"e" is for Automobile Electronics

Tim Jarvis
Aftermarket electronic products must carry the e-mark, but attaining approval can be confusing for those accustomed to the CE mark.

For more than 30 years, the automotive industry has been accustomed to the e-mark and type approval of motor vehicles sold into the European Union. Since October 1, 2002, however, manufacturers of aftermarket electronic

Figure 1. Example of an e-mark.

products intended for use in vehicles were also required to obtain formal type approval for these products prior to placing them on the market. All such products are required to carry an e-mark and to comply with European vehicle legislation in order to be legally fitted inside a vehicle.

The testing and approvals regime required for e-marking is very different from that normal for CE marking. This has come as a surprise to some electronics manufacturers more familiar with the latter. This article attempts to provide a basic understanding of e-marking automotive electronic products, principally to readers already familiar with the process of CE marking electronic goods, particularly information technology (IT) products.

Background

Since 1970, cars and trailers sold in Europe have required formal type approval and e-marking.¹ On January 1, 1996, European Commission Directive 95/54/EC extended these requirements to cover in-vehicle electronic subassemblies (ESAs) by defining acceptable

1 Germany 2 France 3 Italy 4 The Netherlands 6 Belgium 9 Spain 11 United Kingdom 13 Luxembourg 18 Denmark 21 Portugal 23 Greece 24 Ireland

Table I. Codes for member states granting the type approvals.

performance criteria for the electromagnetic emissions and susceptibility of such devices.²

In the directive, ESAs are defined in two distinct groups: (i) components and (ii) separate technical units (STUs). STUs are specific to a given make of vehicle and are usually fitted by the vehicle manufacturer. Aftermarket electronic products fall into the category termed components that may be fitted to differing makes and models of vehicles. Article 2 of 95/54/EC requires all ESAs to comply with the legislation and to be appropriately e-marked.

Since October 1, 2002, member states may refuse the sale or entry into service of noncompliant or improperly marked products. Specifically in the UK, national implementation of 95/54/EC into the nation's Construction

e-marking CE marking Compliance with 95/54/EC* Compliance with 89/336/EEC4 1 EMC limits and test methods all defined in legislation 95/54/EC. 2 Develop compliant product. 3 Submit product for formal type approval. 4 Apply e-mark and sell. 1 Select appropriate EMCcriterion, identify applicableEMC standards, decide on route to compliance. 2 Develop compliant product. 3+ Submit TFC, submit for third-party testing. 4 conformity, apply CE mark, and sell. * 95/54/EC is the core directive for vehicles. Additional European legislation applies to some classes of vehicles; for example, for motorcycles apply 97/24/EC, for tractors apply directive 2000/2/EC. + Optional, depending on chosen route to compliance.

Table II. Comparison of the CE marking and e-marking processes for EMC.

and Use regulations has made it a criminal offense to use a car with a non-e-marked part fitted. It is essential, therefore, for manufacturers of aftermarket electronic products intended for vehicle use to comply with the regulations.

The e-Mark

An example of an e-mark is shown in Figure 1. It comprises the lowercase letter e followed by a country code number in a square box. The number identifies the EU member state granting type approval (see Table I). The major difference between e-marking and CE marking is the route to approval. For CE marking, manufacturer self-declaration is the norm, whereas the only route to compliance with the Automotive Directive is formal type approval. Each of the EU member states listed in Table I has a national body responsible for issuing type approval certificates. In the UK, for example, this body is the Vehicle Certification Agency (VCA).³

The number placed adjacent to the e-mark in Figure 1 is the base approval number issued by the approval body and printed on the type approval certificate. The first two digits indicate the sequence number

assigned to the most recent amendment of directive 72/245/EEC.⁴ For the Automotive Directive 95/54/EC, the sequence number is 02.

Figure 2. ESA EM radiation limits.

The EEC type-approval mark must be affixed to the main body of an ESA in such a way that it is clearly legible and indelible. It need not be visible once the ESA is fitted within a vehicle.

The e-Marking Process for ESAs

Vehicles must comply with myriad legislation covering such areas as environmental impact, road safety, and crash protection (to name but a few), whereas e-marking for aftermarket ESAs is all about electromagnetic compatibility (EMC). Table II contrasts the processes of compliance between the Automotive and EMC Directives.⁵

Compliance with the EMC Directive usually involves selecting and applying standards from the ever-growing, ever-changing list published in the Official Journal of the European Communities. For the Automotive Directive, the process is simplified (in principle) because the directive contains within it all the test methods and limits necessary for compliance.

Manufacturers and their representatives should note, however, that 95/54/EC compliance does not absolve them from compliance with other applicable European legislation. For example, a product that can be fitted in a vehicle or used at home must comply with both the Automotive and EMC Directives. If a product contains a radio transmitter, it must also comply with the R&TTE Directive, and so on. In all cases, a product must be safe when used as intended and should not impair a driver's ability to properly control the vehicle.

Testing and Test Methods

Contained within the body of the Automotive Directive are the test methods and limits applicable to the electromagnetic emissions and susceptibility of both vehicles and ESAs. The test method for ESAs is based on CISPR 25 and differs considerably from the familiar EN 55022 and EN 55024, as the following comparison shows.^6–8

Conducted Emissions. The Automotive Directive 95/54/EC contains no requirement to test the conducted emissions from automotive ESAs. This is because the legislation for ESAs is derived from that for vehicles and vehicles don't normally feature connecting cables. The argument that an

95/54/EC EN 55022 EUT placed 50 mm above fixed table with conducting surface, 1 m above reference plane (ground plane). EUT placed on nonconducting,rotating table 800 mm abovereference plane (ground plane). Measurement antenna site 1 m from EUT. Measurement antenna site10 m from EUT (or 3 m and results extrapolated to 10 m). EUT fixed throughout test. EUT rotated 360° during test. Measurement antenna height fixed throughout test. Measurement antenna height varied from 1 to 4 m during test. Measurement antenna pointed at midpoint of wiring loom connected to EUT. Measurement antenna pointed at EUT. Single product class. Two product classes (A and B) with different limit lines. Two emission types (broadband and narrowband) with different limit lines. Same limit line applies to allemissions. Average and quasi-peak detectors used for narrowband and broadband emissions respectively. Quasi-peak detector only used. Power supplied via AN (artificial network). Power supplied via LISN (line impedance stabilization network).

Table III. Major differences in test methods between 95/54/EC and EN 55022.

ESA may pollute the wiring it is connected to is apparently ignored. Given that a vehicle wiring loom is an electrically noisy item, vehicle manufacturers have traditionally employed the opposite logic, namely that ESAs should demonstrate a high level of immunity to conducted interference.

Radiated Emissions. Testing of ESAs is described in Sections 6.5 and 6.6 of the Automotive Directive. Specifically, Annexes VII and VIII describe the test method. The differences compared with EN 55022 are shown in Table III.

The limit line used for ESA-radiated emissions (see Figure 2) not only differs from that used for vehicles but also differs considerably from the EN 55022 limit line. The directive gives two limits, one for broadband and another for narrow-band emissions. The concept of broadband and narrowband emissions may be foreign to those unfamiliar with automotive testing. Individual emissions are classified as either narrowband or broadband, depending on their appearance. Broadband emissions are those characteristic of spark ignition systems, and narrowband emissions are those characteristic of microprocessor-based systems. More detail can be found in CISPR 25 Clauses 4.1.4 and 4.1.5.

Unlike EN 55022, no distinction is made between EUT classes, and so the same EUT in the same configuration is tested first against the broadband and then the narrowband limit.

Narrowband emissions are very spiky and occur at the harmonic frequencies of the clocks used in the circuitry contained within the ESA. This author uses the rule that any other type of emission is broadband. A

Figure 3. Example of ESA being tested at RadioCAD's Cowden facility.

manufacturer wishing to be cautious will ensure that all emissions fall well below the narrowband limit, irrespective of classification. Generally, free-running, switched-mode power supplies are the most common source of broadband-type emissions in aftermarket ESAs.

Manufacturers of ESAs with no oscillators operating above 9 kHz can use Clause 8.1, exempting them from narrowband testing altogether. Figure 3 shows an example of an ESA being tested.

Conducted Susceptibility. Clauses 8.4 and 8.5 exempt ESAs from having to meet any specific conducted susceptibility or electrostatic discharge (ESD) criteria. The legislation places the onus on vehicle manufacturers to ensure that ESAs are properly immune to conducted interference. Traditionally, vehicle manufacturers have placed strict immunity requirements on component suppliers as part of the procurement process.

Manufacturers of aftermarket ESAs might be tempted to skip susceptibility testing altogether because there is no legal requirement for it. However, it is foolhardy to do so. The wiring loom of the average car frequently generates conducted phenomena far more severe than those encountered by domestic and commercial electronic products. In the sidebar on page 42, a good practice is recommended for self-respecting manufacturers to adopt.

Radiated Susceptibility. The radiated immunity requirement for automotive ESAs is considerably more onerous than for IT products. For example, the field strength specified for free-field immunity testing is 30 V/m compared with the 3 V/m for EN 55024 Class B. Given the difficulty in generating uniform fields of this intensity, 95/54/EC gives three other alternative test methods: stripline, transverse electromagnetic (TEM), and bulk current injection (BCI). Clause 1.2.1 of Annex IX allows manufacturers full discretion in choosing which methods to apply, provided that the full frequency range from 20 MHz to 1 GHz is covered by the chosen method or combination of methods. Of these methods, stripline is most popular. It is relatively easy to generate high field strengths over a wide frequency range with comparatively small amplifiers. Most manufacturers of aftermarket ESAs dispense with all radiated susceptibility testing by employing Clause 8.3 of the directive, which states that:

ESAs whose functions are not involved in the direct control of the vehicle need not be tested for immunity and shall be deemed to comply with paragraph 6.7 of Annex I and with Annex IX to this Directive.

It should be noted, however, that vehicle manufacturers procuring components that do affect the driver's control of the vehicle, (e.g., antilock braking systems [ABS]), generally insist on much higher field strengths. Immunity to 100 V/m is a common requirement.

A Design Example: 12-V Transient Suppression   ISO 7637 defi es that the test pulse shown on the left in Figure 1 be introduced to the EUT battery input via a series resistor Ri where 0.5 W < Ri < 4 W.12 This example has added a transorb and an in-line current limiting resistor RL. Figure 1. Transorb design example. A transorb must be chosen with a break voltage higher than the charging voltage of the 12-V system. The nominal alternator charging voltage for a 12-V system is 13.5 V. A 12-V transorb has a break voltage of 13.3 V. Therefore, a 13-V device is chosen with a break voltage of 14.4 V to provide some headroom. For severity level III, Vp (the peak transient voltage) is 80 V (Vs + 13.5). For a 13-V transorb, Vc (the clamping voltage) is about 21.5 V, specified at a current Ic derived from the device's peak power Pp (see Table I). Note that the protected electronics still must safely handle voltages up to the clamping voltage. Device Pp@1 mS Ic (max) Pp@100 mS RL>(Vp–Vc)/Ic SMAJ13 400 W 18.6 A 25 W 3.15 W SMBJ13 600 W 27.9 A 38 W 210 W SMCJ13 1500 W 69.8 A 94 W 0.83 W Table I. Comparison of surface-mount transorbs. Next, Ic is used to determine a suitable line resistance. Assume the EUT takes nominally 1 A and can tolerate a voltage drop of up to 2.5 V across RL. Therefore, RL = 2.2 W might be chosen, and an SMBJ13 transorb is specified. The remainder of the example demonstrates that the selection just made affords woefully inadequate protection. Firstly, the transient peak power Pp is given approximately by:   This approximation (Pp) is good if the pulse rise and fall waveforms are roughly linear. In a system where the fall time dominates, a linear approximation is more severe than a logarithmic one, giving some safety margin in the design. Secondly, notice that transorb manufacturers quote Pp for a 1-millisecond pulse. The ISO 7637 load-dump pulse duration, however, varies from 40 to 400 milliseconds and, therefore, the transorb power handling has to be derated using the manufacturer's data. For example, at td = 100 milliseconds, the SMBJ13's peak power rating falls to just 38 W. To compensate, rearrange Equation 1, and plug in a value of Pp = 38 W to obtain Ri:     The problem with this arrangement is that ISO 7637 states that Ri must be 4 W or less. Reducing Ri will increase the peak transient power beyond the absolute maximum limits of the selected transorb. This, in turn, leads to device rupture and the ESA becoming unprotected from the remainder of the lethal transient. To achieve proper protection, the example requires not one but two 1500-W SMCJ13 devices in parallel. This gives a combined Pp @ 100 milliseconds of 188 W and allows Ri to fall as low as 1.1 W without overstressing the transient protection circuit. There is some room now to trade the values of Ri and RL to reduce the undesirable voltage drop developed across RL during normal operation. If nothing else, the exercise demonstrates just how rugged ESAs must be to survive in the automotive environment.

Submitting an ESA for Type Approval

Once manufacturers have an ESA they believe is compliant with 95/54/EC, the formal type-approval process must begin before that product can be sold into the EU market. The following assumes that the United Kingdom VCA has been selected to issue the approval. The UK VCA publishes guidance notes for manufacturers seeking type approval. To start the type approval process, manufacturers may contact the VCA directly.⁹ Most manufacturers, however, start by contacting an independent test laboratory recognized by VCA.¹⁰ These may or may not be designated by VCA as a technical service. Although any competent test house or consultant can guide a manufacturer through the approval process, only a technical service is authorized to act on behalf of VCA.

A designated technical service or a VCA official will witness formal testing at the selected laboratory and will ensure that the test sample and its arrangement are set up for the worst-case condition. Worst-case selection, particularly with products offered in a number of variant configurations, alleviates the need to test all possible permutations. Manufacturers may be required to present their own EMC test results to assist in worst-case selection.

Manufacturers must submit product documentation in support of an application. The documentation required is listed in Annex IIB of the directive 95/54/EC. For successful applications, VCA issues a type approval certificate containing the base approval number that the manufacturer will use with the e-mark. VCA asks for evidence of conformity of production, ensuring that the sample tested is representative of the mass-produced product. Generally, production by an ISO 9002– certified factory is taken as sufficient evidence, although control plans dealing with specific issues may be needed in addition.

Some manufacturers also ask for approval to UN-ECE Regulation 10.02, which is practically identical to 95/54/EC and is accepted by a large number of countries outside Europe. The difference in marking is that the UN mark is a capital E whereas the European mark is a lowercase e. VCA or a designated technical service can advise on this and other applicable vehicle-legislation issues.

Conducted Transients

As previously mentioned, the Automotive Directive doesn't stipulate any requirements for the conducted susceptibility of ESAs. Although not a legal requirement, self-respecting manufacturers will want to ensure that their products are fit for use inside a vehicle. They should, therefore, design and test for immunity to conducted transients and ESD. The International Organization for Standardization (ISO) publishes suitable standards for automotive transients and ESD immunity.

With discharges up to 25 kV, ISO 10605, the ESD standard, is significantly more severe than the generic IT standard.¹¹ Anyone who has slid in or out of the driver's seat on a dry day can testify, however, that vehicle static discharge is a prevalent menace. Therefore, ISO 10605 seems, to this author at least, reasonable.

ISO 7637, the conducted-transient standard, is split into three parts:¹²

Parts 1 and 2 are almost identical, with the same set of transient tests defined in both, except that the 24-V transients are more severe and part 2 adds an extra test pulse 1b. Test pulses 1 and 2 are pretty straightforward switching transients. Pulse 3 is a fast transient such as what might be generated in mechanical switching. It is applied directly to the dc power source in Parts 1 and 2, and indirectly via a coupling clamp in Part 3. The author, experienced with designing for immunity to transients, found that none of these pulses presented much of a problem in designing for immunity. Textbook EMC techniques can readily be used to provide sufficient immunity.

Pulse 4 simulates battery dip during starting. Most ESAs containing microprocessors reset during the application of this pulse. For Class C ESAs, a reset during exposure is not a failure, provided that the EUT recovers following the test. Annex A of ISO 7367 describes the functional classes and severity levels.

The author recommends that severity level III, Class C, be applied in the absence of a specific requirement. Note, however, that some continental vehicle manufacturers are now suggesting that ISO 7637, severity level IV, be called up in revisions currently under consideration to 95/54/EC.

Pulse 5, the load dump pulse, is a notoriously severe overvoltage surge peculiar to ISO 7637. The pulse simulates a car battery being disconnected from a spinning alternator and the resulting long-duration, high-voltage surge introduced into the 12-V supply line. It is, therefore, both a reasonably realistic transient and one that often spectacularly destroys unprotected ESAs.

Pulses 6 and 7 apply to electronic ignition systems and historic vehicles, respectively. These are generally ignored.

Load Dump Protection. Fitting transorb transient suppressors to the supply line as it enters the ESA is the usual method of providing load dump protection. Because the load dump pulse is very high energy, it is essential to design transient suppression circuits carefully (see example in the sidebar). An alternative method (used successfully by the author) is the rider circuit. This method disconnects the ESA electronics from the supply for the duration of the pulse, which effectively rides out the storm—hence, its name.

Acknowledgment

Many thanks to Terry Beadman of MIRA UK for his assistance in producing this article. He can be reached at terry.beadman@mira.co.uk.

References

1. 1970/156/EEC, "Council Directive on the approximation of the laws of member states relating to the type approval of motor vehicles and their trailers," Official Journal of the European Communities (OJ), no. L 42, 23/2/1970.

2. 1995/54/EC, "Commission Directive adapting to technical progress Council Directive 72/245/EEC and amending Directive 70/156/EEC," OJ, no. L 266, 08/11/1995.

3. Vehicle Certification Agency, http://www.vca.gov.uk/noframes/ about_us.htm.

4. 72/245/EC, "Council Directive on the approximation of the laws of member states relating to the suppression of radio interference produced by spark-ignition engines fitted to motor vehicles," OJ, no. L 152, 6/7/1972.

5. 89/336/EEC, "Council Directive on the approximation of the laws of the member states relating to electromagnetic compatibility," OJ, no. L 139, 3/5/1989.

6. CISPR 25, "Radio disturbance characteristics for the protection of receivers used on board vehicles, boats, and on devices—Limits and methods of measurement," International Electrotechnical Commission CISPR 2002-08.

7. EN 55022:1998, "Information technology equipment—Radio disturbance characteristics—Limits and methods of measurement," CENELEC, Brussels, 1998.

8. EN 55024:1998, "Information technology equipment—Immunity characteristics—Limits and methods of measurement," CENELEC, Brussels, 1998.

9. "Automotive Type Approval for Electromagnetic Compatibility," VCA Pub TA045, Revision 4, May 1, 2002, http://www. vca.gov.uk/publications/vca045.pdf.

10. "Test Facilities Suitable for Automotive EMC Tests to 95/54/EC," VCA Pub TA054, Issue 6, December 11, 2002, http://www. vca.gov.uk/publications/vca054.pdf.

11. ISO 10605, "Road vehicles, Test methods for electrical disturbances from electrostatic discharge" (Geneva: International Organization for Standardization: 2001).

12. ISO 7637, "Road vehicles, Electrical disturbances from conduction and coupling; Part 1: Passenger cars and light commercial vehicles with nominal 12 V supply voltage — Electrical transient conduction along supply lines only," 1990; "Part 2: Commercial vehicles with nominal 24 V supply voltage—Electrical transient conduction along supply lines only," 1990; "Part 3: Vehicles with nominal 12 V or 24 V supply voltage—Electrical transient transmission by capacitive and inductive coupling via lines other than supply lines," 1995 (Geneva: International Organization for Standardization).

Tim Jarvis is an independent consultant. His company, RadioCAD (www.radiocad.com), assists clients in designing electronic products for compliance with European directives and worldwide standards. Jarvis has worked in the electronics industry since 1983 and specializes in RF design and EMC. He can be reached at t.jarvis@radiocad.co.uk.