FR2712934A1 - A coil ignition method and device for a spark ignition engine - Google Patents

Abstract

The method and the electronic ignition device of the invention ensure, between two charging times of the primary (13), the passage of a measurement current in the primary (13) for at least one diagnostic period defined by the calculator (16). The rate of increase of the current in the primary (13) is determined and discriminated with respect to a threshold to deduce information on the presence and / or quality of spark at the spark plug (10) connected to the secondary (12) of the ignition coil (11). The primary current is measured at the terminals of a shunt (15), and the measurements transmitted to the detector (18) with comparator, delivering signals of presence of spark to the computer (16), which controls the switch (14) by a interface amplifier (17). Application to the diagnosis of the ignition of internal combustion engines with positive ignition.

Description

"COIL IGNITION METHOD AND DEVICE,

FOR ENGINE IGNITION CONTROL "

  The invention relates to a method and an electronic ignition device, for a spark ignition internal combustion engine, in each cylinder of the engine, by at least one spark plug in series with a secondary of an ignition coil, of which the primary is supplied with electric current from an electrical source, such as the battery of a vehicle equipped with the motor, by a switch placed in a charging circuit of the primary, and itself controlled by a computing unit and order, including

  advantageously a microcontroller.

  The subject of the invention is more precisely a method and a device of the type mentioned above, making it possible to carry out a diagnosis of the ignition of the engine, thanks to the computing and control unit, or computer, which manages the ignition function and, possibly, that the injection function, when the engine is equipped with an injection fuel supply system, the computing and control unit then being a calculator called "de

engine control ".

  The method and the device according to the invention are intended to be implemented not only on board the vehicles, but also on the control apparatuses and

diagnostic in service station.

  The ignition diagnosis is to give qualitative information about the spark produced

  by the candles in the combustion chambers of the engine.

  This information must make it possible to determine whether there is absence or presence of a spark, and, in the latter case, whether

the spark is correct.

  Numerous methods and devices have already been proposed to provide such a diagnosis, which is obtained either by measuring the current flowing through the spark plug, using a shunt, or by performing a voltage analysis at the same time.

  primary terminals of the ignition coil, during the spark

the actual or presumed one.

  An ignition device, with ignition diagnosis of the second type mentioned above, has been proposed by the applicant in the European patent application EP 559 540: it comprises, on the one hand, means for measuring the current of charge, intended to supply, on an output, a signal in response to the exceedances of a given value, lower than the current required to create the ignition voltage at the secondary coil by breaking the primary current, and secondly means for transposing the primary voltage, to form the image of the secondary voltage and to provide, on said output, a signal when the secondary voltage falls below a determined threshold indicating

  the end of the spark to the candle. These means of transposition

  tion, in series between the terminal of the primary

  connected to the source and the mass, the emitter-

  collector of a transistor and collector and emitter resistors, so that the voltage across one of the resistors is representative of the difference between the

  voltage across the primary and the source voltage.

  A first disadvantage of such a device is that it does not allow, in a simple way, to discriminate sparks of different cylinders at high speed, due to a superposition between the transpositions of the secondary voltages and the primary charges. For example, in a four-stroke four-cylinder in-line engine, the primary charge of the No. 3 cylinder coil can begin while the spark spark of the No. 1 cylinder is not

completed.

  To remedy this drawback, it is necessary to duplicate certain components of the circuit, and / or to provide a diode between the primary of each coil and transposition means that may be common to both.

  cylinders, otherwise to all the cylinders.

  Another disadvantage is that such a device is expensive to implement, because the resistors and diodes it comprises are high voltage components, and occupy a relatively large area on circuits.

printed matter used.

  The object of the invention is to overcome the drawbacks of the aforementioned device known from EP 0 559 540, and to

  propose a method and a device for diagnosing

  which makes it possible to discriminate in a simple way the sparks of the different cylinders of the engine, while being

economic implementation.

  Another object of the invention is to provide a method and a device for controlling the quality of the ignition by detecting whether the spark is of a duration neither too short nor too long, and if the spark is

  of sufficient quality for a sufficient period of time.

  Another object of the invention is to propose

  a method and a device for monitoring

  wear of the candle, or, more generally, the deterioration

  the secondary circuit which has it in series with the secondary coil and a resistive wire suppression,

for example integrated in the candle.

  Another object is finally to propose a method of ignition diagnosis compatible also with the calculation and control of the charging time of the primary, in the manner already known, in particular by the aforementioned document EP 559 540, the method according to the invention. which can be implemented by a device which is structurally little different from known devices for performing this calculation and this

  control of the charging time of the primary.

  The basic idea of the invention is that an image of the quality of the spark at the candle is given by the dynamic impedance of this candle during the spark, and that the observation of the intensity of the spark a current which is circulated in the coil primary during the spark can be used to analyze the rate of increase during at least one diagnostic period, this rate of increase of the intensity of the current primary depending in particular on the dynamic impedance of the candle during the spark.

  In summary, according to the invention, the quality of the stain

  that is determined from the measurement of the rate of increase of the intensity of the current flowing in the

  coil for at least one diagnostic

  determined during the duration of the presumed spark.

  For this purpose, the invention proposes a process for

  electronic mage of the type presented above and comprising the step of cyclically controlling the charge of the primary during defined load times and at times defined by the calculation and control unit, so that the breaking of the load at the end of each charging period, is intended to cause a spark at the spark plug, and is characterized in that it further comprises the steps of: - between two successive charging times, to control the passage of a measuring current in the primary for at least one diagnostic period and at a time defined by the computing and control unit, - determining a rate of increase of the intensity of the current in the primary for at least one duration diagnostic, - to discriminate the high growth rates of low growth rates, and - to derive information on the presence and / or quality of a spark at the candle, after the end of the

previous charging time.

  The discrimination of the rates of increase of the intensity of the current in the primary during each diagnostic period can be ensured by comparison with at least one corresponding threshold, to deliver a signal of presence of spark in case of exceeding this corresponding threshold. but, in a particularly simple variant of implementation, the intensity of the current is measured

  in the primary for at least each diagnosis period

  tic, and growth rates are discriminated by comparing the maximum intensity measured at the end of the duration.

  corresponding to a corresponding intensity threshold.

  laying, to deliver a spark presence signal if

  the maximum intensity measured is greater than the inten-

  site. Advantageously, the method consists in controlling at least two successive diagnostic periods between two successive charging periods, and in deducing therefrom, on the one hand, the presence of a spark if at least one presence signal is delivered in a first interval of time

  predetermined, following the end of the previous charging time.

  re, and ending at the earliest at the same time as the first diagnostic period, and, secondly, a minimum duration of a spark of sufficient quality depending on the

  the number of presence signals delivered.

  In addition, the method may further consist in deducing that the spark is too short if a second presence signal is not delivered in a second predetermined time interval, greater than the first time interval, also following the end of the previous charging time, and ending at the earliest at the same time as the

second duration of the diagnosis.

  In a preferred embodiment, the method also consists in controlling a third diagnostic period between two successive charging periods, and in deducing that the spark plug is short-circuited if a third presence signal is delivered before the end of a third predetermined time interval, also following the end of the previous charging time, greater than the first and, if applicable, the second predetermined time interval, and ending at the earliest at the same time as the third

duration of diagnosis.

  To monitor the wear of the candle and

  secondary circuit which has it in series with the second

  the process may consist in measuring the inten-

  maximum current in the primary for at least one diagnosis period of the same rank after successive charging times, to compare the measurements between them and / or at least a maximum intensity threshold, and to deduce at least one signal testifying to the degradation of the secondary circuit. The invention also relates to an electronic ignition device, of the aforementioned type, further comprising means for measuring the current in the primary, connected in series in the load circuit between the switch and the ground, and connected to detection means, themselves connected to the computing and control unit, as known from EP 559 540, and which is characterized in that: the computing and control unit comprises means generating a signal controlling the switch for the passage of the current in the primary for at least one diagnostic period and at least one time defined by said unit, between two successive charging times of the primary, and - the detection means comprise means discriminating the speeds of increasing the intensity of the current in the primary and delivering a signal to said computing and control unit when the measured intensity of the current in the primary is greater than a threshold,

  less during the diagnostic periods.

  According to a simple and economical embodiment, the switch is controlled, from a control output of the computing and control unit, via an interface amplifier, the means for measuring the current in the primary comprise a shunt in series with the switch, between the latter and the ground, and the detection means comprise at least a first and a second comparator, each receiving the signal of the shunt on an input and comparing it respectively to a first and to a second intensity threshold, received on another input of the corresponding comparator, the output of each comparator being connected to a diagnostic input of the computing and control unit, the second intensity threshold being greater than the first, but lower the current required to create an ignition voltage by breaking the primary, for the

  calculation by said unit of the charging times of the primary.

  Advantageously, to enable monitoring of the secondary circuit, the calculation and control unit comprises a second diagnostic input, connected to the shunt, and transmitting the measurement of the intensity of the current to an analog / digital converter, itself connected. means for memorizing and comparing the maximum intensity values, measured during the diagnostic periods, with each other and / or at a maximum intensity threshold, and means delivering a degradation signal of the secondary circuit.

  Other features and advantages of the invention

  will flow from the description given below,

  nonlimiting title of an exemplary embodiment described with reference to the accompanying drawings in which: - Figure 1 is a block diagram of a device according to a particular embodiment of the invention; FIG. 2 shows the curves representative of the variation, as a function of the time t, of the control signal delivered to the switch, of the current Ip in the coil primary, of the current Is in the secondary of the coil, and of the signal of diagnosis transmitted to the computing and control unit; FIG. 3 shows a possible constitution of the device of FIG. 1; FIGS. 4 and 5 show the variations as a function of time of the currents Ip and Is according to more representative curves and on a larger scale than those of FIG. 2; FIG. 6 represents, on a larger scale, four curves representative of the measuring current during a diagnostic period for four different values of the resistance of the secondary circuit, and

  FIG. 7 represents the four curves corresponding to

  The device, the constitution of principle of which is represented in FIG. 1, is intended for the ignition of an internal combustion engine combustion chamber, equipped with at least one spark plug 10. placed in series with the secondary 12 of an ignition coil 11. The primary 13 of the coil 11 is connected to a power source (vehicle battery in general), voltage Vbat. On the opposite side respectively to the source Vbat and to the spark plug 10, the primary 13 and the secondary 12 have a common terminal connected to the collector of a bipolar transistor 14 of the NPN type, whose emitter is connected to ground by the The shunt 15 is thus in series in the charging circuit of the primary 13 from the source Vbat, between the ground and the ignition transistor 14, serving as a switch or switch of cutoff, which can switch from the blocked state to the saturated state and vice versa, for controlling the primary current Ip, which flows through the primary 13 when the transistor 14 is conducting in the saturated state. This switch 14 is controlled at closing and opening from a computing and control unit 16, via an amplifier 17 serving as an interface between the base of the transistor 14 and the unit 16. ,

  which is a motor control calculator, of the microconon type

  controller, having at least one microprocessor, and having at least one output S giving the control information to

the amplifier 17.

  The shunt 15 serves to measure the primary current Ip, and provides at its terminals a voltage representative of this current. This information coming from the shunt 15, and relating to the intensity of the primary current Ip, is processed, in analog form or in logical or digital form, by a threshold detector 18 connected to at least one input E of the computer 16, making it possible to receive diagnostic information on the presence and quality of a spark produced between the two electrodes of the spark plug 10 and generated by the secondary 12 of the ignition coil 11 by

  the cutting off of the primary charge 13.

  The operation of the device according to the method of the invention is described below with reference to the four curves (a) to (d) of FIG. 2. The curve (a) represents a chronogram of the control signal of the switch 14 applied to the output S of the unit 16. The curves (b) and (c) respectively represent the intensity of the primary current Ip and the secondary current Is as a function of time, and the curve (d) represents the signal applied by the detector 18 on the input E of the unit 16. On its output S, the unit 16 cyclically applies an output signal of the instant t0 of

  closing switch 14 until time tl cutoff.

  During the charging period of primary 13, between these

  instants t0 and t1, the primary current Ip gradually increases

  as shown in curve (b). In about 1 ps after the cut-off time t1, Ip is zero from its maximum value of the order of 6 A, sufficient for its break to give the required starting voltage across the spark plug 10, and the secondary current Is very quickly reaches its maximum value of the order of 60 mA, as shown in curve (c). From this maximum value, Is decreases

  gradually, while Ip is zero.

  Between t1 and the beginning of the next charge duration (t1-t0), the unit 16 applies on its output S a command for closing the switch 14 for three successive diagnosis periods, preferably of the same value, and spaced in the time, t3-t2, t5-t4 and t7-t6. The primary current Ip gradually increases from the beginning t2, t4

  or t6 of each diagnostic period to cancel fast

  cut at t3, t5 or t7 at the end of each diagnostic period, as shown in curve (b). During the spark, during which the secondary current Is gradually decreases as shown on the curve (c),

  Until canceling in t8, so that the duration of the spark

  that T = t8-t1 is of the order of 1.5 ms, under normal ignition conditions, it corresponds to each pulse of a selected positive sign of primary current Ip a pulse of a negative selected sign secondary current Is, which is practically no longer sensitive outside the spark duration T. During this spark duration T, during which the spark is normally present, after the instant tl (at the end the charging time tl-tO, of about 3 ms, during which the charging current flows in the primary 13), the impedance Z of the spark plug 10 is low (of the order of 20 to 100 kΩ ). After the spark

  (after t8), the impedance Z is high (greater than 1 Mn).

  In theory, and during a normal spark, the rate x of increase of the primary current Ip, is expressed by the following formula (1): (1) x = Imax x Z i 2 x L, o, Imax is the maximum current saturation of the primary 13 (of the order of 24 A, for a usual battery voltage of

13 V),

  Z is the impedance of the plug 10 (of the order of 20 k1), r is the transformation ratio between the secondary 12 and the primary 13 (of the order of 100), L, is the self-leakage of the primary set 13 and

secondary 12 (of the order of 1 mH).

  On the other hand, in the absence of a spark, this rate of increase of the primary current Ip, which corresponds to the slope of the curve of Ip as a function of time t, is given by formula (2): (2) t '= Vbat Lp o, Vbat is the voltage of the battery (of the order of 13 V),

  and Lp is the inductance inductance (of the order of 6 mH).

  During diagnostic times t3-t2, t5-t4 and

  t7-t6, and during a normal spark, the speed of

  increase of Ip is of the order of 50 A / ms, whereas in the absence of a spark, this rate of increase x 'is

of the order of 2 A / ms.

  This rate of increase of Ip corresponds to the speed gradients during the diagnostic times, which are calibrated at a constant value of the order of 20 ps for example. This rate of increase of Ip can therefore be expressed by the ratio of the maximum intensity of this current at the end t3, t5 or t7 of each diagnostic period over said diagnostic period, or even more simply,

  by the values of the maximum intensity.

  The strong gradients (50 A / ms) are discriminated from the weak gradients (2 A / ms) by the detector 18, comprising for example a threshold comparator. The threshold Io can be calibrated to a constant and unique value for all the diagnostic periods, and be set for example at 1 A. Each time and as long as Ip is greater than Io, the detector 18 applies to the input E of the unit 16 a non-zero signal, as shown in (d) in Figure 2. The signal transmitted during tl-t'0 corresponds to a current

  primary Ip> Io = l A, during the charging time (tl-tO).

  The signals transmitted during t3-t'2 and t5-t'4 correspond to a primary current Ip> Io, during the diagnostic periods t3-t2 and t5- t4, which took place before the end of the spark in t8 . On the other hand, no signal is given on the input E during the diagnostic period t7-t6 posterior to t8, because Ip remained below the threshold

Io of 1 A during this time.

  The signals received by the input E of the unit 16 during diagnostic periods such as t3-t2 and t5-t4

  are therefore signals of presence of spark.

  The specific diagnostic program implemented by the microprocessor of the unit 16 makes it possible, from the presence of spark signals delivered by the detector

  18, to determine that the spark was of sufficient quality

  during a time interval at least equal to t5-t1, in this example, therefore T> t5-tl, and that the spark is very likely extinguished before t7, so T <t7-tl. In general, if, during a predetermined time interval T1, beginning in t1 and ending at the earliest in t3, but not after t4, the unit 16 does not receive a spark presence signal from the detector 18 , it delivers a diagnosis of spark failure, meaning that

  the circuit of the candle 10 is likely open.

  If the unit 16 receives a first presence signal during T1 (for example t3-t'2), but no second presence signal during a predetermined time interval T2, also starting in t1 and ending at no earlier than t5 (but not after t6), the unit 16 considers that there has been a spark, but that it has not been of a suitable quality for a sufficient duration. Unit 16 delivers

  then a diagnosis of spark too short.

  If the unit 16 successively receives two signals of presence of spark during T2, the first during T1,

  but no third sign of presence during a third

  the second predetermined time interval T3, beginning in t1 and ending not earlier than t7, but not before the beginning (t0) of the next charging time, the unit 16 delivers a spark diagnosis of suitable quality for a period of time.

sufficient time.

  Finally, if the unit 16 successively receives three signals of presence of spark during T3, it delivers a diagnosis of spark too long, meaning that the candle

  is probably shorted.

  For a finer detection of a possible short-

  circuit, the detector 18 can, during the third diagnostic period t7-t6, compare the measured current Ip with a particular short-circuit threshold, for example less than 1 A. In general, the detector 18 can compare the intensity of the measured current, during each duration of

  diagnosis, at a threshold specific to each of the di-

  gnostic, especially if they are not of equal value. The unit 16 can thus diagnose the presence or absence of a spark of suitable quality, and the duration of such a spark included, for example, between 0.4 ms and 2 ms. The measurement of the spark duration can be taken into account by the unit 16 to adjust the value I 2 of the primary current Ip for which the cut is made, in order to guarantee, for example, a duration of spark that is sufficient to avoid unburnt, such a duration for less than 0.4 ms indicating a high probability of open circuit, while a spark duration greater than 2 ms indicating a high probability of short circuit, these two values delimiting a range of spark duration that experience

revealed as normal.

  In FIG. 3, the elements of the analogous device

  those to Figure 1 have been indicated by the same references. The amplifier 17 comprises two bipolar transistors, including an NPN type 19 connected by its base to the output S of the unit 16 via a resistor 21, while its emitter is grounded and its collector connected by a resistor 22 at the base of the other transistor, of the PNP type, whose emitter is turned on by the source Vbat, and the collector connected by a resistor 23 to the base of the ignition transistor 14, which comprises, on its collector connected to the primary 13, an integrated Zener diode

  (and not shown) in known manner.

  The detector 18 comprises an input filter RC 24-

  , filtering the primary current measurements across the shunt 15, and transmitting them to a negative input of a first threshold comparator 26, receiving on an input

  a voltage threshold corresponding to the threshold of inten-

  Io spark detection, and defined, from a logic voltage source Vcc (+5 V for example) by

  via a resistance bridge 27, 28 and 29.

  The output of the comparator 26 is connected in parallel with the voltage source + Vcc through the resistor 30 and at the input of an inverter 31. The output of the filter RC 24- is also connected in parallel with the positive input. a second threshold comparator 32 receiving on its negative input another voltage threshold defined by the resistor bridge 27, 28 and 29 from the voltage source

  + Vcc. As for comparator 26, the output of comparator

  32 is connected in parallel with the source + Vcc through the resistor 30 and at the input of the inverter 31, isolated from the ground by the capacitor 33. The combination of the two comparators 26 and 32 constitutes a OR circuit, which

  attacks the diagnostic input E of unit 16 through

intermediate of the inverter 40.

  The comparator 26 compares the measurements of current at the threshold Io, of 1 A for example, for the delivery of the signals

  of spark presence as described above. The comparison

  The converter 32 compares the measurements of the intensity of the primary current Ip with a second intensity threshold Il, greater than that Io of the comparator 26, and for example equal to 4.5 A, to provide a signal when the charging current Ip in the primary 13 exceeds this second threshold Il, when the current Ip to be cut in order to obtain a spark of satisfactory quality is I 2, for example 6 A, as indicated on the curve (b) of FIG. 2. From the duration ( t0) between the beginning t0 of the charge of the primary 13, fixed by the unit 16, and the instant t'l of exceeding 11, the unit 16 can determine the time tl-t'l necessary to get to the current I2 and, for

  following ignition, determine t0 appropriately.

  The instant of the beginning of the charge can thus be adjusted optimally by the unit 16 without additional output of the detector 18 to the control unit 16.

  The curves of the currents Ip and Is measured on such a device are respectively represented in FIGS. 4 and 5, with an offset of the time origin for Is by

  report to Ip. After a primary charging time of approx.

  3 ms, for which Ip increases from 0 to practically 6 V, there are two peaks of selected positive sign of primary current Ip measured during the first two diagnostics periods of 20 ps beginning at approximately 3.2 ms and 4 ms respectively. At these current spikes Ip, correspond to negative selected sign points of the secondary current Is, which vanishes towards 5 ms, indicating a duration of spark

slightly less than 2 ms.

  The signals for measuring the primary current Ip across the shunt 15 are also transmitted from the output of the RC filter 24-25 in parallel directly to a second input E 'of the computer 16 and applied to an analog / digital converter 34, itself connected to counting means, memories, registers and digital comparators 35 of the computer 16. The primary current Ip, measured during

  the diagnostic times, is used as an analog signal

  diagnostics transmitted to the analog input E 'of the microprocessor of the computer 16. Thanks to the converter 34 and the digital memories and comparators 35, it ensures the comparison of the maximum currents of the primary current Ipmax, as measured during periods of diagnosis, of the same rank (for example always during the first diagnostic period) between successive charging periods, and these values Ipmax are compared with each other and with a maximum threshold Imax of primary current intensity, adapted in according to the order of the diagnostic time considered in the following of these diagnostic times between two consecutive charging times, in order to detect the decrease of the values Ipmax, as a function of the operating time of the installation, which translates

  the wear of the candle 10 and, more generally, the deterioration

  secondary circuit, comprising this candle 10, the secondary 12 and a suppression resistance, in

  series with the candle 10 and / or integrated into it.

  Indeed, the degradation of this secondary circuit is reflected

  by a progressive increase in the equivalent resistance

  of this circuit, resulting in a correlative decrease of the signals of maximum intensity Ipmax, as shown in FIG. 6. This FIG. 6 represents four curves of Ip measured during the first diagnostic period t3-t2, the curve 36 corresponding to a candle 10 in new condition,

  while curves 37, 38 and 39 correspond to the use of

  the same candle 10 after progressively increasing periods of use. It can be seen that the maximum intensity of the primary current Ipmax gradually decreases as the duration of use of the candle 10 increases, from curve 36 to curve 39. The monitoring of Ipmax thus gives an image of the degradation of the secondary circuit, and by comparison with a threshold of maximum intensity Imax, for example of 1 A, we can consider that the curves 36 and 37 correspond to a candle 10 in an acceptable state, since Ipmax is greater than the threshold Imax, whereas the curves 38 and 39 correspond to a candle 10, or, more generally, to a secondary circuit too degraded, requiring by

example the change of the candle.

  FIG. 7 shows the four curves of the secondary current Is 40, 41, 42 and 43, obtained during the sparks during which the curves 36 to 39 of Ip, measured during the first diagnostic period, were respectively measured. correspond to the negative sign peaks current Is visible and superimposed for the four curves 40 to 43 of Figure 7. It is found that the curves 42 and 43, corresponding to sparks obtained with a candle 10 in a degraded state, reflected by the curves 38 and 39 of Figure 6, correspond to sparks of too short duration, while the duration of the sparks curves and 41, obtained with a candle 10 in good condition, as reflected by the curves 36 and 37 of the figure 6, is sufficient to obtain a spark of suitable quality

for a sufficient period of time.

  It should be noted that the diagnostic method described above can be implemented by a device (see FIGS. 1 and 3), which is at the same time structurally little modified compared with that described in EP 0 559 540, and of much cheaper than the latter, since

  does not require high voltage components.

Claims (11)

  1. An electronic ignition method for a spark-ignition internal combustion engine in each cylinder of the engine, by at least one spark plug (10) in series with a secondary (12) of an ignition coil (11) of which the primary (13) is supplied with electric current from a source (Vbat) by a switch (14) placed in a charging circuit of the primary (13), and itself controlled by a unit (16) calculation and control, comprising the step of cyclically controlling the load of the primary (13) during defined charging times (tl-tO) and at times (tO) defined by the calculating and controlling unit (16) , so that the cutting of the charge at the end (tl) of each charge period (tl-tO) is intended to cause a spark at the spark plug (10), characterized in that it further comprises the steps of : - between two successive charging times (tl-tO), to control the passage of a measuring current in the primary (13) pe at least one diagnostic period (t3-t2) and at least one instant (t2) defined by the computing and control unit (16), - determining a rate of intensity increase (Ip) of the primary current (13) for at least one diagnosis period, - discriminating the high growth rates of the low growth rates, and - deriving information on the presence and / or quality of a candle sparkle (10) after the end of   the previous charging time (tl-tO).   2. Method according to claim 1, characterized in that it consists in discriminating the rates of increase of the intensity (Ip) of the current in the primary (13) during each diagnostic period (t3-t2) compared with at least one corresponding threshold, and to deliver a signal of   presence of spark in case of exceeding the threshold corresponding to   laying. 3. Method according to claim 2, characterized in that it further comprises the step of measuring the intensity (Ip) of the current in the primary (13) for at least each diagnostic period (t3-t2), and discriminate growth rates by comparing intensity   measured during the corresponding diagnostic period.   to a corresponding intensity threshold (Io), to deliver a spark presence signal if the intensity   measured maximum is greater than the intensity threshold (Io).   4. Method according to one of claims 2 and 3,   characterized in that it furthermore consists in controlling at least two successive diagnosis periods (t3-t2, t5-t4) between two successive charging periods (t1-t0), and in deducing therefrom, on the one hand, the presence of a spark if at least one presence signal (t3-t'2) is delivered in a first predetermined time interval (T1), following the end (tl) of the previous charge duration, and ending at the most early at the same time (t3) as the first diagnostic time   (t3-t2), and, on the other hand, a minimum duration of one   that of sufficient quality according to the number of signals issued.   5. Method according to claim 4, characterized in that it further consists in deducing that the spark is too short if a second presence signal (t5-t'4) is not delivered in a second time interval. predetermined value (T2), greater than the first time interval (T1), also following the end (tl) of the previous charge duration, and ending at the same time at the same time (t5) as the second duration of diagnosis (t5-t4).   6. Method according to one of claims 4 and 5,   characterized in that it also consists in controlling a third diagnostic period (t7-t6) between two successive charging periods (t1-t0), and in deducing that the spark plug (10) is short-circuited if a third presence signal is delivered before the end of a third predetermined time interval (T3), also following the end (tl) of the previous charging time, greater than the first (T1) and, if appropriate, the second interval of time (T2) predetermined, and ending at the same time at the same time (t7) as the third diagnostic duration (t7-t6).   7. Process according to any one of the claims   1 to 6, characterized in that it further consists in measuring the maximum intensity of the current (Ip) in the primary (13) during at least one diagnostic period (t3-t2) of the same rank after charging times (tl-tO) successive, to compare the measurements between them and / or at least a maximum intensity threshold (Imax), and to deduce therefrom at least one signal testifying to the degradation of the secondary circuit comprising the candle (10) and the secondary (12) of the coil (11).   8. Process according to any one of the claims   1 to 7, characterized in that it consists in calibrating the diagnostic periods (t3-t2, t5-t4, t7-t6) on a constant value and the corresponding thresholds (Io) of intensity of   current (Ip) to a value that is also constant.   9. An electronic ignition device for a spark-ignition internal combustion engine in each cylinder of the engine, by at least one spark plug (10) in series with a secondary (12) of an ignition coil (11), whose primary (13) is supplied with electric current from a source (Vbat) by a switch (14) placed in a charging circuit of the primary (13), and itself controlled by a computing unit and control (16), means (15) for measuring the current (Ip) in the primary (13) being connected in series in the load circuit, between the switch (14)   and the mass, and connected to detection means (18), themselves   same connected to the computing and control unit (16), characterized in that: - the computing and control unit (16) comprises means generating a control signal of the switch (14) for the passage of the current (Ip) in the primary (13) for at least one diagnostic period (t3-t2) and at least one instant (t2), defined by said unit (16), between two charging durations (tl-tO ) of the primary (13), and - the detection means (18) comprise means (26) discriminating the rates of increase of the intensity of the current (Ip) in the primary (13) and delivering a signal to said computing and control unit (16) when the measured intensity of the current (Ip) in the primary (13) is greater than a threshold (Io), at least during diagnostic times.   10. Device according to claim 9, characterized in that the switch (14) is controlled from a control output (S) of the unit (16) for calculation and control, via an amplifier d interface (17), the current measuring means (Ip) in the primary (13) comprises a shunt (15) in series with the switch (14), between the latter (14) and the ground, and the detection means (18). ) comprise at least a first (26) and a second comparator (32), each receiving the signal of the shunt (15) on an input and comparing it respectively to a first (Io) and a second intensity threshold (II) , received on another input of the corresponding comparator (26, 32), the output of each comparator being connected to a diagnostic input (E) of the computing and control unit (16), the second intensity threshold ( Il) being greater than the first (Io), but smaller than the current (I2) required to create a primary cutoff ignition voltage (13) , for the calculation by said unit (16) of the charging times (tl-tO) of the primary (13).   11. Device according to claim 10, characterized   characterized in that the computing and control unit (16) comprises a second diagnostic input (tl), connected to the shunt (15), and transmitting the current intensity measurement (Ip) to an analog / digital converter digital (34), itself connected to means (35) for storing and comparing the maximum intensity values (Ip), measured during the diagnostic periods (t3-t2, t5-t4, t7-t6), between them and / or at a maximum intensity threshold (Imax), and means delivering a degradation signal of the secondary circuit comprising the candle (10) and the secondary (12) in case of intensity (Ip) measured lower lower audit threshold (Imax).