Designing Fire Prevention into Equipment4 Safety Newsletter

Product Safety Newsletter • 1

What's Inside

Vol. 7, No. 3 May-June 1994

Chairman's Message

Continued on page 20

Chairman's Message .................................. 1

Officers of the PSTC's ............................... 2

Letters to the Editor ................................... 3

Designing Fire Prevention into Equipment4

Area Activities ........................................... 5

Technically Speaking ................................ 6

News and Notes ......................................... 7

Prevention of Fire in Equipment ............... 8

Institutional Listings ................................ 22

Employment Wanted ................... back page

Defending The Line

Congratulations to the Santa Clara ValleyChapter of the EMC Society for anotherexcellent colloquium. The event was

well attended and an impressive number of ex-

hibitors were present. It was a good time forexchange of ideas and information. Whichbrings me to the point of this column. Some ofthe presentations and hallway conversationsraised some interesting questions when juxta-posed upon each other.

1. I spoke with one of the central EMC Societyleaders about promoting product safety papersfor next years EMC Symposium. Product safetyhas not had much of a history or presence in theSymposia records over the years. Half a sessionat this years Symposium in Chicago is devotedto biological effects, a great topic. But thequestion arises, what technologies do productsafety people have to report on? Are we thepresenters or the audience listening to expertsfrom other disciplines or sciences presentingsafety-related information.

TheProductSafetyNewsletter


The

ProductSafetyNewsletter

Officers of the Product SafetyTechnical Committees

The Product Safety Newsletter is pub-

lished bimonthly by the Product Safety

Technical Committee of the IEEE EMC

Society. No part of this newsletter

may be reproduced without written

permission of the authors. All rights to

the articles remain with the authors.

Opinions expressed in this newsletter

are those of the authors and do not

necessarily represent the opinions of

the Technical Committee or its mem-

bers. Indeed, there may be and often

are substantial disagreements with

some of the opinions expressed by

the authors.

Subscriptions are free. To receive a

subscription, send your request to:

PSN Subscriptions,

Dave McChesney

1865 Farndon Avenue

Los Altos, CA 94024

fax: (408) 296 3256

Comments and questions about the

newsletter may be addressed to:

The Product Safety Newsletter,

Roger Volgstadt (Loc. 55-53)

c/o Tandem Computers

10300 North Tantau Avenue

Cupertino, CA 95014

Fax No. (408) 285 2553

Editor: Roger Volgstadt

News Editor: David Edmunds

Activities Editor: John Reynolds

Page Layout: Ken Warwick

Subscriptions: Dave McChesney

Institutional Listings: Ervin Gomez

Distribution: John McBain

Local

Central

Chairman Brian Claes (510) 659 6574(510) 659 6852 (fax)

Vice-Chairman Richard Pescatore (408) 447 6607Secretary-Treasurer John McBain (408) 746 5016

(408) 746 5258 (fax)Standards Paul McDonald (415) 592 5111

(415) 592 6052 (fax)Symposium Murlin Marks (408) 985 2400

ChicagoChair John Allen (708) 238 0188

(708) 238 0269 (fax)

Orange County/Southern CaliforniaChair/Secretary Charlie Bayhi (714) 367 0919Vice-Chair/Treasurer Ercell Bryant (714) 589 0700Program Co-Chair Michael Dastmalchian (310) 604 8739

PortlandChair Jim Pierce (503) 626 6694Secretary-Treasurer Scott Varner (503) 656 8841

Santa Clara ValleyChairman Mike Campi (408) 987 6527Vice/Program Chairman Murlin Marks (408) 985 2400 X2353Treasurer Mark Montrose (408) 247 5715Secretary Parviz Boozarpour (510) 527 7593

SeattleChairman Walt Hart (206) 356 5177Membership Chairman John Quigley (206) 226 1660

Texas (Central)Chairman Vic Baldwin (512) 990 6145


Letters to the Editor

Laying out design criteria in regard to Ap-proval, then testing hand made or machinedprototypes has proven to be successful for manyof our clients.Best Regards,John R. Allen

Dear John,I appreciate your kind and encouraging re-marks on the article “Product Safety as a De-sign Parameter”. I would comment on youritems concerning safety “over-kill” and non-technical management’s lack of appreciation ofsafety considerations in design objectives. Whenthe four elements listed under “How Safe isSafe Enough?” are satisfied, a reasonably safeproduct has been obtained. I would view anyadditional costs in the name of safety as possi-bly “over-kill” unless a convincing case can bemade that the additional costs in fact enhancethe safeness level of the product.

On the matter of non-technical management’slack of appreciation of safety considerations, Ifind that both non-technical and technical man-agers need a better appreciation of mandatedsafeness levels, criteria for accepting safetyrisks, regulatory requirements and other safetyaccountability stipulations these managers haveknowingly or unknowingly assumed. I coverthese items in a second text, “Managing Prod-uct Safety Activities” which the PSNL maywish to include in this series.

Sincerely,Paul W. Hill ❏

Dear Editor,The article "Designing Safety into Products"(Vol 6, Issue 3) was refreshing.

Much of our business is assisting product de-velopers and designers from concept. In everycase, except those in which marketing changesdirection, the product passes first time through.

I should add, however, designing for safety canlead to over-kill and/or custom components,driving the cost of the product [up] (ie: is a TCOreally needed in a heating device or can theheater be designed to open safely before there isa risk of fire or electric shock?).

The ultimate proof is in testing, testing, testing.

Although the up-front cost and time may seemunreasonable to a non-technical executive, de-signing Approval (and safety) into the productfrom concept will undoubtedly save many thou-sands of dollars in re-tooling, re-layout of PWBs,etc.


Designing Fire Prevention intoElectronic Equipment

by Mike GraboisAscom Timeplex

Abstract: Malfunctions in electronicequipment are the primary source ofoffice fires. One of the best ways to

prevent fires in electronic equipment is to in-terrupt the circuitry’s access to the power source.Fuses in circuits and some power supply fea-tures may do just that. There are also someuseful design tips that may reduce fire hazard.

Why Does Fire Happen?It is well known that electronic systems have atendency to fail occasionally. A majority of thefailures are caused by thermal stress. Thermalstress implies temperatures high above the nor-mal operating level. High temperatures on com-ponents may ignite the plastic of the packagingand in turn the PCB material. This kind ofoccurrence is difficult to predict and prevent,however, with appropriate selection of highquality components the risk may be minimized.

How Does Fire Sustain Itself?The section above shows how fire most com-monly starts. Other possible causes include: abroken air conditioner that allows the roomtemperature to rise or arcing on an intermittentcontact of high voltage components or brokenwire. Another question is why sometimes, oncestarted, the burning process subsides and re-sults in component or PCB damage, or some-times sustains itself, propagates to other com-


ponents, system modules, furniture, and even-tually creates a major catastrophe.

The source of any thermal stress in electronicdevices is a dissipation of power on resistivecomponents in the form of heat during thepassage of electric current. Most of the materi-als composing the PCB or semiconductor com-ponents are, or at least should be, fire retardant.This means that once the source of the heat isremoved, the fire will extinguish itself auto-matically within some time period. Therefore,when current path is intact and electric currentis present, heating up of the conductive ele-ments and burning process can be sustained.Once current path is interrupted, the poweraccess to the heating elements is removed andburning should subside and eventually stop.

My observations indicate that in order for a fireto last, it needs to be constantly fed by a powersource. The start of the heating and burningprocess may be caused by any component or asignal trace which are overheating and causingthe melting of the epoxy on PCBs or insulationon wires. The burning process, once started,produces a short circuit between the powerrails. The power layers in the multilayer boardsare especially susceptible to this kind of event.Once the material in the vicinity of the shortcircuit is burned out, the short circuit and burn-ing process propagate toward periphery, therebysustaining it further. The burning of the PCBingredients produces strong flames that mayheat up the PCB of neighboring modules start-ing a similar process in them, until the whole


Area Activities

interface between the equipment manufacturerand the facility needs to include.

Central Texas -The February 23 meeting featured a presentationon ‘Normas Oficiales Mexicanas, OfficialMexican S tandard NOM Procedures fo rElectrical Consumer Products’ by Mr. GeorgeJurasich of TÜV. Location: ROLM Corp., 2205Grand Avenue Parkway in Austin. The socialwas at 6:30; meeting at 7:00 pm.

The March 23 meeting was on the subject ofBABT approvals. Vic Baldwin reported sendingout 80 faxes in an attempt to publicize theirupcoming meeting, but unfortunately, only 8people showed up. The poor showing may havebeen due to bad weather . A nominat ioncommittee was formed, with Bob Hunter as thechairman, to nominate officers for the chapter.Voting will occur in May. Bob Hunter of BobHunter Associates spoke on BABT EquipmentApprovals Process and Vic Baldwin of ROLMspoke on the BABT Factory Approvals Process.Vic Baldwin and George Jurasich of TÜV Co-Chair the TC-8 committee.

Upcoming events:April - The next Central Texas get-togetherwill be a luncheon meeting and will occur onSaturday, April 30th at 12:00 noon in Arguile,Texas. This meeting will be a joint meetingwith the Dallas/Fort Worth EMC Society. Thefirst hour will cover European product safety

by John Reynoldsvoice: 415-390-1344fax: 415-962-9439e-Mail: [email protected]

Portland-The speaker for the March 15th meeting wasChuck Mello of Electro Test Incorporated (ETI).Chuck is the area Manager and Director ofTraining and Standards in the Oregon office.Chuck has been the Chairman of the City ofPortland Electrical Appeals Board since 1991and was President of the Oregon Chapter of theIn te rna t iona l Assoc ia t ion of E lec t r i ca lInspectors (IAEI) for 1993.

The subject was Fault Currents, WithstandingCapabilities and Grounding. The presentationbriefly addressed the facility as it relates tofault currents and grounding and what the



by Richard NuteCopyright 1994

Technically Speaking

Probably the single most frequently occur-ring and most misunderstood issue inelectronic product safety is electrically-

caused fire. I thought I would review fire pro-cesses in plastics materials (the most commonflammable construction material in electronicproducts).

The kind of fire that we are concerned with canbe defined as an uncontrolled combustion pro-cess, usually accompanied by flames.

There are two combustion processes, flamingand flameless. This discussion is limited toflaming combustion. Flaming combustion is theone with which we are most familiar, whereluminescent flames appear directly above thefuel material.

Flaming combustion is a chemical chain-reac-tion process which takes place exclusively inthe gas phase. Solids and liquids do not burn.Only gasses burn. If you look carefully at anyflame, you will see that it appears ABOVE thefuel material. This demonstrates that a gas isburning. However, the gas is evolved from thesolid or liquid by heating the solid or liquid toa “suitable” temperature.

Let’s examine the process by which a solid orliquid contributes to a fire. Consider ignitingpaper with a match (presuming the match isalready lit).

To initiate the fire (flaming combustion), thesolid material (paper) must be heated by asource (the match) whose temperature is greaterthan the combustible material’s “flame-igni-tion temperature”.

As the material is heated and the material tem-perature increases, chemical decomposition (py-rolysis) occurs. Pyrolysis is a chemical change(decomposition) within the material and is in-dicated by a change in color and by evolution ofsmoke. Typical products of pyrolysis are flam-mable gasses, non-flammable gasses, liquids,and carbonaceous residues.

As heat is continued to be applied to the evolvedgas, the pyrolytic gas temperature continues torise to the “flame-ignition temperature”. Atthis temperature, the flammable gasses, mixed



News and Notes

by Dave Edmundsfax: (716) 422-6449

VDE Guide to the EMC Directive89/336/EECThe VDE Guide to the EMC Directive 89/336/EEC is now available in the U.S. throughEuroPort. The guide consists of 308 pages andexplains in detail the European CommunityDirective on EMC. As a guide to relevant stan-dards, it has a section on interpreting Emissionand Immunity Standards as well as sections onHarmonized Standards, BSI Publications Re-lating to EMC, Principal VDE EMC Specifica-tions, and Principal US Commercial EMC Stan-dards in addition to an action plan for achievingcompliance.

Other VDE publications (in English) and inter-national standards from IEC, ISO, DIN and

others are also available from EuroPort. Call508-526-1687 for more information and pricequotations.Information submitted by R. Volgstadt

USNC Nominates Candidate for IECPresidentOn January 24 the USNC officially named Mr.Bernard Falk, former NEMA President, as acandidate for the IEC President for the 1996-1998 term.

TC 76 ChairmanThe Committee of Action (COA) of the IEC hasreaffirmed Dr. B. Kleman term of office asChairman of IEC TC 76 until March 1997.

IEC 825 ChecklistUL has copies of a checklist for IEC 825 firstedition with Amendment 1 dated 1990. Thischecklist will be required when submitting aproduct that contains a laser to IEC 950. Clause4.3.12 requires compliance to IEC 825.

Happy BirthdayCanadian Standards Association (CSA) waschartered in 1919, our congratulations on thisoccurrence of their 75 anniversary.

UL announces Generic Component ProgramIn a Subject 1950 letter dated January 31, 1994,UL announced several significant changes tothe way they approach safety of low-voltagedevices such as disc, tape and CD-ROM drives.In a 3 phase program, that has already started,UL has begun to evaluate low-voltage compo-



Prevention of Fire in Equipment

Fire tends to involve many individuals in thegeneral area of the incident because of smoke,evacuation of a building or spread of the fire toadditional areas. Secondly, fire is responsiblefor property loss while electric shock rarelyresults in a significant loss of property. Inaddition, Consumer Product Safety Commissionstatistics indicate there are nearly as manydeaths from fire as electric shock, but firescause many more injuries.

General Concepts for Prevention of Fire inEquipmentMany people have a general misconceptionabout the intent of safety standards concerningthe flammability of equipment. The generalmisconception is that equipment must beessentially fireproof. In fact, in any generalconflagration the equipment is not expected tosurvive intact or in any minimal operationalstate.

Product Safety standards require the followingkey fire prevention attributes in equipment:1. Components and insulation materials within

the equipment are adequately rated for theapplication.

2. Components are protected from overstressing current, voltage and thermalconditions.

3. Equipment enclosures are capable ofcontaining ignitions within the equipment,should ignition occur.

4. The fire containment integrity of theequipment enclosure is maintained bylimitations on the number and size of

by Paul W. Hill & Associates© 1990

[We are grateful to the author for providinganother installment condensed from his book"Product Safeness As A Design Parameter",2nd Edition, 1990. The text is a registeredcopyright of Paul W. Hill & Associates, Inc.,and is reproduced with permission. Detailsabout the purchase of the book may be obtainedby calling (407) 368 2538.

Because of the amount of information in thisarticle, only part of the article will fit in thisissue. The second part will follow in the nextissue of the Product Safety Newsletter. - Ed]

Fire as a hazard differs from electric shockin several important ways. First, fireusually involves more individuals per

incident than electric shock. Electric shockalmost always involves a single individual.


enclosure openings.Causes of Fire in EquipmentThe major cause of ignition in electronicequipment is over stressed components.Charred resistors and scorched areas of printedcircuit boards are common evidence of this.The component causing the ignition is eitherunderrated for the application, or it was notadequately protected from over voltage, overcurrent , or excessively high temperatureconditions in the equipment.

Arcing is also a source of ignition in equipment.Arcing which is sustained or is of short durationbut of high energy content can ignite adjacentmaterials if such materials do not have resistanceto ignition characteristics. For this reason manystructural parts and insulating materials havespecial ignition resistance requirements.

Temperature excurs ions which br ing theequipment opera t ing ambien t to l eve l sexceeding thermal ratings increase the risk ofignition. Product safety standards specifyvarious abnormal operating conditions toevaluate the ability of over stress protectivedevices to prevent components and electricalinsulation materials from exceeding thermalratings.

Another factor which must be considered is thequantity of flammable consumables within theequipment such as lubricants, paper productsand chemicals. These combustibles contributeto the "fuel load" once ignition has occurred.Fuel loads influence the intensity and durationof burning and increase the difficulty ofcontaining, extinguishing or controlling a fire.

Adequate Component Ratings

Because over stress of components is a majorcause of ignition in equipment, designers mustselect components which are adequately ratedfor the following characteristics:1. Voltage operating range.2. Current handling ability.3. Component temperature limits.

Voltage operating range is the upper and loweroperating voltages specified by the componentmanufacturer. These are the voltage limits forwhich the component is designed for continuousoperat ion. Safety cer t i f icat ion tes ts foradequacy of component ratings is a series ofequipment operating tests performed with themains input voltages increased to ten percent ofthe rated input voltage. The rated input voltageis that specified by the manufacturer on theequipment rating plate. For these tests thenominal frequency of the input voltage is notvaried.

Components, wires and printed circuit boardconductors must be capable of handling thecurrents likely to be encountered. This mustinclude safe conduction of any fault currentthat is likely to be encountered until protectivedevices sense and react to the over currentsituation. Some faults may dwell at levels justbelow the activation point of fuses and thermalcut-out devices for long periods of time. This"cooking" of a component should be consideredin its rating for the application.

Undersize conductors will heat rapidly whenattempting to conduct currents not appropriatefor the cross sectional area of the conductor.Product safety standards specify minimumacceptable conductor sizes or cross sectionalarea for various current conduction levels.


The thermal ratings of components should begiven the same consideration as voltage andcurrent ratings. Thermal ratings of componentsshould be established by continuous operationin still air. This will establish the inherentthermal capability of the component whether itis cooled in operation by convection only orforced air.

Applicability of thermal ratings given by thecomponent manufacturer should be verified inthe proposed application. The influence offactors such as proximity to heat generatingdevices or the ability of the cooling mechanismto maintain components within thermal limitsmay be considerably different from staticthermal performance.

Care should be t aken to de te rmine thetemperature of cooling air passing aroundcomponents in the enclosure. Cooling air whichhas been preheated by components up-stream inthe air flow, could be considered as a possiblethermal derating factor for temperature sensitivecomponents.

Good design practice is to verify the adequacyof thermal ratings of components as used in theactual equipment application. The constructionof a prototype enclosure, even as simple as awood mock-up or model, with simulated coolingair or vents and internal partitions, can yieldva luab le in format ion on the thermalperformance of components in operat ingequipment.

Most product safety standards will set limits onthe operating temperature of components whichpose a risk of:

a) personal injury from burns, if parts areoperator accessible, or

b) degradat ion of d ie lec t r ic s t rength ofinsulation materials, or

c) ignition of consumable materials withinthe equipment.

Item a) concerns the surface temperature ofparts accessible by operators and servicepersonnel. The concern is not only the risk ofskin burns but also the risk of injury resultingfrom involuntary or startle reactions wheninadvertent contact is made with a hot surface.

Item b) concerns the operating temperature ofparts and materials relied upon for protectionfrom electric shock. These include insulatingmaterials and structural parts support inguninsulated conductors and providing requiredspacings between conductors. Included in thesethermal requirements are minimum softeningtemperatures for plastic parts.

I tem c) concerns f lammable l iquids andlubr ican ts used in the opera t ion of theequipment. Requirements for these materialsare minimum flash point, quantities permittedwi th in the equipment and prevent ion ofaccumula t ions o f exp los ive a i r -vaporconcentrations.

Temperature l imits may be expressed as"temperature rise" above an assumed cold startambien t o f 25°C. Tempera ture r i semeasurements are calculated from 25°C to thetemperature obtained when the equipment hasreached thermal stability. Thermal stability istaken to mean, after four hours of continuousequipment operation or when three consecutivetemperature readings taken at 15 minute


intervals show no significant change intemperature. The permissible temperature riselimits are given in safety standards for variousmaterials, components and accessible surfaces.

Temperatures in product safety standards arenormal ly considered surface temperatureobtained by a thermocouple. However, forsituations in which the thermocouple can not beplaced at the point of measurement, such as theinterior of a coil windings in transformers, thetemperature measurement method may bespecified as the change in resistance from acold start to thermal stability. This is anappropriate measuring method when the internaltemperature within the coil is more critical thanits surface temperature.

Protection of Components From Over StressComponents must be protected from abnormaloperating and fault conditions which forcecomponents to dissipate energy beyond theirinherent ability to do so. Over stress protectionfor components can be provided by:a) fusing,b) current limiting impedance in a circuit,c) over current and over voltage sense and

control circuitry, andd) thermal cut-out devices.

It is unlikely that only one of these approacheswi l l be sa t i s fac tory fo r comprehens iveprevention of component over stress protection.A combination of methods is recommended,one for slow rise faults and a second for fast risefaults. A common example of this dual approachis the placement of a thermal cut-out in one ofthe coil windings of an input transformer and afuse immediately ahead of the input winding.The fuse will respond to the overcurrent

conditions, and the cut-out responds to the overtemperature conditions.

A fuse is the most common approach to detectionof overcurrent conditions and interruption ofpower to the components it is protecting. As asimple one element device it is consideredextremely reliable in response to overcurrentconditions. It is often the reference whencompar ing the dependabi l i ty o f o therovercurrent sense and control methods.

Disadvantages of the fuse are that it must bereplaced after each rupture and the type andrating of the replacement is not fully controlled.Test and certification agencies will require thata fuse replacement warning be near the fuselocation and repeated in service manuals. Thewarning must state that a risk of fire exists if thereplacement fuse is not the fuse rating and typespecified.

Differences between the North American andIEC fuse lengths and rupture characteristicsmay be a problem in certifying equipment.Certification difficulties can be eased if thefuse is not operator accessible, replaceable onlyby authorized servicing or is of the cartridgetype which accepts either length fuse.

In some countries fuses must be sand filled orhave ceramic bodies, in certain fuse ratings.This is the result of glass body fuses shatteringwhen there is no substance to quench the heat ofthe arc and failure of glass casings to withstandthe thermal shock when the fuse ruptures.

All fuses, circuit breakers and similar interruptdev ices a re requ i red to be approved orrecognized components.


requirements; the second hour will cover EMCrequirements. Contact Vic Baldwin at 512-990-6145 for directions. Vic urges those in thecentral Texas area to support their local chapter.

Santa Clara Valley Group -The February meeting was held on Weds, Feb.23 (rather than the usual 4th Tuesday) at AppleComputer in Cupertino. About 30 peopleattended, including folks from UL, CSA, TÜVRheinland and the FDA. The speaker was Mr.Gene Panger of TÜV Product Service of Boulder,Colorado. Gene is a very enthusiastic speakerand we enjoyed his talk and lively discussionswith the audience.

Gene’s topic was Comparing the US’s Food andDrug Administration and Europe’s NotifiedBodies. The similarities are that both werecreated to keep unsafe products from beingmarketed. The differences are the different focusand accred i ta t ion l inkages (government ,accred i ta t ion body , ce r t i f i ca t ion body ,manufacturer and consumer) that comes fromdifferent approaches to manufacturing and sales.In general, he said that the US tends to thinksale of a product is OK as long as it does noharm. In Europe, the product must do no harmbut must additionally do what the manufacturerclaims.

The third party certifier must be neutral andtechnically competent so that a trust can beestablished and maintained between the certifierand other interests such as manufacturing,professional groups, government, consumersand others.

Circuits which have permanent impedanceswhich limit the current in a circuit are acceptableprotective measures. Such circuits will beevaluated by abnormal operation and simulatedfault tests to determine the degree of protectionprovided. The circuit impedance must beinherent in the circuit and not rely solely on oneresist ive element for the total protectiveimpedance.

Circuits which detect and control over voltageor overcurrent conditions are an acceptablemeans of component over stress protection.Abnormal tests and simulated faults are used toevaluate the degree of protection provided. Thereliability of the over stress sense and controlcircuits will be a major concern in equipmentsafety certification. The reliability of suchcircuits is expected to approach that of a fuse.The main issue is the overall reliability of thecircuits composed of a number of components,contrasted to the single element fuse, and thefail safe attributes of the sense and controlcircuits.

Thermal cut-outs are those devices whichinterrupt power to a circuit when the cut-outdevice reaches a predetermined temperature.These devices may be non-reset, operating onlyonce and require replacement, or manual resetand usable repeatedly. They may be a separatecomponent or embedded in a component windingor enclosure. Automatic reset thermal cut-outswould be considered a thermostat and may notbe viewed as a thermal protective device bytesting agencies.

_______________________________

Part II of this Paul Hill article on Preventionof Fire in Equipment will continue in the nextissue of the PSN ❏

Area ActivitiesContinued From page 5


standard, UL1950/CSA 950, 3rd edition.

CSA 950, 2nd Edition - published March 31st.

UL 1950 Bul le t ins to Power SupplyManufacturers - Power supply manufacturersare not automatically included in distributionof UL1950 bulletins. Contact UL to ensure youreceive these updates.

Opto Couplers - UL has issued a Bulletin(Subjec t s 1577/1950) which reques t smanufac ture rs o f R/C Opt ica l I so la to rsvoluntarily provide UL with clearance distances,creepage distances and distances throughinsulation of their devices.

UL Bul le t in Subject 508B/1950 - UL ispublishing an outline of Investigation for RelaysUsed in ITE, subject 508B.

UL Bulletin Subject 1950 - UL announces ULGener ic Component P rogram. At amanufacturer’s request, UL will provide ageneric description in the Procedure for specifictypes of components. The R/C Directory will beexpanded to inc lude addi t ional da ta forcomponents. A manufacturer will be free tosubst i tute any component as long as theinformation in the R/C Directory is within therestrictive limits of the description in theProcedure.

Nordic Acceptance of TÜV - According toinformation provided by Hal Keeling, the Nordiccount r ies a re now accep t ing TÜV.Corresponding to the terms of the EuropeanEconomic Area (EEA), the foreign manufacturermust s ign a declara t ion s ta t ing tha t h isequipment meets the compulsory requirementsfor safety and EMC, based on data and/or reports

If you would like more information, pleasecontact Gene Panger, TÜV Product Service,(612) 631-2487.

The March meeting was held on Wednesday(this is becoming a habit!) , March 23 atWindham Hotel in Santa Clara. This was a jointmeeting with the IEEE System Software Society.The presentation was on Software Safety andReliability by Dr. Norman Schneidewind,Professor of information sciences at the NavalPostgraduate School.

Upcoming events:May 24 - Location to be announced. BrianClaes, LAM Research (and Chairman of theProduct Safety Technical Committee) will speakon Forensics. Elections for Central Committeeand the Santa Clara Valley Group.

Orange County/Southern California Group -The following topics were discussed at theFebruary meeting:Power Cords - UL and CSA to revise ITEstandard to allow SP and SV types for floorstanding and wall mounted units.

Job Openings - There were three openings listedin the newsletter. Please contact Charles M.Bayhi at 714-367-0919, for details.

Insulating Tubing - UL and CSA have rescindedthe requirement to surface mark insulatingtubing.

UL IAC Meeting - There was a UL/CSA meetingof ITE/Telecom joint Bi-National Standard onMarch 14-15. They reviewed the proposed


from a recognized testing body or “competentlaboratory”.

Elections - Congratulations to the recentlyelected officers of the Southern California/Orange County Group. They are as follows -Chairman/Secretary - Charlie BayhiVice Chairman/Treasurer - Ercell BryantPrograms - Michael Dastmachian

Upcoming events:April 3 - “ISO 9000 as applied to ProductSafety, EMI and RFI”. Speaker, Gilbert Walterof Safety & Compliance Eng. For details contactErcell Bryant Tel: (714) 589-0700.

Chicago-John Allen called to report that only has ahandful of individuals that are interested inholding Product Safety Technical Committeemeetings. Any help or advice in getting theChicago group going would be appreciated. Hecan be reached at 708-238-0188. [Perhaps aportion of a future column can be dedicated topublishing those good ideas and suggestions -Ed.]

_______________________________

I would like to thank all the different groupsthat send me their meeting announcements andnewsletters. They not only enable the writing ofthis column but they are encouraging to all theother groups.

John ReynoldsArea Activities Editor ❏

nents as commodities rather than as individualproducts. Eventually, a PC manufacturer willbe able to specify that any R/C drive that drawsno more than a defined maximum amount ofpower and that has certain minimum flamma-bility characteristics can be used in a certainlocation in its product.

The first phase will allow component manufac-tures to voluntarily have their products tabu-lated in the RCD. In the second phase UL willmail a list of these component manufacturers tothe end-use manufacturers who use them. Thelast phase will allow the end-use manufacturersto change their Reports to reflect the new de-scriptions.

Under the new system, a typical descriptionmight look like:"Hard Disk Drive - R/C (NWGQ), input voltage5/12 V d.c., loading current max. 1.5/0.5 amps,F/R minimum 94V-1, minimum clearance fromuninsulated live parts 4.0mm.”The RCD wi l l be changed to re f lec tmanufacturer’s name, model number of the com-

News and NotesContinued From page 7

For Pamela & Roger Volgstadt

A Baby!

Rebecca Danae Volgstadt

born February 26, 1994


NRTL clients in Germany before granting NRTLstatus to TÜV Rheinland of North America(TÜVNA), the wholly owned subsidiary of theGerman Government’s notified body, TÜVRheinland; (2) filing suit against Robert B.Reich as Secretary of Labor to have all OSHANRTL references to “Underwriters Laborato-ries, Inc. and Factory Mutual Research Corpo-ration” declared invalid as they, among otherthings, do not comply with the requirementsimposed by OSHA on other NRTL’s such asMET. This legal action follows numerous peti-tions to OSHA for “equitable treatment” byvarious NRTL’s. It is apropos to state that theACIL (American Council of Independent Labo-ratories) has filed similar comments to OSHAon the TÜVNA NRTL application, and OSHAhas granted additional time for ACIL to detailstill other arguments against granting the NRTLstatus. ❏

ponent drive, electrical rating, flame rating(typically of the bezel) and laser class, if appli-cable.

In the same letter, UL also announced expan-sion of its Listed Field-Installed Accessory Pro-gram. The expanded program will address drivesthat are intended for installation by non-servicepersonnel (typically, retail store employees orconsumers).

If you’d like more information about this bulle-tin or about the Generic Component Program/Expanded Listed Field Installed Accessory Pro-gram, please contact Tom Burke at UL, SantaClara (408-985-2400, X2286) or your local of-fice.

The following material was obtained from theDecember, 1993 edition of the TMO Update,published monthly by The Marley Organiza-tion. This publication is a confidential newslet-ter providing subscribers with information inthe fields of government and non-governmentstandardization. Contact The Marley Organi-zation at 203-438-3801 for subscription infor-mation.

Equal Treatment under NRTLANSI Accredited Committee Z34 Chairman andMET Laboratories Inc. (MET) President ,Leonard Frier, as he has successfully done inthe past, has become the advance advocate forU.S. independent conformity assessment third-party organizations by: (1) Filing a requestedcomment with OSHA’s Nationally RecognizedTesting Laboratory (NRTL) program that theyassure equal treatment, (ie., through the U.S.Trade Representative), for the product’s of U.S.

Please send any safety related articles to:Dave LorussoEMC Corporation171 South StreetHopkinton, MA 01748508-435-1000, x2130 (phone) or508-435-5067 (fax)


with atmospheric oxygen, will self-ignite, andflaming combustion occurs. For the purposes ofthis discussion, the flame-ignition temperatureof paper is taken as 451° F or 232° C.

Note that, for ignition, the GASSES must attainthe “flame-ignition temperature,” not the solidmaterial (although the solid material must beheated sufficiently to evolve flammable gas-ses). Consider that gasoline produces gasses atroom temperature, but those gasses do not burnspontaneously. Instead, the gas above a con-tainer of gasoline must be heated to its “flame-ignition temperature” before a fire occurs. Sucha gas is very dangerous since a small spark issufficient to achieve “flame-ignition tempera-ture.”

To sustain flaming combustion, the productionof flammable gasses (pyrolysis), must be sus-tained. Heat for pyrolysis must be continuallysupplied to the solid or liquid material, eitherfrom an external source or from the flame.

Sustaining flaming combustion after the exter-nal heat source stops generating heat (i.e., thematch is taken away) depends on the heat of theremaining flame.

If the heat from the flame exceeds the heatnecessary for pyrolysis, flaming will be sus-tained. Likewise, if the heat from the flame issmall, there will be less pyrolysis, the tempera-tures will decrease, and the fire will die.

So, a sustained fire is dependent on sustainedthermal feedback from the flame to the solid

material.

Considering the situation in terms of the law ofconservation of energy, pyrolysis is an endot-hermic (heat-absorbing) chemical process; flam-ing is an exothermic (heat-producing) chemicalprocess. If the exothermic reaction producesmore energy than that required by the endother-mic reaction, and there is adequate thermalcoupling between the two processes, then firewill be sustained.

In addition to producing heat, the flame alsoproduces combustion products, usually carbonmonoxide, carbon dioxide, water vapor, othergasses produced by pyrolysis and flaming of theparticular material, and solid carbonaceous resi-dues. Some of the solids are carried off byconvection currents, and some are left behindas ash.

Now, let’s consider ignition of a plastic mate-rial from a non-flame source, i.e., electricalheating.

Electrically-caused fire is electrical heating ofa material to ignition temperature followed bycombustion. So, we’ll now consider the situa-tion of non-flame heating and ignition.

In the case of plastics, as the material is heated,first the plastic softens, often changes color,then produces smoke, then boils.

Published decomposition temperatures of com-mon plastic materials range from 200° C forPVC to 500° C for PTFE, with most others inthe 300° C range. My measurements suggest adecomposition temperature of 225° C for poly-ester.

Technically SpeakingContinued From page 6


If heating is continued, and since we have noflame from an electrical heat source, the com-bustible gasses evolved from the heated plasticwill “self-ignite” at a temperature slightlyhigher than the “flame-ignition temperature.”Typical plastic self-ignition temperatures rangefrom about 350° C for polyethylene and polypro-pylene to 580° C for PTFE.

Sustained flaming combustion after the exter-nal heat source stops generating heat (e.g., thefuse opens) depends on the heat of the remain-ing flame.

For plastic materials rated HB, most of thepyrolytic gasses are flammable. The heat in theflame of such plastic materials usually is morethan the heat necessary for sustained pyrolysis.After ignition, when the external heat source isremoved from the material, flaming is sus-tained.

For materials rated V2, V1, V0, or 5V, most ofthe pyrolytic gasses are non-flammable. Theheat in the flame of the remaining flammablegasses of such plastic materials is less than theheat necessary for sustained pyrolysis. Afterignition, when the external heat source is re-moved from the material, the rate of combus-tion will decrease until extinction occurs.

The fire process is:1. Fuel material.2. Pyrolysis (an endothermic reaction requir-

ing heat).3. Pyrolytic products:

a. Combustible gasses.b. Non-combustible gasses.c. Liquid products.

d. Solid products.4. Mixing of pyrolytic products with atmo-

spheric oxygen.5. Flame (an exothermic reaction producing

heat).6. Combustion products.7. Sufficient thermal feedback from the flame

to the fuel material to sustain pyrolysis.

(Smoke is an indicator of incomplete combus-tion. Incomplete combustion is due to insuffi-cient pyrolysis or insufficient oxygen.)

With this background in the process of fire, wecan address control and prevention of fire inelectronic products.

_______________________________

AcknowledgmentsSome of the material presented here is theresult of a collaboration of Dave Adams, RayCorson, Kevin Cyrus, Richard Pescatore, andBrady Turner, all of Hewlett-Packard, and BobDavidson and Don Mader of Underwriters Labo-ratories.

A good basic explanation of fire is a videotapeentitled “Fire Concepts and Behavior,” FL-46VH, available from the National Fire Protec-tion Association, Batterymarch Park, Quincy,Massachusetts.

A good reference on burning processes of plas-tics is the “International Plastics FlammabilityHandbook ,” Second Edi t ion , by JurgenTroitzsch, published by Hanser Publishers,Munich. Distributed in the USA and Canada byOxford University Press, New York, ISBN 0-19-520797-1. ❏


Designing Fire Prevention into Electronic EquipmentContinued From page 4 The last statement has always been of great

concern to me as I have observed a number offires originating on the backplane. In order toeliminate this source or at least to alleviate thepossibility, the area covered by the power dis-tribution should be minimized. That will re-duce the possibility of a short circuit betweenthe two power rails in case of the overheatingand melting of the backplane material.

Another important aspect of the backplane de-sign is the proximity between the signal tracesand the power distribution areas. A majority offires start because of some kind of componentfailure. Components may cause short circuits orotherwise overload the signal traces connectedto them. Therefore, close proximity of signaltraces to the power distribution on the backplaneis undesirable. It is preferable to deliver poweralong the periphery of the board without anytraces above or below the distribution areas inthe multilayer structure of the PCB. The refer-ence layers, required for the control impedancelines, may be all of the same potential, such asGround layers. Thus, the risk of a massive shortcircuit between the different voltage layers inthe backplane will be greatly reduced.

I would not recommend a fuse or a circuitbreaker on a power supply rail common to thesystem. It has a much larger effect on systemperformance and significantly reduces the reli-ability of the whole system. A better featurewould be a shut down of the power supplyoutput in case of overload. The benefit of thispower supply characteristic is even more ap-parent in the case of the redundant power sup-ply. However, this feature should include cer-tain intelligence:

system is consumed.

Prevention TechniquesBased on how fire starts and sustains itself,limiting or interrupting the current in the eventof a short circuit failure is the best fire preven-tion method.

One of the best ways to eliminate the possibil-ity of sustaining the burning process is to pro-tect the load from short circuit or overcurrentconditions. A simple fuse on all of the powerentries to all circuit modules will be very effi-cient. The burning process on any componentor PCB may start with a relatively small amountof current. However, any further developmentand spreading of the process will require asignificant amount of current. That is why afuse would be good protection. However, thereare number of precautions and issues that shouldbe considered:

- A fuse has resistance and will cause a certainvoltage drop which should be considered in theoverall voltage budget.

- A fuse may interrupt power to a single modulewhile the others are on line. The backplane busdrivers/receivers should not cause a bus hangup.

- Value selection of the fuse and its placementon the module should be done with consider-ation to the local temperature conditions.

- Module fuses do not provide protection forfailure on the backplane or wiring.


- The power supply should shut its output oroutputs down only after a significant voltagedrop on the common rail, indicating a massiveshort circuit.

- The power supply should allow overload topersist for a few seconds before activating theshut down in order to permit a destruction of ashorting device or material, if they are smalland localized, or filter out the glitches.

Incorporating Smoke Detection Equipmentinto Electronic DesignI once have investigated a peculiar applicationof smoke detectors used in common fire alarms.The conclusion was that it is much easier, morecost effective and efficient to use fuses in com-bination with power supply shut down feature.However, I would like to share some findings,should anyone contemplate “improving” a sys-tem or device in this manner.

There are two types of smoke detection deviceswhich may be incorporated into the electronicequipment in order to prevent or alleviate a firehazard: smoke detectors with photo-electriccomponents or with an ionization chamber.Those with photoelectric action require a cer-tain amount of smoke in order to react. Theionization type devices perform better in casesexhibiting low smoke levels or flame presencesince they react to the presence of ionizedparticles. The smoke detector with an ioniza-tion chamber also is better suited to the appli-cation in electronic equipment. Among the ACand low voltage DC powered devices, the lowvoltage DC type is probably more convenientfor electronic equipment applications. Thesedevices may be powered from the secondary

power supply voltages (+12 V). Such a deviceshould have independent contacts that may beused to trigger a circuit breaker type devicewith remote trip coil or a relay.

To minimize equipment damage, fuses are rec-ommended over smoke detectors. Fuses typi-cally produce much better protection since theymay prevent fire at a very early stage, andtherefore, significantly lessen damage. Forsmoke detectors to react, they require some fireto be present. That means that at least one of thecircuit modules, including the PCB, or sometimes even more, will be destroyed. In addition,smoke detectors have the potential for falsealarms and require periodic maintenance.

During my investigations, I have entertainedsome thoughts about applying current and volt-age detectors for accurate indication of a poten-tially dangerous malfunction. Although I neverused this approach, in combination with veryaccurate gauging of the current consumptionsby the modules and overall system, it mayresult in a more accurate and more efficientdamage prevention technique. In conjunctionwith appropriate software support, it may alsoprovide an accurate failure detection tool forremote or local applications.

SummaryThe requirements of fire prevention in elec-tronic equipment are often overlooked or ne-glected by design engineers. However, the li-abilities may be significant, both the legal as-pects and in terms of the impact the reputationof the product and the company. The design offire prevention measures may be difficult —there is a large number of “what if” cases that


are not easy to cover and it is impossible toeliminate any damage completely. However,with due caution and thought to the design andproper use of fuses or other overcurrent protec-tion devices, it is relatively easy and inexpen-sive to create a healthy and safe system.

_______________________________

Mr. Grabois is a staff engineer at AscomTimplex. Mike has a BSEE from Moscow Insti-tute of Railway Engineering and a MSEE fromNew York University. He has worked in digitaland analog design at Ascom Timeplex since1985 and has been involved in design of equip-ment power distribution, power supplies andhigh-speed backplanes. ❏

to product design and/or product evaluationrequirements. On the other hand, they weredefended out of concern that exclusively high-level requirements would result in a plague ofwidely varying interpretations. Additional con-cerns were expressed for effects of varyinginterpretations made by test houses. Is contin-ued use of prescriptive standards that draw anarbitrary line separating what is “good” andacceptable from what isn’t the best practice?

3. I regret I was unable to attend Rich Nute’spresentation on “The Myth of Accessibility”.However, his slides in the colloquium recordappear to take a very interesting tack: ratherthan deal with accessibility using a collectionof prescriptive, arbitrary (though widely ac-cepted) requirements, he based his approach interms of a principle or collection of principles.

While it may seem like a stretch relating all theabove together, to me it came down to thefundamental issue of what we safety profes-sionals do. How much time do we spend enforc-ing arbitrary lines that someone else devel-oped, lines that we can’t well defend as opti-mum from either an engineering or technologi-cal standpoint? Why do we presume interpreta-tions as to what is safe enough would differ sowidely? And, assuming we can know why,what can we do about it?

Product Safety Practice—Brokering?Let’s shift gears, applying considerable imagi-nation in comparing product safety practice tothe legal profession. In both cases, the practi-tioners, equipped with their familiarity withcodes and standards, serve as experts arguingclaims or resolving differences, frequently be-fore independent third parties or judges.

2. I attended a presentation given by a long-time veteran of the IEC standards scene on thelatest from the world of international informa-tion technology safety standards. It was anexcellent update that was occasionally punctu-ated with apologies for the standard’s inad-equacies, this after many years of gestation.Eventually, the discussion migrated to compar-ing alternatives for the very nature of stan-dards, from very prescriptive (as we have to-day) to very generic (“products must be safe”)to hazard-based standards.

During the session, prescriptive standards werecriticized, on one hand, as stifling innovationby being very specific about applying the past

Chairman's MessageContinued From page 1


Summarizing some of the questions posedabove:1. Why do prescriptive requirements, even if

arbitrary and technologically indefensibleas optimum, seem to be preferred over high-level or generic requirements?

2. Can we promote a principle-based agree-ment reflecting what is safe enough?

3. To what extent is safety widely taught as acore discipline in our engineering schools,thoroughly integrated in the same way asother key engineering principles?

4. What core technologies are substantiallyunique to product safety practice?

5. Is our primary role to migrate safety-relatedtechnology or science-based enhancementsfrom other disciplines into mainstream en-gineering practice?

Maybe we should consider a change in our jobdescriptions. Maybe we should be promotingsafety in terms of engineering and managementprinciples, implicit and explicit contracts to bemaintained between producer and customer,rather than focusing on fine-tuning rules havinglong-forgotten origins that arbitrarily dividethe acceptable from the unacceptable. As areflection of this, perhaps our Symposium pa-pers should address how we can promote thereturn of safety practice to the core of engineer-ing and management practice, particularly if itbecomes clear that we do not have a technol-ogy-based core around which our engineering ispracticed.

I hope to address some suggestions in the nextissue. In the meantime, I’d like to hear yourcomments on the subject, particularly if youhave a different perspective or have ideas onwhat improvements should be made and how tomake them. ❏

Is our role to bridge the gap between safety-related science and technology and the classicengineering and management disciplines? Doesproduct safety practice have a technology basisor core? Or are we arbitrating someone else’sarbitrary technical requirements? With all thenegative press about how the legal profession isperceived today, are we as highly valued by ourorganizations or customers as lawyers are? Inwhat ways do we uniquely add fundamentalvalue to the product or service?

Product Safety Practice—Principles of SafetyLet’s take a step back and look at a biggerpicture. From a societal standpoint, laws in ademocracy represent a contract between theindividual, various groups and the society as awhole in order to fairly maintain order andbalance. When there is general widespreadagreement as to what is right and wrong andwhat is acceptable and unacceptable, historyhas shown the collection of laws can be at afairly high and generic level and be both simpleand brief. However, as societal cohesivenessbreaks down, as a primary set of consistentprinciples are no longer widely followed andstandards of conduct splinter, laws become morecomplex as various interests fight for their“rights” and privileges. These laws changeover time according to who’s in power, what’sfashionable, etc. The point is this: rules prolif-erate and become more prescriptive as prin-ciples become less understood and agreed to.

Does proliferation of prescriptive requirementsmerely attest to the fact that there is littleagreement as to what is safe enough? If this isthe case, will a large body of specific rulesreally help in attaining the goal of suitably safeproducts?


Institutional Listings

NCB Laboratory for Global Certification

YourListingHere?


We are grateful for the assistance given by these firms and invite application for Institutional Listingsfrom other firms interested in the product safety field. An Institutional Listing recognizes contributionsto support publication of the Product Safety Newsletter of the IEEE EMC Society Product Safety

Technical Committee. Please direct inquiries to:

Ervin Gomez at (408) 553-7684 (phone) or (408) 553-7694 (fax)-

YourListingHere?


BULK RATEU.S. POSTAGE PAIDCUPERTINO, CAPERMIT NO. 138

c/o Tandem Computers Incorporated10300 North Tantau Avenue, Loc 55-53Cupertino, CA 95014Attn: Roger Volgstadt

TheProductSafetyNewsletter

ADDRESS CORRECTION REQUESTED

As a free service to our readers, the Product Safety Newsletter willperiodically list Regulatory Compliance professionals who are availablefor employment. Those with employment opportunities are encouraged tocontact the following individuals directly.

Seeking employment as a Regulatory Engineer:

Employment Wanted

Donald E. Shaffer

Ithaca, New York

(607) 277-0629

Bogdan M. Matoga

Hollister, CA

(408) 636 8182

Naftali Shani

Brampton, Ontario

(905) 846 0930

Wayne L. Glover

Tampa, Florida

(813) 887 2534

Documents

Designing Fire Prevention into Equipment4 Safety Newsletter