clicks, as discussed in the next section.

USB clock mismatch

USB specifi cations require a tolerance budget

and a limit on the USB clock frequency as a way

to achieve immunity to radio interference. The

USB clock with the allowed tolerance budget

results in reduced audio quality if there is a USB

clock mismatch.

The real-time streaming audio samples must

arrive at precise, regular time intervals so that

the DAC can convert the digital samples to an

analog signal with the expected constant rate at

which it is confi gured.

The DAC clock that expects and receives

the audio samples at a particular sample rate

cannot miss even a single sample. A missing

sample manifests as a subtle click for the listen-

er, since the DAC fails to generate an accurate

representation of the streamed audio signal.

On a microcontroller or microprocessor with

an embedded USB module, the USB clock is

sourced from an independent clock such as an

on-chip PLL with an external crystal oscillator

of specifi c value. Since the clock is not sourced

from the USB interface, the mismatch in clocks

introduces buff er over-run or under-run, caus-

ing audible clicks as shown in Figure 10.

An easy solution for the audio data under-

run or over-run issue as related to audio-quali-

ty degradation, is to use a good asynchronous

sample rate converter (ASRC), where the input

sample rate is estimated with jitter attenuation,

and the internal fi lters are dynamically tuned

for a new sample rate. However, a good ASRC

is very expensive and the system still requires a

DAC for analog conversion.

As an eff ective low-cost solution, the USB

audio packets are buff ered and the clocks of

the codec or DAC can be tuned to prevent

under-run or over-run using a feedback mech-

anism. The feedback mechanism monitors

the buff er level and ensures it stays within an

acceptable range, while achieving at least the

same quality achieved by an expensive ASRC.

The reference clock output with USB-PLL

clock as its source can be used to generate and

tune the required master clock, as discussed

in the ‘Tuning reference clock output’ section

above. This capability prevents buff er under-run

and over-run while maintaining an acceptable

DAC sample rate with a swing range of 0,2%.

This is the lowest-cost solution compared to

the other two, while still achieving high-quality

audio.

Audio accessory board and application

demonstration

As an example, all of the features discussed in

this article are implemented in the PIC32 USB

Headset application project available for the

PIC32 USB Digital Audio Accessory Board.

For more information contact Tempe

Technologies, +27 (0)11 452 0530.

Dataweek, 25 July 2012www.dataweek.co.za

feature 23

The TPA3116D2, TPA3118D2

and TPA3130D2 feature

a very high program-

mable switching rate and

wide power supply range,

enabling audio output that

is both high-quality and

highly effi cient.

The devices are pin-to-

pin and software compat-

ible and can support a

single-layer printed circuit

board (PCB), enabling cost

savings by allowing designers to reuse a single

layout for multiple designs.

A supply range of 4,5 V to 26 V enables si-

multaneous operation of line power and battery

power on the same board for longer battery life

in soundbars and portable audio docks. Ther-

mally effi cient FETs allow 30 W x 2 operation

without the need for a heatsink, or 50 W x 2 with

a heatsink (TPA3116D2 only).

Programmable switching rates of 400 kHz to

1,2 MHz allow inductor fi lter-free operation and

interference avoidance for clear audio with no

distortions. The ICs’ closed-loop design provides

excellent power supply rejection ratio (PSRR)

Class-D audio amplifi ers

CML Microcircuits recently introduced the

DuraTALK family of digital voice products,

consisting of a wide range of fl exible

technologies supporting voice-data

generation, coding, transcoding, decod-

ing, storage and scrambling functions for

digital communication systems.

The CMX608, CMX618 and CMX638 ICs

are designed for RALCWI, a CML-owned,

low bit rate (2400 bps) vocoder technol-

ogy that facilitates the transmission and

reception of highly compressed voice

over inherently noisy narrowband radio

channels. RALCWI off ers fl exible, highly

integrated, high-performance vocoders,

with near toll-quality voice at very low bit

rates.

The CMX8341 digital PMR baseband

processor system-in-a-chip is aimed at

low-cost, licence-free digital PMR (dPMR)

radios conforming to the ETSI standard, TS

102 490, and analog PMR (legacy mode).

Seen as the digital successor to

PMR446, digital PMR is now a proven

technology targeting highly functional

solutions by using low-cost and low-com-

plexity technology. dPMR is a narrowband

(6,25 kHz) FDMA technology that is 100%

digital and off ers many forms of voice

and/or data applications. RALCWI in digi-

tal PMR will provide savings in cost, size

and power consumption.

For CVSD/PCM/ADM, fl exible coders

and decoders are available for analog-to-

digital and digital-to-analog voiceband

conversion. The CMX639 and CMX649

address this market segment, providing

robust and fl exible coding algorithms, low

cost, low power and small size, ideal for

use in a wide variety of consumer, military

and business digital voice applications.

The CMX7261 is a single-chip G.711/

G.729A/CVSD full duplex multi-transcoder

device, capable of encoding analog voice

into linear, μ-law or A-law PCM, CVSD

or G.729A data formats. It is capable of

decoding PCM, CVSD and G.729A back to

analog voice. It can also transcode data

between PCM, CVSD and G.729A.

The fl exible IC creates toll quality

speech for applications such as software-

defi ned radio (SDR), voice-over IP (VoIP),

wireless private branch exchange (PBX)

and privacy type digital voice communica-

tions.

For more information contact Avnet Kopp,

+27 (0)11 809 6100.

ICs for digital voice communication

A fast laser diode

from OSRAM Opto

Semiconductors has

been used to provide

light for a new 3D camera

from Swedish manu-

facturer Fotonic. The

component was devel-

oped especially for 3D

cameras and facilitates

range fi nding for moving

objects with good depth

resolution at video rate.

The infrared laser di-

ode provides high optical

power of about 0,5 W, even when in continu-

ous wave operation. With a wavelength of

about 845 nm the light is hardly visible to

the human eye, and yet it is well within the

sensibility range of the camera chip.

Three-dimensional cameras superimpose

a two-dimensional picture with information

on the range and surface profi le of an object.

Industrial applications for such systems are,

for example, the gauging of fl uid levels or the

grading of objects according to size or shape.

Three-dimensional sensors also provide

reliable stereoscopic information, e.g. for the

control of robot arms.

IR laser diode used in 3D camera

featureAudio & video electronics

Texas Instruments has introduced three new analog-input, class-D stereo amplifi ers

for use in soundbars, after-market automotive solutions, portable audio docks and

LCD televisions.

and slower edge rates for waveform switching

to improve electromagnetic interference (EMI)

performance. RDS(on) of 120 mΩ at 30 W x 2

improves effi ciency by decreasing resistance

during operation.

An evaluation module and SPICE model are

available for the TPA3116D2, TPA3118D2 and

TPA3130D2, allowing designers to evaluate

system-level performance and quickly integrate

the amplifi ers into product designs.

For more information contact Dirk Venter, Arrow

Altech Distribution, +27 (0)11 923 9600.

The Fotonic C70

camera is based on a

CMOS sensor and makes

use of the principle of

time-of-fl ight measure-

ment. A laser diode,

which is integrated into

the camera, transmits

a light pulse which the

target refl ects back. Each

of the 120x160 pixels of

the camera chip records

the distance to the object

by measuring a respective

phase shift compared to

the emitted light pulse.

The precision of this distance measurement

and the range of the sensor increases with

the modulating frequency of the light source.

Due to their fast switching times of only a few

nanoseconds, laser diodes are particularly well

suited for this application.

Within the range of 0,1 to 7 metres, the

camera measures distances with an accuracy

of 3 up 30 millimetres.

For more information contact Ryan Hunt,

OSRAM Opto Semiconductors,

+27 (0)79 525 1779.

Dataweek, 25 July 2012 www.dataweek.co.za

feature24

Historically, testing of acoustic response in

audio playback devices has required criti-

cal performance analysis conducted by

specialised audio engineers. In addition

to design verifi cation, acoustic response

testing is also important in production

line testing of multimedia devices.

Currently, many multimedia devices

such as MP3 players, mobile phones and

car stereos utilise speakers. Accurate

frequency response measurement and

detailed harmonic distortion analysis

during speaker production testing require

high resolution dynamic signal acquisition

modules with fl exible analog input/output

function and high dynamic range.

AO output, for example, generating a

multi-tone or stepped sweep sine signal,

may be employed as a stimulus for a

speaker. To convert speaker output to a

voltage signal, a calibrated microphone

may be employed as a sensor. AI input can

then acquire the signal output from the

microphone and perform a fast Fourier

transform (FFT) to analyse the measured

audio signals, providing frequency and

impedance response and harmonic distor-

tion analysis.

The ADLINK PXI-9527 features a 2-

channel, 24-bit simultaneous sampling

analog input. The Sigma Delta ADC pro-

vides up to 432 KSps sampling rate and

high resolution, ideally suiting it for higher-

bandwidth dynamic signal measurement.

The sampling rate can be adjusted by

setting the onboard DDS clock source to

an appropriate frequency. All channels are

sampled simultaneously and accept inputs

in the range of ±40 V to ±0,316 V. The

analog input supports software-selectable

AC or DC connection, and 4 mA bias cur-

rent for integrated electronic piezoelectric

(IEPE) sensors.

For more information visit

www.adlinktech.com

Signal acquisition card for speaker testing

Housed in 3 x 3 mm packages, the devices

reduce board space compared to fully discrete

solutions and are ideal for space-constrained

applications, such as rear-view cameras.

The ADA4830-1 (single-channel) and

ADA4830-2 (dual-

channel) video dif-

ference amplifi ers,

and ADA4432-1

(single-ended

output) and

ADA4433-1 (dif-

ferential output)

analog video fi lter

amplifi ers, provide

designers with a

complete chipset

for transporting

analog video sig-

nals over shielded

and unshielded

twisted pair

cables.

The ICs can

also be used in-

dependently to provide receiver (ADA4830-1/2)

and transmitter (ADA4432-1/33-1) functions

as part of a video signal chain. They feature

integrated short-to-ground protection, fault

detection, wide common-mode voltage ranges

and robust ESD performance.

The ADA4830-1 and ADA4830-2 video

diff erence amplifi ers are designed for use as

receivers of CVBS (Composite Video Blanking

and Synchronous) signals in standard-defi nition

PAL/NTSC video applications. In addition to

Video amplifi ers

Analog Devices has introduced a 16-chan-

nel audio D/A converter (DAC) capable of

improving audio system performance and

reducing power consumption for profes-

sional, ‘prosumer’ and automotive audio

equipment applications.

The 24-bit ADAU1966 provides 118 dB

SNR (signal-to-noise ratio) performance at

32 kHz to 192 kHz sampling rates. This is

achieved with power consumption of less

than 300 mW for 16-channel operation by

running the digital portion of the chip at a

lower voltage.

The digital supply can be generated on-

chip via a linear regulator, allowing the IC

to operate off a single analog supply. Alter-

natively, the regulator can be bypassed and

a separate 2,5 V to 3,3 V supply provided to

the chip.

feature Audio & video electronics

Hi t i ll t ti f ti i

H

r

s

a

A

c

f

a

(

o

A

f

a

Operating the ADAU1966 with full-

scale diff erential output voltage of 3 Vrms

reduces the need for downstream gain and

preserves a high SNR against electrically

coupled system noise. By using the device’s

onboard PLL (phase-locked loop) to derive

the internal master clock from an external

left/right clock, the IC eliminates the need

for a separate high-frequency master clock

and can be used with or without a bit clock.

The ADAU1966 is qualifi ed for AEC-100

automotive applications and operates over

a -40°C to +105°C temperature range. It also

features an SPI/I2C port which allows an ex-

ternal microcontroller to adjust volume and

read the chip temperature to within ±3°C.

For more information contact Arrow Altech

Distribution, +27 (0)11 923 9600

16-channel audio DAC

Analog Devices has introduced a series of high-speed video amplifi ers that

integrate short-to-battery protection and fault detection for analog video

systems in automotive safety and infotainment applications.

integrated short-to-battery (input) protection up

to 18 V on a single 3 V to 5 V supply, the devices

include a short-to-battery fl ag that indicates the

presence of a fault condition at the input.

The ADA4830-1 and ADA4830-2 also have

ESD tolerance up

to ±8 kV Human

Body Model

(HBM), a wide

input common-

mode voltage

range that ex-

tends to -8,5 V be-

low ground and

55 dB common

mode rejection at

7 MHz.

The ADA4432-

1 (single-ended

output) and the

ADA4433-1 (dif-

ferential output)

analog video

fi lter amplifi ers

are designed

as transmitters for automotive analog video

fi ltering applications. They feature the same 18 V

short-to-battery protection (output), fault fl ags

and ±8 kV HBM tolerance as the ADA4830.

Additionally, they feature continuous output

short-to-ground protection and include a high-

order video fi lter with a -3 dB cutoff frequency at

10 MHz and 45 dB of rejection at 27 MHz.

For more information contact Avnet Kopp,

+27 (0)11 809 6100.

Dataweek, 25 July 2012www.dataweek.co.za

News 25featurefeatureAudio & video electronics

Silicon TV tuner ICs are rapidly displacing

legacy mixer oscillator phase-locked loop

(MOPLL) CAN tuner technologies to reduce

cost and size and improve performance.

Silicon tuner IC adoption began before 2007

and gained real traction during 2010 when

fl at-panel TVs and set-top boxes (STBs) saw a

signifi cant increase in sales.

Designing silicon TV tuners that matched

MOPLL performance levels was the main

hurdle for adoption, but once semiconductor

suppliers met this performance standard the

road was cleared for silicon tuner IC ship-

ments to accelerate.

Several suppliers now off er a wide variety

of silicon TV tuner products. Navigating the

competing options can be a complex process

given the wide range of issues surrounding

TV tuner IC selection. Putting these technical

issues into perspective will help simplify the

TV tuner selection process.

State of the market

TV market demand has grown to an amazing

230 million units (Mu) per year across a wide

range of regions. The broader TV market

breaks down into two segments relevant to

TV tuner ICs: integrated digital television

(iDTV) platforms (approximately 160 Mu)

and analog-only TVs (approximately 70 Mu),

which consist of both fl at-panel TVs and

analog-processed CRT TVs.

Analog-processed CRT TVs employ an

older image processing technology. In

contrast, fl at-panel TVs incorporate advanced

DSP-based image processor SoCs supplied

by companies such as Broadcom, MediaTek,

MStar, Novatech and Sunplus Technology.

Some major TV brands also have their own

internal semiconductor teams that build

captive fl at-panel SoCs.

Analog-processed CRT TVs use less

capable microcontrollers and cannot easily

accommodate silicon tuner IC programming

and confi guration. These analog-processed

TVs are likely to retain the older MOPLL

CAN tuner implementations, which often

require more than 150 separate components

(see Figure 1).

Flat-panel platforms employ modern

high-performance processors that run a

substantial amount of fi rmware so they are

well positioned to confi gure modern digital

devices such as silicon TV tuner ICs.

Many TVs must support both analog

and digital transmission standards to meet

consumer demands. Examples of analog

transmission standards include NTSC and

PAL/SECAM. In this case, the word ‘analog’

describes the transmission modulation

format and not the specifi c processing

technology used to decode it.

Many countries are in various stages of

executing their analog switch-over plans

where the terrestrial analog broadcasts

will be terminated and replaced with

higher-quality digital broadcasts. Yet even

in countries that have theoretically made

the hard cut over to digital, the analog

transmissions remain active, and consum-

ers continue to demand support for these

transmissions.

Japan is perhaps an exception, having

achieved a true conversion to all digital, but

TVs sold in virtually every other region are

forced to retain analog reception capability

to deal with low-power analog broadcasts,

slow-to-upgrade cable networks and legacy

consumer electronics devices.

Tuner support for analog transmissions

will be required by all major TV brands

for TVs outside of Japan for at least fi ve

more years and likely for ten more years.

Numerous TV platforms accept content

from both terrestrial and cable sources.

As a practical matter, these platforms must

support analog transmissions as long as

the multi-subscriber operators (MSOs)

continue to transmit analog content on their

cable networks.

Cable and terrestrial STB segments add

an additional 175 million systems of demand

per year. A recent forecast of silicon tuner

ICs from IMS Research (Figure 2) projects the

market growth for the overall number of TV

tuners for demand TV, STB, cable modem,

satellite receiver and other segments. Sev-

eral multipliers are applied to account for

multi-tuner IC applications, generating the

roughly 700 million units of overall tuner IC

market opportunity shown.

The rapid adoption of silicon tuner ICs is

evident from this data and is even stronger

when focusing on just the TV and cable STB

segments where silicon tuner ICs have made

deeper inroads.

Opportunities going forward

The high-growth markets of mainland China

and Taiwan are taking the lead in new silicon

TV tuner IC opportunities. TV makers in

Figure 1. Example of MOPLL CAN tuner, which requires 150+ components.

Figure 2. IMS Research forecast shows growing market demand for silicon tuners.

Silicon tuner ICs in mainstream television designs Information from Silicon Laboratories.

Dataweek, 25 July 2012 www.dataweek.co.za

News26 feature Audio & video electronics

China have rapidly increased TV production

over the last three years, and cost-optimised

silicon TV tuner ICs help them deliver high-

quality TV models at competitive price points.

According to the analyst fi rm Display-

Search, China became the world’s largest

market for LCD TVs in 2011. DisplaySearch

also predicts that fl at-panel TV shipments

(both LCD and plasma models) in China

will nearly double from the 31 million units

reported in 2009 to 59 million units in 2014,

reaching a compound annual growth rate

of 14%.

Top-tier TV brands are accelerating their

adoption of silicon tuners to save

cost, improve reliability, reduce

component inventories and support

thinner TV form factors. These same

advantages are driving signifi cant

opportunities in the second- and

third-tier TV brands where the

adoption of silicon TV tuners is just

starting.

Because today’s silicon tuner ICs

must support all worldwide terres-

trial and cable broadcast standards,

these smaller TV brands benefi t

from the intensive work already

accomplished by the top brands.

Connected TVs expand the range

of program sources beyond just the

antenna and cable, and are gaining

signifi cant traction as ‘apps’ make

their way to large-screen TVs.

New communication interfaces

for TVs such as Wi-Fi, Ethernet and

LTE (cellular) are delivering rich,

diverse content beyond the tradi-

tional broadcast sources. These

interfaces present special technical

challenges for over-the-air (OTA)

broadcast tuners. For example, a special Wi-Fi

immunity circuit is needed to protect against

this interference.

Silicon TV tuner selection challenges

Tuners perform a complex task in that they

must cleanly receive one low-power trans-

mission in a crowded spectrum full of other

content. This puts a great deal of pressure

on the RF front-end of the silicon TV tuner IC.

The complexities present at the geographic

boundaries of two diff ering transmission

standards add additional challenges.

Add to this the architectural complexi-

ties of popular DTV and STB SoCs, and it

becomes clear that a single silicon TV tuner IC

cannot meet the needs of the entire market.

Fortunately numerous tuner IC architec-

tures are available today that address these

varying platform needs. The challenge is to

confi dently select the right device for the

application.

Channel reception: The keys to achieving

clear TV broadcast reception are sensitivity

and selectivity. These characteristics in turn

are based on low noise fi gure (NF), high RF

front-end linearity and high-quality RF front-

end fi ltering.

Low noise fi gure means that the RF

front-end adds very little unwanted noise to

the incoming signal, while high sensitivity

means that there is ample video signal-to-

noise ratio (VSNR) even with weak input

signals to correctly receive the broadcast.

But TV tuners almost always need to

receive a single broadcast in the presence

of many other signals. This is where the

silicon tuner IC’s selectivity – the ability to

block or exclude content on nearby ‘blocker’

channels – is essential for clean reception.

To achieve high selectivity in a cost-eff ec-

tive solution, a silicon tuner IC needs both

high linearity and high-quality fi ltering in the

RF front-end. As a result, silicon TV tuner ICs

with high sensitivity and excellent selectivity

performance, as shown in Figure 3, add up

to more channels received under real-world

conditions.

Field testing: Major TV brands deliver

products adhering to the offi cial broadcast

standards, but they must also address the

many real-world exceptions to these broad-

cast standards or face consumer complaints

and costly product returns.

Unusual broadcast conditions and excep-

tions occur more frequently than one might

realise. This is especially true at the geo-

graphic boundaries between one broadcast-

ing standard and another, where issues can

get very complex. For example, in Europe

numerous analog and digital standards can

be received across international borders by

a single TV.

To address these non-standard reception

conditions, the TV tuner and demodulator

must fi rst detect and then compensate for

the non-standard transmissions. These condi-

tions are corrected on a case-by-case basis

and require extensive fi eld testing to perfect.

Without the extensive control and confi gura-

tion capabilities of the silicon TV tuner ICs,

these increasingly challenging requirements

could not be addressed.

TV manufacturers put TV tuners through

rigorous fi eld testing, highlighting both

standard performance as well as performance

to anomalous broadcasts. TVs destined for

cost-conscious regional markets may choose

a lower-cost TV tuner, but TVs destined for

broader export markets will certainly require

robust worldwide performance.

Flexible architectures reduce cost: Greater

choice in TV tuner architectures has been a

positive development for iDTV and STB sup-

pliers. Several silicon tuner IC variations are

now available to meet the specifi c architec-

ture needs of iDTV and STB platforms.

Modern TV and STB platforms have three

main circuit functions: the silicon TV tuner

(see Figure 4 for example), the demodulators

and the audio/video and processing graphics

SoC. The output of the tuner drives the input

of the demodulator, and the output of the

demodulator drives the input of the audio/

video processing and graphics SoC.

TV and STB platforms have varying needs

for analog and digital tuners. For example, a

simple narrowband cable access box (cable

STB) may need a single digital-only tuner. In

contrast, a high-volume mainstream iDTV will

likely support analog and digital terrestrial

broadcast as well as cable input sources.

Figure 3. The Si21x6 TV tuner demonstrates exceptional VSNR performance in the presence of a blocker

(red trace) compared to alternate tuners (blue trace).

Dataweek, 25 July 2012www.dataweek.co.za

News 27featurefeatureAudio & video electronics

For the larger TV brands, these tuners must

support all analog and digital transmission

standards worldwide.

TV tuner ICs are now available sup-

porting numerous architectural confi gura-

tions such as digital only, multi-channel

digital only, analog and digital tuners,

analog and digital tuners with analog TV

demod ulator, as well as various receiver

confi gurations with a tight linking of tuner

and demodulator.

Most modern TVs support at least one

tuner that can tune both analog and digital

channels. Separate demodulators for the

analog and digital channels are needed. To

support varying digital demodulator stan-

dards, there may even be multiple demodu-

lators such as DVB-T and DVB-T2 within the

same platform.

Demodulators vary in complexity with

ATSC (North America) and DVB-T (digital ter-

restrial broadcast for Europe and many other

regions of the world) being relatively low-

complexity standards compared to DTMB

(China’s new digital broadcast standard)

and DVB-T2 (a new higher-density digital

standard in Europe).

In the case of digital demodulators such

as DVB-S2, DVB-C2 and DVB-T2, the demodu-

lator is typically contained in a standalone IC.

Less complex demodulators can be found

in the audio/video processing and graphics

SoC, or they may be combined with other,

more complex digital demodulators.

STBs support a cable input and in some

cases a broadcast antenna input. Unlike TVs,

the SoCs used by many STB platforms usually

lack an analog TV (ATV) demodulator and are

not well suited to analog broadcast reception.

In these cases, the ATV terrestrial broadcast

demodulator will be external to the SoC, if

such a demodulator is required for the STB.

Demodulators for analog TV (ATV

demodulators for NTSC, PAL and SECAM)

Figure 4. Example of state-of-the-art silicon TV tuner architecture.

are especially prone to poor behaviour

when presented with the many real-world

anomalous transmissions. In practice,

actual demodulator performance is highly

dependent on the experience gained from

prior generations of eff ort. This experience is

concentrated in specialised control fi rmware

running on modern silicon TV tuner ICs and

demodulators.

Given the many nuanced interactions

between tuners, demodulators and image

processing SoCs, TV manufacturers must

perform extensive fi eld testing to validate

specifi c combinations of tuners, demodula-

tors and SoCs. This means that in some

cases, a tuner with a higher-performance

internal ATV demodulator will be chosen

even if the SoC integrates its own ATV

demodulator.

The key question for the designer is to

determine which demodulator has been

adequately fi eld-tested relative to the TV’s

or STB’s intended market and use. If fi eld

testing for a specifi c tuner/demodulator

combination has not been completed, it may

be necessary to adopt an integrated and

more thoroughly tested tuner/demodula-

tor combination to meet the development

timeline and cost targets.

Cable STB applications can present a

special challenge for tuner selection. In most

cases, the demodulator will be supported

by the STB SoC. Generally, the transmission

environment for cable is considered less

variable than that of terrestrial broad-

cast. Signal strength and other factors are

reasonably well controlled within a cable

network.

Tuner performance metrics are gener-

ally relaxed for cable STBs applications

using this easoning. However, the ability

to control the cable plant depends on the

specifi c cable provider or multi-subscriber

operator (MSO).

In some cases a given MSO will have

issues with ‘channel tilt’ (diff ering signal

strength across the channel spectrum),

external signal ingress (most likely from local

terrestrial broadcast stations that interfere

with the cable transmissions) or a mixed

transmission mode network that re-transmits

an array of channels from diff erent sources

with diff erent transmission characteristics.

A lower-performance silicon TV tuner IC

may not be adequate in these situations, and

a higher-perform ance TV tuner designed for

terrestrial TV broadcasts may be needed to

receive the full range of channels.

Silicon Labs’ product

off ering

As a practical matter, what

types of TV tuner options

are available for use? A

review of Silicon Labs’ latest

generation of silicon TV

tuner ICs can help answer

this question.

This new TV tuner

family off ers a range of

architectures that combine

easily with popular TV and

STB system-on-chip (SoC)

devices. The Si21x6 TV

tuner family includes the

high-performance Si2176

silicon TV tuner IC with ana-

log demodulator, as well as

the Si2136 complete analog

receiver.

The Si2146 digital-only TV tuner IC is

ideal for cable set-top boxes and multi-tuner

HDTVs. The high-performance Si2156 TV

tuner IC provides a popular confi guration

(with just the digital and analog tuners) with

features similar to the cost-eff ective Si2155

tuner. The Si2155/56 devices easily combine

with popular TV and STB SoCs. These devices

apply equally well to iDTV, cable platforms

and STBs.

The entire Si21x6 family, including the

Si2155 TV tuner IC, delivers exceptional RF

performance and design simplicity off ered

by Silicon Labs’ current-generation TV tuner

ICs that are in high-volume production at fi ve

of the world’s top seven TV brands.

The Si2155 TV tuner is designed to

reduce system costs in high-growth regional

markets such as mainland China and Taiwan

where there is substantial variability in the

technical requirements for TV and STB plat-

forms depending on the needs of regional

and export platforms. TV tuner architectural

fl exibility is especially important to reducing

costs and simplifying designs in today’s

rapidly growing TV markets.

For more information contact

Gary de Klerk, NuVision Electronics,

+27 (0)11 608 0144.

Dataweek, 25 July 2012 www.dataweek.co.za

feature Interconnection, switches, relays, cables & keypads28 regular General

Global concerns over climate change require

using our limited energy resources more

effi ciently. For a cleaner and sustainable

future, energy effi ciency plays a key role.

Approximately 20% of the global electrical

energy is consumed for lighting applications.

The transformation to new, intelligent LED light-

ing technologies for residential and commercial

lighting allows signifi cant energy savings.

With an increasing number of countries

banning incandescent lamps in favour of

energy-saving alternatives, 230 V operated LED

lamps are emerging as the favoured replace-

ment. The charm of these so-called retrofi t LED

lights is that they require minimal or even no

change or modifi cation of the original socket

and control gear.

This advantage is generally accompanied

by the customers’ requirement that the overall

performance of the LED retrofi t must be equal

to or even better than the original incandescent

solution. What this means for the LED driving

solution is that:

• The complete circuitry must be designed in a

space-saving way, to fi t into a standard socket;

• High effi ciency is desired to minimise thermal

design eff orts;

• Easy to design, for a fast time to market.

For the consumer who installs the LED

retrofi t at home, two additional requirements

are important:

Especially for 230 V LED bulb retrofi ts, good

LED lighting solution for lamp retrofi tsdimming performance is one of the most (if not

the most) important features;

The LED solution must not fl icker when there

is a variation in voltage at the supply rail.

Infi neon is off ering innovative solutions

to deal with these challenges. The newly

introduced off -line LED driver ICL8002G

sets high standards with respect to inte-

gration, performance, features and total

system cost. It employs quasi-resonant

operation modes with integrated

power factor control (PFC) for off -

line (230 V) residential LED lighting

applications in particular for 40 W

/ 60 W / 100 W incandescent bulb

replacements.

Applied primary control

techniques result in system ef-

fi ciencies of 90% while external

component count is kept low.

The resulting compact driver design

can easily be embedded within the

standard screw-in form factor. The ICL8002G

outperforms current regulatory requirements

and is prepared for the future. Together with an

external CoolMOS switch, it is easily scalable and

presents a solution to implement multiple LED

bulb designs with one controller.

When it comes to ease of designing, Infi neon

off ers designers an online design tool in the form

of Lightdesk (accessed via www.infi neon.com/

lighting). It is available for the whole ICL(S)8xxx

family and helps customers to design the

transformer, which is the most critical external

component in such a system. Furthermore, the

tool gives guidance on the other external

components and lets users share their

results and designs with others.

As good dimming performance

is a key requirement, Infi neon has

put a strong focus on this issue

during product development. The

ICL8002G has broad compatibility

with leading- and trailing-edge

phase cut dimmers found in

most homes.

To provide smooth dim-

ming, without distortion in the

light output, the IC has a con-

tinuous dimming curve, similar

to dimming an incandescent

lamp. It also incorporates

integrated circuitry to ensure a

stabilised light output when trailing-

or leading-edge dimming is active.

The stabilisation of the light output also

accounts for very good line regulation. With

integrated digital fold-back current correction,

variations in the supply voltage will not have a

visible eff ect on the light output.

For more information contact EBV Electrolink,

+27 (0)21 402 1940, Electrocomp,

+27 (0)11 458 9000.

NXP Semiconductors announced the

SSL2115x series of high-effi -

ciency GreenChip driver

ICs for low-cost LED

retrofi t lamps. The new

SSL21151 and SSL21153 are

designed for non-dimmable

5 W and 10 W LED lamps

respectively, providing good

current control in fl yback and

buck-boost confi gurations.

The SSL2115x enables

small form-factor LED drivers,

using primary-side sensing to

regulate current, and off ering

integrated LED open/short pro-

tections, an internal 700 V MOSFET,

and a circuit enabling startup di-

rectly from the rectifi ed mains voltage.

With the use of an additional valley fi ll

circuit, the series can off er a power factor

of approximately 0,9.

For more information contact Avnet Kopp,

+27 (0)11 809 6100.

LED drivers for retrofi t lamps

T,

tage.

ey fi ll

er factor

Dataweek, 25 July 2012www.dataweek.co.za

featurePower & energy 29regularGeneral

Instrument Plastics’ range of light control fi lm is

designed for a variety of purposes and is avail-

able with a range of options.

For privacy viewing, the fi lm’s microlouvres

allow for controlled viewing so an unauthorised

observer’s line of sight is blocked. It is ideal for

confi dential viewing on CRTs, automatic teller

machines (ATM) and anywhere privacy viewing

is desired.

In terms of sunlight readability, the microlou-

vres help block out annoying off -axis light while

maximising the transmission from the display

to the viewer. The result is improved display

contrast with minimal loss of brightness. Light

is directed to where it is needed or away from

where it is not. This eliminates night-time win-

dow refl ections in automotive and aeronautical

applications. It can also be used to hide the light

source in incandescent lighting applications.

Contrast is improved and glare reduced on

electronic displays by blocking out annoying

off -axis light, making the display more comfort-

able to read. The microlouvres maximise the

transmission from the display to improve the

image with little loss of brightness.

Light control fi lm

TDK has expanded its portfolio of SMD power

inductors with new CLF7045-D types that can

withstand temperatures of up to 150°C.

The new types, which have dimensions of

6,9 x 7,2 x 4,5 mm, off er rated inductance values

ranging from 1 μH to 470 μH. Types are available

for rated currents of 0,43 A to 8,9 A, based on

a 10% drop in initial inductance, and their DC

resistance ranges from 9,6 mΩ to 1,42 Ω.

For more information contact Electrocomp,

+27 (0)11 458 9000.

SMD power inductors

Vectron’s new TX–801 is a TCXO off ering

±100 ppb stability in a 5 x 3,2 mm foot-

print. The device operates from a 3,3 V

supply over an output frequency range of

8 to 26 MHz.

IP-based applications have inherited

legacy timing requirements, such as

Stratum 3 levels of accuracy. Services that

are now carried over packet switched

networks need to comply with levels of

performance that used to be delivered

over the circuit switched networks of the

past.

The TX–801 is optimised for support-

ing timing requirements over modern

IP-based networks and meets the reduced

cost and footprint expectations set by

emerging timing applications.

In addition to the TX–801 TCXO,

Vectron also off ers a range of precision

oscillators designed to support timing

requirements over IP-based networks,

along with innovative clock path technol-

ogy including OCXOs, TCXOs, XOs VCXOs,

VCSOs and jitter attenuators that provide

low-jitter timing solutions.

For more information contact Andrew

Hutton, RF Design, +27 (0)21 555 8400.

Stratum 3 TCXO

The product range off ers choices for

customers in terms of optical quality level,

louvre angle, viewing angle, abrasion/solvent

resistance and surface fi nish.

Instrument Plastics, though its agent Actum

Electronics, is able to handle anything from a

one-off prototype to bulk orders of 100 000

or more. Lead times are typically three to four

weeks.

For more information contact Kevin Klaff ,

Actum Electronics,

+27 (0)11 608 3001.

Dataweek, 25 July 2012 www.dataweek.co.za

feature Interconnection, switches, relays, cables & keypads30 regular General

Since the introduction of the Subscriber

Identifi cation Module (SIM) card technology,

the size of smart cards has greatly shrunk

to pack more on-

chip memory for

provider applications;

increased security,

control and data

processing func-

tions; and to fi t newer

but smaller future

devices.

The migration of

developing markets

from 2G and older networks to advanced

SIM-based 3G networks led to the develop-

ment of a new SIM card

format that enabled

smartphones, tablet

PCs or other devices

to transfer digital

photos and huge data

fi les over high-speed

3G networks. This

format is known as

3FF (3rd form factor),

Mini-UICC (Universal

Integrated Circuit Card)

or more popularly as

micro-SIM.

Measuring only 12

by 15 mm, the micro-

SIM card is 52% smaller

than its 15 by 25 mm

predecessor, the Mini-

SIM (or 2FF) card. It re-

tains the same electric

contacts and circuitry

of previous SIM cards,

has added functional-

ity and is backward

compatible with larger

SIM card holders and

readers via additional

surrounding plastic

cut-outs.

Ideal for ultra-slim

smartphones, tablet

Micro-SIM card sockets

The 55116 connector from Bernier is cylin drical,

has 5 or 6 contacts and is connected with a

bayonet system. This connector is designed for

the transmission of low-intensity signals, and is

often used in the fi eld of military and civil radio

transmissions and on data transmission systems.

It is also used for measurement apparatus,

sensors and testers.

The connectors comply with MIL C 55116

standard. The shells are constructed from

stainless steel and aluminium alloy and insu-

lated with Thermoset. With self-cleaning, gold

Bayonet connectors

Accord has developed a RoHS-compliant keypad

with blue backlighting. The AK-207-N-SSL-WP-

MM-BL has a contact rating of 20 mA, 24 V

d.c. and consumes an LED forward current of

240 mA. Contact resistance is 200 Ω maximum

and the keys can survive 2 million cycles each.

The keypad measures 72 mm high by 56 mm

wide and can withstand operating temperatures

in the range of -20°C to +60°C.

For more information contact Sivan Electronic

Supplies, +27 (0)11 887 7879.

Blue-backlit keypad

PCs, GSM/UMTS modems, PC cards, WLAN

(wireless LAN) cards, as well as machine-to-

machine (M2M) systems, Molex’s micro-SIM

card sockets are

developed for

PCB real estate

and vertical space

savings.

A distinguishing

feature of Molex’s

micro-SIM card

sockets is their

built-in anti-short

design. The risk

of electrical shorting is particularly high

when users of larger SIM cards attempt to

force-fi t these into the shape of a micro-SIM

card by trimming around the plastic edges

of the contact pads (the actual working

area of all SIM cards). This may leave the

pared SIM card with an exposed pad which

may potentially cause shorts with all other

contact pads when inserted into the metal

socket.

Shorting is prevented in Molex’s series

78646 socket by a shell design feature that

confi nes any lateral movements of the SIM

card when inserted, even if improperly cut.

The series 78727 socket has raised plastic walls

that act as contact barriers to insulate the pads

of the inserted SIM card from the metal shell

of the socket.

The socket also has a card polarisation fea-

ture to prevent the card from being inserted

in the wrong direction; and a detect switch to

enable card detection when fully inserted. The

sockets’ round contact terminals provide ex-

cellent electrical contact; their reverse terminal

design with gradual lead-in enables smooth

card insertion and withdrawal.

The sockets have LCP housings that can

withstand high temperatures and feature wide

‘fi nger’ areas for push-pull actions on the SIM

card. Available in 6-or 8-circuit options, these

sockets are ELV- and RoHS-certifi ed.

For more information contact Avnet Kopp,

+27 (0)11 809 6100.

plated contacts, the connectors can withstand

temperatures between -40°C and +85°C, can

be submerged under 2 m of water for 48 hours,

and can mechanically endure 3000 operations.

Voltage rating is 60 V maximum; current

rating is 0,5 A; contact resistance is ≤50 mΩ

mated; dielectric strength is 500 V a.c. RMS;

and insulation resistance is 5000 MΩ under

500 V d.c.

For more information contact Kevin Klaff ,

Actum Electronics, +27 (0)11 608 3001.

Dataweek, 25 July 2012www.dataweek.co.za

featurePower & energy 31regularGeneral

Vishay recently released the fi rst

devices in its next-generation D Series

of high-voltage power MOSFETs. The

new 400 V, 500 V and 600 V N-channel

devices combine low specifi c on-resis-

tance with ultra-low gate charge and

currents from 3 A to 36 A in a wide

range of packages.

Based on a new high-voltage

stripe technology, the MOSFETs’ stripe

design − with a smaller die size and

terminations − lowers the total gate

charge by 50% compared to previous-

generation solutions, while increasing

switching speed and reducing on-resistance

and input capacitance.

The 400 V, 500 V and 600 V devices feature

on-resistance down to 0,17 Ω, 0,13 Ω and

0,34 Ω, respectively. This translate into low

conduction and switching losses to save energy

in high-power, high-performance switchmode

applications, including server and telecom

power systems, welding, plasma cutting, battery

chargers, ballast light, high-intensity discharge

N-channel power MOSFETs

(HID) lighting, semiconductor capital equip-

ment and induction heating.

With gate charges down to 9 nC for the

400 V devices, 6 nC for the 500 V devices

and 45 nC for the 600 V devices, the D Series

MOSFETs off er gate charge times on-resistance

down to 7,65 Ω-nC, 15,6 Ω-nC and 12,3 Ω-nC,

respectively.

For more information contact Dirk Venter,

Arrow Altech Distribution, +27 (0)11 923 9600.

Key features of the

new models include

high power densities,

no external tuning

components (plug

and play), low profi les,

high effi ciencies and

land grid array (LGA)

footprint. They are suit-

able for harsh telecom

and broadcast mast-

mounted applications,

as well as industrial,

test and measurement,

routers and medical

equipment.

The DOSA-2 product

line is comprised of four series with output

current/power ratings as follows, iCF (3 A /

16,5 W), iCG (6 A / 33 W), iBF (12 A / 66 W) and

iAF (20 A / 110 W). All models operate from an

input voltage range of 4,5 V to 14 V and fea-

ture adjustable output voltages from 0,7 V to

5,5 V with conversion effi ciencies up to 94,5%.

The devices range in size from 12,2 x 12,2

x 4,45 mm to 33 x 13,46 x 7,75 mm. Optional

EPC (edge plated castellation) layout is also

available. Featuring fully optimised control

loops, all models within the series provide

excellent transient response without external

loop-tuning components.

These PoL converters operate at a constant

SMT DC-DC converters

switching frequency of 600 kHz, which pro-

vides a low noise profi le and facilitates easy

EMI fi ltering. All models are safety approved

to UL/CSA60950-1 and VDE CB scheme per

IEC/EN60950-1.

Remote on/off , remote sense and power-

good signal are standard. In addition, all

models can operate in ambient temperatures

from -40°C up to +115°C and include a 3-year

warranty.

For more information contact

Accutronics, +27 (0)11 781 2645,

Vepac Electronics, +27 (0)11 609 7122.

TDK-Lambda has released the iCF, iCG, iBF and iAF series of non-isolated SMT DC-DC

converters that comply with the latest 2nd-generation DOSA standards for point-of-

load (PoL) converters.

Dataweek, 25 July 2012 www.dataweek.co.za

feature Interconnection, switches, relays, cables & keypads32

HotHot Chips

32

ON Semiconductor off ers two new multiple-

channel diff erential switching ICs in single-pole,

double-throw (SPDT) switch confi gurations

for high-frequency signals such as PCI Express

3.0 and DisplayPort 1.2 I/O. The NCN3612B is a

6-channel diff erential SPDT switch that supports

8 Gbps data transfer with an on-state capaci-

tance (CON

) of 2,1 pF and an on-state resistance

(RDSON

) of 8 Ω. This is complemented by the

NCN3411, which is a 4-channel diff erential

switch also capable of supporting 8 Gbps opera-

tion. It has CON

of 2 pF and RDSON

of 7,5 Ω.

The NCN3612B has a power supply voltage

range of 3 to 3,6 V and draws a supply current

of just 250 μA, while the NCN3411 has a power

supply voltage range of 1,5 V to 2,0 V and draws

a supply current of 200 μA.

Avnet Kopp, +27 (0)11 809 6100

The BGU8006 from NXP Semiconductors is

a GPS low-noise amplifi er packaged in a tiny

WLCSP package with a footprint of just 0,65 x

0,44 x 0,2 mm and requiring only two external

components. Featuring a low noise fi gure of

0,60 dB, the device off ers enhanced reception

for weak GPS signals by dynamically suppress-

ing strong cellular and WLAN transmit signals.

It uses adaptive biasing techniques to instantly

detect any output power from jammers, and

compensate by temporarily increasing the

current. With the BGU8006, adaptive biasing

improves linearity with a 10 dB better IP3 under

-40 to -20 dBm jamming conditions and pro-

vides eff ective GPS output with jammer power

up to -15 dBm.

Arrow Altech Distribution, +27 (0)11 923 9600

RFMD’s new

RF5836 provides a

complete inte-

grated solution in

a single front-

end module for

Wi-Fi 802.11a/n

systems. The

small form factor

and integrated

matching minimise layout area and reduce

the number of external components. The

IC integrates a power amplifi er, single-pole

double-throw switch, and a power detector

coupler for improved accuracy. It is provided

in a 3 x 3 x 0,5 mm, 16-pin package. The chip

operates from a single supply voltage of 3,0 V to

4,8 V, and a control voltage above 1,6 V. Power

output is 15,5 dBm (11a, 54 Mbps at 4% EVM) or

14,5 dBm (11n, 65 Mbps at 2,8% EVM).

RF Design, +27 (0)21 555 8400

Multiple-channel switching ICs

Low-noise amplifi er for GPS

Wi-Fi front-end module

Texas Instruments introduced the fi rst in a

family of multi-chemistry, multi-cell battery

management gas gauge circuits with propri-

etary ‘Impedance Track’ technology which

maintains up to a 94% accurate capacity mea-

surement for the entire life of the battery. The

bq34z100 supports a wide range of lithium-ion

and lithium iron phosphate chemistries in 2- to

16-cell battery packs, extending battery run-

time in applications like medical instruments,

power tools, e-bikes and uninterruptible power

supplies. TI also plans to introduce new gauges

this year to support lead acid, nickel cadmium

and nickel metal hydride chemistries.

EBV Electrolink, +27 (0)21 402 1940

Battery gas gauge

AMD intro-

duced the

latest entry in

its embedded

G-Series pro-

cessor family

with the

G-T16R Accelerated Processing Unit

(APU). It is targeted at cost-sensitive

embedded designs that require a

combination of x86 compatibility

and graphics. The chip boasts power

consumption of just 2,3 W on average, or 4,5 W

thermal design power (TDP). It fi ts into small

form factor boards by implementing a two-chip

platform – the APU and its companion control-

ler hub. Legacy I/O card support is based on a

full 32-bit PCI interface and an ISA bus solution

with DMA support. The APU supports a full

range of display technologies, with analog VGA

and LVDS support for legacy applications and

DVI, HDMI and DisplayPort interfaces for the

latest display technology.

www.amd.com

Low-power embedded processor

International Rectifi er has introduced its third

generation of SupIRBuck synchronous buck

voltage regulators. The new single-output

devices feature a newly patented modulator

scheme claimed to generate the industry’s

smallest, jitter-free pulse widths. By reducing

jitter for pulse widths as small as 50 ns, the tech-

nology enables higher closed-loop bandwidth

for better transient response and lower output

capacitance. In addition to switching frequency

up to 1,5 MHz, pre-biased startup, input volt-

age aware enable, tracking and programmable

power-good, the new devices off er reference

voltage accuracy of 0,5%, enhanced line/load

regulation with input voltage feed-forward and

thermal compensation for current limit.

Future Electronics, +27 (0)21 421 8292

Synchronous buck regulators

Dataweek, 25 July 2012 IBC

* denotes advertisement

QUICKFINDCompany Telephone Pagese-mail Website

More suppliers, more products, more info:

www.ebg.co.za

www.dataweek.co.za

Accutronics +27 (0)11 781 2645 sales@accutronics.co.za www.accutronics.co.za 15,31

Actum Electronics +27 (0)11 608 3001 kevin@actum.co.za www.actum.co.za 29,30

Arrow Altech Distribution +27 (0)11 923 9600 info@arrow.altech.co.za www.arrow.altech.co.za 6,8,23,24,31,32

Avnet Kopp +27 (0)11 809 6100 sales@avnet.co.za www.avnet.co.za 9,23,24,28,30,32

Centurion Micro Electronics +27 (0)12 666 9066 sales@cme.co.za www.cme.co.za 7*

Comtest +27 (0)11 608 8520 info@comtest.co.za www.comtest.co.za OFC*,10,11,12,13,14

Conical Technologies +27 (0)12 345 1545 info@conical.co.za www.conical.co.za 15

CST Electronics +27 (0)11 608 0070 sales@cstelectronics.co.za www.cstelectronics.co.za 4

Current Automation +27 (0)11 462 4253 ca@switches.co.za www.rectifi er.co.za Stitched insert*

Denver Technical Products +27 (0)11 626 2023 denvertech@pixie.co.za www.denvertech.co.za 13*,16,17

EBV Electrolink +27 (0)21 402 1940 capetown@ebv.com www.ebv.com 28,32

Electrocomp +27 (0)11 458 9000 cbrand@electrocomp.co.za www.electrocomp.co.za 28,29

Electrocomp Express 0860 10 20 20 sales@eexpress.co.za www.eexpress.co.za 6

Electronic Touch Systems +27 (0)11 782 3346 touch@global.co.za www.electouch.co.za 9*

Electronica +27 (0)11 486 2775 mvanniekerk@germanchamber.co.za www.electronica.de 5*

Future Electronics +27 (0)21 421 8292 marian.ledgerwood@futureelectronics.com www.futureelectronics.com 8,32

Keystone Electronic Solutions +27 (0)12 460 4135 info@kses.net www.kses.net 5

Lambda Test Equipment +27 (0)12 349 1341 chris@lambdatest.co.za www.lambdatest.co.za 11*

Logica Technologies +27 (0)12 654 4174 info2@logica.co.za www.logica.co.za 14*

M/A-COM Technology Solutions +27 (0)11 908 6679 grant.annandale@macomtech.com www.macomtech.com 7

Mantech Electronics +27 (0)11 493 9307 sales@mantech.co.za www.mantech.co.za 12*,29*

Mateq Systems 0861 373 777 info@mateq.co.za www.mateq.co.za 15*

Newelec +27 (0)12 327 1729 sales@newelec.co.za www.newelec.co.za 30*

NuVision Electronics +27 (0)11 608 0144 gdeklerk@nuvisionelec.co.za www.nuvisionelec.co.za 25,26,27

OSRAM Opto Semiconductors +27 (0)79 525 1779 r.hunt@osram.com www.osram.co.za 23

RF Design +27 (0)21 555 8400 sales@rfdesign.co.za www.rfdesign.co.za 7,8,29,32

Sivan Electronic Supplies (SES) +27 (0)11 887 7879 elecsupp@global.co.za 28*,30,31*

Technews Publishing +27 (0)11 543 5800 malckey@technews.co.za www.technews.co.za OBC*

Tempe Technologies +27 (0)11 452 0530 willem.hijbeek@telkomsa.net 18,19,20,21,22

TRX Electronics +27 (0)12 347 4419 info@trxe.com www.trxe.com IFC*

Vepac Electronics +27 (0)11 609 7122 sales@vepac.co.za www.vepac.co.za 31

Webb Industries +27 (0)11 719 0000 prichards@webb.co.za www.webb.co.za 15