Unlike other ATEX types of protection, intrinsic safety as defined by IEC 60079-11 does not intend to isolate the electrical circuits from the potentially explosive atmosphere, but ensures that the apparatus is intrinsically not dangerous because of sparks or thermal effects, even in situations resulting from possible faulty operations or component failures.
For the risk due to sparks, intrinsic safety is based on the physical phenomenon that a spark can not cause the ignition of an explosive atmosphere if the energy involved by the spark is below the level specified for the given group of gas. Consequently, the standard provides curves and tables to help to determine with a safety factor, the maximum acceptable voltages and currents, in particular when capacitors and inductors are implemented because they store energy.
Constructional requirements
The IEC 60079-11 standard specifies the constructional requirements and, the faults and failures to be taken into account to determine the assessment criteria in order to demonstrate by tests or calculations, that the safety limits and safety factors can never be overstepped even in situations resulting from multiple faults and failures.
For the risk due to thermal effects, similar methods of assessment are carried out to determine the maximum surface temperatures of the components in contact with the potentially explosive atmosphere, in order to ensure the compliance with the temperature classification of the apparatus.
Methods of assessment and constructional requirements result directly from some basic principles described in the standard :
- two conductors are considered to be short-circuited if the separation distance between them, is below the minimum distance required by the standard.
- an electrical connection is considered to be open-circuited, unless it satisfies the constructional requirements of the standard. In this case, it is called an "infallible connection".
- an electronic component must be considered failing in such a way that the most critical case might result for the safety. Nevertheless, the standard defines for some simple components (transistors, diodes, resistors, ...) the actuel failures to be taken into account for the assessment, and can even consider that they are infallible if the components are manufactured and used in accordance to the requirements ; in this case, they are called "infallible components". Please note, that a component is never regarded as infallible by nature,it is the way in which it is used in the electrical circuit, which can make it considered as infallible.
- assessments are carried out by considering the simultaneous occurrences of the most critical faults and failures.
Transposing into an ATEX model
If the constructional requirements are not taken into account, the assessment is based on the transposition of the electrical schematic into an ATEX model representing the risk due to sparks.
The conductors being likely to be in short-circuit resulting to the worst configuration, the semiconductors and integrated circuits being also likely to have internal short-circuits, the more critical case for the risk due to sparks would be furthermore to see all the capacitors therefore connected in parallel and all the inductors connected in series. It does not matter if this possibility is realistic or not, if the constructional requirements are not taken into account, this even very unlikely configuration is considered as the actual basis for the assessment.
Therefore, the method consists in replacing the electrical schematic by a capacitor equivalent to the sum of the capacitances (tolerances included) and by an inductor equivalent to the sum of the inductances (tolerances included).
The next step consists in assessing that resulting ATEX model for compliance with the maximum values given by the curves and tables of the standard, when considering the highest voltages and currents that might appear in the circuit even in case of faults or failures. For example : a step-up DC/DC converter delivers a theoretically infinite voltage when the feedback loop is failing !!!
Rather frequently, information given by the standard is not sufficient, voltages or currents are out of the given curves or tables, assessment of capacitor-inductor coupling are impossible ; in those cases although well-justified simulations and extrapolations may be allowed to assess for compliance with the standard, the final evidence will be brought by a test with a spark test apparatus(*) carried out by the certifying laboratory.
For the risk due to thermal effects, it is considered that all the electrical power of the circuit, is dissipated by the electronic component which is because of its package, the most likely to reach the higher temperature.
The simple fact of, imagining a power of 1 Watt (or more) dissipated in a small transistor (SOT23-package or smaller), or considering that the maximum capacitance value allowed by the standard at a voltage of 10 V for the gas Group IIC is only 3 uF, shows quickly the limits of such a global approach.
Then it becomes obvious, that from the point of view of the risk due to heating effects the use of components increasingly smaller and the race to the miniaturization are a nonsense, that from the point of view of the risk due to sparks another approach is also necessary, and that designing an apparatus intented to be used in potentially explosive atmospheres is a matter of compromises and structuring, on which all the electronic design must be based.
An apparatus which is not designed from the beginning for a use in potentially explosive atmospheres, has thereafter very little chance to be able to be adapted to this context
(*) spark test apparatus : test equipment strictly defined by the standard as well as how to use it, consisting of an explosion chamber filled with an explosive gas mixture, in which a mechanical system can produce short-circuits (consequently sparks) at a given point of the electrical circuit in test.
Structuring into "islands"
Structuring consists in gathering the non-dissociable components in "islands", so called because they are isolated from each other with "infallible components" which limit the voltage and split the total electrical power as well as the stored energies, therefore decreasing the risk due to sparks.
Inside each "island", the electrical circuit is transformed in an ATEX model as described in the previous chapter, and the whole electrical circuit is then represented as an assembly of capacitors and inductors, separated by "infallible resistors" or protected by "infallible devices". In each "island", any electrical problem due to faults or component failures is circumscribed in such a way that any electrical effects to the other "islands" are fully controlled .
In the example below, an imaginary intrinsically safe apparatus is connected to a sensor. The system thus made up, will be intrinsically safe if the electrical parameters calculated at the output of the apparatus are compatible with the safety parameters given by the ATEX certificate of the sensor.
The electrical circuit is broken down into 3 "islands", with "infallible components" (represented by triangles) added to limit the voltages, the currents and the injected powers at the input of each "island".
Inside each "island" is specified the maximum value of the voltage even in case of any fault or failure, the maximum injected power, the sum of the capacitances et the sum of the inductances. For each "infallible component" the power rating is also specified, so that it can be easily verified by calculation that the electrical circuit is designed in accordance with the requirements of the standard.
Doubling the "infallible components" characterizes the level of protection "ib" applied to intrinsically safe apparatus intended to be used within Zone 1 ; the intrinsically safe apparatus intended to be used within Zone 0 require the level of protection "ia" characterized by the tripling of the "infallible components".
Some transients have also to be taken into account (for example, during the clearing time of a fuse), so as to be sure of the reliability of the components used as "infallible components". Datasheets have to be read very carefully, and even sometimes completed by some more confidential information given by the manufacturer of the component, or some specific tests on samples.
Characteristics of some Zener diodes - Measurements carried out with a pulsed current
Thermal analysis of a component
The impact of the printed circuit board design is so predominant (tracks, power planes, material, distances, thickness, ...), that most of the time no information given by the manufacturer of the component should be taken into account for the assessment of the rise in temperature of the package.
Except for the cases where the thermal resistance of the package is defined well above the minimum calculated requirement, measurements must always be made with the component soldered on the final printed circuit board by injecting a power into the component and by measuring the temperature at the hottest point of the package. By subtracting the room temperature during the test, a function T = f(P) is established, where T is the increase of temperature and P the power injected. From this point, knowing the maximum possible power in the "island" where the component is used, and the maximum operating temperature, it is then very simple to verify the compliance with the temperature classification assigned to the apparatus.
Thermal reports of (SOT23, SOT223, SOD87) components on a given printed circuit board
In the event that an internal junction of the component breaks down while conducting the test, the curve is extrapolated. Indeed, it is not possible to ensure that the component will always break down in the same way at the same injected power. The test does not consist in measuring the highest temperature, but it does consist in determining the thermal resistance of the component when it is soldered on the printed circuit board.
