chairman's message john mcbain 1 product...

Product Safety Newsletter •

TheProductSafetyNewsletter

®

Vol. 3, No. 3 May/June 1990

Chairman’s Message

Continued

In this issue Page

Chairman's Message John McBain 1

Technically Speaking Rich Nute 3

News and Notes Dave Edmunds 11

CSA Compliance Nick Maalouf 12

For Your Information Chuck Robertson 15

What Do Your Think R.C. Maheux 16

Guest Editorial John McBain 17

Area Activity Reports 18

Letters to the Editor 21

Institutional Listings 25

Calendar 26

Rich Pescatore

Rich Pescatore has asked me to fill infor him this time, since he has beenout of the country for about six weeksout of the last eight. I’m sure you willjoin me in extending our sympathy tothis poor fellow who has been forcedto spend his time drinking Germanbeer and French wine while touringEurope. What - not even a littlesympathy?

Last issue Rich wrote about theProduct Safety Newsletter. Previ-ously he has discussed the localgroups that regularly have technicalmeetings on topics of Interest toproduct safety professional s. An-other type of communication we aresponsoring is called by severaldifferent names: symposium, collo-

quium, convention or conference.

This year we are participating for thefirst time in the IEEE InternationalSymposium on ElectromagneticCompatibility, which is being held inWashington, D.C., from August 21 to23. This participation takes the formof a “Product Safety Special Ses-sion” scheduled for Tuesday, August21, starting at 2:00 p.m. The EMCSymposium Advance Programdescribes it as follows:

“All electric products have thepotential of being a source of electricshock. It is important in the design ofelectrical products to be aware of thevarious means for protection tominimize electric shock. The meth-ods Include: grounding, doubleinsulation, shielding, ground faultcircuit Interruption and polarization.This tutorial by the Product SafetyTechnical Committee will provide abrief overview in defining eachmethod and the advantages of each.A more in depth study will be made ofone or more of the methods as timeallows. .

This year would be an excellent timeto attend the EMC Symposium and

drop in at the Product Safety SpecialSession. Just in passing, anothertime to remember is that of the IEEEEMCS TC-8 (product safety) meeting,which is scheduled for Wednesday,August 22, from 7:00 to 9:00 p.m..Hope you can make it to Washington!

A cooperative effort like organizing aSpecial Session requires more thanthe presence of the presenters,although, of course, that is essential.Those people who devote hours oftheir time to arranging all the tediousdetails, who make sure things workwithout sharing the spotlight on thepodium, must be mentioned andpraised. Perhaps this sounds a lotlike your regular job: to make sureyour company’s products are safeand radiated emissions "don’t bringdown low flying airplanes", while themarketeers closing multi-milliondollar deals get the glory. Why notask any space shuttle passengerwhether he prefers the glory ... orgaskets that work? Keep up thatattention to details!

Sorry about the harangue. It wasinspired both by the upcoming



Chariman's MessageContinued

The Product Safety Newsletteris published bimonthly by theProduct Safety TechnicalCommittee of the IEEE EMCSociety. No part of this news-letter may be reproducedwithout written permission ofthe authors. All rights to thearticles remain with the au-thors.

Opinions expressed in thisnewsletter are those of theauthors and do not necessar-ily represent the opinions ofthe Technical Committee or itsmembers. Indeed, there maybe and often are substanttialdisagreements with some ofthe opinions expressed by theauthors.

Comments and questionsabout the newsletter may beaddressed to the ProductSafety Newsletter, Attention:Roger Volgstadt,Loc. 55-53, c/o Tandem Computers,10300 N. Tantau Avenue.,Cupertino, CA 95014. fax no. (408) 285-2553.

EditorRoger VolgstadtNews EditorDavid EdmundsAbstracts EditorDave LorussoAssistant EditorJohn McBainSubscriptions ManagerJohn McBain (acting)

Vol. 3, No. 3 May/June 1990 Symposium and by the President’sMessage in the last issue (Spring1990) of the EMC Society Newsletter.If you haven_ read it yet, then rightnow would be a good time. He isurging people to participate, to getinvolved, which will help the Societyand themselves. Does this topicsound a little familiar to readers of thiscolumn? And this leads me back tothe topic of symposia.

A number of people are contributing to the Product Safety Special Ses-sion at the Washington Symposiumthis year, including the sessionchairman, John Knecht, from Under-writers Laboratories in Northbrook,Illinois. This kind of participation isgreat, and we need more of it. But ifparticipating at the international level isa bit too intimidating for a first effort,have I got a deal for you!

The Santa Clara Valley Chapter ofthe EMC Society is sponsoring alocal colloquium to be held on June 12to 13, 1991. "IEEE SCV 91 EMC: AProduct Compliance Colloquium" willinclude an afternoon session onproduct safety. Speakers are beingsought now for this event. If you areinterested, please call Franz Gisin,Adm Systems Inc., at 408-492-3543.

I am hoping that other local productsafety groups may take advantage oftheir association with the EMCSociety and participate in localconferences like this one. Anytakers?

Best regards

John McBainSecretary-Treasurer

In this article, I have shown that thefuse current rating cannot be selectedon the basis of normal-mode current;it must be selected on the basis offault-mode current.

I have also shown that the fusecurrent rating must be selected toprevent that overcurrent which wouldresult in unacceptable overheating.

Acknowledgments. References forthis article are Bussmann’s “Elec-

tronic Designer’s Protection Hand-book,” Bulletin EDPH, Littlefuse’s“Designer’s Guide,” Form NumberEC 101, SAN-O Industrial Corpora-tion’s catalog, UL Standard 198G,and IEC Publication 127.

Your comments on this article arewelcome. Please address your com-ments to the Product Safety Newslet-ter, Attention: Roger Volgstadt, c/oTandem Computers Incorporated,10300 N. Tantau Avenue, Loc. 55,Cupertino, California 95014-0708.

Technically Speaking

(continued from page 10)


Technically SpeakingRich Nute

Selecting a Fuse Value

Rich Nute

Continued

Copyright 1990 by Richard Nute

Hello from Vancouver, Washington,USA:

Recently, a colleague asked, “Howdo I select the fuse rating for a newproduct?”

His cord-connected product has aswitching-mode power supply andconsumes 170 watts, with maximumnormal-mode input current of 2.74amperes.

One source suggested a rule-of-thumb that the fuse rating should beabout 1.5 times the maximum normal-mode input current. If he used thisrule, he would use a 4.2-ampere fuse.

On the other hand, he had another,similar product, also with a switching-mode power supply, consuming 180watts. Its maximum normal-modeinput current was 3.2 amperes, and it

used a 3-ampere fuse!

Naturally, with this gross differencebetween two similar products, mycolleague was confused. Let’s try toreduce the confusion.

I’m limiting my discussion to “small-dimension” fuses. Generally, small-dimension fuses are those that do notexceed 13/32-inch in diameter and 1-1/2 inches in length. Small dimensionfuses include the popular 5 x 20 milli-meters and 1 /4 x 1-1 /4 inches dimen-sions. In the parlance of UL 198G,these fuses are U miniature” and"micro.“

Before you can select the fuse currentrating, you must first answer thequestion, “What is the function or pur-pose of the fuse?”

The purpose of the fuse is to preventoverheating or fire in the event of afault on the load side of the fuse.

Electrical heating (power dissipation)is one result of electrical energytransfer. Electrical power dissipation(electrical heating) is of the form E*I orI*I*R or E*E/R.

To control overheating, we must havesome means of controlling or limitingpower dissipation. To control or limitpower dissipation, we must control orlimit voltage (E), current (I), or resis-tance (R).

In most cases, the circuit is a voltagesource so we cannot control thevalue of E.

Since we are dealing with fault cond-itions in the load, the value of resis-tance (or impedance) is obviouslyout of control; so, we cannot control orlimit the value of R (or, more gener-ally, impedance (Z)).

To control overheating due to electri-cat power dissipation, we must limit orcontrol the value of I.

Most overheating is prevented bylimiting the maximum value of I in thepower equations, E*I and I*I*R.

A fuse prevents overheating by auto-matically reducing the value of I tozero when the current increases to avalue that would cause overheating.

(Current limitation must be automaticbecause the circuit may not becontinuously attended by someonewho will manually disconnect thepower, and because overcurrent con-ditions are not necessarily immedi-ately apparent.)

An increase in current to a value thatwould cause overheating is pre-sumed to be a circuit fault-modecondition. The fuse rating or value isrelated to the maximum acceptableheating due to fault-mode current.


Technically SpeakingContinued

Continued

The fuse rating or value is NOT re-lated in any way to the maximum nor-mal-mode current of a product.

Before you can select the fuse currentrating, you must first have some ideaof how a fuse works and what itsspecifications and ratings mean interms of reducing the value of I tozero.

A fuse is comprised of a fusible linkencapsulated in an enclosure andconnected to contact terminals. Thelink is a metal which melts as a func-tion of current and time.

As with any metal conductor, at nor-mal temperatures the link has a lowbut finite value of resistance.

Current through a resistance results inpower dissipation in that resistance.The power dissipation is directly pro-portional to the resistance and to thesquare of the current, as shown by theequation, P = I*I*A. The result ofcurrent in the link is power dissipationand heating of the link.

As with most metals, the link is a posi-tive-temperature-coefficient device.That is, the resistance of the link in-creases with increasing temperature,and temperature increases with in-creasing current.

Nominally, the current rating of thefuse is the maximum current for whichthe link temperature is stable with

time.

So, then we have our first fuse pa-rameter, the current rating.

For currents greater than the fusecurrent rating, the link temperature isnot stable (i. e., the temperature in-creases continuously), and, at sometime the link will melt. Upon melting,the structure fails, the link breaks, andthe current goes to zero.

Electrically, for all currents up to itsrated current, the resistance of the linktypically is less than 0.1 ohm. Atcurrents greater than its rated current,the resistance of the link increasesnon-linearly with current and time to avery high value (hundreds ofmegohms) , which effectively reducesthe current to zero.

Whenever a circuit is opened, as theconductors separate, an arc occurs.Electrical power of the form E*I isdissipated in the arc. This powerheats the ends of the broken link; theends continue to melt back, thuscreating a continuously enlargingdistance between the two ends.Eventually, the distance becomessufficiently large that the voltagecannot sustain the arc and the airbecomes an insulator. (Some fuseconstructions use a granular insulat-ing material which, after the link melts,fills the space formerly occupied bythe link, effectively displacing the airwith a solid insulating material, thusquenching the arc.)

Now, in addition to the fuse currentrating, we have three time parame-ters for fuse performance: meltingtime, arcing time, and total time (whichis the sum of the melting time and thearcing time). These three times are afunction of the value of the current, I.As users of fuses, melting time andarcing time can be ignored; theparameter we need is total time as afunction of current Fuse manufactur-ers commonly publish 1- T curves foreach fuse current rating, where T isthe total time. This is our second fuseparameter.

(When the current is very high, themagnetic field, during both meltingand arcing periods, imparts energy tothe highly mobile molten metal, pro-pelling it to the wall -- enclosure -- ofthe link where it cools to its solidstate. This accounts for the silver,mirror-like deposits on the inside ofglass enclosures following a high-current fault.)

Note that the fuse operates by meansof heating, over a period of time, of aparticular conductor -- the link -- byelectrical power dissipation in theforms I*I*R (melting) and E*I (arcing).

Electrical power dissipation over aperiod of time is thermal energy, andis expressed in watt -seconds, eitherI*I*R*t or E*I*t We cannot expect afuse to dissipate unlimited thermalenergy; therefore, fuse ratings include



Continued

a maximum current (I) rating whichproduces the maximum thermalenergy which the fuse can safelydissipate. This maximum currentrating is known by several names:interrupting rating, breaking capacity,and short-circuit rating.

Now, we have a third parameter,interrupting rating.

The interrupting rating applies to afuse or other overcurrent device. InEurope, the terms “circuit prospectivecurrent” and “prospective short-circuitcurrent apply to the circuit on thesupply side of the fuse. Thesecurrents are defined as the maximumcurrent available from the supply intoa short-circuit.

The fuse interrupting rating must beequal to or greater than the circuitprospective current. More about thislater.

The thermal energy dissipation limitfor a fuse is established by both itsvoltage rating and its interruptingrating in the energy equations, I*I*R*tand E*I*t

A few years ago, I had the opportunityof participating in an experimentwhere we compared the performanceof two fuses, both with 10,000-ampereinterrupting rating, but with differentvoltage ratings. One was rated 250volts, the other rated 600 volts.

We set up a short-circuit (less than 0.1

ohm) and applied 480 volts throughabout 5 feet of wire from a utility pole-pig transformer. (We did this test at asite belonging to the local electricutility.)

With the 600-volt fuse, the onlyconsequence of applying the 480volts was the open fuse.

On the other hand, with the 250-voltfuse, three inches of flame shot out ofthe fuseholder, the fuseholder wasdestroyed, board conductors werevaporized, and arcs jumped betweenthe load-side board conductors andother unrelated board conductors. Inaddition, the connections from wiresto the unit under test showed signs ofsignificant arcing.

In the case of the 250-volt fuse, sincethe applied voltage exceeded therating, the energy the fuse wasrequired to dissipate was twice itsrating. The fuse resistance did notincrease, but, instead, let the energythrough where it was ultimatelydissipated by the various circuitcomponents.

When you exceed the fuse voltagerating, the fuse does not reduce thevalue of I to zero. This failure of thefuse to reduce the current to zeromeans that electrical energy iscontinued to be applied to the fault,and the safety of the situation iscompromised.

(Note, however, that the fuse voltage

rating is related to the fuse interrupt-ing rating. The voltage rating can beexceeded, and still get acceptablefuse operation, when the circuitprospective current is very muchless than the fuse interrupting rating.One fuse manufacturer suggests thefuse voltage rating can be exceededwhen the circuit prospective currentis no more than ten times the currentrating of the fuse. Generally, thiswould be the case for most high-voltage secondary circuits in elec-tronic equipment.)

Now, we have a fourth fuse rating: thevoltage rating.

In selecting a fuse, we must considermost of the fuse parameters:

1) Current rating.2) Current-time curve ratings.3) Interrupting current rating.4) Voltage rating.

From the circuit, we know the circuitvoltage, so we can readily choosethe fuse voltage rating.

Before choosing the interruptingrating, we need to know the prospec-tive short-circuit current of the supplycircuit. Fortunately, in North America,at least, for typical supply circuits(120/240 V and 120/208Y) the pro-spective short -circuit currents do notexceed 10,000 amperes; all ULListed and CSA Certified fuses --including 5 x 20 mm fuses -- rated 125



V or 250 V have 10,000-ampere inter-rupting ratings at 125 V. So, for the mostpart, we need not address interruptingratings as all North American fuseshave sufficient interrupting ratings.

However, IEC 5 x 20 mm fuses do nothave 10,000-ampere interrupting ratings.IEC 5 x 20 mm fuses are either 1,500-ampere or 35-ampere interrupting rat-ings.

(Be careful! There are TWO kinds of 5 x20 mm fuses: some 5 x 20 mm fuses areUL-Listed to UL standards, while other 5x 20 mm fuses are UL-Recognized toIEC standards. The UL-Listed 5 x 20mm fuses are 10,000-ampere interrupt-ing rating, while the UL-Recognized 5 x20 mm fuses are either 1 ,500- or 35-am-pere interrupting rating.)

If supply circuits have prospective short-circuit currents up to 10,000 amperes,then what good is an IE C fuse? On theother hand, IEC fuses have been usedfor years with no history of detrimentaloperation. Why?

At the load end of an 18 AWG 2-meterpower cord, the maximum current islimited by the impedance of the powercord, the plug and socket contact andother wiring contact resistances, and thebuilding wiring impedance back to thedistribution transformer, and by the size(impedance) of the distribution trans-former. To get 10,000 amperes, the totalimpedance of the supply and returnconductors must be less than 12 mil-

liohms. The contact resistance ofeach wire termination is of the or-der of 10 milliohms, with at least 12terminations in the circuit. Ignoringinductance, the system has 120milliohms which limits the prospec-tive short-circuit current to 1,000amperes. (I believe that theinductance is more significant andlimits the current to about 100 am-peres, but I don’t have the refer-ences readily available to prove itright now.)

Since the wiring of a practical in-stallation limits the prospectiveshort -circuit current to about 1,000amperes, the IEC 1,500-ampereinterrupting rating is adequate inmost cases, and we do not seesigns of detrimental fuse operation.

Finally, we come to selecting thefuse current rating and the current-time curve rating.

The fuse current rating selectionprocess is really quite simple:First, the fuse current rating must begreater than the normal current ofthe load. Second, the fuse currentrating must be less than that currentat which unacceptable heating oc-curs in the load.

The first complication is that ofdetermining the current at whichunacceptable heating occurs in theload.

The second complication is that ofdetermining normal current as a func-tion of time.

Let’s first look at the current at whichunacceptable heating occurs in theload.

Overheating results from electricenergy being converted to thermalenergy when current exceeds normalcurrent. Since we are dealing withenergy, and since energy includesthe dimension of time, we are dealingwith current exceeding normal currentfor some period of time. That is, weare not dealing with overcurrent situ-ations of short duration -- say lessthan a few seconds. Such shortdurations, from a practical point ofview, do not usually result in sufficientenergy to cause ignition. So, we arelooking for a fault that results in asteady-state value of overcurrent thatcauses overheating.

In our search for the value of steady-state overcurrent which causes over-heating, we must first identify thoseparts which could dissipate powerand which therefore could overheat.Only those parts which can dissipatepower can overheat

Parts which dissipate power includeresistors, heaters, transformers, andsemiconductors. Other parts, such aswire, connectors, line filters, inductors,and switches, which don’t normallydissipate power, will dissipate power

Continued


Continued


under fault conditions. Both insulationand circuit component faults cancause excessive power dissipationin all of these kinds of parts. Note thatthe part which overheats is not thepart which has the fault.

On the other hand, some parts whichincur an internal fault will then dissi-pate power because of that fault.Capacitors and semiconductorsexhibit faults which will cause thecapacitor or semiconductor tooverheat. Fortunately, such faults arein the range of 0.5 to 1 ohm andtherefore can be simulated with aresistor.

Once you identify candidate power-dissipating parts, you must introducethe fault which will cause the part toover-dissipate more-or-less as asteady-state condition. (This is doneeither with the fuse shorted out, or thehighest available rating fuse in thefuseholder.) Then, you measure the“fault-condition” input current. Thefuse current rating must be less thanthis current.

Faults which cause unacceptableoverheating are not necessarilyshort-circuits. In fact, short-circuitsoften cause some power dissipatingcomponents such as resistors andsemiconductors to immediately failopen with no consequent overheat-ing.

More often, the faults which cause

unacceptable overheating are thosethat are of IOY/, but finite resistance.For DC circuits, we use electronicloads as fault simulators. For AC cir-cuits, a suitable load is a big 1:1transformer supplied by a big vari-able transformer. The output of the1:1 transformer is connected out-ofphase to the AC being subjected toan overcurrent condition. Both ofthese schemes give a continuouslyvariable load to the circuit under test.

This process applies to both primaryand secondary circuits and to bothlinear and switching-mode powersupplies. In some cases, you willneed both primary and secondaryfuses.

For many switching-mode power sup-plies, the parts which can fault do sowith relatively low impedance or re-sistance. For example, the forwardresistance of a bridge rectifier diodeis usually 1 ohm or less. Such faultimpedances result in very highcurrents; under such conditions, thevalue of the fuse current rating is notcritical.

However, there usually is one powerresistor in the primary circuit of aswitching-mode power supply. It isthe resistor in the snubber circuit.Often, it is in series with a capacitor.Often the capacitor is prone to short-circuit.. In such an event, the powersupply continues to operate, but thepower resistor must dissipate exces-

sive power, and may cause over-heating of nearby materials. Thefuse current rating should be se-lected based on the current thatoccurs with the snubber capacitorshorted.

In general, for a switching-modepower supply, the fuse current ratingshould be selected at the nextconvenient current greater than themaximum input current of theswitcher.

For some circuits - usually low-power circuits — the difference be-tween normal current and and faultcurrent is less than the increment ofavailable fuse current ratings.Transformers rated 50 watts andless, when subjected to fault condi-tions, often will overheat without in-creasing the current sufficiently tocause a fuse to protect the trans-former. In such cases, a thermalswitch is called for.

Now, let’s first look at the problem ofdetermining normal current as afunction of time.

Most loads have turn-on currentsthat are higher than the steady-statecurrent. The ideal way of selecting afuse would be to measure the loadcurrent as a function of time from turn-on to steady-state. Using this curve,we would overlay candidate fusecurves until we found one that every-where, from turn-on to steady-state,



Continued

Product Safety Newsletter • Continued on page 2


was just greater than the load curve.

There are two major problems withthis technique. First, the conventionalmeans of making time-dependentmeasurements is with an oscillo-scope. The scope measures peakcurrents, while the fuse responds torms currents. Second, electronicloads are usually non-linear; that is,the current is zero during a significantportion of the voltage cycle. Thismeans that the current peaks are notin the ratio of 1.41:1 peak-to-rms aswould be the case for a linear currentwaveform. Therefore, it is difficult torelate the measured time-currentcurve to the fuse time-current curve.

Since all fuses are thermal energy-operated devices, typical fuse time-current curves indicate the 10-millisecond current at 5 to 10 timesthe steady-state current. Typicaltime-delay or siow fuse time-currentcurves indicate the 10-millisecondcurrent at 10 to 50 times the steadystate current.

Let’s now look at “crowbar” circuits.The crowbar circuit I’m familiar withis an electronic short-circuit across thepower supply output terminals. Thecircuit deliberately blows a fuse whenthe voltage of a power supply outputgoes too high. It’s purpose is toprotect IC’s from catastrophic failurefrom high voltage in the event of ashort-circuit of the series-pass tran-sistor in the voltage regulator circuit.

I’m a bit uncomfortable with deliber-ately creating an overcurrent situation-- but I can’t justify my position with anengineering explanation. If thecrowbar circuit is designed to carrythe circuit prospective short-circuitcurrent, then the crowbar should beokay.

Another anecdote: The original crow-bar circuit worked as designed.Then, the circuit layout was changedand the SCA was relocated. As apart of the change, a connector wasadded to the circuit, and the wire sizereduced. When the series-passtransistor shorted, the SCA fired, butthe fuse did not blow. The reason:too much resistance had beenadded to the SCA circuit with theconnector and the smaller wire. As aresult, the crowbar circuit caught fire!

Some caveats:

Be aware that the world has TWOsmall-dimension fuse ratingschemes: (1) the North America (readUL and CSA) scheme, and (2) the IECscheme. To get the same perform-ance from either scheme, you wouldneed to specify two different fuserating values, one for a UL-CSA fuse,and the other for an IEC fuse. SeeFigure 1, Table 1, and Table 2.

Be aware that a fuse is a tempera-ture-dependent device; its ratings arefor nominal room temperatures. A

fuse should be physically located inthe equipment at a point where thereis little temperature rise. Or, select alarger current rating using the manu-facturer’s temperature de-ratingcurves.

Be aware that a fuse is a tempera-ture-dependent device; time-currentcurves change with physical sizeand, within the same physical size,with link constructions. Time-currentcurves are not of constant shapewithin fuse types. The curves willalso change as a result of the heat-sinking of various kinds of fusehold-ers.

Be aware that the typical slow” fusehas a minimum time rating at 200% ofits current rating, whereas the “nor-mal” fuse has no minimum time ratingat 200% of its current rating. A “slow”fuse will not blow in less than 5seconds at 200% rating for up to 3amperes, and will not blow in lessthan 12 seconds at 200% rating forfuses rated greater than 3 amperes;all other characteristics are the sameas a normal fuse. See Figure 2.

Be aware that most electronicproduct input current waveforms arenon-linear; to measure input currentfor the purpose of deciding a fusevalue, you must measure the currentwith a true RMS current meter.

Well... I hope this gives you someguidance for selecting a fuse value.


TC 76 Plenary

A plenary meeting of TC 76 (LaserSafety) is planned for Boulder CO theweek of June 4th. The major docu-ments for discussion relate tochanges to I EC 825 to incorporateFiber optics and high power materialprocessing laser.

IEC 950 Second Amendment/Edition

The 2nd amendment of IEC 950 isplanned to be published and avail-able late 1990. This amendment willinclude approximately 13 CentralOffice documents.

The 2nd Edition of IEC 950 is plannedfor late first quarter 1991 and incorpo-rate approximately 26 Central Officedocuments. There has been specu-lation this 1991 date is optimistic.

CBEMA had circulated list of IECTC76 documents, the clause it affectsand the CO document number.

Electromagnetic Fields

Several articles on the possiblehealth hazards of Electromagneticfield which may be of interest are:

News and NotesDave Edmunds

Dave Edmunds

Occupational Health & Safety News,February 8, 1990 - Some ScientificConcerns over possible HealthHazards of Electromagnetic Fields”

PC magazine, January 1990 “TheDangers on Your Desktop” U.S.Department of Energy, BonnevillePower Administration, Portland, OR,“Electrical and Biological Effects ofTransmission Lines: A Review”

Telecommunication

CENELEC has called for a vote, toits members, on document PREN41003 dated January 1990 entitled“Particular electrical safety require-ments for equipment to be con-nected to telecommunicationnetwork.” This document build onthe requirements of EN 6950.

EN 60950

All European countries will withdrawIEC 380 & 435 by 1 September 1990,and replace it with EN 60950. Allcountries have established a dateafter which the sale of equipmentevaluated to the IEC 380/435 will notbe permitted.

CSA 950

CSA has several documents for usewith CSA 950. One is a checklist, theothers are a side by side compari-son documents. One of thesecompares CSA 950 and UL 1950,then compares CSA 950 to UL 1950to IEC 950. The format and structureof the CSA checklist is similar in andlayout the ECMA checklist for IEC950. However, the CSA list has anexplanatory statement for each

clause. Mr. L Bahr of the Rexdaleoffice has issued these documentsand requested comments.

VDE 950 Checklist

VDE has a document titled “IEC 950Test Report”. This document issimilar in format to the ECMA IEC 950checklist with these differences: a)explanatory statement has beenadded for each clause, b) Parts 3 & 4have been incorporated into appro-priate clause, c) the Summary of Testcheck sheet has been removed.

IEEE Spectrum

A special issue of Spectrum isplanned which will survey the statusof consolidation efforts, the majorcorporate players and implications ofthe EC plan for a single market aswell as the political changes inEastern Europe.

Trade Associations Compared

The May 1990 issue of I EEE Spec-trum has an article comparing twoleading US electronic trade associa-tions EIA (Electronic Industry Asso-ciation) and AEA (American Elec-tronic Association). This articlebriefly reviews the goals and historyof each and how they aid the elec-tronic industry.

ANSI/IEC

To facilitate the coordination of ANSI/USNC and IEC, ANSI now has anindividual in the IEC office.

Send your news items to:Dave Edmunds (MS 843), c/o XeroxCorp., 800 Phillips Rd., Webster, NY,14580 (or fax 716-422-7841)


CSA ComplianceNick Maalouf

Continued

PRODUCT COMPLIANCE, ISTHERE A BEST APPROACH?

Many questions are now beingraised about the various approachesused by certifying agencies forensuring product compliance. In achanging world, all these questionsare valid. But, in order to appreciatethe different ways of achievingproduct compliance and be able toanswer these questions, we must firstunderstand the manufacturingprocess itself.

A manufacturing process follows thesteps shown in Fig. 1.

Idea_____ Specifications___ Design____ Development___ Production____ Outgoing Product

FIG. 1

First, a need is recognized frommarket research or customer de-mand. Then, specifications arecreated which incorporate customer’sas well as other requirements. Adesign is conceived to meet thesespecifications; and after experimenta-tion and development, a product isdeveloped. A production processthen manufactures the item, andfinally, the product is delivered forcustomer use.

Given such a process, the questionis: How can we ensure that theproduct delivered to the marketcomplies with the applicable codesand standards?

There are basically three ways to

control product compliance in themanufacturing process:

1. By controlling the outgoingproduct; or2. By controlling the system thatproduces the product; or3. By controlling both; the productand the system.

Which approach is best? Before weanswer this question, let’s examinewhat each approach entails, andthink of the advantages and disad-vantages of each of them.

CONTROL OF THE PRODUCTONLY - THE FOLLOW-UP AP-PROACH

This approach focuses primarily oncontrolling the outgoing product anddoes not concern itself with thesystem that produced the item.Control is achieved by the followingmeans:

��First, compliance is assessed bytesting and evaluating one or moresamples of the population, and bydocumenting the construction of theconforming product;

��The manufacturer then confirmsthat the item will be manufactured ina manner so as to ensure that theoutgoing product is same as thattested;

��Following which, formal certificationis granted by the certifying agencyfor the entire production line;

��And the actual control of continuingcompliance is achieved by conduct-ing unannounced follow-up inspec-tions at the factory. During such

inspections, a sample of the outgoingproduct is examined, and its con-struction Is compared to that of thetested sample.

Does this approach assure continualcompliance of the outgoing productat all times? An interesting questiontop be kept in mind.

CONTROL OF THE SYSTEM -THE SYSTEM APPROACH

This type of compliance control ispracticed in Europe under variousforms. The “manufacturer’s declara-tion” approach is the one variationmost widely used.

The concept borrows from QualityAssurance fundamentals, and isbased on the premise that compli-ance of the product can be assuredby effective control of the system thatproduces it. In simple terms, this isaccomplished by ensuring that thefollowing criteria are met:

��All related standards and coderequirements are referenced indesign specifications of the productin question;

��An acceptable Quality AssuranceSystem is implemented at thefactory;

��The product Is tested by themanufacturer who issues a declara-tion confirming that it meets theapplicable requirements.

��If an agency is involved, theagency carries out an annual audit ifthe QA system, and mayor may notperform a field product audit.


CONTROL OF THE SYSTEMAND THE PRODUCT: THECSA CATEGORY CERTIFICA-TION APPROACH

Category Certification is a CSAconcept of certification that assurescontinual product compliance bycontrolling the product as well as thesystem.

This concept was pioneered by CSAtwenty years ago with an implacablesafety record.

The superiority of the “CategoryConcept” is based on the fact that ittakes the best elements from the“product/follow-up approach” andcombines them with the best from the“system approach”. The result is aunique concept that integrates both:cure and prevention.

This seemingly impossible missioncan only be accomplished byserious dedication and commitmentto this objective and to product safetyby the two parties involved: CSA andthe Manufacturer.

Why commitment is important?Because, as mentioned earlier,continual compliance of the productunder Category Certification isachieved not only by controlling theoutgoing product, but also by control-ling the system that produces theproduct. Here is how this is done.

First, the manufacturer’s track recordand experience in codes and stan-dards is evaluated. Only those with aproven track record in product safetyare considered further; and, in orderto qualify, they must meet strict

CSA ComplianceContinued

requirements that only the mostresponsible of manufacturers canfulfill. One such requirement is thatthey must affirm their commitment toQuality and Product Safety by imple-menting a formal and documentedcompliance control system spanningthe entire manufacturing cycle; fromdesign conception, all the waythrough to shipment of the finalproductAssurance of continual compliance isachieved by frequent CSA Audits,and by the active involvement of CSAStaff in each and every step of themanufacturing process not only ascontrollers, but also as trainers andeducators of manufacturer’s staff.Let’s now retrace these steps andsee how they are controlled.

At the specification and design stage,control is achieved by:

��Ensuring that all pertinent safetyrequirements are incorporated in thespecs;

��Ensuring that product designersare familiar with the applicable codesand standards;

��CSA Staff reviewing all initialdesigns and also approving all futuredesign changes.

��As the design concept translates intoa prototype, control at this stage ofdevelopment is achieved by:

��CSA Staff evaluating the construc-tion and recommending the appropri-ate tests;

��Ensuring that the manufacturer hasan acceptable testing facility andqualified staff with experience in

product safety and testing to safetystandard;

��Performing the required tests;

��Ensuring that all records anddocumentations of product construc-tion and testing are maintained andcontrolled;

During production, compliance ismaintained by:

��Effective control of purchasedcomponents;

��Verifying production methods andprocedures to ensure that theprocess is capable of producing theproduct with a high degree ofconsistency;

��Verifying that the applicableproduction tests are carried out asrequired;

��Effective control of drawings andchanges;

��Actual inspection of the outgoingproduct by CSA Staff at the factory.

��Field audits of the product byCSA’s Corporate Audits & Investiga-tions Group. This Group also worksclosely with the Inspection Authori-ties by monitoring and acting onincidents in the field.

It is by now quite evident that the CSA“Category Certification” approachnot only provides very effectivecontrol of the product, but it alsostresses prevention. The conceptreflects a philosophy and a funda-mental quality engineering principlethat safety cannot be ensured by

Continued


TWO APPROACHES:CSA CATEGORY CERTIFICATION VS. PRODUCT/FOLLOW UP

Deals with prevention and cure Deals with cure only.

Provides assurance that safety No assurance.is designed into the product.

Initial certification of Initial Certification ofproduct category is granted tested model is granted by certifyingby certifying agency. agency.

Continuing involvement of No involvement.agency in product design anddevelopment.

Frequent audits of product No audits.during development.

Formal (Quality) Control Not required.Program required.

Total documentation control Descriptive report offrom design to delivery product.

Legal contract with agency Contract only.and formal commitment byMfg’s Senior Management toa total system of control.

Team approach to safety No training.(Mfg & Agency) resulting inongoing training of Mfg’sStaff .

System & Product audits. Product audits only.

Authorized Mfg’s staff verify Not allowed.compliance of product variationswithin Certified category.

Follow-up expert staff. Follow-up bygeneralists.

Third party certification. Third party certification.

Maximum confidence level in Minimum.product compliance.

testing and factory follow-up inspec-tions alone. Safety must also bedesigned into a product by designersfamiliar with the applicable standardsand the safety concerns expressed inthem. In addition, a quality manage-ment system must be put in place toensure continual commitment to, andunderstanding of product safety.Only then can we speak with confi-dence about product compliance.

Now back to the question posedearlier: Which of the three ap-proaches provides the best compli-ance control?

Rather than answer the question,here is a comparison table betweenthe CSA “Category Certification”Approach, and the “Product/Follow-Up” Approach which is the mostcommon. You, the readers andexperts in this field, be the judge.Your comments and input will also beappreciated.

Nick Maalouf, P. Eng.DirectorEngineering, International and Quality AssuranceCANADIAN STANDARDS ASSO-CIATION(416) 747-4236

CSA ComplianceContinued


A new specification, IEC 950, pres-ents a major change in suppressioncapacitors safety requirements.European safety agencies, includingVDE, SEV and SEMKO, specify thatoffice equipment manufacturersseeking approval to the new IEC 950specification must use either X-2capacitors with SEV approval, or theheftier X-1 capacitors. This changeaffects all equipment, includingcomputers and computer peripher-als, telephone equipment, copiers,fax machines, etc.

Although IEC 950 does not specifi-cally spell out the requirements forsuppression capacitors, it referencesIEC 384-14 which does. Two interpre-tations of 384-14 presently existamong the European Safety Agen-cies because of a vague descriptionof surge voltage requirements.However, a new version of 384-14clearly defines those requirements.

As it stands now, VDE says that X-2suppression capacitors are accept-able for IEC 384-14 compliance ifalso approved by SEV, the SwissSafety Agency. Sweden and otherScandinavian countries, on the otherhand, say that X-1 types must beused. But a unified position amongthe European Safety Agencies isexpected with the release of the newversion of IEC 384-14 because thesurge voltage requirements, nowprecisely stated, can only be met withX -1 capacitors.

As currently written, IEC 384-14 statesthat an X capacitor must withstand asurge voltage of 2500 V AC but it doesnot state the number of times thatsurge voltage is to be applied to thecapacitor. SEV requires 3000 VACsurges applied three times, which

For Your Information

some agencies say meets thecurrent IEC 384-14 specification. Butthe new 384-14 requires at least threepulses of 2500 VAC per hour for 1000hours, which means the X capacitormust withstand a total of 3000 appli-cations of the surge voltage. This ismuch more severe than the SEVrequirement and can only be met withX-1 types.

Once the new version of I EC 384-14is issued, manufacturers of officeequipment seeking new approvals toIEC 950 will have to use X-1 typesmay involve repacking because X-1types are substantially larger than theequivalent X-2 type. In addition, X-3types with SEV approval are avail-able up to 2.2uF, but the highestcapacity X-1 capacitor is 1.0uF.

These suppression capacitors areused in power supplies to filter RFinterference (AFI) from either enteringor exiting the equipment via thepower lines. Government agencies,including the FCC, require filtering any

RFI to prevent its conduction throughthe power lines. Suppressioncapacitors and other components aidin meeting this requirement.A typical suppression applicationuses one X and two Y capacitors. Xtypes, called across-the-line capaci-tors, are used across the equip-ment’s power line input. Y capacitorsconnect from each side of the powerline to ground; they are known as line-to-ground capacitors.

Power supply designers must beaware that R FI suppression is afunction of more than just the capaci-tance value.. There is no guaranteethat a 1.0uF X-1 can replace the samevalue X-2. Successful AFI suppres-sion depends on part size, leadlength and distance between theleads. Therefore, designers mustallow time to test and evaluatecircuits using X-1 capacitors.

- Chuck Robertson, RodersteinElectronics, Inc. [Reprinted withpermission from “Power Conversionand Intelligent Motion., March 1990.]

Calls for Papers

Two groups have issued a “Call forPapers” for conferences to be held in1991, one in California and one inTexas. Both involve safety of electri-cal products and may be of interest.

The Santa Clara Valley Chapter ofthe IEEE EMC Society will host It IEEESCV 91 EMC: A Product ComplianceColloquium” from June 12 to 13, 1991,in San Jose, California. The empha-sis is on EMC, but one completesession will be devoted to productsafety, as well as part of the introduc-tory compliance process overview.

For more information about attendingor participating, call Franz Gisin atRolm Systems, (408) 492-3543.

The System Safety Society is holdingits “Tenth International System SafetyConference” from July 18 to 22, 1991,in Dallas, Texas. The theme empha-sizes society objectives of contribut-ing to the advancement of safety anddisseminating state-of-the-art knowl-edge about safety engineeringpractices. Request more informationfrom the Tenth International SystemSafety Conference Committee, P.O.Box 292455, Lewisville, Texas, 75029-2455. Tel. (214) 492-9005


What Do You Think?

WHAT? STANDARDS?

Question: What is the most oftenquoted reaction to the apathy andshort attention span shown by somegovernment bureaucrats. seniormanagement. designers. manufactur-ers and marketers. towards stan-dards and standardization?

Answer: “THEY” don’t understand.

Maybe “THEY” don’t understandbecause the word” standard” hasnever been explained to THEM”. (Ichose the word “standard” instead ofstandardization because it is shorter.It should be obvious that standardiza-tion is the development and implem-entation standards.)

“Standard” is made up of the follow-ing:

In the world of Product Safety andCertification, there are many pitfallsfor the unwary. If you have a problemthat seems insoluble, then it’s time toask Doctor Z! He has the answers,derived from his many years oftraining and experience in the Sci-ence of Product Safetiology. Pitfallshold no fearers for Dr. Z, since he is ona first name basis with most of them.Any resemblance to persons, places,products, agencies, or good adviceis purely coincidental, but don’t letthat stop you. Write to Dr. Z today!

Ask Doctor Z

St: This is the recognized standardabbreviation for “saint”. One has tobe a “saint” to work on standards. Itrequires a dedicated discipline, acharitable attitude, self-denial andpatience, patience, patience andmost of all, recognition from above.

and: This is the written symbol for“addition” . It signifies that somethingelse is coming, another work, anotherphrase, another clause, anotherstandard. It is a clear Indication thatwork on standards Is never finished, itis a continuing commitment.

rd: This is the recognized abbrevia-tion for third”. It may signify the nextstep to the two legal actions” on theone hand....then on the other hand”which Imply a “third” hand, namely:by the way, what standard does the

product comply with? Some say italso means that standards comethird, after design cost and profitabil-ity, in “THEIR” mind.

The astute observer will have noticedthat the second” a” is missing. It hasfallen between the proverbial cracks.It may signify the beginning, the“alpha”, of many other cracks that willopen up if “THEY” continue to denyadequate support and dedicatedresources to standards and stan-dardization.

All that remains is to find out if"THEY” now understand. Apply thestandard test for “understanding”:ask for “THEIR” commitment.“THEY” aIl ready know “THEY” haveours.

R.C. Maheux

[Dr. Z is presently on sabbatical. Herecently has been tarred ... that is,tired with his busy schedule anddecided to take a trip by train. Atleast he said something about ridingout of town on the rails. Anyway, hisfaithful assistant Fred is standing byto answer any questions sent in (oreven some that aren’t). The standardset by Dr. Z for irrelevant, Incompre-hensible and Irritating answers is sureto continue to be met. - Ed.]


You may have noticed something alittle different about this issue of theProduct Safety Newsletter. No, notthat it Is a month or two late. Unfortu-nately, that is not different, althoughsomeday... The difference is that wehave not had the assistance of the

Guest EditorialJohn McBain

Tandem Computers GraphicsDepartment this time, and boy do wemiss it!

Now don’t stop reading Just becauseyou suspect the next item is a plea forhelp from our readers. You may beperfectly correct, in fact you areperfectly correct, in your suspicion,but we do need help. And the morepeople who offer some of their time,the less of their time is needed. Butthe best part of participation is that itis a rare example of “something fornothing” because you can get outmore than you put in.

Besides the various tasks witheditors already involved, we have

several other possible jobs. If noneof these seem attractive, invent yourown position and we will see how Itworks.1) Institutional Listings, 2) IssueLayout and Graphics, 3) WordProcessing, 4) Mailing List Updates, 5) Back Issues and Reader Directo-ries, 6) Area Activities and Calendar,7) Reproduction and MailingIf I picked my top three they wouldprobably be numbers 2, 4 and 6.

So please get in touch with the editor,Roger Volgstadt, as soon as pos-sible, or even sooner. Or give me acall at 408-447-0738. Thank you foryour support.

John McBain

Damage to the components precludedreuse of anything except the fuses.

What should the Consulting Engineerdo? Obviously, use onlycombinations that have been fullytested. These combinations arelisted in the UL Yellow Book, “REC-OGNIZED COMPONENT DIREC-TORY”under “Circuit Breakers - SeriesConnected” .

William E. May, P .E.

Correction:

Dear Sir,

The withdrawal dates of productsgiven by Helmut Landeek in his letterpublished in the Nov IEEE 1989 issue(Vol. 2, No.5, page 20) are identicalto those given by me in the TUVseminar 1992 ONE. EUROPE held inDallas. Your reporter, Ercell Bryant,quoted me incorrectly in his article onthat seminar published in the July -Oct. issue (Vol. 2, No.4, page 17). Ihope that this sets the record straight.

Dr. Steven KraemerTUV Rheinland

Letters(Continued from page 24)


SANTA CLARA VALLEY

The March 27th meeting had apresentation by Mr. Jim Howard andMr. Bob Cameron of AmericanSeating Company.

We were given a brief history of thecompany; Founded in 1886 makingclassroom furniture, by 1892 had theworld market, and then went on toother types of seating. They havebeen using ergonomics since 1886 todesign the most comfortable andmost useful seating available. Anumber of sample chairs weredemonstrated. They were designedto fit 90% of the population.

The April 25th meeting hosted apresentation by Mr. JosefKirchdorfer of Weber Limited.

We were given a presentation andhandouts that covered in detail thedesign of circuit breakers. The topicscovered were the following:

1. An introduction.2. The basics of protection.3. The means of protection.4. Protection by fuses.5. Overload protection by circuitbreakers.6. Short circuit interruption by asingle breaker.7. Short circuit interruption by acascade of CBs.8. Comparison of protective de-vices.9. The selection of circuit breakersbased on performance requirements.10. Protection against electric shock.11. The differences in productstandards.12. Definitions according to IEC 934.

Area Activity Reports

All very interesting material for designengineers and other.

The May 22nd meeting hosted apresentation by Mr. GrantSchmidbauer of Canadian StandardsAssociation, Pacific Region.

We were given a presentation andhandouts that covered:1. An Overview of CSA’s structure,activities, and operations.2. A time table for the implementa-tion of CSA Standard C22.2 No. 950.The key dates are 9/93 and 3/2000.After 9/93 submittals must be to 950and 154 production must end. After3/2000 the D3s may not by used fornew products.3. A listing of some of the majordifferences between CSA 220 and950 covering the more stringentconstruction, marking, and testrequirements.4. A copy of Certification Notice No.671 covering 950 vs 220 differencesin detail. 5. A copy of the new CSAdraft report for 950 products.

The meeting ended with a questionand answer session on the newstandard. For more information onthe activities of the Santa ClaraValley chapter, please contact DavidMcChesney at 408-985-2400, exten-sion 2771 or John Reynolds at 408942-4020.

SOUTHERN CALIFORNIA

Charlie Bayhi opened the meeting.Four members were in attendance.

The May meeting featured a fantasticMagic Show presentation by one ofour own members, Mr. Gabriel Roy,aka” Gabriel” . The Great Gabriel

provided a wonderful evening ofentertainment with his professionalrepertoire. I am still trying to figure outhow he enticed me to “donate” adollar bill, cut it in half, burn half,and then reach into his pocket and pullout the “burned” half!! Thanks again,Gabe, we all enjoyed it!

Meetings are held the first Tuesdayof each month at MAI/Basic FourCorp. The next meetings are sched-uled for 6:00 PM Tuesday June 5,July 10 (2nd Tues.), and August 7,1990. the June meeting will be a testequipment demo by Dean Marxer ofValhalla Scientific. The programscoming up are a UL presentation ofISO 9000 series quality specifica-tions, medical standards by KonradKobel of TUV America, and pitfalls inEN60950 by Ed Spooner of TUVRheinland.

The EMC Society meetings are heldon the third Tuesday of the month.Meetings are held at LeChuga’s FineFoods of Mexico in Los Alamitos.Call Michele at 714-732-5691 forreservations. For details about theSouthern California chapter meetings,please contact Paul Herrick atGradco Systems, Inc., 7 Morgan,Irvine, CA 92718, phone 714-770-1223, 714-768-6939 FAX

NORTHEAST

The most recent meeting of theNortheastern Chapter of the ProductSafety Technical Committee washeld on Wednesday, April 25, 1990 atthe Sheraton Boxborough Hotel inBoxborough, Mass. Charles Kolifrathof Dash, Straus and Goodhuechaired the meeting in Bill von

Continued


Achen’s absence.

Committee Reports:

Constitution Bylaws Committee: TonyNikolassy continued his discussionfrom last month concerning therelationship between the localChapter and the Product SafetyTechnical Committee of the IEEEEMC Society. According to Tony, hiscontact with Rich Pescatore and Dr.Schindler (Chairman of IEEE inBoston) indicates that a structuredaffiliation would prove beneficial.

Technical Presentation Committee:Tony Nikilassy stated that more helpis needed in this area. Tony askedthat anyone having ideas for techni-cal presentations, including topicsand / or speakers, submit these eitherto him or at the May meeting.

Legislative Committee: NancyAraway and Fred Grund discussedthe issue of mutual acceptance of testdata between U.S. and Europeansafety agencies, especially concern-ing flammability of plastic materials.Their report stated that this dataexchange is currently minimal, atbest. For example, UL flame ratingsfor plastics may be sufficient to obtainsome National safety marks inEurope, but several Europeanagencies will not accept this data.Furthermore, even if the data isaccepted by one of the EuropeanNational safety agencies, that datacannot be used in obtaining otherNational safety marks, as would dataactually compiled by one for theEuropean National safety agencies.

Liaison Reports:

Area Activity ReportsContinued

UL: Doug Kealey reported that aFebruary 1990 bulletin from UL con-cerning the standards for powersupplies, battery chargers and directplug-in transformers will allow for areduction in electrical spacingrequirements for these productcategories. Doug also reported thatan upcoming UL bulletin will discusssupplies to be tested in accordancewith the standard to which the endproduct will be approved. Furtherdiscussion by attendees of themeeting revolved around similaritiesbetween this upcoming bulletin andCSA Bulletin 1402C,

CSA: Paul Smith reported that CSAwill be offering a seminar on June 25-27, 1990 dealing with various issueswhich are pertinent in today’s manu-facturing industry. Paul pointed outthat one major segment of thisprogram deals with assisting manu-facturers in the quality managementprocess. Further information aboutthis program can be obtained bycontacting CSA at (416) 747-4128.

Paul also reported that Alan Knightof CSA would like to conduct a focusgroup locally to discuss thosetechnical services most required byexporters of computer equipment.Those interested in participating inthe focus group should contact Alanat CSA (416) 747-4128.

EIA: Bruce Langmuir posed thequestion as to whether the majorsafety agencies allow for equipmentconnected to a branch circuit to relyon the branch circuit protection (i. e.,circuity breaker) for protection of thatequipment. Several different opin-ions on this subject were offered byvarious attendees. Bruce also

expressed an interest in gatheringdata which might prove useful in newdesigns for circuit breakers. In ournext meeting, Bruce will issue a formfor gathering data relating to therelationship between in-rush currentand steady-state current of theequipment which is produced by thevarious companies represented atour meetings.

CBEMA/NEMA/NFPA: Jeff Tuthillreported that there is currently aneffort to form a common body forsafety testing and approval inEurope. This body will be known asthe European Organizational TestingCommittee (ECTC). This committeewill consist of a signatory from eachof the European countries involved,and is expected to be in full operationsome time during 1993, after severalintroductory and experimentalstages.

IEC: Fred Grund reported on severalCENELEC amendments to variousEuropean safety documents anddistributed copies of the summary ofthese changes at our meeting.

Technical Presentation:

Mr. Roy Clay of Rod-L Electronicsgave a presentation “Critical Factorsof Hi-pot Testing.” Mr. Clay dis-cussed the reasons for performing hi-pot and ground continuity testing, andoutlined the specific listing agencyrequirements mandating hi-pottesting parameters, and highlightedfactory safety and economic consid-erations for performing hi-pot testing.

The presentation was followed by abrief question and answer period.

Continued


For more information please contact:

William von AchenChairman, Northeastern Chapter(508) 263 - 2662

CHICAGO

Our last meeting was held on May 15.The subject was Product Safety -Its Impact on Computer Software. Thespeaker was Joseph Faitler, Interac-tive Systems Corp., and Paul A.Phillips, Argonne National Labora-tory.

The previous meeting was held on 3/20/90. Mr. Tony Tutins, Perma PowerElectronics, Inc. spoke on TransientVoltage Surge Suppression. As youare aware, many of the product safetyissues we talk about relate to otherSocieties in the IEEE. The ReliabilitySociety, in Chicago, has expressedan interest in combining effortsregarding meeting topics. Althoughwe will not revolve all meetingsaround Reliability, we will pursue thismatter to see what can be done and ifit will be beneficial to our Chapter andthe PSTC. It should be noted that ournext meeting is being co-sponsoredby six (6) societies (Computer,Reliability, Communications, Medi-cine and Biology, EngineeringManagement, and the IndustrialElectronics Societies). This showsthat product safety is a concern of alldisciplines, and should be takenadvantage of to gain exposure andincrease membership and aware-ness.

The Chicago Chapter will not meetagain until September, 1990.

NORTHWEST

The Portland Section held its lastmeeting on May 15 at the PortlandGeneral Electric Company building.The featured speaker was Walt Hart.Walt is Product Safety Manager atJohn Fluke Mfg. Inc. and has beeninvolved in product safety standardsfor the past 15 years. He participatedin development of UL 1224, ANSI/ISAS82 (formerly ANSI C39.5). CSAC22.2 No. 231, IEC 990, and IEC 1010.Walt’s talk was on Clearance,Creepage Distance and Test Volt-age for Equipment as called for inIEC 664 and 664A. Of particularinterest was methods for clearancereduction under conditions where“controlled overvoltage” and “homo-geneous field construction” tech-niques are employed in the equip-ment. These methods will receivefinal review for test, measuring,process control and laboratoryequipment (IEC 1010 Amendment 1)at the October, 1990 IE C SC66Emeeting in Beijing, China.

In April, Bob Wersen of Panel Com-ponents Corp. tad us about the newpower connector standards and theimportance of the using the IEC 320connectors on exported electricalproducts. As an active Europeantraveler and importer for the last 20years, Bob gave his forecast of thechanges that will occur with Europeanunity in 1992.

The March meeting featured DebbieTinsley from UL She explained howUL Is working with CSA and VDE toprovide a coordinated testing servicefor U.S. manufacturers wantingrecognition by all three agencies.This should substantially reduce the

cost. The discussion with Debbiewas very interesting.

Want a trip to Washington, DC?Provide a paper for 1990 IEEEInternational Symposium on Electro-magnetic Compatibility. Contact JohnKnecht at Underwriters Laboratories708-272-8800. The Portland Sectionwill have its meeting on Tuesday,June 19, 1990, at 7:30 PM at thePortland General Electric Co. building(14655 SW Old Scholls Ferry Road,Beaverton at the south end of MurrayRoad).

The featured speaker will be BobCogan, a transformer design engi-neer with MultiComp of Beaverton,Oregon. Prior to MultiComb, Bob wasinvolved with transformer design anddevelopment at Western Transform-ers and for many years at Tektronix,Inc.

Bob’s topic will be “Product safety asit relates to transformers”. Construc-tion t recognized insulation systems,testing and the certification processare some of the topics Bob will coverin his presentation.

Come and visit with Bob at the no-host dinner at the Cattle CompanyRestaurant (3800 SW Cedar HillsBlvd., Beaverton) at 5:30 PM. Every-one (including spouses) is welcometo join us. We’ve had substantialturnouts the last two months but themore the merrier.

This will be the last meeting untilSeptember. Have a good Summer.

Please call Art Henderson if you wantto be deleted from his mailing list.

Area Activity ReportsContinued


Letters to the

Editor

[The opinions expressed in theseletters are those of the authors onlyand are presented as such. Wewelcome letters of comment butreserve the right to edit them. - Ed.]

Circuit Breaker Safety articlerebuttal:

I have had the opportunity to view thevideo tapes that Mr. Franklin referredto in his “Product Safety Newsletter”[article]. and I have followed hiswritings for the past several years.What I see and read is a collection ofuncontrolled experiments conductedto prove the author’s contention thatall residential circuit breakers shouldhave 5 times instantaneous triplevels.

One does not need to have thisfamiliarity with Mr. Franklin’s work tospot some of the more obvious flaws.On page 10 of the Newsletter, Mr.Franklin has 100 Amperes flowingthrough a 14 gauge copper wire forover an hour. I have tried to duplicatethis test to no avail. The insulationmelts completely off of the wire afterabout 35 seconds!

Likewise, fires caused by overcur-rents are called a myth by Mr. Fran-klin. There is no data or studiesoffered to back up this statement, andit is in direct conflict with publishedreports of actual field studies. I referthe reader to the January / February1990 “Fire Journal” article titled“What Causes Wiring Fires in Resi-dences”.

Finally, there is never any Informationpresented on the European electricalsystems which are cited as havingsuperior circuit protection means.

Surely the reader is curious as to thedifferences in systems voltages,wiring methods, plug protected cordsets, etc. According to the author,ignorance is the answer. “As of thiswriting, the Europeans are not awarethat their 5X circuit breakers preventarcing short circuit fires so well,because apparently no one therehas ever measured the current inhousehold short circuit arcs either.”

I know that the author has beenchallenged before on the abovepoints. It is disappointing to see thesame presentation again and againwithout some effort to clear up thequestions which surety must impactthe credibility of the work.

William E. May, P.E.

The author’s response:

In his letter, Mr. May wants to argueabout some minor points, but fails toaddress the major issue, which is thatmost present American circuitbreakers are not sensitive enough toprevent 20 percent of the fires in thiscountry, in my opinion. Mr. May’simmediate supervisor is the chairmanof the NEMA task force which for fouryears has been studying my recom-mendation that American circuitbreakers be changed. His name isJohn Halferty, and at a meeting atNEMA in Washington D.C. on May 18,1990, Mr. Halferty stated that allAmerican 15 and 20 ampere circuitbreakers will be redesigned to tripmagnetically at 150 amperes, +/- 20percent. I believe this result speaksfor itself.

Sincerely,

Frederick F. (Rick) Franklin, P.E.

Technically Speaking ques-tions:

I have a comment on Richard Nute’sarticle on the use of overcurrentprotection in plug and socket con-nected equipment.

Rich states in the beginning of thearticle that plug and socket con-nected Class 2 transformers do notcomply with IEC 950 clause 2.7.1because they do not have a fuse inthe primary. His entire article isbased on this premise. However,then I read clause 2.7.1 I found that itstates that the protective device maybe part of the building installation. Asa matter of fact, the second para-graph of clause 2.7.1 specificallystates: “Where single-phase equipmentis to be connected to standardsupply outlets, the building installa-tion shall be regarded as provid-ing this protection in accordancewith the rating of the wall outlet. ...”.

My reason for my writing is not tocriticize Rich Nute’s article but ratherto find out if I may have missedsomething in my interpretation ofclause 2.7. 1.

Best Regards,

Gabriel R. Roy


In my Technically Speaking columnin the March-April Newsletter, Iimplied that IEC 950 and IEC 601-1require fuses in the product. Mr. Roypoints our that IEC 950, Sub-clause

Continued


Continued

2.7.1 states that

"... protection devices shall beincluded either as integral parts ofthe equipment or as part of thebuilding installation...”

Mr. Roy therefore questions whetherIEC 950 requires a fuse in the product.

My point in citing Sub-clause 2.7.1and 2.7.3 was that IEC 950 requires afuse — regardless of the site of thatfuse.

The direction of this article was toshow that, for some kinds of construc-tion, the required fuse — regardless ofsite — cannot and does not providethe intended protection againstoverheating.

The column identifies what buildingfuses protect, what equipment fusesprotect, and shows that for somelowpower kinds of equipment, a fusedoes not provide any protection — yetthe standards require a fuse.

With best regards,

Richard Nute

More Tech/Speak questions:

The article by Rich Nute uses theterms I*I*E and E*I to representheating losses in a transformer. Idon’t understand what the differenceis. Does one refer to conductorresistance and the other to ironlosses (eddy current and hysteresis)in the core? I would think that bothcould be characterized as an equiva-lent resistance. Am I missing some-thing or is this just a convenient way

to distinguish between loss in aconductor and loss in a component?

I’m pleased I discovered a safetygroup within IEEE and look forward tolooking at the past newsletters.

Sincerely,

Robert O. Diedrichs, P.E., C.S.P.


Dear Mr. Diedrichs:

Roger Volgstadt forwarded yourletter to me for response to yourtechnical question regarding theterms I*I*E and E*I to representheating losses in a transformer.

The paragraph in my article was:“Two kinds of output terminal loadingsimulate the worst -case transformeroverheating. One load is the outputshort...circuit. This maximizes I*I*Eheating within the transformer. Theother load is the output maximumpower. This maximizes the E*Iheating within the transformer."

Looking back into a transformer’soutput terminals, we can create aThevenin equivalent circuit of avoltage source with a series resistor.The objective is to load the trans-former output terminals such that thepower, P, in the Theveriin seriesresistor is maximized. We don’t needto know whether the power dissipa-tion is in conductor resistance or iniron losses.

I agree with your statement that allof the transformer losses” could becharacterized as an equivalent

resistance.” Indeed, the Theveninequivalent circuit was what I had inmind when I wrote the article.

Interestingly, in considering yourcomment, I drew the Theveninequivalent circuit and wrote the powerequations for the Thevenin equiva-lent resistance for both the short-circuit case and the maximum powertransfer case. The equations are:

E**2P = ----------- for the short-circuit. R

(E/2)**2p = ------------- for the maximum R power transfer.

where E is the Thevenin equivalentvoltage source, and R is theThevenin equivalent resistance.

Note that the condition of short-circuitALWAYS create the worst-casepower dissipation in the transformer.

The only reason for not testing withthe transformer output short-circuitedis when the winding is either pro-tected by a fuse or other overcurrentdevice, or the winding opens. Inthese cases, we would want to loadthe transformer to the point just beforesuch an event occurs. which likelyhas no relation with maximum powertransfer.

Sincerely,

Richard Nute

Even more about Tech/Speak:

L e t t e r sContinued

Product Safety Newsletter • Continued

In this issue’s Technically Speakingcolumn, Rich Nute claims that theinductance of a branch circuit andpower cord is significant and limitsthe fault current to about 100 am-peres. His claim that fault current is inthe vicinity of 100 amperes may becorrect, but it is not due to circuitinductance.

I have measured the resistance andinductance of several branch circuitcomponents with the following results(the reactance is computed from themeasured level of inductance):

[See Table 1 below. - Ed.]

It seems from these results that thefault current is limited primarily byresistance rather than by inductance.

In the last issue of the Product SafetyNewsletter, the Technically Speakingcolumn states that .Operation of afuse in the event of an insulation faultto ground necessarily requires theground return circuit impedance to beabout the same as the supply circuitimpedance. I n North America, theground wire and the neutral wire arethe same size... This gives reason-able assurance that the ground circuit

return impedance is about the sameas the supply circuit impedance..

It is clearly the intent of the NationalElectrical Code that the groundcircuit return impedance [should be]about the same as the supply circuitimpedance, II but it was not always so,and many installations (both commer-cial_ and residential) still are in usewhere the ground impedance may besignificantly higher than the supplyimpedance.

As recently as 1958, a cable assem-bly did not contain a ground wire butit may have had a very thin layer oflead impregnated into the wovenouter covering. When the cable wasclamped into a metal autiet box, thebox was considered “grounded”.Unfortunately, this outer covering didnot always remain conductive (itsometimes flaked off in handling andwas sometimes attacked by mice - itwas, after all, impregnated intonatural fiber), nor was it a particularlylow impedance. I n either case, thegrounding system1s integrity becamesuspect. Later revisions of the codeno longer permitted this type of wiringto serve as the ground system.

Later (ca. 1964), a 12 AWG Type NMcable assembly was commonly builtwith a 14 A WG ground. The resis-tance of a 14 A WG wire is 1.6 timesthe resistance of a 12 AWG wire.

More recently, the NEC has requiredthat the ground wire be sized to matchthe circuit conductors, if it exists at all.Metal conduits of various types,cable trays, and other forms ofgrounding enclosures are alsoaccepted as grounding conductors.When properly installed and notsubjected to damage, these gen-erally do provide a good low-imped-ance path, but there has never beena clear specification of what consti-tutes proper installation, and hencethere may be a wide variation ofgrounding system impedances.Furthermore, the ground wires (whenused) are often connected with crimp-on or slip-on connectors (which canloosen over time) whereas the circuitwires are connected with screw-onconnectors or solder.

We, as product safety professionals,need to be careful to avoid puttingtoo much faith in the integrity of thebuilding ground system to which ourproducts are connected. There aretoo many variables of components,installation methods, and aging for usto reasonable expect very lowgrounding impedances in all installa-tions.

Ray Corson, P. E.

Support/General Comments:

At the recent TUV Rheinland seminarin Boston [Nov., 89], you approached

TABLE 1:

Component Resistance Inductance Reactance (60 Hz)

1 00 feet of 12-2 (w/gr) 0.32 ohm 14 uH 0.053 ohmType NM cable (loop)

2 meter 18-3 0.15 ohm 1.6 uH 0.006 ohmpower cord (loop)

20 A circuit breaker 0.0 ohm 0.2 uH 0.00075 ohm0.05 ohm 0.2 uH 0.00075 ohm0.22 ohm 0.2 uH 0.00075 ohm



me with the suggestion to publish mypaper, IEC 950 - Safety of ITE, in theProduct Safety Newsletter. I have noobjection to this as long as you sendme a draft for approval beforepublication.

If you are interested in the formationof a Toronto area Chapter, let meknow, as I am sure this can easily beset up.

I think that you are doing a good joband the newsletter is an interestingforum. In order to avoid misunder-standings, I suggest that, if anyone isquoted, you send them a copy of thearticle for approval before publica-tion.

Regards,

Dr. Steven KraemerTUV Rheinland

Series Combinations of CircuitBreakers and Fuses:

Several technical papers werepresented at the IEEE’s IndustrialApplications Society meeting in SanDiego last October thatadvocate a potentially dangerouspractice. I’m referring to, what hascome to be called, the Up-Over-and-Down” method of sizing fuseand circuit breaker series combina-tions.

Series combination of short circuitdevices is not something thatcan be taken lightly. Basically, theengineer is relying on theupstream protective device to limitthe energy in a fault to sucha level that a lesser rated down-

stream protective device can beused. This is usually done for rea-sons of economy. UL recognizesthe seriousness of this application byrequiring all such proposedcombinations to be short circuittested and witnessed.

The “Up-Over-and-Down” method,proposed by fuse manufacturers,tells the engineer that through com-paring let-through energies thatare expected in a down-streamdevice’s operation, the up-streamcurrent limiting fuse can be chosenanalytically. This method isdetailed in several manufacturers’catalogs and brochures.

The IEEE papers referred to above,and these brochures, arepresented to unsuspecting consultingand application engineers who,considering the IEEE framework ofthe forum in which the paperswere presented, or the assumedreputation of the fusemanufacturers, accept what ispresented as truth. This can be apotentially dangerous mistake.

Modem Molded Case Circuit Break-ers use repulsion contact designsto increase opening speed andthereby increase interruption levels.A simple comparison of theseinterrupting levels available 15 yearsago, to what the present day productsare capable of doing, willprove the point. Today’ s MCCB's aredynamic short circuitinterrupting mechanisms. Interruptiontimes as low as 2 - 4milliseconds are common, and thelet-through energies are reducedaccordingly. It is these characteristicswhich must be taken into

account when proposing a seriescombination with another MCCB ora fuse.

Consulting Engineers cannot beexpected to be expert in the design,or even the testing, that goes intothese products. In fact, eventhe engineers involved in the actualdesign are hard pressed toaccurately predict the performance ofan untested product. That iswhy UL insists on controlled, wit-nessed testing on the discretedevices, and especially on seriescombinations of these devices.This testing must be successfullydone before the devices can belabeled or listed.

Up to this point, this letter should haveno more credibility thanthe questioned papers or brochures.However, we have taken the onlystep that should matter. We havebeen testing thesecombinations. Successfully testedcombinations are submitted to ULand are listedin the Yellow Book.

Using one fuse manufacturer’s stepby step application guidelines,several unlisted fuse-breaker combi-nations were selected which,according to the fuse manufacturer’sapplication literature, shouldhave been vallexx. These combina-tions were applied on a three phase,65 KA, 480 V short circuit. All failed.In fact, the same threefuses were used on four different testswithout ever melting. Inall cases the Molded Case CircuitBreakers interrupted the currenteven though it was significantlyabove their published ratings.


Continued on Page 17


The Product Safety Techincal Committee of the IEEE EMC Society is grateful for me assistance given by the firms listed below and invites applica-tions for Institutional Listings from other firms interested in the product safety field.An Institutional Listing recognizes contributions to support the publication of the Product Safety Newsletter of the IEEE EMC Society Product SafetyTechnical Committee. Minimum rates are $100.00 for listing in one issue or $400.00 for six consecutive issues. Inquiries, or contributions madepayable to the Product Safety Technical Committee of the IEEE EMC Society and instructions on how you would like your Institutional Listing toappear, should be sent to:

PSTC Product Safety Newsletter, c/o John McBain (MIS 42LS), Hewlett-Packard, 19447 Pruneridge Avenue, Cupertino, CA 95014.

Institutional Listings


CalendarThe Product Safety Technical Committee of the IEEE EMC Society

Chicago Activities

There are no meetings planned for theChicago product safety group untilSeptember.Contact: John Allen

(708) 827-7520

Los Angeles Activities

Monday, July 9, 1990Subject Technical Services forExportersSpeaker: Allen Knight, CSATime: 6:30 pmLocation: Harman Electronics

8500 Balboa Blvd.Northridge, CA

Contact: Rolf Burckhardt(818) 368-2786

Monday, August 6, 1990Subject Quality -ISO 9000Speaker: Rich Lee, ULTime: 6:30 pmLocation: Harman Electronics

8500 Balboa Blvd.Northridge, CA

Contact: Rolf Burckhardt(818) 368-2786

Portland Activities

There are no meetings planned forthe Portland product safety groupuntil September.Contact: Art Henderson

(503) 635-8445

Seattle Activities

There are no meetings planned forthe Seattle product safety group untilSeptember.Contact: Art Henderson

(503) 635-8445

Northeastern Activities

Wednesday, July 25, 1990Subject: To be determinedSpeaker: To be determinedTime: 7:00 pmLocation: Sheraton

Routes 111 and 495Boxborough, Mass.

Contact: Bill Van Achen(508) 263-2662

Wednesday, August 22, 1990Subject: To be determinedSpeaker: To be determinedTime: 7:00 pmLocation: Sheraton


Contact: Bill Von Achen(508) 263-2662

Wednesday, September 26, 1990Subject: To be determinedSpeaker: To be determinedTime: 7:00 pmLocation: Sheraton


Contact: Bill Van Achen(508) 263-2662

Santa Clara Valley Activities

Tuesday, July 24, 1990Subject: How to Handle UL Projects More EfficientlySpeaker: UL Santa Clara StaffTime: 7:00 pm

This could be yourbusiness card!See page 25 for details.

Institutional listingsContinued

Continued


CalendarContinued

Location: Apple Computer20705 Valley Green DriveCupertino, CA

Contact: Hugh Hagel(415) 962-9400 x619

There are no other meetings plannedfor the Santa Clara Valley productsafety group until September.

Orange County/Southern Califor-nia Chapter

Tuesday, July 10, 1990Subject: Technical Services for ExportersSpeaker: Allen Knight, CSATime: 6:00 pmLocation: MAl Basic Four

14101 Myford Rd.Tustin, CA

Contact: Paul Herrick(714) 770-1223

Tuesday, August 7, 1990Subject: ISO 9000 (not finalized)Speaker: Rich Lee, ULTime: 6:00 pmLocation: MAl Basic Four



Tuesday, September 11, 1990Subject: Med GV & IEC 601 -

MedicalSpeaker: Mr. Konrad Kobel, TUV America,Time: 6:00 pmLocation: MAl Basic Four



Tuesday, October 2, 1990Subject: Summary of SeptemberCBEMA MeetingSpeaker: Charlie BayhiTime: 6:00 pmLocation: MAl Basic Four



PSN READERS -

We need your help to make this newsletter complete! The items in"News and Notes", the articles in PS Abstracts", the events in the"Calendar", and the letters in "Letters to the Editor", must come fromyou. The better the input the better the output. So, the next timeyou find an article, news item, or conference with an aspect relatingto product safety, tell us about it. Then we can pass it along to ourcolleagues through the pages of the Product Safety Newsletter.Thanks for your help!



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

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

BACK ISSUES

The Product Safety Newletter’s firstissue, Vol. 1, No.1, was ten pagespublished in Feb. 1988. We nowhave available two years (1988 and1989) of past PS Newsletters, with anarticle index, for those who missedthem. To order a set, please send acheck for $20.00 payable to theProduct Safety Technical Committeeto the add ress given above for thePS Newsletter.

READER DIRECTORY

Our Reader Survey confirmed thathundreds of you are interested incontinuing to read the Product SafetyNewsletter. We have prepared anaddress directory, organized byname and by postal code, to help putreaders in touch with each other. Toorder a copy, please send a checkfor $15.00 payable to the ProductSafety Technical Committee to theaddress given for the PS Newsletter.

[if you want your address deleted from thedirectory, please contact us.]

chairman's message john mcbain 1 product...

Documents