H ow l on g can a b a tt e r y r e a l l y l a s t ? P age 1 0
Eval uat i ng opt i ons f or USB T y pe-C C onnec tions Page 33
NOVEMBER 2018
POWER &
ENERGY EFFICIENCY HANDBOOK
final cover – P&EE HB 11.18.indd 1
11/8/18 8:46 AM
FEATURED MANUFACTURERS
181001_AutoR_DW_US.indd 1 Digi-Key 11-18.indd 13
9/27/18 12:45 3:38 PM 11/7/18 PM
Discover Your Power
Flexibility to Choose the Desired Power Solution As a leading supplier with a comprehensive power management and monitoring portfolio, Microchip gives you the power, flexibility and confidence to choose the right solution for your design. Managing your system’s power usage is crucial to achieving the performance that your design requires. Our product portfolio of power monitoring devices allows you to accurately measure active, reactive and apparent power, Root Mean Square (RMS) current and voltage, line frequency, and power factor. Our broad selection of power management devices including DC-DC controllers and regulators, MOSFETs and MOSFET drivers, voltage supervisors and references, and power modules enables you to efficiently design a solution to manage the power requirements of your system. From reference designs to evaluation boards to simulation tools, you’ll reduce design-in time and minimize risk with the comprehensive support Microchip can offer.
Discover your power at microchip.com/PowerSolutions The Microchip name and logo and the Microchip logo are registered trademarks of Microchip Technology Incorporated in the U.S.A. and other countries. All other trademarks are the property of their registered owners. © 2018 Microchip Technology Inc. All rights reserved. 8/18 DS20006065A
microchip 1 – P&EE HB 11.18.indd 1
11/7/18 12:53 PM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
Don’t let the LED blue light blues get you down TAKE A LOOK at some of the websites devoted to alternative medicine and you’ll likely come across headlines like, “The Dangers of LED & Blue Lights Will Blow Your Mind!” or, “LEDs: The Blue Light Risk,” or “Artificial lighting and the blue light hazard.” These all stem from the belief that light in the blue part of the spectrum causes medical problems, everything from insomnia to promoting macular degeneration. And LEDs are supposedly a major source of blue light. Here’s what one website says on the subject: …..the most efficient LED bulbs of the day are almost exclusively emitters of predominantly blue light, though there are some exceptional products that mix all colors of the spectrum—but not many are available or affordable for house use. Scary. And wrong, at least as far as we can determine. The pronouncements above come from a website that aligns itself with the herbalism. We tend to be skeptical of the alternative medicine crowd. So when we came
LEE TESCHLER • EXECUTIVE EDITOR
2
DESIGN WORLD — EE NETWORK
commentary – P&EE HB 11-18 V2 LT FINAL.indd 2
across passages such as the one cited above, we decided to run our own tests of LEDs. It’s not difficult to check claims about the wavelengths of LED output. To do so, we obtained a light meter designed for use by architects and room-lighting specialists who need to set light levels. To this we added an optical bandpass filter which only passes light in the bluewavelength range, about 450 to 490 nm. To decide what LED bulbs to check, we went to Consumer Reports. As with other product categories, they rank LED bulbs. So for samples, we used five 60-W-equivalent LED bulbs Consumer Reports puts at the top of its rankings. They included bulbs from Sylvania, Philips, and EcoSmart (made by Feit). We also got hold of two incandescent 60-W bulbs, one with frosted glass, the other without. For testing, we put each bulb in a completely dark room and measured its illuminance into the light meter sitting one-foot away, with and without the blue bandpass filter. That let us calculate the percentage of blue light output coming from from each of them. We’ll cut to the chase. All of the 60-W equivalent LED bulbs we measured had less than 2% of their output in the blue light range. That’s
11 • 2018
a long way from being “emitters of predominantly blue light.” Moreover, both incandescent lights also had under 2% of their output in the blue wavelength range. And lest you think we tested super-pricey bulbs that few people can afford, one of our Sylvania bulbs cost a little over a dollar on Amazon. To give the alternative medicine crowd the benefit of the doubt, we also tested an old LED bulb that came out before the DoE held its “L Prize” event in 2011. The L Prize is credited with kicking LED bulb development into high gear. So, we reasoned, perhaps bulbs made before then had the problematic blue light output. But the results were the same. We had one pre-2011 LED bulb on hand for testing. It, too, emitted less than 2% of its output in the blue wavelengths. We aren’t done testing. The five bulbs at the top of the Consumer Reports list all put out relatively warm light. There are others whose light is colder, in the 3100 to 4500 K range. Eventually we’ll check a few of those as well. But so far, it looks as though you’re no worse off lighting your rooms with LEDs than with ordinary incandescent bulbs. It’s best to take all those scary scenarios painted by alternative medicine websites with a grain of salt.
eeworldonline.com | designworldonline.com
11/7/18 3:15 PM
e
Fly me to the moon Let me play among the stars FRANK SINATRA
BEST SOURCE FOR LITHIUM COIN CELL RETAINERS AND BATTERY CAGES
W W W. B AT T E R Y H O L D E R S . C O M mpd – P&EE HB 11.18.indd 3
11/7/18 12:46 PM
design world FOLLOW THE WHOLE TEAM ON TWITTER @DESIGNWORLD
Editorial
Creative Services
VP, Editorial Director Paul J. Heney pheney@wtwhmedia.com @dw_Editor
VP Creative Services Mark Rook mrook@wtwhmedia.com @wtwh_graphics
Managing Editor Leslie Langnau llangnau@wtwhmedia.com @dw_3Dprinting
Art Director Matthew Claney mclaney@wtwhmedia.com @wtwh_designer
Executive Editor Leland Teschler lteschler@wtwhmedia.com @dw_LeeTeschler
Graphic Designer Allison Washko awashko@wtwhmedia.com
Senior Editor Miles Budimir mbudimir@wtwhmedia.com @dw_Motion
Videography Services
Senior Editor Mary Gannon mgannon@wtwhmedia.com @dw_MaryGannon Senior Editor Lisa Eitel leitel@wtwhmedia.com @dw_LisaEitel Associate Editor Mike Santora msantora@wtwhmedia.com @dw_MikeSantora
Director, Audience Development Bruce Sprague bsprague@wtwhmedia.com
WTWH Media, LLC 6555 Carnegie Ave., Suite 300 Cleveland, OH 44103 Ph: 888.543.2447 FAX: 888.543.2447
Videographer Bradley Voyten bvoyten@wtwhmedia.com @bv10wtwh Videographer Derek Little dlittle@wtwhmedia.com Videographer Graham Smith gsmith@wtwhmedia.com
Finance Controller Brian Korsberg bkorsberg@wtwhmedia.com Accounts Receivable Specialist Jamila Milton jmilton@wtwhmedia.com
Digital Marketing
Events
Digital Marketing Director Virginia Goulding vgoulding@wtwhmedia.com @wtwh_virginia
Events Manager Jen Kolasky jkolasky@wtwhmedia.com @wtwh_Jen
Webinar Manager Lisa Rosen lrosen@wtwhmedia.com
Events Marketing Specialist Christina Lograsso clograsso@wtwhmedia.com @wtwh_christina
Webinar Coordinator Halle Kirsh hkirsh@wtwhmedia.com Digital Marketing Coordinator Lexi Korsok lkorsok@wtwhmedia.com @wtwh_Lexi Leadlift Onboarding Specialist Mike Ulanski mulanski@wtwhmedia.com @wtwh_mike
Digital Media/Web/ Development Web Development Manager B. David Miyares dmiyares@wtwhmedia.com @wtwh_WebDave
Senior Digital Media Manager Patrick Curran pcurran@wtwhmedia.com @wtwhseopatrick
Production Services Customer Service Manasger Stephanie Hulett shulett@wtwhmedia.com Customer Service Representative Tracy Powers tpowers@wtwhmedia.com Customer Service Representative JoAnn Martin jmartin@wtwhmedia.com Customer Service Representative Julie Ritchie jritchie@wtwhmedia.com
Digital Production Manager Reggie Hall rhall@wtwhmedia.com Digital Production Specialist Brian Furda bfurda@wtwhmedia.com Multimedia Specialist Jaeden Froese jfroese@wtwhmedia.com
DESIGN WORLD does not pass judgment on subjects of controversy nor enter into dispute with or between any individuals or organizations. DESIGN WORLD is also an independent forum for the expression of opinions relevant to industry issues. Letters to the editor and by-lined articles express the views of the author and not necessarily of the publisher or the publication. Every effort is made to provide accurate information; however, publisher assumes no responsibility for accuracy of submitted advertising and editorial information. Non-commissioned articles and news releases cannot be acknowledged. Unsolicited materials cannot be returned nor will this organization assume responsibility for their care. DESIGN WORLD does not endorse any products, programs or services of advertisers or editorial contributors. Copyright© 2018 by WTWH Media, LLC. No part of this publication may be reproduced in any form or by any means, electronic or mechanical, or by recording, or by any information storage or retrieval system, without written permission from the publisher. Subscription Rates: Free and controlled circulation to qualified subscribers. Non-qualified persons may subscribe at the following rates: U.S. and possessions: 1 year: $125; 2 years: $200; 3 years: $275; Canadian and foreign, 1 year: $195; only US funds are accepted. Single copies $15 each. Subscriptions are prepaid, and check or money orders only. Subscriber Services: To order a subscription or change your address, please email: designworld@halldata.com, or visit our web site at www.designworldonline.com POSTMASTER: Send address changes to: Design World, 6555 Carnegie Ave., Suite 300, Cleveland, OH 44103
4
DESIGN WORLD — EE NETWORK
STAFF & CONTENTS PAGE AUTO HB 8-18 V1.indd 4
11 • 2018
eeworldonline.com | designworldonline.com
11/7/18 12:31 PM
inside TH E POWE R & E N E RGY E FFICI E NCY HANDBOOK DESIGNWORLDONLINE.COM | EEWORLDONLINE.COM
NOVEMBER 2018
10
23
02 Don't let the LED blue light blues get you down 06 Prepare for IEC 62368-1, the new hazard-based safety standard Get ready to evaluate safety measures in a different way. 10 How long can a battery really last? It pays to know how a testing regime can reveal whether specific types of lithium batteries will last as long as application demands. 14
18
Maximizing efficiency and safety with rechargeable batteries Some rechargeable lithium battery chemistries need more TLC than others when it comes to safety precautions. Optimizing the energy efficiency of converters and controllers Here's how to evaluate efficiencies associated with components that make up widely used power conversion circuitry.
23 Capacitors in renewable energy It pays to know the distinguishing features of electrolytic, film, and super capacitors when planning applications where energy efficiency is important.
eeworldonline.com | designworldonline.com
STAFF & CONTENTS PAGE AUTO HB 8-18 V1.indd 5
41
45
27 Transforming ac to dc without transformers Substituting special capacitors for transformers helps realize ac-to-dc power supplies that are both efficient and compact. 32 Evaulating options for USB Type-C connections Port controllers, transcievers, and special dc/dc converters now simplify the task of adding USB type-C functions. 36 GaN FETs boost efficiency in electric vehicle chargers The role gallium-nitride FETs play in the design of an EV charging circuit help illustrate how these widebandgap semiconductors facilitate energy efficiency. 41
Don't be led astray on three-phase power measurements Measurements made on one power analyzer may differ from those on another because different manufacturers employ different settings and wiring assumptions.
45 Energy efficiency, smart lighting, and the IoT Sure, smart lights will save energy, but they'll also be the backbone of systems that can help minimize mini-catastrophes.
11 • 2018
DESIGN WORLD — EE NETWORK
5
11/7/18 12:31 PM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
Prepare for IEC 62368-1, the new hazard-based safety standard JEFF SCHNABEL • CUI
Get ready to evaluate safety measures in a different way. THE FAMILIAR 60950-1 AND 60065 SAFETY STANDARDS for information and communication technology (ICT) and audio-visual (AV) equipment are being phased out in 2020, in favor of a single, harmonized standard based on a more modern approach to safety and testing. OEMs must act now to be sure their products are compliant by the deadline. The new electrical safety standard IEC 62368-1 supersedes IEC 60950-1 and IEC 60065. Coordination between the North American and European certification bodies, UL/CSA and Cenelec, means regionalized 62368-1 standards will take over fully in each territory from December 20, 2020. Manufacturers producing products covered by 60950-1 or 60065 should be preparing now to handle the transition.
More widespread adoption of 62368-1 is ongoing. Authorities in Australia/New Zealand and Japan have already published their own equivalents, Mexico has introduced NMX-I-623681-NYCE-2015 on a voluntary basis, and bodies in China, Korea and South America are at the evaluation stage. There are several drivers behind the move to the new standard. The outgoing 60950-1 and 60065 standards apply to ICT and AV equipment respectively, but the distinction between these two categories has become increasingly blurred as new technologies and new markets have developed. 62368-1 unifies and consolidates the two. Clause 1 of the documentation describes the scope of 62368-1, and Annex A presents a nonexhaustive list of product types covered.
These include computers and networking products, consumer electronics, office equipment like copiers or shredders, telecom products and audio/video, information and communication technology. However, the change goes beyond harmonization or consolidation. 623681 replaces the prescriptive approach common in older standards, which tend to specify safety features and the way they are implemented, in favor of HazardBased Safety Engineering (HBSE). HBSE takes a performance-oriented approach that offers greater flexibility for standards-makers and designers, while also enhancing safety for end users. It lets manufacturers find innovative ways to ensure the safety of their products and at the same time relieves the need
CUI has already upgraded a large portion of its external power supply line to the 62368-1 standard, with the intention of moving its entire power portfolio to the new standard well before the deadline.
6
DESIGN WORLD — EE NETWORK
CUI – P&EE HB 11-18 V3 FINAL.indd 6
11 • 2018
eeworldonline.com | designworldonline.com
11/6/18 11:37 AM
PREPARE FOR IEC 62368-1
T H E T H R E E - B LO C K M O D E L for frequent updating of the standard to keep pace with technological advances. Rewriting standards is expensive and time-consuming, and usually behind the curve of progress. In addition, the adoption of this approach is expected to minimize differences between regional or national interpretations of the IEC reference standards. HBSE principles As a safety-science discipline, HBSE has developed over the last 25 years and has shaped the content of standards such as ECMA-287 Safety of Electronic Equipment, which was published in 2002. The IEC Technical Committee, TC108, which created IEC 62368-1 and is also responsible for IEC 60950-1 and IEC 60065, has been committed to HBSE for a similar period. HBSE looks at hazards as energy sources capable of causing pain or injury to an operator. It calls for product designers to identify energy sources, quantify the energy produced, and classify according to the severity of the hazard. By identifying how the energy can be transferred to a body part of the user, it then requires designers to determine appropriate safeguards for people and property and to measure the effectiveness of the safeguards. IEC 62368-1 defines several types of safeguards, including equipment safeguards such as insulation or protective earthing, installation safeguards, instructional safeguards, precautionary safeguards and skill
The three-block model allows flexibilit y in the design and implementation of safeguards. Here’s the model as applied to energy transfer.
Here’s how the three-block model is used in analyzing a safeguard.
safeguards that rely on the knowledge and experience of skilled operators. To support this approach, energytransfer mechanisms and safeguards are each assessed according to a threeblock model. In keeping with HBSE principles, IEC 62368-1 contains clauses that refer specifically to various forms of energy, including electrical, thermal, chemical, kinetic and radiated energy such as optical, acoustic or others. Many items of ICT/AV equipment will contain more than one of these types of energy sources. For electrical sources, the energy contained depends on both the voltage and the current. Moreover, the limits specified for class 1, 2 and 3 also depend on the frequency. Up to 1 kHz, the ES1 limit is 30 Vrms, 42.4 Vp and 60 Vdc. The ES2 limit is 50 Vrms, 70.7 Vp and 120 Vdc.
Energy Source
Effect on the body
Effect on combustible materials
Class 1
Not painful, but may be detectable
Ignition not likely
Class 2 Class 3
Painful, but not an injury
Ignition possible, but limited growth and spread of fire
Injury
Ignition likely, rapid growth and spread of fire
In a similar way to the outgoing 60950-1 and 60065 standards, 623681 defines several categories of users, namely: ordinary person, skilled person, and instructed person. An ordinary person should not be exposed to energy sources capable of causing injury. At least one safeguard must be provided against class 2 energy sources, and a double or reinforced safeguard for class 3 sources. A skilled person is expected to be capable of protecting themselves against class 2 and class 3 energy sources, although some safeguards may be required. An instructed person is directly aided by a skilled person and may access energy sources in higher classes than are permitted for an ordinary person but cannot be exposed to energy levels capable of causing injury. As the least hazardous category, class 1 energy sources require no equipment safeguards to prevent access by any category of user. The published 62368-1 standard describes the rules on electrically caused injury in Clause 5 of the document. Subsequent clauses cover electrical fire, thermal burn injury and radiation.
Energy sources are classified into three levels according to potential for harming the user and fire risk.
eeworldonline.com | designworldonline.com
CUI – P&EE HB 11-18 V3 FINAL.indd 7
11 • 2018
DESIGN WORLD — EE NETWORK
7
11/26/18 2:45 PM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
S I Z I N G U P E N E R GY S O U R C E S
Making the transition The fact that UL/CSA and Cenelec have coordinated their changeovers helps OEMs manage the transition efficiently and ensure their products remain marketable on both sides of the Atlantic after the old standards are withdrawn. As far as the EU is concerned, Dec. 20, 2020 represents the Date of Withdrawal of 60950-1 and 60065 and is also (following a June 18, 2018 announcement by the EU that harmonized the two deadlines) the date for cessation of presumptions of conformity with other standards that reference the outgoing standards – such as the Low-Voltage Directive (LVD). From this time, the European version, EN 62368-1, will be the only acceptable standard for the categories of products covered. In North America, the UL/CSA joint body refers to Dec. 20, 2020 as the Effective Date, from which UL 62368-1 will supersede UL 60950-1 and UL 60065 and they will be withdrawn. Although new submissions will be tested according to 62368-1, legacy products that comply with the outgoing standards will not be subjected to an Industry File Review.
8
DESIGN WORLD — EE NETWORK
CUI – P&EE HB 11-18 V3 FINAL.indd 8
Classification of electrical energy sources in IEC 62368-1 in relation to their potential to cause injur y. The diagram also shows that electrical energy sources are classified according to either the voltage or current but need not meet limits for both simultaneously.
Currently, 62368-1, 60950-1 Other useful references and 60065 are effective, both in include the document IEC/TR the EU and North America. While 62368-2, Ed. 2, “Audio/video, this eases the transition, OEMs can information and communication and should engage with 62368-1 as technology equipment – Part 2: soon as possible and begin testing Explanatory information related their products in accordance with to IEC 62368-1”. Available from the new standard. the IEC store and elsewhere, this As with 60950-1 and 60065, provides extra background and 62368-1 applies to certain explanations of sub-clauses. components and subsystems within In addition, the UL the product as well as to the endorganization has published a product itself. This includes built-in transition toolkit, which contains power supplies, or external power the UL standard document, adapters that ship in the box with white papers and brochures, the product. To help OEMs manage their existing and access to training resources 60950-1 and 60065 certified parts inventories, the including webinars and regional latest edition of 62368-1, Edition 2, contains the IEC 62368-1 workshops. following clause, known as clause 4.1.1: Companies that have not “Components & subassemblies that comply already begun their 62368-1 with IEC 60950-1 or IEC 60065 are acceptable journey may find that a good as part of equipment covered by this standard place to begin is the CUI without further evaluation other than to give white paper on “IEC 62368-1: consideration to the appropriate use of the An Introduction to the New component or subassembly in the end-product.” Safety Standard for ICT and TC108 has also amended 60950-1 and 60065 AV Equipment,” for more to permit newer 62368-1 certified parts to be used background information on in products compliant with the outgoing standards. the changes. While these allowances ease the transition, it must be noted that clause 4.1.1 is temporary. It will be withdrawn in the EU on Dec. 20, 2020. UL/CSA is awaiting a decision by TC108 on REFERENCES whether to extend the clause into Edition 3 of 62368-1, which is expected to UL Toolkit for IEC 62368-1 be published sometime in 2019. www. 62368-ul-solutions.com/ To be sure of having the necessary toolkit.html approvals in place ready for the December 2020 deadline, the time to act is now. As White paper, An Introduction to the deadline approaches, test houses are the new safet y standard for ICT likely to be under pressure and lead times and AV equipment, could be long. www.cui.com/catalog/resource/ Studying the standard is essential. iec-62368-1-an-introduction-toClause 0 of the currently published version t he-new-safet y-standard-for-ictoffers an overview of the standard and its and-av-equipment.pdf history, objectives and approaches.
11 • 2018
eeworldonline.com | designworldonline.com
11/6/18 11:38 AM
Call us at 1.800.463.9272
newark.com
Discover Over a Million
Engineering Products from Suppliers You Know and Trust
§ 900+ new products added each week § Custom services such as kitting, panel meters, enclosures, and many more § Market-leading online community of over 500,000 engineers
newark – P&EE HB 11.18.indd 9
11/7/18 12:47 PM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
How long can a battery really last? S O L JAC O B S • TA D I R A N B AT T E R I E S
It pays to know how a testing regime can reveal whether specific types of lithium batteries will last as long as an application demands. DESIGNERS MUST UNDERSTAND how their end device will consume power before they can choose a battery. Will the battery be depleted in a matter of hours, as for an army manpack communications radio? Will the device spend most of its time in sleep mode and transmit data bursts several times daily, as is the case with most wireless sensors? Does the device require a 10+ year operating life? Answers to these and other key questions will help determine the ideal power supply. You may want to run your own tests -- engineers at Tadiran can assist you in establishing the proper test procedures. Make sure you run ‘apples to apples’ comparison tests using similar types of batteries, as comparing totally different types of batteries could yield highly inaccurate results. For example, it is unrealistic to compare cells designed to supply moderate rates of current with those designed to supply low average daily rates of current. Among primary lithium batteries, there is a natural trade-off between two kinds of cells: those designed to work for the long-term with a low usage rate and low-selfdischarge; and those offering short-term operation with a higher usage rate but higher self-discharge. Certain lithium batteries, including lithium thionyl chloride (LiSOCl2) chemistry with spirally wound construction, lithium sulfur dioxide, and lithium manganese dioxide, can deliver energy at a high rate for
I m a g e c o u r t e s y Ta d i r a n B a t t e r i e s
10
DESIGN WORLD — EE NETWORK
tadiran – P&EE HB 11-18 V4 FINAL.indd 10
11 • 2018
eeworldonline.com | designworldonline.com
11/7/18 3:19 PM
HOW LONG CAN A BATTERY LAST
B O B B I N - ST Y L E L i S O C l 2 relatively brief periods. Spirally wound LiSOCl2 cells have electrodes with large surface areas which support higher rates of output. However, the downside of spiralwound construction is a significantly higher self-discharge rate. By contrast, bobbintype LiSOCl2 cells feature much less electrode surface area, supporting a lower rate of output along with a lower rate of self-discharge. A simple analogy between these two types of batteries would be an 8-oz. (capacity) glass of water. Over time, the large opening of the glass allows the water to evaporate faster (higher self-discharge). But the larger opening allows the water to pour out rapidly (higher rate of output). The wider opening serves up energy immediately in applications thirsty for large amounts of energy. However, if you want to keep that water around for an extended period of time, the wider opening becomes problematic, allowing the water to evaporate more quickly. Certain bobbin-type LiSOCl2 cells, including the Tadiran 59XX IXTRA Series, were designed to supply a moderate rate of discharge over a relatively long operating life with a low annual selfdischarge rate ranging from 1 to 2% a year. In storage, these batteries would use up approximately 10-20% of their initial capacity over ten years. This performance has analogies to a soda can with a moderate opening -- the smaller opening lets the soda evaporate more slowly than from a large glass with the trade-off that the liquid can’t pour out as fast. Tadiran manufactures the XOL EXtended Operating Life cells (49XX family), which features the lowest selfdischarge rate of any lithium battery commercially available, which allows you to draw a low rate of current while featuring a very low rate of annual self-discharge, just 0.7% per year. Think of it as a bottle of water with a small opening and a straw that allows only a sip of water, while limiting evaporation to a slow rate. To create a bobbin-type LiSOCl2 battery able to deliver high current pulses while offering a low rate of annual
eeworldonline.com | designworldonline.com
tadiran – P&EE HB 11-18 V4 FINAL.indd 11
self-discharge, Tadiran developed PulsesPlus batteries. They contain a patented hybrid layer capacitor (HLC) to store and release high pulses of energy without a higher rate of self-discharge. The analogy is a plastic bottle: Squeeze it when you want to squirt the water (high pulse). When you are not squeezing the selfdischarge is still extremely low, similar to the XOL cell. With any battery that uses a capacitor, the capacitor should never get fully exhausted of energy while suppling the pulse. This has much to do with the capacity of the capacitor in Farads. If capacitor charge capacity is depleted during the generation of the pulse, the battery might have to supply some of the pulse energy. In cold temperatures or toward the end of the battery's useful life, the resulting current drain could cause the battery voltage to drop too low and cause the device to shut-off. HLCs manufactured by Tadiran have significant energy storage capacity starting at about 100 F. With so much capacity, there is little chance the battery will ever be called upon to directly supply more than the low current necessary to charge the HLC. By contrast, standard capacitors can have a much smaller capacity ranging from a few hundred millifarads to about 10 F. Comparative testing Consider our analogy again: It takes longer to pour 8 oz. of water through a small straw than from a glass. Similarly, it takes much longer to run a test that shows how much water is in the can than one that shows the amount in the glass. You also need to make sure that you are
11 • 2018
A b o b b i n - st y l e L i S O C l 2 b a t te r y i s c h a ra cte r i ze d by a cy l i n d e r i c a l c a t h o d e s u r ro u n d e d by a n o d e material. This st yle of batter y construction is e a sy to m a n u fa ct u re, h a s a l ow s e l f- d i s c h a rg e ra te, a d n e e d s n o s a fet y f u s e.
comparing straws to straws and cans to cans. Otherwise it will look like the cans and glasses have more capacity than the bottles, which is not true. Also, how do you properly measure the evaporation (self-discharge)? Needless to say, properly testing a battery is time-consuming and intensive process that demands appropriate test methods to yield valid results. Tadiran has been designing and manufacturing lithium primary batteries for over 30 years. The company tests its batteries under various conditions, covering almost every possible scenario. In addition, it collects data based on factors such as cell size, temperature, load size, and so forth, under different loads. This knowledge has been organized into a comprehensive matrix of battery status information covering many different application requirements.
DESIGN WORLD — EE NETWORK
11
11/7/18 3:21 PM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
When recommending a battery for use in a specific application, Tadiran draws upon a vast library of historical test points, including OC (over charge) voltage, voltage under load, annual self-discharge rate, lowvoltage point at load, and temperature. This comprehensive data matrix is then used to calculate long-term dissipation of available cell capacity based on application-specific requirements. These calculations are based on actual long-term test data, not theoretical projections. Here is a brief synopsis of the various types of testing methods that Tadiran employs: Actual long-term testing – Tadiran’s lab has batteries still being tested after 25+ years and which are still operating. These batteries are being tested under different loads and test profiles. Accelerated testing - using a method called Arrhenius test (two-fold increase of reaction rate for every 10°C) we can reduce the amount of time needed to simulate longterm operation. These tests take place at 72°C, which is the equivalent of about 32 times the theoretical lifetime of a battery at 22°C. It is important to properly run and interpret results from this test, as an inferior cell that suffers from passivation could indicate a false positive result. For this reason, it is essential not only to store the cell at 72°C prior to testing, but to actually test the cell during storage at 72°C. Here is one example of this type of testing: Some cells are tested at 72°C for one month at the one-month discharge rate (depleting the cells). Others are tested for two months at the two-month discharge rate. Still others are tested for 3-,4-,5-, 6-month and one-year discharge rates. At the one- and two-month discharge rates, Tadiran XOL cells have demonstrated low capacities because they are not designed to discharge quickly. (Recall a lithium battery can have a high usage rate with high selfdischarge or low usage rate with low selfdischarge rate, but NOT both). By contrast, Tadiran IXTRA cells, as well as competitive products, will show higher capacities at
12
DESIGN WORLD — EE NETWORK
tadiran – P&EE HB 11-18 V4 FINAL.indd 12
fast discharge rates. But these tests also indicate that their self- discharge rate will also be higher. Starting with month three, testing at the three-month discharge rate at 72°C (the equivalent of eight years at 22°C) will show competing cells to have higher self-discharge rates, indicating that Tadiran cells can deliver Spiral Wound Bobbin Primary Cell a longer operating life Primary Cell thanks to lower selfdischarge. The longer S p i ra l -wo u n d c e l l s h a ve a l a rg e c o m m o n s u r fa c e a re a the test proceeds with between the anode and cathode, giving them a high c u r re nt c a p a b i l i t y. B o b b i n - st y l e c e l l s h a ve l e s s c o m m o n a low usage rate, the a n o d e/c a t h o d e s u r fa c e a re a w h i c h l i m i t s t h e i r d i s c h a rg e better Tadiran cells will ra te, b u t a l s o l i m i t s t h e i r s e l f- d i s c h a rg e ra te. perform because of their lower self-discharge. To confirm this, Tadiran has run this test on cells for 90 months at the lithium anode in the battery after 72°C (the equivalent of hundreds of years of specific test conditions (e.g. partial continuous operation). discharge, temperature soaking and Calorimeter testing: A battery is so on). surrounded by water and then calorimeters An example might be after a are used to look for a rise in temperature battery has been tested for several caused by self-discharge. This test runs only months under elevated temperature after a battery has completely stabilized for and various discharge currents. It is a period of one year prior to being tested. then cut open to dissolve the remaining Field results: Laboratory test results lithium. Results from the titration can continue to be proven and confirmed by predict annual battery self-discharge as actual results from the field. Tadiran often a function of the applied currents and/ checks random samples of batteries in or temperature, because the higher long-term use to see if the amount of lithium the self-discharge rate, the less lithium left in the battery coincides with the results remaining in the cell. from the lab. Another valuable metric is the Competitive Testing: Tadiran number of Failures In Time (FITs) measured routinely runs lab tests on competing in billions of device operating hours in batteries. But first, we typically operate the field. Tadiran’s historical average FIT the battery for one year, which allows is between 5 and 20 batteries per billion, any impurities in the electrolyte to show extremely low compared to the industry up in the test results. average. Lithium Titration: There are specialized applications with unusual requirements where sufficient data points may not REFERENCES be available (temperature extremes, prolonged high current pulses, short life Tadiran Batteries time applications etc.). In these cases, www.tadiranbat.com Tadiran measures the remaining content of
11 • 2018
eeworldonline.com | designworldonline.com
11/7/18 2:51 PM
Get Ahead of the Thermal Curve
Measure Power, Monitor Temperature and Turn on Your Fan Before it Gets Too Hot! How many parts does it take to accurately measure temperature and manage power? With our high-side current sensors, it could be as few as one. For example, the EMC1701/2/4 family has one current sensor and can monitor one, two or four temperature channels respectively. Power consumption has long been a leading indicator for thermal management. Measuring diverse power sources with multichannel chips closes the thermal information gap. For example, the PAC1933 can simultaneously measure a 1V Field Programmable Gate Array (FPGA), USB Type-C™ at 20V and a memory rail. Review our entire offering of high-side current sensors and DC-power monitors, including 36-hour on-chip accumulators and 16-bit precision multirail monitors.
microchip.com/DC-Power-Monitor The Microchip name and logo and the Microchip logo are registered trademarks of Microchip Technology Incorporated in the U.S.A. and other countries. All other trademarks are the property of their registered owners. USB Type-C and USB-C are trademarks of the USB Implementers Forum. © 2018 Microchip Technology Inc. All rights reserved. 10/18 DS20006057B
microchip 2 – P&EE HB 11.18.indd 13
11/7/18 12:55 PM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
Maximizing efficiency and safety
TO M B L A H A • M E M O RY P R OT E C T I O N D E V I C E S
with rechargeable batteries Some rechargeable lithium battery chemistries need more TLC than others when it comes to safety precautions.
RECHARGEABLE BATTERY TECHNOLOGY has come a long way, continually boosting performance with fewer annoying memory effects that severely limited the useful operating life of earlier generation cells. The most common rechargeable battery chemistries commercially in use include nickel cadmium (NiCd), nickelmetal hydride (NiMH), lead acid, lithium ion (Li-ion), lithium ion polymer (Li-ion polymer), and reusable alkaline cells. NiCd batteries suffer from low energy density (45-80 Wh/kg) but deliver long life and high continuous discharge rates. These cells contain toxic materials. Common NiCd applications include two-way radios, biomedical equipment, video cameras, and power tools. NiMH cells feature higher energy density (60-120 Wh/ kg) but a shorter operating life. These cells contain no toxic materials. Common NiMH applications include mobile phones and laptops. Lead-acid batteries have a low energy density (30-50 WH/kg), which is acceptable for large equipment where the added size and weight are often not critically important. These applications include automobiles, wheelchairs, emergency lighting, and uninterruptible power supplies (UPS). Lithium-ion (li-ion) chemistry has become widely used for delivering relatively high energy density (100-130 Wh/kg) that aids in miniaturing products. Consumer grade Li-ion batteries can operate for approximately five years and 500 full recharge cycles. Industrial grade Li-ion batteries can operate for up to 20 years and 5,000 recharge cycles and feature an extended temperature range. Li-ion polymer batteries present this chemistry in an ultra slim form factor that is a good candidate for low-profile applications such as notebook computers and cell phones.
14
DESIGN WORLD — EE NETWORK
MPD – P&EE HB 11-18 V4 FINAL.indd 14
11 • 2018
An example of a batter y holder for A A A A batteries is MPD’s part number BE-2A A A A-PC. It sits 8.50 mm above the PCB and comes in 100-piece bags. The holder is black ABS, and the spring contacts are nickelplated spring steel.
Reusable alkaline batteries have an initial energy density of 80 Wh/kg and help to address the growing environmental challenge surrounding alkaline battery disposal. Alkaline rechargeable batteries have severe performance limitations, including a short lifespan and the ability only to be recharged to 50% of their initial capacity. When deciding among these various chemistries, design engineers must review such parameters as battery life, load qualities, self-discharge, and initial price versus long-term cost of ownership. Invariably, the battery specification process involves some form of compromise that balances the strengths and weaknesses of each chemistry. Lithium-ion battery safety Despite the growing use of Li-ion batteries, their safe operation requires vigilance. Designing for max battery runtime requires packing more active materials into each cell, forcing the separator to become
eeworldonline.com | designworldonline.com
11/26/18 2:43 PM
…
EE-P&E-THiNK Lighthouse+CAFH 11-18p.indd 1 keystone – P&EE HB 11.18.indd 15
10/9/18 11/7/18 11:50 12:46AM PM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
Crack open a batter y pack for a laptop computer and you will find both batter y cells and a built-in protection circuit module, similar to those visible here. High-energy batteries such as those in laptops and other consumer electronics goods t ypically incorporate these protective circuits.
thinner. Too many microscopic particles converging on one spot can cause a short circuit, resulting in a sizable flow of current between the positive and negative plates. This action could result in thermal runaway, where one failing cell begins a chain reaction that compromises surrounding cells. For this reason, Li-ion battery packs often incorporate dividers to protect adjacent cells. There are two common types of lithium-ion chemistry: cobalt and manganese (spinel). Manganese offers superior thermal stability, sustaining temperatures of up to 250°C (482°F) before becoming unstable. Cobalt cells feature a low internal resistance, thus delivering higher current on-demand for power tools and medical devices. The trade-off is lower energy density, as a cathode made of pure manganese offers about half the capacity of cobalt. Li-ion batteries often mix the two metals, then provide a safety margin by limiting the amount of active materials and by including three additional levels of safety protection: an electronic protection circuit (PTC) that inhibits high current surges; a circuit interrupt device (CID) that opens the electrical path if an
16
DESIGN WORLD — EE NETWORK
MPD – P&EE HB 11-18 V4 FINAL.indd 16
excessively high charge voltage raises the internal cell pressure to 10 Bar (150 psi); and a safety vent that allows a controlled release of gas in case of a rapid rise in cell pressure. The typical PTC contains a solidstate switch that automatically switches on if the charge voltage reaches 4.30 V. A fuse also cuts the current flow if the skin temperature of the cell approaches 90°C (194°F). To prevent the battery from over-discharging, a control circuit cuts off the current path at approximately 2.5 V/cell. A patented printed circuit board (PCB) can offer added safety protection by limiting the voltage range from 2.5-4.25 V. Under normal circumstances, a Li-ion battery will simply power down in the event of a short circuit. However, certain defects such as microscopic metal particle contamination may go undetected, in which case the PTC may not be able to stop the thermal runaway once it begins. Other notable safety concerns involve static electricity or poorly made battery chargers that can destroy the PTC, causing the solid-state switch to remain stuck in the on position without
11 • 2018
the user’s knowledge. Another potential concern is cold temperature charging, as consumer grade Li-ion batteries cannot be charged below 0°C (32°F). Although they may appear to be charging normally, metallic lithium can plate onto the anode while on a sub-freezing charge, and the plating is permanent and cannot be removed. If done repeatedly, such damage can compromise the safety of the pack. The battery will become even more vulnerable to failure if subjected to impact, crush, or high-rate charging. By contrast, industrial grade Li-ion batteries can be recharged and discharged at extremely cold temperatures. Choosing a battery holder Battery holders can accommodate the extended length of a cylindrical cell with a built-in PTC. For example, standard battery holders are available to securely hold between one and four 18650 Li-ion cells, with similar holders available for other widely used sizes. Custom battery holders are also available that can accommodate circuit protect modules (CPMs) that are not integrated into the battery. Problems often arise when hobbyists try to assemble their own battery packs without considering battery safety. For example, consumers often manipulate cell phone batteries to extend their operating life, sometimes with devastating results.
eeworldonline.com | designworldonline.com
11/6/18 2:19 PM
MAXIMIZING EFFICIENCY AND SAFETY
For this reason, most laptops contain embedded code that only permits the use of an OEM-approved battery pack. Design engineers and procurement professionals must be wary about specifying inferior-quality Li-ion batteries that could contain high levels of impurities, lack a properly designed or manufactured PTC, or omit other time-proven safeguards. Battery safety is a growing concern, especially for airplane travel. So be cautious and always specify a reputable brand of Li-ion battery with built-in PTC protection, then secure it with the right battery holder. Cylindrical-shaped rechargeable Li-ion batteries come in various sizes, including ½ AA, AAAA, AAA, AA, A, B, C, Sub-C, D, F, N, A23, A27, BA5800, Duplex, and 4SR44. Rechargeable Li-ion cylindrical cells are designed to be slightly larger than Alkaline batteries to accommodate the internal protection circuitry that prevents over-discharging and short-circuit damage. Button cells and coins cells have a distinctive disc-like shape that is usually wider in comparison to height. The most common type of coin cell is the CR2032 Li-ion cell. Li-ion polymer cells are constructed with a series of flat layers to maintain a low profile and typically require the use of a wire harness instead of positive and negative terminals. All rechargeable cells can be configured into battery packs that combine multiple cells into a self-contained unit. The choice of battery chemistry, the size and shape of the cell, and the size and packaging of the device all dictate the need for a battery holder. Standard battery holders are available for virtually every type of rechargeable battery, including cylindrical cells, rectangular batteries, coin cells, button cells, and many common battery packs. Custom battery holders are also available to meet specific application requirements. It is critical to choose a battery holder that properly secures the battery against shock and vibration, thus ensuring a rugged and reliable electrical connection.
eeworldonline.com | designworldonline.com
MPD – P&EE HB 11-18 V4 FINAL.indd 17
Examples of batter y holders designed specifically for Li-ion cells include the BK-18650-PC2 and PC4 with pressure contacts. The holders sit 21.44 mm above the PCB and have a temperature range of -40 to +180°C. They have flammabilit y ratings of UL94V-0. The contacts are nickel-plated stainless spring steel. The holders are designed for cells with built-in protection circuit modules (PCMs). The parts are RoHS2 compliant for worldwide distribution and surpass ANSI/EIA standards.
The shape and packaging of the device, the space available for the battery, and special requirements for miniaturization can impact the choice of battery holder. If the application involves exposure to harsh environments, including extreme heat and humidity, the battery holder must be constructed from materials that can endure such extreme conditions. For example, gold-plated contacts are highly recommended to provide greater corrosion resistance for battery holders used in hot, humid climates. Application-specific requirements will dictate the ideal rechargeable battery and its accompanying battery holder.
REFERENCES Memor y Protection Devices www.memor yprotectiondevices.com
11 • 2018
DESIGN WORLD — EE NETWORK
17
11/6/18 2:19 PM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
Optimizing the energy efficiency of converters and controllers Here’s how to evaluate efficiencies associated with components that make up widely used power conversion circuitry. IN SOME WAYS, selecting a module, converter or controller for power conversion resembles deciding whether to take a flight, board a train, or drive from point A to B. Once you’ve decided an airplane is your mode of transportation, for example, it’s easy to pick an airline to fly. Let’s start with the definition of each device type: A controller can be used with external field-effect transistors (FETs), inductors and capacitors. A converter is a controller plus integrated FETs that uses external inductors and capacitors. A module integrates both the FETs and the inductor but uses external capacitors. It’s clear that with the progression of integration from controllers to modules, the main heat dissipaters (FETs, inductor) are brought in from the outside. The obvious advantage with integration is smaller size and ease of use. The clear disadvantage is the thermal challenge; you’ll need to optimize the devices to handle the thermals and still operate at a high ambient temperature. Because thermals are directly related to power loss – which translates to efficiency – we’ll focus on efficiency improvement methodologies, or “knobs,” for each device option: modules, converters and controllers. As you progress toward more integration, there are fewer knobs available to optimize efficiency. That’s because converter and module manufacturers try to
18
DESIGN WORLD — EE NETWORK
texas instruments – P&EE HB 11-18 V3 FINAL.indd 18
11 • 2018
F R A N K D E STA S I , M AT H E W JA C O B TE X AS INSTRU M E NTS INC.
CONTROLLER
CONVERTER
MODULE The knobs available for optimizing efficiency in controllers, converters and modules.
eeworldonline.com | designworldonline.com
11/6/18 11:52 AM
e
No place to replace a battery.
PROVEN
40 YEAR OPERATING
LIFE
*
Highly remote locations call for Tadiran batteries. Battery replacement is costly and often dangerous work. Reduce the risk with Tadiran bobbin-type lithium thionyl chloride (LiSOCl2) batteries. With an annual self-discharge rate of just 0.7% per year, Tadiran LiSOCl2 batteries enable low power consuming wireless devices to operate for up to 40 years on a single battery, up to 4 times longer than the competition. Our batteries also feature the highest capacity, highest energy density, and widest temperature range of any lithium cell, plus a glass-to-metal hermetic seal for added ruggedness and reliability in extreme environments.
ANNUAL SELF-DISCHARGE TADIRAN
COMPETITORS
0.7%
Up to 3%
Take no chances. Take Tadiran batteries that last a lifetime.
* Tadiran LiSOCL2 batteries feature the lowest annual self-discharge rate of any competitive battery, less than 1% per year, enabling these batteries to operate over 40 years depending on device operating usage. However, this is not an expressed or implied warranty, as each application differs in terms of annual energy consumption and/or operating environment.
Tadiran — IoT HB 04-18.indd 19
Tadiran Batteries 2001 Marcus Ave. Suite 125E Lake Success, NY 11042 1-800-537-1368 516-621-4980 www.tadiranbat.com
11/7/18 12:48 PM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
optimize the internals for efficiency over a wide range of operating conditions. Realistically, the knobs on a controller will be numerous because the internal parameters of the FET constitute knobs that affect switching losses. It is possible to get much better efficiency with similar on-resistance (or RDS[on]) FETs if the other internal parameters of the FET that affect switching losses are lower. In general, the lower the capacitance or charges inside the FET, the less the loss. The most widely used knobs for converters generally include the operation frequency, the switching FET Rdson_HS and Rdson_LS, and the inductance value and series resistance of the coil used to store switching energy. Now we’ll examine each of these knobs in detail and see how they affect efficiency. Each knob affects losses, so start with the loss equation for a buck topology (Pcond and PSW):
Here, D = The switch duty cycle of converter. This is the proportion of the total period that the high-side MOSFET is on; FSW = Converter switching frequency = 1/period; IO = Load current of the dc/dc converter; L = Inductance value of power inductor, h; PCOND= Conduction losses of the system. These losses are due to current flowing in resistances; PCORE = Power loss due to eddy currents and hysteresis in the core of the power inductor; PCOSS = Power loss due to the capacitance of the power MOSFETs, W; PSW = Switching losses. These losses are due to the switching nature of the dc/dc converter and are
20
DESIGN WORLD — EE NETWORK
texas instruments – P&EE HB 11-18 V3 FINAL.indd 20
usually proportional to FSW; QG = Gate charge required to turn on the power MOSFETs; RHS = Resistance of high-side MOSFET (RDSON), Ω; RL = Resistance of the power inductor, Ω; RLS = Resistance of low-side MOSFET (RDSON) , Ω; Tt = Transition time of the switch waveform, sec.; VG = Gate drive voltage for the power MOSFETs, V; VIN = Input voltage to the converter, V; VOUT = Output voltage of the converter, V; and ΔI = Peak-to-peak magnitude of the inductor ripple current. The losses are split into two groups: Pcond and PSW. The total loss is the sum of the two. Pcond represents the conduction losses from current flowing in the various resistances. For a buck converter, this current is Io, the output current. Notice that the conduction loss is proportional to the square of the output current. This loss can grow large for higher-load-current applications. Both the MOSFET and inductor contribute resistances to the converter. In the previous PCOND equation, the term within the square brackets represents the effect of the ripple current in the inductor, ΔI. Because the inductor current contains both dc and ac components, this term accounts for the rise in root-mean-square (RMS) current flowing through the loss-producing resistances. The value of ΔI depends both on the inductor and the switching frequency. It’s the nature of a dc/dc converter that causes switching losses in a buck regulator. A rectangular waveform appears on the switch node of the regulator. This signal switches from nearzero volts to near VIN, with fast transition times. Because the inductor will hold the current constant during these transitions, there will be an overlap between current and voltage, and therefore a power loss. The first term in the previous PSW equation accounts for this loss, with Tt equal to the rise and fall time of the switch waveform. The loss is also proportional to the input voltage and load current. The MOSFETs must switch on
11 • 2018
and off at the switching frequency rate, which requires some energy from the gate driver supply. This gate drive loss is represented by the second term in the previous PSW equation. It depends on the value of the gate driver supply voltage, VG, and the total gate charge for both MOSFETs, QG. PCORE is the core loss of the inductor. Because the inductor has an ac component of current flowing through it, there will be an ac component of core flux. This will cause power loss in the core due to hysteresis and eddy current effects. All of the terms in the previous PSW equation are proportional to the switching frequency, except Pcore. Pcore will usually depend on switching frequency to a higher power than one. The total power loss is then the sum of Pcond and PSW. Frequency Recall that the switching frequency comes into the total loss equation in the switching-loss terms and sets the inductor current ripple along with the inductor value. From the PSW equation, it’s clear that a higher frequency will create more switching losses. These losses arise both in the MOSFET and in the core of the inductor.
I N D U CTO R LO S S
How loss in an inductor varies with switching frequency.
eeworldonline.com | designworldonline.com
11/6/18 11:52 AM
ENERGY EFFICIENCY OF CONVERTERS & CONTROLLERS
You may think that use of the lowest switching frequency would offer the best efficiency. This would be true if the inductor size did not depend on the switching frequency. As will become clear, the trade-off between inductor size and switching frequency is important. The power inductor is a key component in any dc/dc converter. In conjunction with the output capacitor, the inductor smooths the rectangular waveform at the switch node and provides the regulator’s dc output. The inductor is sized based on the desired ripple current, ΔI, using the equation:
It’s typical practice to select a ripple current of about 30 to 40% of the converter’s full-load current. From the above equation, there’s an inverse relationship between switching frequency and inductor value. This implies you can reduce the switching frequency to improve efficiency and choose a larger inductor value to keep the same ripple current. Unfortunately, there are several tradeoffs that get in the way. First, a larger inductor value means a physically larger inductor, occupying more printed circuit board (PCB) area. More important, a larger inductor value will require more turns on the core. More wire in the inductor means more dc resistance for a given physical size. The higher dc resistance will produce more losses and lower the efficiency. The RL term in the PCOND equation models this effect. There will usually be an optimum switching frequency that will provide the lowest losses. As you change the frequency, you adjust the size of the inductor to keep the same ΔI. At high frequencies, the switching losses will rise; at low frequencies, the inductor will get larger and produce more losses from dc resistance.
eeworldonline.com | designworldonline.com
texas instruments – P&EE HB 11-18 V3 FINAL.indd 21
The core losses of the inductor will also depend on switching frequency, sometimes in a complex way, but always rising with frequency. The type of core material plays a large role in core losses. Typically, a powdered-iron core will have more loss at switching frequencies greater than about 1 MHz compared to a ferritetype core. The best way to estimate inductor core loss is to use the graphs or calculators on the inductor manufacturer’s website. Careful choice of the switching frequency and inductor will go a long way toward minimizing overall losses. Setting the switching losses equal to the conduction losses is a widely used approach that provides good results in the majority of cases.
M O S F E T C H A R ACT E R I ST I C S
MOSFET parameters vs. MOSFET size
MOSFETs With a controller, designers must choose the power MOSFETs as well. Depending on the type of controller, at least one high-side FET and perhaps one low-side FET are necessary. In a dc/dc converter, the power FETs serve as switches. In a controller, the parameters important for efficiency are the RDS(on) and the charges required to turn the FETs on and off at the switching frequency. In the PSW equation, these terms are represented as RHS, RLS, QG and Pcoss. The Pcoss term represents the loss when charging and discharging the drain-to-source capacitance of the FETs. The conduction losses from the MOSFETs also depend on the steady-state duty cycles, D and 1-D. It would seem intuitive, then, to choose FETs with the lowest RDS(on) and lowest QG and Pcoss. Unfortunately, the RDS(on) and FET charge parameters are inversely related.
11 • 2018
DESIGN WORLD — EE NETWORK
21
11/6/18 11:53 AM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
HIGH-SIDE FET
The use of parallel FETs for a low-dut ycycle case and a high-dut y-cycle case, respectively. Below, power loss in each low-side FET for parallel combinations. Left, power loss in each high-side FET for parallel combinations. A note of caution when reading these figures: The total loss is the loss of each FET multiplied by the number of FETs. You will reach a point of diminishing returns when the total loss exceeds the loss in a single FET. The key point is that parallel FETs primarily help with thermal management.
LO W- S I D E F E T As the size of the MOSFET grows, the RDS(on) and conduction loss will diminish. However, the gate charge, drain-to-source capacitance, and switching losses will rise. So, again, you will need to trade off RDS(on) and switching losses to minimize overall loss. As with the inductor selection, try to choose a MOSFET so the switching losses equal conduction losses. That method also works well when selecting the relative sizes between the high-side and low-side FETs; size them so the MOSFET conduction losses are equal. From the PCOND equation, you can derive:
Because you know the operating duty cycle, you can choose either the high- or low-side FET and determine the size of the other. Having looked at the various knobs available to optimize efficiency or reduce losses in a buck dc/dc stage, it’s clear there is a lot more leeway to optimize a controller design. It’s a big plus to have control of the FET and inductor sizing.
22
DESIGN WORLD — EE NETWORK
texas instruments – P&EE HB 11-18 V3 FINAL.indd 22
There are some instances where use of parallel FETs in controller design help keep the power dissipation in each FET under control. A rule of thumb is that anytime you use FETs in parallel, losses rise. Even if you reduce the conduction losses, the rise in switching losses dominate, and the total losses go up. Switching losses rise because of REFERENCES the added capacitance when connecting FETs Texas Instruments in parallel. However, the ability to distribute www.ti.com/ww/en/powerthe loss across multiple FETs helps limit the training/login.shtml temperature rise in each FET. In a nutshell, there are various knobs available for designers when choosing a powerconversion device. The knobs specifically affect efficiency. Whether you choose a controller, a converter or a module, a little care in the selection of switching frequency, inductor and MOSFET can ensure that your design is optimally efficient and with minimal temperature rise.
11 • 2018
eeworldonline.com | designworldonline.com
11/6/18 11:53 AM
e
CAPACITORS IN RENEWABLE ENERGY
Capacitors in renewable energy M O R R I E G O L D M A N • I L L I N O I S CA PA C I TO R
It pays to know the distinguishing features of electrolytic, film, and super capacitors when planning applications where energy efficiency is important.
Power film capacitors come in multiple series and are widely used in inverter circuits.
A B A S I C S U P E R CA P A P P RENEWABLE ENERGY SYSTEMS all have a common need for highefficiency power circuits. Whether the original source of energy is light, heat, or mechanical motion, ancillary circuitry will probably either store energy, convert it from dc to ac or vice versa, or even from one level of dc to another. Capacitors are a common ingredient in all of those scenarios. Capacitors are one of those components that engineers often A simplified supercap-based energy har vesting take for granted. Unlike integrated circuit, while not a practical complete system, circuits or power semiconductors, illustrates the concept. On the far left, the their technology hasn’t changed input voltage could originate from a PV cell, a radically every few years. That piezoelectric device or other mechanical energy conversion device. The supercap C1 stores energy. said, capacitors have been steadily Diode D1 prevents stored supercap energy from evolving to perform better and being fed back to the source. Before connecting last longer. And some of the to the load, some t ype of over-voltage protection newer technologies, such as is needed. In a practical application, additional supercapacitors (also called doublecircuitr y is needed to more efficiently capture and regulate the power. layer capacitors) have become much less expensive in recent years, making them affordable in more applications than ever. Green energy power sources are usually thought to be designed as lowmaintenance systems because they are often off-grid resources, in remote locations. Because of the demands placed on these circuits, and the difficulty of performing field service, component reliability is even more important than in mainstream applications. Solar Photovoltaic (PV) cell arrays represent the most common type of solar power generation. The cells produce power by pulling or “knocking” electrons loose from absorbed solar energy, to create an electron flow that is captured as dc current. The dc power is converted to ac via inverters. The inverter technology may be in the form of either a single microinverter connected to each solar panel or a string inverter, which converts the accumulated power of multiple solar panels, wired
eeworldonline.com | designworldonline.com
illinois capacitor – P&EE HB 11-18 V4 FINAL.indd 23
11 • 2018
DESIGN WORLD — EE NETWORK
23
11/6/18 1:43 PM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
Having capacitance values from 110 µF to 2.7 F, the Cornell Dubilier DCMC Series provides a high capacitance that allows it to ser ve in applications such as power supply filters and in energy storage applications such as welding equipment, UPS systems, and computer hold-up power.
in a series configuration. Within each inverter, film capacitors or long-life electrolytics find use in the dc link, snubber and ac output filters. One of the fastest growing sectors of the PV solar market is “offgrid power.” Off-grid power setups are entirely self-sufficient solarpowered supplies for lighting, remote monitoring and other applications. For example, a remote parking lot can be lit by high-efficiency LED lights and powered by solar without any connection to a power grid. This application usually employs microinverters. Another type of solar power generation is a solar furnace, which concentrates sunlight on a focal point. Temperatures at that focal point can reach 3,500°C and provide a means to power a steam–powered ac generator, as in a conventional power plant. To operate the solar furnace at maximum efficiency, the mirrors that redirect the sunlight must constantly be repositioned. Electric motors, powered by rechargeable batteries or supercapacitors, handle the repositioning. A key benefit of solar furnace technology is that its use of time-tested steam power generators makes it safe and clean – it releases no pollutants, even in a disaster. France is home to the largest solar furnace power station at Odeillo. It covers about 2,000 m2. Most are considerably smaller.
24
DESIGN WORLD — EE NETWORK
illinois capacitor – P&EE HB 11-18 V4 FINAL.indd 24
Wind Wind turbines have a pitch system that adjusts the angle of the props depending on wind conditions. The pitch-control system may be hydraulic or electric though the majority are electric. Here, electric motors powered by batteries or supercapacitor modules alter the pitch angle. It’s well known that the performance of rechargeable batteries degrades with a rising number of charge/discharge cycles. Additionally, batteries may lose capacity in cold weather and experience reduced lifespans when it’s hot. Published literature suggests that proppitch batteries need replacement about every five years. Unlike batteries, supercapacitors don’t degrade much with each charge/ discharge cycle. A significant additional benefit is that they should last two to four times longer than battery packs, reducing long-term service costs. Supercapacitor arrays are already in use by major wind power manufacturers, demonstrating the attractiveness of this approach.
Energy Harvesting Remote monitoring systems, such as IoT devices that transmit complex data or simple on/off detection, all require reliable sources of power under varying conditions. Here, energy harvesting transducers often serve as primary or secondary power sources. Examples include vibration sensors, piezoelectric transducers, small wind turbines or small PV cells. Supercaps can be an excellent way to store energy for these applications. The ability of a supercap to quickly supply a burst of power makes them good candidates for remote sensors. In such circuits, supercapacitors may be the sole power source or supplement a battery when the remote sensor burst-transmits data back to a monitoring station. Capacitors in green power apps Power Film Capacitors are a specialized family of film capacitors intended primarily for high voltage, high-current applications, such as dc-links and power semiconductor snubbers. They are
...capacitors have been steadily evolving to perform better and last longer. And some of the newer technologies, such as super capacitors (also called doublelayer capacitors) have become much less expensive in recent years... 11 • 2018
eeworldonline.com | designworldonline.com
11/6/18 1:43 PM
e
m
CAPACITORS IN RENEWABLE ENERGY
Illinois Capacitor modules combine supercapacitors with appropriate balancing circuitr y for higher voltage operation and higher capacit y.
especially well suited for the types of high-frequency inverters used in green energy power conversion. Power film capacitors can provide high reliability and long life. That said, high-performance power applications are characterized by harmonics, irregular signals, complex waveforms and continuously changing signal levels. These conditions -- along with working voltage, operating frequency and temperature, and ESR –are considerations when selecting power film capacitors because ignoring them can lead to greatly shortened device life. Power film capacitors are available with various types of terminations and package styles. Terminations range from standard two-leaded axial and radial types to multiple leads, lugs or screw terminals. Snubber capacitors are specialized types of film capacitors placed across semiconductor devices for protection and to improve circuit performance. In high-power modules, the voltage across the power semi can exceed the dc bus voltage. When this happens, the snubber capacitor absorbs the excess energy. Snubbers can reduce or eliminate voltage or current spikes from highspeed switching, limit dI/dt or dV/dt, shape the load line to keep it within the safe operating area (SOA), and transfer
power dissipation from the switching device to a resistor or a useful load. (Only the capacitor is needed in some applications.) They can also reduce both total losses due to switching and EMI by damping voltage and current ringing. Depending on the circuit configuration, a snubber capacitor may mount directly onto the output terminals of a power semiconductor or PCB mount. As is widely known, supercapacitors have incredibly high energy density as compared with other types. While individual devices have single-digit voltage ratings, supercaps can be banked to make up modules for virtually any required voltage. In energy storage applications, their most significant advantage over batteries is their ability to be charged and discharged continuously without losing capacity. Supercapacitors may serve either as solo energy storage devices or be connected in parallel with batteries. In the latter application, supercapacitors respond quickly when energy bursts are required. Unlike batteries, they can charge and discharge fast while batteries are better at supplying bulk energy for longer periods. Supercapacitors have a construction With capacitance to 470 F, Illinois that differs drastically Capacitor's DGH Series is part from that of other of a new wave of lower-priced capacitors. Supercaps supercapacitors. are generally based on carbon (nanotube) technology. The design creates a large surface area with an extremely small separation distance between layers. Conventional capacitors have two metal electrodes separated by a dielectric
eeworldonline.com | designworldonline.com
illinois capacitor – P&EE HB 11-18 V4 FINAL.indd 25
11 • 2018
material. In contrast, supercaps separate the electrodes with a physical barrier made from activated carbon. When an electrical charge is applied to the material, a double-electric field is generated which acts like a dielectric. The thickness of the electric double layer can be as thin as the diameter of a molecule. As a result, the surface area of the activated-carbon layer is extremely large, several thousands of square meters per gram. So, in concept, it behaves like a capacitor that has a physical size many times larger. Individual supercaps can be found in small energy harvesting applications; modules are generally required for larger applications that involve conventional ac line voltages. The construction of these modules requires balancing circuitry, to keep the voltage across each cell approximately equal, and may also include protection devices. When a supercapacitor has no stored charge, it “looks like” a dead short to a charging device. Fortunately, PV cells are not bothered by this condition. Other charging circuits must be designed to meet that challenge. Electrolytics are a good fit for many renewable energy inverter applications. With much higher power density than film caps, they can be a good choice in circuits below 600 Vdc. Inverters typically require high capacitance to handle peak load requirements that electrolytics handle well. Electrolytics generally cost less than film capacitors and have higher capacitance for their size but are not available for bus levels exceeding 550 Vdc. Circuits with levels above this would need aluminum electrolytic capacitors placed in series which reduces the
DESIGN WORLD — EE NETWORK
25
11/6/18 1:44 PM
P O W E R & E N E R G Y EF F IC IEN C Y H AND B OOK
- OUT NOW! RUTRONIKER 2018 re!
ledge and mo Technology, Know 8 com/rutroniker201 Visit www.rutronik.
Illinois Capacitor MPX/MPXB Series X2-class interference suppressor capacitors pass the 85/85 THB test, accelerated testing that subjects a part to 85°C heat with 85% relative humidit y and an applied bias voltage (240 Vac) for 1,000 hours. This demanding test simulates the extraordinarily harsh operating conditions encountered by many commercial and industrial products.
Electronics
High-Tech Components Worldwide for Your Innovations
As a leading distributor of electronic components we are able to offer you a wide portfolio of products, expert technical support for product development and design-in, individual logistics and supply chain management solutions as well as comprehensive services. Semiconductors Passive Components Electromechanical Components Displays & Monitors
Boards & Systems Storage Technologies Wireless Technologies
For more information about RUTRONIK: sales-na@rutronik.com Committed to excellence
www.rutronik.com illinois capacitor – P&EE HB 11-18 V4 FINAL.indd 26
capacitance. Their higher voltage ratings often make power film capacitors more economical when bus voltages exceed 550 Vdc. While electrolytics are typically thought of as having a much shorter lifespan than film caps, proper selection can yield a life roughly equal to the expected life of the inverter. Also, the latest electrolytics offer lower ESR and longer life. (To see how the life of any electrolytic can be greatly extended by derating, experiment with Illinois Capacitor’s life calculator at: www.illinoiscapcitor.com/tech-center/life-calculators.aspx) Both electrolytics and film capacitors can be used to filter out noise generated within switching power supply circuitry and to keep out external conducted noise. EMI suppression capacitors are divided into two classifications, X and Y. X class capacitors connect across the output line. X capacitors are further subdivided into three subcategories X1, X2 and X3. X1 capacitors are used where the capacitors see a peak voltage greater than 2,500 V and less than 4,000 V. Class X2 are for applications where the peak voltage is REFERENCES below 2,500 V. X3 types are for peak voltages less than Illinois Capacitor 1,200 V. X2 capacitors are the www.illinoiscapacitor.com most common. Y capacitors are connected from line-toCornell Dubilier ground and typically have a www.cde.com low capacitance value. X class capacitor dielectric materials include film, ceramic and paper, while Y capacitors typically have ceramic or paper dielectrics. All in all, it’s good to be aware of the options available for converting and storing green energy. The right capacitors will yield greater efficiencies and longer device life.
26
DESIGN WORLD — EE NETWORK
11 • 2018
11/7/18 10:25 AM
m
TRANSFORMING AC TO DC
Transforming ac to dc without transformers J E F F S O R E N S E N • H E L I X S E M I C O N D U C TO R S
Substituting special capacitors for transformers helps realize ac-to-dc power supplies that are both efficient and compact.
POWER SUPPLIES that convert ac to dc have been a part of nearly every electronics-based product since the popularization of the vacuum tube radio. While the voltage levels have changed, the conversion process has always included a transformer to provide two functions: voltage conversion and safety isolation between the user and the mains power. For safety, ac-dc conversion circuits must be galvanically isolated from the mains voltages to prevent a shock hazard. Transformers are suited for this purpose because the primary side and secondary side can easily be isolated from each other and the voltage reduced or raised by the transformer’s design. But there are downsides to using transformers – especially when it’s important to improve the power efficiency or reduce the physical size. Today, capacitors can also isolate circuits from mains power while exhibiting lower inherent losses than transformers. Voltage-level conversion can also take place using highfrequency, switched-capacitor charge pumps, so transformers can also be eliminated for that purpose as well. Thus, capacitors can provide the basic functions that transformers deliver.
T Y P I CA L T R A N S F O R M E R L E S S C I R C U I T C O N F I G U R AT I O N
A block diagram of the Helix transformer-less concept board used as a vehicle for testing the concepts of a new integrated two-chipset using Helix MuxCapacitor and other proprietar y technologies.
eeworldonline.com | designworldonline.com
helix – P&EE HB 11-18 V3 FINAL.indd 27
11 • 2018
DESIGN WORLD — EE NETWORK
27
11/6/18 2:59 PM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
The concept board used for testing the Helix MuxCapacitor technology. This board was constructed for UL testing of touch current and isolation.
a conductor. Any other conductors or components nearby can have an induced current (noise) from the unshielded magnetic field that surrounds the transformer. Magnetostriction loss is from grain movement in the core material. Leakage inductance arises from the imperfect coupling between the primary and secondary inductances of the transformer. It is worth examining transformers and their properties more fully. Transformers have several disadvantages compared to capacitors. First, they are large, usually the largest component in the power supply. They are heavy and add considerably to the overall weight. And they are a major cause of inefficiency in power supply designs. Transformer losses can be divided into two categories: dc resistive losses in the transformer windings and core losses caused by the magnetic material and physical qualities of the core. I2R power loss arises because of the resistance in the primary and secondary wire. This parameter can be easily calculated knowing the resistance of the wires and the currents they will carry. Though 60-Hz transformers have relatively high dc losses, high-frequency switched-mode transformers can mitigate them somewhat. The transformer core losses are more complicated and arise from a combination of sources: hysteresis loss, eddy current loss, magnetostriction loss, and leakage inductance.
28
DESIGN WORLD — EE NETWORK
helix – P&EE HB 11-18 V3 FINAL.indd 28
Hysteresis loss comes from the inherent memory in the core material when it becomes magnetized and, when the field must reverse, “remembers” the previous direction of the magnetic field. It takes energy and therefore power to overcome the residual magnetic field each time it reverses. Eddy current losses arise from the induced current circulating in the core because the core is also electrically
Advantages of capacitive coupling Capacitors are relatively small compared to transformers and can be made smaller through the careful selection of the values and types used. The higher the operating frequency, the lower the capacitance needed for a given ac impedance (to a point). Safety capacitors are normally used in products for controlling the conducted and radiated emissions from high-frequency circuitry, and “Y” types are certified by UL and other standards bodies for operation
O U T P U T E F F I C I E N CY V S . C U R R E N T
The output efficiency vs. output current at various barrier switching frequencies for the MaxCapacitor concept board submitted to UL.
11 • 2018
eeworldonline.com | designworldonline.com
11/6/18 2:59 PM
TRANSFORMING AC TO DC
The heat dissipated in the inductors on the concept board becomes evident in a thermal image. The inductors are the red and yellow areas and the drivers are the four dark areas running vertically in the center of the picture. All the other components are near ambient temperature.
across the isolation barrier. Y1 capacitors are rated for a maximum 500 Vac working voltage and a breakdown voltage of 8 kV. Y2, Y3, and Y4 have progressively lower voltage ratings. Since these capacitors are already intended and approved by UL for use across the isolation barrier, they are also an obvious choice for coupling the power. Safety issues such as “touch current” must be considered when using highervalue capacitors across the isolation barrier. “Safety” capacitors are a bit of a misnomer because too much capacitance across the isolation will also allow more ac mains current to pass. Using a highcarrier frequency allows the use of capacitors with smaller values. Also, it is better to use higher voltages, therefore lower current, to keep the reactive impedance losses as low as possible. MuxCapacitor is a proprietary charge pump circuit that can either increase or
eeworldonline.com | designworldonline.com
helix – P&EE HB 11-18 V3 FINAL.indd 29
decrease the input-to-output voltage (and current) levels at 99% efficiency. The final chipset will have two separate MuxCapacitor blocks, one on either side of the isolation barrier. To get the best efficiency across the isolation barrier, the voltage on the primary side must be as high as possible. To do this, the primary-side rectified line voltage is regulated to keep the voltage as high as possible before the power is sent across the barrier. Because the circuit is intended to be used worldwide, this input voltage can vary from 85 to 220 Vac, which after rectification is approximately 120 to 310 Vdc. A primary-side MuxCapacitor stage is used to sense the input voltage and adjust its output to keep the input to the drivers or “modulator” as close to 300 V as possible. It can therefore operate at different levels of gain/attenuation to keep the V+ supply at a constant level.
11 • 2018
On the secondary side, the output from the diode bridge demodulator is stored on a capacitor (denoted C3 in the nearby drawing, Typical transformerless circuit configuration). This is the input to the secondary side MuxCapacitor stage that reduces the voltage on C3 from about 300 to 20 Vdc (nominally) using a variable gain step-down MuxCapacitor. Both the primary and secondary side MuxCapacitor stages can adjust for varying input power and output load conditions. Referring to the nearby concept circuit, the modulator operates at approximately 125 kHz. There are two phases to each charging cycle. Mains power is applied to the input of the primary side diode bridge and charges capacitor C4, which connects across the input of the primary side voltage conditioning circuitry. This stage outputs the voltage to supply the V+ to internal switches Q1 and Q3. For the first phase, Q1 and Q4 are turned on and V+ current flows across the barrier L1/C1. The coupled current then forward-biases diode D1 to charge C3, which connects across the input of the secondary side MuxCapacitor circuitry. The return path of C3’s isolated ground is through D4, then back across C2/L2 and through Q4 to the primary side ground. After the phase-one charge has been transferred to C3, switches Q1 and Q4 are turned off. The phase-two cycle begins, and switches Q2 and Q3 are turned on. V+ current flows through Q3 across L2/C2 and D3, and again continues charging C3. The return path is now through D2, across C1/ L1 to Q2 to the primary side ground. After transferring the charge, switches Q2 and Q3 are turned off, and the cycle repeats. The secondary side bulk capacitor is charged in a similar manner to a conventional transformer-based fullwave bridge power supply – the main difference being the charge is coupled in pulses through two capacitors instead of a transformer. Selection of the capacitive isolation barrier soft-switching inductors (L1, L2) and capacitors (C1, C2) define key power supply performance specifications,
DESIGN WORLD — EE NETWORK
29
11/6/18 2:59 PM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
O U T P U T V O LTA G E V S . C U R R E N T
The output efficiency vs. output current at various barrier switching frequencies for the MaxCapacitor concept board submitted to UL.
including maximum power transfer, optimal modulation frequency, safety ratings, and touch current. The concept board design shown here is limited by primary and secondary side electronics to 10 W operation at ~125 kHz switching frequency. However, the basic topology used for the capacitive isolation barrier would support a maximum power transfer of 65 W. Adjustments would need to be made to support higher power and voltages. Switching frequencies and the capacitors, inductors, and drivers will need to change. In realizing this test board, silicon using existing lowervoltage MuxCapacitor stages was used. Because of voltage and power limitations of the silicon available, circuit operation doesn’t entirely represent what a final product would be able to do. The voltage ratings for the MuxCapacitor devices used limit the acceptable voltage across the isolation barrier. The MuxCapacitor blocks use existing silicon made with the X-FAB XDM10 (primary side) and XT018 (secondary side) processes. The maximum voltage allowed for some components on the secondary side is 60 V. The final product will use a 400-V process and 300 V across the barrier. For the isolation barrier testing, the lower-voltage secondary side components were bypassed so voltages up to 200 V could be used. The prototype board was designed to be a vehicle for UL approval of the touch current and isolation. The maximum input voltage that can be used is limited, so testing consisted of applying 152 V to the input of the modulator and bypassing the voltage conditioning that will be included in final silicon. The output is taken on the secondary side across the dc bulk
30
DESIGN WORLD — EE NETWORK
helix – P&EE HB 11-18 V3 FINAL.indd 30
11 • 2018
capacitor after “demodulation” by the full wave bridge. The clock frequency can be varied so the effect of barrier frequency on the operation can be tested. The component values for the isolation coupling consist of two 2.2-nF Y1 capacitors in series with two 1-mH inductors (one capacitor and one inductor in each leg). These have a resonant frequency of 107 kHz. The circuit performs best when using zero-voltage switching (ZVS). Without the inductors, more power will dissipate in the drivers. The switches have three pertinent parasitics that reduce efficiency: on resistance Rdson, output capacitance Coss, and gate capacitance Cg. The switch timing depends on Rdson and Coss, which determine the required dead time. Without the inductor, the switching losses would be 4CossV+2f, where f is the switching frequency. This loss can be eliminated using ZVS where the inductor recovers the charge on Coss. The gate has similar losses that are less significant because the gate switching voltage is much lower. With the capacitor and inductor values used, the operation and efficiency can be adjusted by changing the frequency of operation. For ZVS to work properly, the net impedance of the L/C network must be inductive, so the best operating frequencies are somewhat above the resonance of the coupling network. The observation that the efficiency can be affected by changing the frequency suggests that adjusting the barrier frequency is a way to optimize the overall efficiency.
eeworldonline.com | designworldonline.com
11/7/18 2:56 PM
TRANSFORMING AC TO DC
At 150 kHz the best overall efficiency (95%) is measured at 45 mA. From 50 to 65 mA the highest efficiency (95.25%) is at 125 kHz. Reducing the barrier frequency nearer the resonance point reduced the peak efficiency, and the efficiency across the different loads became flatter. Though the circuit is not operating at resonance, the capacitive and inductive reactance mostly cancel in the L/C network but with some inductive reactance remaining. Operating with only the capacitor would not have this benefit and the losses would be higher. As the barrier frequency rises above the resonant point, the output voltage drops with rising output current. Operation above 150 kHz did not help the performance and significantly reduced performance at 175 kHz compared to performance at 125 and 150 kHz, especially the output voltage vs. output current. When operating the barrier near the resonant frequency, the output voltage and efficiency are relatively flat over the range of currents tested, but the efficiency is lower than when operating
at frequencies somewhat higher than the resonant frequency. This means the barrier frequency can be set to optimize the overall efficiency. Another benefit of ZVS is that the board operates without significant component power dissipation, i.e., heat. The inductors are the warmest components on the board but are only about 50°C at 13 W. ZVS reduces the losses in the other components but boosts the dissipation of the inductors. Using inductors with a lower ESR and higher Q should improve the efficiency. In all, clearly it is possible to efficiently pass power and provide isolation using capacitors. Additional work is continuing to build a 65-W high-power board and to optimize the components used. Operation at higher voltages and barrier frequencies are planned to boost performance. Passing UL requirements was the first big hurdle and now circuit optimization is underway to make this technique the future of isolation and power transfer.
REFERENCES “High Voltage Capacitive Voltage Conversion” Presented by Neaz Farooqi, Helix Semiconductors, 13:55 17-2 Randall L. Sandusky, Helix Semiconductors, United States; Alexander Hölke, X-FAB Sarawak Sdn. Bhd., Malaysia, 2018 ISPSD, 30th International Symposium on Power Semiconductor Devices and ICs Mitchell Kline “Capacitive Power Transfer ” Technical Report No. UCB/ EECS-2010-155 www.eecs.berkeley.edu/Pubs/ TechRpts/2010/EECS-2010-155.html
XEL50xx Series
Low Loss Power Inductors Extremely low DCR ... and ultra-low AC losses 25 inductance values ... from 0.1 to 22 µH Optimized for high frequencies ... up to 5+ MHz Superior current handling ... up to 44.0 Amps Compact footprint: 5.28 x 5.48 mm
helix – P&EE HB 11-18 V3 FINAL.indd 31
Free Samples @ coilcraft.com 11/7/18 2:56 PM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
Evaluating options for USB Type-C connections Port controllers, transceivers, and special dc/dc converters now simplify the task of adding USB Type-C functions.
DA N I E L L E I H • M I C R O C H I P T E C H N O LO GY I N C .
CONSIDER THE VIRTUES of the USB-C connector. In addition USB-C connector in most new models. Initial USB -C and Power to its 10 Gbit/sec (Gbps) bandwidth and Alt Mode video Delivery designs were complicated, involving many external capabilities, two attributes in particular make this form factor components and software configuration tools. Now, new ICs incredibly valuable: a reversible plug and intelligent, high-power take the guess work out of USB-C designs. capability. The value of the reversible plug is obvious: Finally, we In any product design, the first step is to define the desired feature set. This is especially true in a USB-C system with can all simply plug in our devices without having to flip the plug Power Delivery because the number of PD features have a over (often twice). It is the intelligent power aspect, however, direct impact on the cost of the system. PD itself adds cost to that makes the USB-C connector so useful. the system, so the end product must benefit from the power USB has always had the ability to provide power, as long delivery capability to justify the added expense. as 5 V and less than 1.5 A was enough. That limited the previous USB-C is versatile and Type-A and Type-B form factors to supports data types beyond powering small electronics devices USB-C PORT those of traditional USB. like thumb drives or keyboards, or Designers must keep data types to trickle-charge devices such as cell phones. With USB-C comes a in mind when selecting the appropriate USB-C components. new standard, Power Delivery (PD), If the product is a storage device which allows the source and the sink to negotiate power up to 100 or a battery charger, there is no need to spend money on W at voltage levels from 5 – 20 V. This means that a little USB-C plug firmware to implement Alt Mode video. Conversely, if the system can energize many more products is a monitor that will connect to than previously possible, including external storage devices, phones, a DisplayPort-enabled laptop, the design must include specific PCs, power tools, medical devices port controllers and components. and countless others. With 100 W on The USB-C port elements include tap, just about anything you can fit the PD/USB protocols, data and in your electric car can be charged power, so the design will include with a USB-C port (but not the car both a USB-C Power Delivery itself – sorry). The PC and mobile phone Port Controller and analog and industries have quickly adopted power components. USB-C. Most notably, the iPhone One of the simplest USB-C supports Power Delivery through implementations is the chargethe Lightning Connector, and only port. In this case, the system Typical simplified block diagram for a USB-C port Android phones implement the is designed to power and/or ser ving as a charge-only power source.
32
DESIGN WORLD — EE NETWORK
microchip – P&EE HB 11-18 V4 FINAL.indd 32
11 • 2018
eeworldonline.com | designworldonline.com
11/7/18 11:56 AM
e
m
TRANSFORMING AC TO DC
charge the devices that connect to it. An example of this type of system is a rear-seat charger in a car, a power tool battery or a wall charger in a house. In this example the list of major ‘Bill of Materials’ (BOM) components is relatively short: USB-C port controller – Negotiates the connection and power contract. DC/DC converter – Converts the input voltage into the Vbus voltage required by the PD contract. Load switch – Delivers 5 V to Vbus at plug in, connects the appropriate Vbus voltage once the PD contract is established. Sometimes combined with the DC/DC Converter. LDO – regulates voltage to the port controller because the DC/ DC may be requested to supply from 5 V – 20 V. USB-C connector The port controller in this example must be able to handle all of the negotiation with the connecting device. Modern standalone controllers such as those from Microchip include, at a minimum, the following features: USB-C connector support with connection detection and control USB Power Delivery 3.0 compliant MAC Pre-programmed Power Delivery Firmware Support for all standard Power Delivery profiles (15/27/45/60/100 W) Integration of select analog components that reduce the BOM cost and design complexity. Examples required for connection include: VCONN FETs with Rp/Rd switching Dead battery Rd termination Programmable current sense for overcurrent conditions Voltage sense for overvoltage conditions Appropriate temperature support for the application Because this is a charge-only app, it requires no other system controllers. Though some suppliers offer programmable devices, the logical choice for a charge-only design is a pre-programmed product with no software requirements. System configurations take place through simple device straps (connections to ground or Vcc). If the controller is PD 3.0-compliant, users will have access to all standard power profiles: 15 W/27 W/45 W/60 W/100 W. The choice of dc/dc converter type mostly depends on the input voltage. The power supply must always be able to provide an output voltage from 5 to 20 Vdc for full PD compliance. A basic buck topology can usually handle a design having a 24-Vdc input or voltage greater than 20 Vdc. Lower-voltage dc or offline acpowered systems will need alternate topologies. Another common configuration is that of a USB-C connection behaving as a power source and handling USB 2 data. In this case, the designer provides USB 2 host support for data transfer because the
eeworldonline.com | designworldonline.com
microchip – P&EE HB 11-18 V4 FINAL.indd 33
product contains a microcontroller with native USB 2 support. Note that the port controller does not need any connection to the USB 2 data path. No additional components are necessary, and the USB-C port BOM is the same as the charge-only design. USB 3 could also be added by including a USB 3 mux (to enable the USB-C plug reversibility), provided the MCU/system controller supports USB 3. In this example, use of a standalone USB-C pre-programmed port controller is also the simplest way of adding a single USB-C port to an existing product offering. At the top of USB performance architectures is the hub-based system. The hub-based design offers the most flexibility and performance of any USB architecture while removing the burden of communications from the central processor. This type of system is common in PC docks and monitors, automotive center consoles, and in any application that needs multiple USB connections. In a PC use-case, video signals will likely pass through the USB-C port, so the Alt Mode functions must be supported. The port controller for this use case must be capable of supporting Alt Mode functions, and the design must contain the circuitry to manage direction and interpretation of the protocol passing through the Alt Mode channels. The use of a multi-port “SmartHub” in this system offers designers a more efficient system-level design. The designer could simply purchase a more feature-rich port controller and leave the functions separate, but the use of the controller within the hub as the port controller reduces cost and processing overhead. This is especially true in multi-port systems where coordination of data movement or power usage is important.
P O W E R A N D DATA A t ypical simplified block diagram for a USB-C connection ser ving as a power source and handling USB-2 data.
11 • 2018
DESIGN WORLD — EE NETWORK
33
11/7/18 11:57 AM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK T Y P I CA L U S B H U B S E T U P
A USB Hub-based architecture with Type-A ports, Type-C ports, and Alt Mode support.
An evolved version of port control is increasingly prevalent as USB-C becomes more available natively within controllers and processors. All USB-C functions -- such as Port Policy Management, Power Delivery, Alt Mode support and Billboard support -- reside within the hub. In this architecture, the standalone port controller is replaced with a transceiver. The transceiver contains the physical layer of the USB-C interface, much in the way Ethernet networks are designed. To support the Alt Mode function, the design includes an external crossbar mux which redirects the video data to a DP connector for display on an external monitor. To address today’s increasing concerns about data and network security, this design contains a security IC that enables secure updates to the system firmware. Highly secure devices such as Microchip’s ECC608A enable the designer to ensure safety of code through use of NIST, SHA-256 and HMAC hash, and AES-128 encryption, without the manufacturer even knowing the owner’s key. Additions to the system BOM examples above include: USB multi-port SmartHub – Contains controller and multiple USB connections.
34
DESIGN WORLD — EE NETWORK
microchip – P&EE HB 11-18 V4 FINAL.indd 34
Crossbar mux – Diverts various data channels to different locations. DP Connector – Connects to video display. Type-A connector(s) Type-A power source Security IC – Enables secure code updates to the hub. USB-C transceiver for each port DC/DC converter for each USB-C PD port
A USB SmartHub with integrated Power Delivery also enables other system-level features. Advanced systems that contain HostFlex Technology, wherein any Type-C port can become the system host, let users take over displays and output functions without regard to which port they connect to. Power Balancing functions also let a userdefined algorithm apportion some fraction of total power to various subsystems. Users can decide if power is delivered in the order of connection, based on device type, based on the number of devices connected, or via some combination of those criteria. The enabling technology for these features is the Microchip SmartHub, which orchestrates platform-level management of all concurrent USB-C PD port connections. Microchip has demonstrated system-level features such as HostFlex, MultiHost (concurrent host capability), and power balancing on the newest line of USB 3.1 multiport SmartHubs with integrated PD. USB-C is the connector that finally enables multiple types of data and multiple power levels to coexist within a single connector. Advanced system features such as HostFlex and Power Balancing can be easily implemented using a SmartHub design, while basic charging circuits can be implemented with simple, strappable port controllers. Future devices will make use of higher levels of integration and will further ease implementation. Designers need not fear the task of adding USB-C to their designs. REFERENCES Semiconductor companies like Microchip are now producing unique and highly capable port controllers, transceivers and Microchip SmartHub IC Design Center companion dc/dc converters, as well as www.microchip.com/design-centers/ the support needed to make the design usb/product-families/smar t hub job simple and low-risk.
11 • 2018
eeworldonline.com | designworldonline.com
11/7/18 11:57 AM
e
PUSH PERFORMANCE TO THE TOP
PCB Terminal Blocks and Connectors for Power Electronics • • • •
Compact dimensions and high current carrying capacity – up to 85 A Fast, easy connection with Push-in CAGE CLAMP® Optimal handling for any application Comprehensive range covering 24 - 4 AWG
www.wago.us/powerelectronics
Pluggable PCBHB - Design World35 - Print.indd 1 wago – P&EE 11.18.indd
10/16/18 12:48 3:06 PM PM 11/7/18
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
GaN FETs boost efficiency in electric vehicle chargers The role gallium-nitride FETs play in the design of an EV charging circuit help illustrate how these widebandgap semiconductors facilitate energy efficiency. THE PHASE-SHIFT FULL BRIDGE (PSFB) is a classic topology for applications that must accommodate a wide range of operating voltages, as with battery chargers. A PSFB converter generally uses four power switches (MOSFETs or IGBTs) to form a full bridge on the primary side of an isolation transformer. Diode rectifiers or MOSFET switches perform synchronous rectification on the secondary side. A phase shift between PWM signals driving the two legs of the full bridge determines how much energy gets transferred to the load. A control algorithm manipulates this phase shift to regulate and maintain the output voltage at the desired level. An attraction of this topology is that it permits zero-voltage switching (ZVS) of its switching devices. This lowers switching losses and makes the converter efficient. An advantage of this circuit over an LLC topology is that it does not need either a variable switching frequency or a variable dc-link voltage to regulate the battery voltage. However, there is one imperfection in PSFBs: soft switching is possible at high loads, but the devices inevitably enter hard switching under low loads. It can be particularly advantageous to devise PSFBs with GaN FETs rather than with traditional silicon power devices. Reasons include the ability to realize a higher switching frequency and a higher power density, and the topology can work at a larger phase-shift angle at high efficiency. The soft-switching region is also larger, and when hard switching is necessary, the losses are lower. These benefits result from a low output charge (Qoss) and low switching losses in the GaN power devices. To further improve efficiency, an active snubber can replace the conventional RCD snubber. The snubber is necessary to squelch secondary parasitic ringing that
36
DESIGN WORLD — EE NETWORK
transphorm – P&EE HB 11-18 V4 FINAL.indd 36
11 • 2018
FENG QI • TRANSPHORM
S I M P L I F I E D S C H E M AT I C, P S F B
TOP: A protot ype of the 200-kHz 3.3kW PSFB using a TPH32 12PS GaN FET. BOTTOM: A simplified schematic of the 200-kHz PSFB.
eeworldonline.com | designworldonline.com
11/7/18 9:53 AM
e
m
GaN FETS BOOST EFFICIENCY
A CT I V E S N U B B E R C O M P O N E N T S
Real-time plots of I L1 and V rec
arises with high output voltages. Without some kind of a snubber, the ringing would put a lot of voltage stress of the rectifier diodes. A traditional RC damping snubber isn’t practical because it is lossy in high-voltage high-frequency applications. An RCD clamped snubber circuit is generally added to keep the secondary voltage down. In one case, a prototype PSFB employed the 72-mΩ Transphorm GaN FET (TPH3212PS). Its total resonant inductance is about 2.7 µH and arises from the combination of leakage and resonant inductance. This resonant inductance allows di/ dt as high as 150 A/µsec with a 410-V dc-link. The high di/dt significantly increases the maximum phase-shift angle, reducing power loss caused by the freewheeling time. This trick can easily push
K E Y PA R A M E T E R S O F T H E P S F B DC link voltage (V)
380 ~ 410
Battery Voltage (V)
250~ 450
Maximum Power (W)
3300
Maximum Current (A)
11
Switching frequency (kHz)
200
Transformer turns ratio
1:1.18
Leakage inductance ( µH)
1.0
Resonant inductor ( µH)
1.7
Output inductor ( µH)
65
DC blockinh capacitor ( µF)
5
eeworldonline.com | designworldonline.com
transphorm – P&EE HB 11-18 V4 FINAL.indd 37
the angle over 0.9π (the higher the angle, the less free-wheeling time required). The high phase angle allows the turns-ratio of transformer to be reduced and further improves the efficiency. KEY COMPONENTS OF TH E ACTIVE SNUBBER Using this technique a 450-V battery can be charged with a C0 ( µF) 0.5 410-V dc-link and 0.93π phase L0 ( µH) 10 shift. The high di/dt lets the circuit C1 ( µF) 0.2 reach 26 A of current in 175 nsec at a power level of 3,600 W. L1 ( µH) 2700 The voltage spike on D0 SCS206 the output (Vrec in the nearby VIPER06HS, orfigure labeled Plots of IL1 and S1 NCP1060AD100 Vrec) is clamped to around 540 V by the D1, D2 ES1J active snubber. The snubber is controlled with offline converter ICs such as the STMicroelectronics VIPER06HS or ON Semiconductor NCP1060AD100. The active snubber has three terminals connected across Vrec and Vo. It is composed of two parts, an input filter and switcher cells. In the simplified schematic of the active snubber, the switcher, S1, is configured to work with a constant peak current. In this design, the snubber is comprised of three switcher cells and can pump 100 mA depending on the requirement from the parasitic capacitance of the transformer and rectifiers on the secondary side. The junction capacitance of D0 in the active snubber schematic also contributes to the parasitic capacitance. In this design, the four rectifying secondary diodes are 650-V SiC Schottky SCS210 devices, and D0 is a SCS206 diode. By tuning the compensation resistor in the switcher circuits, the overshoot caused by the parasitic components can be clamped at a desired Vrec. The dead time must be less than 175 nsec to give a 0.93π phase shift. Switching details C O M PA R I S O N O F O U T P U T In the simplified PSFB C H A R G E A N D E N E R GY schematic, Q1 and Q2 are R on(tpy) C o(tr)_400V the leading phase leg, which (mΩ) (pF) starts its transition before TPH3212PS 72 225 power transfer. Q3 and Q4 IPP60R090CFD7 75 751 are the lagging phase leg, which starts its transition after power transfer. IL1 is always higher at the end of power transfer, so it is always easier for the lagging leg to soft switch. In other words, the leading leg loses soft switching at higher power than does the lagging leg.
11 • 2018
DESIGN WORLD — EE NETWORK
C o(er)_400V (pF) 142 73
37
11/7/18 9:53 AM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
Insight into the circuit operation comes from examining the switching transition after Q1 has turned-off for both soft and hard switching. Consider an ideal soft switching waveform along with a simplified circuit diagram as depicted nearby (labeled Ideal soft-switching waveform). Q1 turns off at t1, and Q2 turns on at t2. There is a zerovoltage soft switching transition during the dead time tdb. During this process, Coss of Q1 charges from 0 V to Vdc, and Coss of Q2 is discharged from Vdc to 0 V. In other words, Qoss_0ToVdc and Eoss_0ToVdc are injected in Q1, and Qoss_VdcTo0 and Eoss_VdcTo0 are removed from Q2. Meanwhile, Qoss_VdcTo0 x Vdc are energized to the level of the dc-link because the discharging current of Q2 flows through the dc-link. The process is powered by the inductor current, IL1. Equations (1) and (2) give the relationship between IL1 and the charge and energy during the transition.
B O U N DA RY O F S O F T S W I TC H I N G IL1_1 is the current at t1, and IL1_2 is the current at I L1_1 I L1_2 t db (nsec) (A) (A) t2. In (2), average current during the transition is TPH3212PS 70 5.2 0 approximated by the IPP60R090CFD7 128 9.4 0 arithmetic mean, which TPH3212PS 87.5 5.3 -1.2 is smaller than the actual IPP60R090CFD7 87.5 10.1 3.63 value. If there is no restriction on tdb, the minimum IL1_1 to maintain soft switching happens when IL1_2 is 0 A. If tdb is targeted to a value, such as 87.5 nsec in this case, the minimum IL1_1 can also be derived from the equations. The corresponding values are calculated at a 400-V dc-link. They define the boundary between soft switching and hard switching. When IL1_1 is lower than the boundary values, Q1 and Q2 start hard switching. Assuming Q2 is turned on with Vx remaining on its drain, the nearby figures give the waveform and simplified circuit diagram for two typical cases of hard switching. For both cases, before turn-on of Q2, Qoss_0To(Vdc-Vx), has charged to Q1, and/or Qoss_VdcToVx, has been discharged from Q2. Correspondingly, Eoss_0To(Vdc-Vx) has been stored in Q1, Eoss_VdcToVx has been removed from Q2, and Qoss_VdcToVx x Vdc has been stored in the dc-link. In other words, Qoss_(Vdc-Vx)ToVdc will be charged into Q1 during the turn-on of Q2. The build-up of the remaining charge causes turn-off loss in Q1, Eoss_VxToVdc, and turn-on loss in Q2, Vdc x Qoss_(Vdc-Vx)ToVdc - Eoss_VxToVdc. On the other hand, the turn-on of Q2 also suffers from losses caused by IL1_2, which is left out of this discussion. For both cases, equations (1) and (2) can be rewritten as equations (3) and (4) by simply replacing the Qoss and Eoss and assigning 87.5 nsec to the dead time, tdb. Because Qoss and Eoss are voltage dependent, the following analysis is calculated
I D E A L S O F T- S W I TC H I N G WAV E F O R M
Ideal soft-switching waveform and simplified circuit diagram showing the current-flow waveform and current flow before t 1 , after t 1 , before t 2 , and after t 2 .
38
DESIGN WORLD — EE NETWORK
transphorm – P&EE HB 11-18 V4 FINAL.indd 38
11 • 2018
C U R R E N T F LO W BEFORE t1
C U R R E N T F LO W AFTER t1
C U R R E N T F LO W BEFORE t2
C U R R E N T F LO W AFTER t2
eeworldonline.com | designworldonline.com
11/7/18 9:53 AM
GaN FETS BOOST EFFICIENCY
I D E A L H A R D - S W I TC H I N G WAV E F O R M W I T H V X C U R R E N T F LO W AFTER t1
C U R R E N T F LO W BEFORE t2
C U R R E N T F LO W AFTER t2
C U R R E N T F LO W BEFORE t1
C U R R E N T F LO W AFTER t1
C U R R E N T F LO W BEFORE t2
C U R R E N T F LO W AFTER t2
SIMPLIFIED CIRCUITS
Ideal hard switching waveform with V x and simplified circuit diagrams showing current flow before t 1 , after t 1 , before t 2 , and after t 2 .
C U R R E N T F LO W BEFORE t1
I D E A L H A R D - S W I TC H I N G WAV E F O R M W I T H D I F F E R E N T V X
with the C – V curve instead of Co(tr) or Co(er) defined at fixed voltages. With an experimental C - V curve, Qoss vs. Vds and Eoss .vs Vds curves of devices are calculated and shown in a nearby figure (labeled Qoss vs. Vds and Eoss .vs Vds ). Curves for Vx vs. IL1_1 and IL1_2 vs. IL1_1 are calculated with (3) and (4) in another figure (labeled Vx vs. IL1_1 and IL1_2 vs. IL1_1).
eeworldonline.com | designworldonline.com
transphorm – P&EE HB 11-18 V4 FINAL.indd 39
SIMPLIFIED CIRCUITS
Ideal hard switching waveform with a different V x and simplified circuit diagrams showing current flow before t 1 , after t 1 , before t 2 , and after t 2 .
From Vx vs. IL1_1, it is worth noting that Vx of both GaN and Si stay below 50 V at an IL1_1 higher than 4 A. Once IL1_1 becomes less than 4 A, the Vx of GaN rises gradually whereas that of Si jumps dramatically and becomes steady around 400 V. This difference can be explained by the Qoss - Vds curve. The charge of GaN is more uniformly distributed across the whole voltage range whereas most of the Si charge resides in the low voltage region. From IL1_2 – IL1_1, as can be expected, for GaN, a tdb longer than the transition time required by the minimum IL1_1 will make Vx slightly pass the valley of Q2 Vds and end up with a negative IL1_2. If fine-tuning of tdb can be practically implemented in nanoseconds, Vx would be able to stay right at the valley of Q2 Vds. As for Si, Q2 will be turned on around 3 A IL1_2 from 10 A to 3 A IL1_1. The remaining charge in Q1, Qoss_(Vdc-Vx)ToVdc, and the loss of a phase leg caused by it, Vdc x Qoss_(Vdc-Vx)ToVdc, are presented in a nearby image. As discussed above, this loss is distributed in Q1, Eoss_VxToVdc, and Q2,
11 • 2018
DESIGN WORLD — EE NETWORK
39
11/7/18 3:45 PM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
P R OTOT Y P E P S E F F I C I E N CY AT 3 . 3 KW O R 1 0 0 % LOA D
Prototype efficiency at 3.3 kW or 100% load.
P R OTOT Y P E P S E F F I C I E N CY AT 0. 3 KW O R 1 0 % LOA D
Prototype efficiency at 0.3 kW or 10% load.
Vdc x Qoss_(Vdc-Vx)ToVdc - Eoss_VxToVdc. It is obvious that GaN significantly reduces the remaining charge and the loss caused by it, especially in light load conditions with IL1_1 lower than 3 A. At 1 A IL1_1, the loss caused by the remaining charge of GaN is only 25% of that of Si. In other words, GaN cuts 75% of the need for heat dissipation and improves power density by eliminating the need for a significant portion of heat sinking normally needed for Si. The PSFB is tested across a battery voltage range from 250 to 450 V and in different load conditions. The efficiency of the PSFB rises with rising battery voltage or phase shift angle and it is above 96% over the whole battery
Q OSS V S V DS
Q oss vs V ds curve from 0 to 400 V
I L1_2 V S I L1_1
IL1_2 – IL1_1 curve from 0 to 10 A.
40
DESIGN WORLD — EE NETWORK
transphorm – P&EE HB 11-18 V4 FINAL.indd 40
Prototype temperature rise at full load and light load conditions.
voltage range from 250 to 450 V. Notably, the arithmetic mean of efficiency is above 97% from 250 to 450 V. Converter efficiency is still around 90% at only 10% load. Benefiting from extremely low switching loss, the GaN devices generate a minimal amount of heat and cause little temperature rise during hard switching at 10% load.
REFERENCES
E O S S V S V D S C U RV E
E oss vs V ds curve from 0 to 400 V
REMAINING CHARGE IN Q1
Remaining charge in Q1, Qoss_(Vdc-Vx)ToVdc plotted against IL1_1.
11 • 2018
P R OTOT Y P E T E M P E R AT U R E R I S E , 1 0 0 % LOA D A N D L I G H T LOA D
Transphorm charger design guide www.transphormusa.com/en/document/ design-guide-200-khz-phase-shift-full-bridgefor-3-3kw-electric-vehicle-on-board-charger/
I L1_1 V S V X
IL1_1 curve vs. Vx for hard switching from 0 to 10 A.
Q O S S _ ( Vd c -V x ) ToVd c V S I L 1 _ 1
The loss in a phase leg caused by the remaining charge, Vdc x Qoss_(Vdc-Vx)ToVdc.
eeworldonline.com | designworldonline.com
11/7/18 9:53 AM
e
m
THREE-PHASE POWER MEASUREMENTS
Don’t be led astray on three-phase power measurements K E N J O H N S O N • T E L E DY N E L E C R OY
The MDA permits display of the filtered Sync signal with a measurement period overlay (at bottom) for easier understanding.
Measurements made on one power analyzer may differ from those on another because different manufacturers employ different settings and wiring assumptions.
eeworldonline.com | designworldonline.com
teledyne lecroy – P&EE HB 11-18 V3 FINAL.indd 41
POWER ANALYZERS have been manufactured for many years and are used in a wide variety of applications. Often, the settings and preferences dialed in from the front panel can significantly impact measurements. Or the instrument has “legacy” capabilities that may not be well understood. In that regard, it can be useful to review how to set up Teledyne LeCroy’s Motor Drive Analyzer (MDA) to see results consistent with a Yokogawa Power Analyzer (YPA) and to explain why the results will be different in some cases. First, for proper power analysis, one must determine a measurement period within which all calculations take place. Both the MDA and YPA use a “Sync” signal to determine the measurement period. In both cases, the Sync signal can be lowpass filtered to reduce the chances of finding an incorrect measurement period. For a nearly sinusoidal (e.g., low-distortion) signal, both instruments should find the same measurement period with the same LPF cutoff selections, though the MDA offers a wider variety of selections.
11 • 2018
DESIGN WORLD — EE NETWORK
41
11/6/18 1:49 PM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
Teledyne LeCroy Real Power P (in W) Apparent Power S (in VA) Reactive Power Q (in VAr)
Yokogawa Type 1
Yokogawa Type 2
Yokogawa Type 3
V × I (instantaneous sample points) V rms × I rms
√(S2−P2)
Power Factor (λ)
P/|S|
Phase Angle (ϕ)
cos−1λ
V rms × I rms
V rms × I rms × sinϕ
The MDA also permits adjustment of hysteresis (band), allowing user-defined control of the software’s ignoring of nonmonotonic factors that interfere with the measurement period calculation. Such control can be useful for signals that have higher distortion (e.g., brushless dc six-step commutated signals) or signals that have high distortion during high stress or failure events.
V rms × I rms
√(S2−P2)
√(Q2+P2) V rms × I rms × sinϕ
A simplified summar y of the Teledyne LeCroy method and the Yokogawa methods for calculating power for a single power cycle. Note the table assumes use of the V rms × I rms equation in the YPA for apparent power calculations.
Note the MDA also permits display of the filtered Sync signal with a measurement period overlay, rendering the Sync signal settings easily understandable. Finally, the MDA applies the Sync filter and hysteresis settings as a postacquisition software process, while most power analyzers apply Sync filtering and hysteresis settings to the waveform during the acquisition. Thus, in the MDA, it is possible to make changes post-acquisition whereas this is not usually possible with a power-analyzer instrument. Combined with the Sync period overlay, one may use MDA post-acquisition processing to fine-tune the filter and hysteresis settings for
Proven integrity AND industry know-how Electrocube is one of the most respected design manufacturers of passive electrical component products for a wide range of standard and custom applications – from aerospace and audio to elevators and heavy equipment – as a capacitor supplier, resistor-capacitor distributor, and more.
Bishop Electronics, Seacor, Southern Electronics, F-Dyne
ELECTROCUBE.COM | 800.515.1112 | INFO@ELECTROCUBE.COM
electrocute | P&EE 11.18HB update.indd 101 teledyne lecroy – P&EE 11-18 V3 FINAL.indd 42
11/6/18 11/6/18 1:48 1:51 PM PM
THREE-PHASE POWER MEASUREMENTS
The three-phase, three-wire (3 V, 3 A) wiring configuration voltage and current associations for the YPA (left) and MDA (right).
best results without having to take a new acquisition and lose the acquired data. The apparent power equation YPA users may measure apparent power (S) by selecting any of the following equations in its setup menu: Vrms× Irms: Same as provided by Teledyne LeCroy. Vmean× Imean: Product of rectified mean values calibrated to the RMS values. Vdc × Idc: Product of simple averages of the voltage and current. Vmean × Irms: Product of the voltage’s rectified mean value and the current’s true RMS value. Vrmean × Irmean: Product of the voltage’s and current’s rectified mean values. The MDA performs only Vrms× Irms of apparent power. To correlate power values from the YPA to the MDA results, set the selection for apparent power calculation on the YPA to Vrms× Irms. Teledyne LeCroy provides one method for calculation of power values whereas Yokogawa provides three different methods that return different results. For perfectly sinusoidal (zero-distortion) waveforms, it is possible to measure the phase angle ϕ between the voltage and current sinusoids. However, it is not possible to measure the phase angle ϕ when the waveforms are distorted (e.g., PWM drive output waveform). Therefore, the Teledyne LeCroy method or Yokogawa’s Type 2 method are the only methods that demonstrably produce accurate results for reactive power (and therefore power factor and phase angle) with distorted waveforms. (Both also produce accurate results with sinusoidal waveform). Yokogawa’s power analyzers offer the Type 3 method on models with the harmonic measurement-mode option. This option appears to enable definition of a fundamental signal from one of the PWM signals using a PLL source. Then, the instrument determines phase angle by comparing the fundamental voltage and fundamental current waveforms, with power determined
eeworldonline.com | designworldonline.com
1:48 PM
teledyne lecroy – P&EE HB 11-18 V3 FINAL.indd 43
for the fundamental and each harmonic through “N” harmonics. This technique should provide a result similar to that from the Teledyne LeCroy Harmonic Filter setting = “Fundamental” or “Fundamental + N,” provided the hardware PLL response in the YPA can accommodate any change in period of the measured signal during the acquisition window. Note that the YPA always calculates real power, P, correctly in all cases. If this is the only power value of interest, then all methods are suitable. However, to correctly calculate S, Q, λ, or ϕ, one must choose the correct YPA measurement method (if more than one is offered). The YPA calculates per-phase power using the same equations as total three-phase power but does so on one phase at a time. The Yokogawa Type 2 method with apparent power setting = Vrms× Irms always correlates (assuming that the YPA can obtain a proper Sync period) with the MDA when using three-phase, four-wire (three voltages, three currents) wiring configurations with lineneutral or line-reference voltage probing. However, if using line-line voltage probing, then perphase power calculations in the YPA will be unbalanced and incorrect, though the three-phase total will be correct. The reasons for the differences are as follows:
The practical impact of the Yokogawa wiring configuration is that the voltage and current pairs all have different phase relationships.
11 • 2018
DESIGN WORLD — EE NETWORK
43
11/6/18 1:53 PM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
The Yokogawa three-phase, three-wire (three voltages, three currents) wiring configuration is natively defined as a two-wattmeter setup. This is beneficial in that switching from a three-voltage and three-current measurement to a two-voltage and two-current measurement requires no re-connection of wires to the YPA. However, it also means that the line-line voltages and currents are incorrectly associated with each other on a per-phase basis. The Teledyne LeCroy three-phase, three-wire (three voltages, three currents) wiring configuration also uses a two-wattmeter method for total threephase power calculations. However, the MDA maintains the correct per-phase vector relationships to obtain proper per-phase calculations. The MDA simply inverts one of the voltage waveforms for the total three-phase power calculation. The voltage associations made by the YPA have no impact on the total three-phase power (real, apparent, or reactive), phase angle, or power factor because the vector (voltage and current) relationships defined in their wiring setup are the correct relationships for the two-wattmeter method used to calculate their total three-phase power. Both instruments can perform a line-line to line-neutral conversion (referred to as a delta-star conversion by Yokogawa, an extra-cost option). However, while the MDA will return accurate perphase power calculations for P, S, Q, λ, and ϕ with this type of conversion, the YPA will return correct per-phase power calculations only for P.
The practical impact of the Yokogawa wiring configuration is that the voltage and current pairs all have different phase relationships (which is what leads to the incorrect per-phase power calculations).
REFERENCES Teledyne LeCroy motor drive analyzer www.teledynelecroy.com/ oscilloscope/mda800a-motordrive-analyzer
Steady-state vs. dynamic events The YPA calculates a mean steadystate power value over a normally short acquisition time (several cycles). There is an option for boosting the update rate for the mean-value calculations, but it is still a mean-value calculation over a defined period of time or number of cycles. The Sync period is determined through a PLL circuit programmed into an FPGA, which has a limited ability to lock onto widely varying speeds, dependent on the PLL loop bandwidth. In general, there is no practical limitation on a steady-state signal with a slowly changing measurement Sync period. Thus, the YPA is good for measuring mean values of power in steady-state (constant load, torque, speed, etc.) operating conditions. The MDA can acquire short records and calculate mean power values for display in a table. Recall the Sync period is determined using a software algorithm. However, it can also make accurate measurements under dynamic operating conditions or loads because it lacks the operating limitations of a power analyzer instrument. If the load is dynamic, it is quite possible a power analyzer instrument cannot provide a meaningful result. First, ensure correlation on a simple, static, steady-state load condition, and then attempt correlation with something more dynamic. If one cannot correlate results during dynamic operation, it is likely because the power analyzer instrument cannot accurately determine or verify Sync period, which is not typically a limitation of the Teledyne LeCroy MDA. It is easy to correlate results from a Yokogawa Power Analyzer instrument to a Teledyne LeCroy Motor Drive Analyzer. The MDA provides more visual feedback on its operation, especially in accurate calculation of the Sync measurement period, which is vital to accurate power measurements. In some cases, a YPA has different settings or makes different wiring assumptions that lead to different results, which can be easily understood once they are known.
Calculated results with and without a line-line to line-neutral (what Yokogawa refers to as a delta-star) conversion for the YPA (blue background) and MDA (black background).
44
DESIGN WORLD — EE NETWORK
teledyne lecroy – P&EE HB 11-18 V3 FINAL.indd 44
11 • 2018
eeworldonline.com | designworldonline.com
11/6/18 1:54 PM
e
m
ENERGY EFFICIENCY, SMART LIGHTING
Energy efficiency, smart lighting, and the IoT C E E S L I N KS • Q O RV O
Sure, smart lights will save energy, but they’ll also be the backbone of systems that can help minimize minicatastrophes.
Image courtesy Qorvo
eeworldonline.com | designworldonline.com
qorvo – P&EE HB 11-18 V4 FINAL.indd 45
THE IoT INCLUDES a rich offering of energy efficiency benefits. Smart lighting is, of course, one of the most common elements of any IoT energy efficiency discussion. Lighting is a feature throughout every room of every building, and every light fixture connects to a power source. The IoT is generally available to every area of the economy. This is a benefit to keep in mind when considering the future of the IoT, particularly for energy efficiency improvements. The IoT’s scalability can bring energy efficiency benefits for all. The IoT promise of “intelligent buildings” offers the ability to view the operations of commercial buildings and receive the data needed to improve efficiency, lower costs and improve the experience of both management and tenants. The ability to get at this data creates a fundamental shift in how commercial buildings are managed. Before the IoT, “conventional” building management involved a set of tools like spreadsheets, monthly utility bills and operations procedures. This approach has historically focused more on tracking the operations inside the walls than on optimizing them. (This applies to homeowners, too, by the way. Traditionally, homeowners have focused on getting bills paid and making sure they were “in line” with expectations.) But the datagathering and processing of the IoT offers building managers the ability to go beyond operations tracking and to make better, informed
11 • 2018
decisions that create efficiencies and save money. (For homeowners the same applies—the IoT offers a new perspective on improving those bills, not just tracking them.) Within the spectrum of IoT-powered features, energy efficiency is crucial to determining intelligent building ROI. Using fewer kilowatt-hours means spending fewer dollars on utilities. Some smart lighting controls may do far more than simply turn lights on or off. Indoor smart lighting can encompass more than just a motion sensor that detects when someone enters or exits a room. It can also determine where and how many occupants sit in the room and whether the room temperature should be adjusted to allow for them. Use of smart lighting controls, optimized air conditioning and heating, and better indoor air quality add up to real, bottom-line savings. (Truly) smart homes IoT-driven savings are available for homeowners as well. The list of smart home energy savers is extensive and includes things like automated schedulers for electronics, lights, and HVAC. But a truly smart power system in a home would go beyond schedules and pre-programmed options. It would monitor and manage how and when power is consumed. It could be used to control the amount of time your kids spend on their electronic devices (and I am sure they will love this feature). It could turn off appliances or power-consuming systems unless they were in use. It might also automatically open and close window shades or curtains to adjust for the sun and the season. For consumers who are on time-of-use rate plans, a smart electrical distribution panel can charge a home power-storage
DESIGN WORLD — EE NETWORK
45
11/6/18 3:26 PM
P O W E R & E N ERGY EF F IC IEN C Y HA N D BOOK
system during the day via solar panels or at night when power is computer, rather than by a light switch or dimmer. Smart lighting less expensive. Then power-greedy appliances can use cheaper also plays a role in “soft security” as an anti-burglary device, electricity from storage instead of from the grid during expensive replacing the electromechanical timers of yesteryear that can give rate times. These systems, controlled by a master app rather the impression that the house is occupied when nobody is home. than by individual controls for each feature, already find use in Smart lighting is also useful on the industrial front. For industrial applications and will soon move to home use. example, large users of artificial light use smart lighting that An effective smart home can make its occupants smarter, too. senses daylight. The sensed natural lighting conditions determine People educated about the power consumption of their appliances whether connected bulbs are on or off, saving energy. are more likely to turn them off when not in use and to use them Additionally, the IoT can bring energy efficiency advantages in more energy-thrifty ways. For instance, it’s much more efficient that are less obvious. Some are in the category of unexpected to run washing machines or dishwashers that are filled to capacity. expenses from accidents and equipment failures. When a water Running appliances half-full heater starts to go bad, for example, wastes power and water, not to the problem typically manifests THE SMART HOME BUTLER mention detergent. itself as a slow leak. This failure can T H E R E A L S M A R T I oT A network of position and be tricky to identify. In many cases, motion sensors enables smart the water heater continues to run, homes to control temperature heating both water for washing and and lighting to accommodate water that is leaking out. This runs rooms in use. For instance, the up the utility bill. The simplest fix is system would be smart enough to install a leak detector that alarms to turn off the lights and the when the tank fails. A better idea is A/C in empty parts of the home to connect that leak detector to a of a family huddled around a smart home network that can alert TV on a hot summer night. It the homeowner and power down the would energize the A/C in the water heater. bedrooms when family members Of course, water heater tanks head off to bed. Because can also rupture and spill gallons of many people prefer cooler hot water, creating costly damage. temperatures for sleeping, the Ditto for a frozen water pipe that system could be smart enough to breaks. A smart home with a waterslowly reduce the temperature at flow sensor can notice water flowing night and then raise it as morning in the pipes when no one is home. It approaches. can warn the homeowner and turn off For another illustration, the water at the main valve, heading imagine someone entering and off flood damage. saying “good evening” to the Insurance companies are Alexa Pod. Alexa replies, turns recognizing that smart homes on the A/C, music, a lamp, and can potentially reduce claims. By Image courtesy Qor vo lowers the blinds. Later, motion preventing a flood, a smart home sensors detect the occupant leaving. The movement matches system can also prevent the mold resulting from water damage—a learned patterns, so the smart home system shuts down the music common home insurance claim. and lights, turning on the bedroom lights as someone enters. The same can be said for natural disasters like earthquakes, The analytics behind the smart energy system are key. The tornados and hurricanes. Imagine being able to extinguish power, system learns from the occupants to predict future behaviors. gas and utilities with a single touch of a smartphone. Even better, Of course, simple overrides are available, but the smart home think of a smart system that recognizes weather alerts and can absorbs the majority of patterns and uses them to enhance automatically disconnect appliances and close storm shutters. settings for comfort, convenience and cost-savings. The effect is Most of us recognize how the smart building can augment like an electronic butler who knows and anticipates habits—and security and convenience and boost energy notices unusual changes in your patterns. efficiency. But truly smart energy management REFERENCES Now consider smart lighting. Smart lighting is one of the most can limit damage from natural and unnatural common elements of any IoT energy efficiency discussion. Smart disasters, reduce insurance costs, educate Qor vo bulbs are becoming more widely available. They connect to the people about their power usage, while making www.qor vo.com internet and are controlled through apps on a mobile device or life more comfortable and convenient.
46
DESIGN WORLD — EE NETWORK
qorvo – P&EE HB 11-18 V4 FINAL.indd 46
11 • 2018
eeworldonline.com | designworldonline.com
11/7/18 3:55 PM
Automation & Control Freak? Same here.
We carry more automation & control brand names than any other distributor in North America. It’s true – we checked.
© Allied Electronics & Automation, 2018
JAN18 Allied-AC— Freak Allied Electronics AutoAd HB(DW).indd 8-18.indd 147
Get your A&C fill at
alliedelec.com
1.800.433.5700 12/15/17 11/7/18 9:07 12:44AM PM
ad index POWER & ENERGY EFFICIENCY
DESIGNWORLDONLINE.COM | EEWORLDONLINE.COM
Allied Electronics & Automation............................................... 47
Microchip .....................................................................................1, 13
Chroma Systems Solutions .......................................................BC
Newark element14...........................................................................9
Coilcraft ........................................................................................... 31
Pico Electronics ...........................................................................IBC
Digi-Key ...........................................................................Cover, IFC
Rutronik........................................................................................... 26
Electrocube ................................................................................... 42
Tadiran ............................................................................................ 19
Keystone Electronic Corp. ......................................................... 15
Wago USA .................................................................................... 35
Memory Protection Devices, Inc. ...............................................3
Sales Jami Brownlee jbrownlee@wtwhmedia.com 224.760.1055 Mike Caruso mcaruso@wtwhmedia.com 469.855.7344 Garrett Cona gcona@wtwhmedia.com 213.219.5663 Bill Crowley bcrowley@wtwhmedia.com 610.420.2433 Michael Ference mference@wtwhmedia.com 408.769.1188 @mrference Mike Francesconi mfrancesconi@wtwhmedia.com 630.488.9029
Leadership Team David Geltman dgeltman@wtwhmedia.com 516.510.6514 @wtwh_david Neel Gleason ngleason@wtwhmedia.com 312.882.9867 @wtwh_ngleason Jim Powers jpowers@wtwhmedia.com 312.925.7793 @jpowers_media Courtney Seel cseel@wtwhmedia.com 440.523.1685 @wtwh_CSeel
Publisher Mike Emich memich@wtwhmedia.com 508.446.1823 @wtwh_memich Managing Director Scott McCafferty smccafferty@wtwhmedia.com 310.279.3844 @SMMcCafferty EVP Marshall Matheson mmatheson@wtwhmedia.com 805.895.3609 @mmatheson
FIND US ON
TWITTER @DESIGNWORLD
48
DESIGN WORLD — EE NETWORK
AD INDEX — P&EE HB 10-18 V1.indd 48
11 • 2018
eeworldonline.com | designworldonline.com
11/7/18 4:13 PM
A59_DesignWrld_9x10_875.qxp_Layout 1 9/18/18 9:38 AM Page 1
DC-DC CONVERTERS
2V to 10,000 VDC Outputs, 1-300 Watt Modules • • • • •
Regulated/Proportional/Programmable • Isolated/Single/Dual Outputs High Reliability Greater than 1,000,000 hrs. Mil Hbk 217F Military Upgrades Expanded operating temperature -55º to +85º C, (no derating required) Environmental Screening Selected Screening, Mil Std 883 ULTRA Miniature From 0.5" x 0.5" x 0.2" Surface Mount/Thru-Hole
Custom Models • US Manufactured • AS 9100 Approved High Reliability AC/DC Power Supplies, Transformers and Inductors
YOUR PROVEN SOURCE.
Over 50 years providing reliable products. Over 2,500 Standard Modules. See PICO’s full line on our NEW Website with Search Wizard
www.picoelectronics.com
PICO ELECTRONICS, Inc. pico – P&EE HB 11.18.indd 1
143 Sparks Ave., Pelham, New York 10803 Call Toll Free 800-431-1064 • FAX 914-738-8225
E Mail: info@picoelectronics.com
11/7/18 12:47 PM
POWER
ADVANCED POWER AND ELECTRICAL SAFETY TEST
AC Power Sources DC Power Supplies Electronic Loads Power Meters Multimeters
SAFETY AND COMPONENT Hipot Testers Safety Analyzers LCR Meters Milliohm Meters Wound Component Analyzers Impulse Winding Testers
AUTOMATED TEST SYSTEMS
EV POWER COMPONENTS | CHARGERS | BATTERIES
Battery (cell, module, pack) EV Supply Equipment On-board Charger Motor Driver SMPS / DC-DC Converter PV Inverter
AUTOMATED TEST EQUIPMENT From Design Validation to End of Line Testing We supply power electronics test equipment and automated test systems to the most innovative R&D and production facilities all over the world. To find out more about our test solutions, visit chromausa.com.
Power Conversion and Electrical Safety Test Instruments and Automated Systems chromausa.com | 949-600-6400 | sales@chromausa.com © 2018 Chroma Systems Solutions, Inc. All rights reserved.
Chroma — T&M HB 06-18.indd 1
11/7/18 12:45 PM