JPH0260868B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a method and apparatus for controlling the ignition timing of an engine, and more particularly to a method and apparatus for controlling the ignition timing by detecting the occurrence of a knock.

Knocks that occur in an engine are accompanied by a knocking sound that lowers the driver's mood, and also causes a reduction in engine output due to the generation of reverse torque or damage due to overheating of the engine. This knock has a close relationship with the ignition timing, and it is known that engine output can be maximized by setting the ignition timing or ignition advance immediately before the knock occurs due to the characteristics of the engine. Therefore, reducing the ignition advance angle in order to avoid the occurrence of a knock will conversely reduce the engine output, so it is necessary to control the ignition timing immediately before the occurrence of a knock.

USP4002115 retards the ignition advance angle when it detects the occurrence of a knock, then gradually advances the ignition advance angle, and when a knock occurs in the process of advancing the advance angle, the ignition advance angle is retarded. A method is disclosed in which the ignition timing is controlled by retarding the ignition timing and repeating such an operation. Furthermore, this technique is also disclosed in Japanese Patent Application Laid-Open No. 54-84142.

In this type of control, the ignition timing is always advanced gradually and retarded when a knock occurs, so the engine torque fluctuates depending on the cycle of the knock, and this torque fluctuation This results in a negative impact on the running performance of the engine. If the angle retarded at the time of knock occurrence is made small in order to minimize this adverse effect, the knock cannot be quickly eliminated, and therefore the adverse effects caused by the knock cannot be sufficiently eliminated.

In addition, Japanese Patent Laid-Open No. 54-103911 proposes a method of adjusting the amount of retardation of the ignition timing in accordance with the knock strength.

However, this method has a problem in that it is very difficult to accurately detect the vibration amplitude because the knock strength is detected by the amplitude of the knock sensor output. That is, in the engine control device, it is necessary to take measures against noise such as ignition noise, and the amplitude of the knock sensor output for which noise measures are taken is attenuated, making it impossible to accurately detect vibrations. In other words, it is not possible to distinguish between the vibration waveform of the knock sensor output and the noise waveform.

SUMMARY OF THE INVENTION An object of the present invention is to provide an engine ignition timing control method that can accurately detect knock strength and obtain highly accurate ignition timing.

A feature of the present invention is that engine rotation information and load information are taken in to generate ignition timing data (θADV) of the engine.
A first step of calculating the knock vibration of the engine (BASE); a second step of converting the knock vibration of the engine into an analog electrical signal by a knock sensor; A third step of converting the analog electrical signal into a pulse signal having a number of pulses corresponding to the frequency of the analog electric signal having a magnitude greater than the knock judgment value; counting the pulse signal for each ignition cycle to obtain a count value (NP); The fourth value detected as the knock strength
a fifth step of determining a retard correction value (ΔθADV1) for retarding the ignition timing based on the count value (NP) representing the knock strength; A sixth step of retarding the ignition timing data (θADV (BASE)) based on the retarded ignition timing data (θADV)
and a seventh step of setting the ignition timing based on the method.

According to this method, by comparing an analog electrical signal that is a knock vibration with a predetermined knock judgment value, it is possible to determine whether the analog electrical signal has a pulse number corresponding to the frequency of vibration of the analog electrical signal having a magnitude greater than or equal to the knock judgment value. The pulse signal is converted into a pulse signal, counted for each ignition cycle, the counted value (NP) is detected as knock strength, and the ignition timing is retarded to retard the ignition timing based on this counted value (NP). Since the angle correction value (ΔθADV1) is determined, it is possible to accurately detect the knock strength and obtain highly accurate ignition timing.

Hereinafter, these inventions will be described in detail based on embodiments shown in the drawings.

FIG. 1 shows an ignition timing control device that controls ignition timing among engine control devices that perform various engine controls. In the figure, a CPU 12 is a central processing unit that performs digital calculation processing of various data such as engine ignition timing, and a ROM 14 is a storage element for storing control programs such as an ignition timing control program and fixed data. RAM 16 is a readable and writable storage element. The input/output interface circuit 20 receives signals from various sensors,
It is sent to the CPU 12 via the bus 18, and also sends an ignition timing signal IGN generated based on the control signal arithmetic processed by the CPU 12 to the ignition coil drive circuit 60. In the case of this embodiment, there are a knock detection device 30 and a crank angle sensor 40 as the sensors that first output a pulse train signal.
In addition, the sensors that output the second analog signal include the battery voltage detection sensor 42 and the load sensor 5.
There is 0.

The output signal of the crank angle sensor 40 is directly input to the input/output interface circuit 20, while the output signals of the battery voltage detection sensor 42 and load sensor 44 are received by the control signal from the input/output interface circuit 20. The signal is converted into a digital signal by a multiplexer 46 in a time division manner by an analog-to-digital converter (hereinafter referred to as ADC) 48, and is input to the input interface circuit 20.

The knock detection device 30 is configured to generate a pulse KNCKP in response to a knock condition occurring in the engine, and to output a time signal OSMP whose level changes for a certain period of time in response to the first generated pulse. For example, the time signal OSMP is produced by a one-shot multivibrator.

Further, the crank angle sensor 40 and the load sensor 50 are provided to obtain information that becomes the basis for calculating the basic ignition timing θADV. The crank angle sensor 40 outputs a reference crank angle signal (REF) P2 and a position pulse signal (POS) P1, and the load sensor 50 in this embodiment obtains a load signal from the suction pipe negative pressure.

The ignition coil drive circuit 60 is composed of an amplifier 62, a power transistor 64, and an ignition coil 66. The ignition signal IGN from the input/output interface circuit 20 is amplified by the amplifier 62 and then input to the transistor 64, thereby causing ignition. coil 6
6 is controlled, and a predetermined ignition timing is obtained.

FIG. 2 shows a specific configuration of a circuit that contributes to ignition timing control in the input/output interface circuit 20. In the figure, the ignition timing data θADV calculated by the CPU 12 and the ignition coil 66
The energization start timing data θDWL is sent to the corresponding advance register (ADV-REG) 202 and dwell register (DWL-REG) 204 on bus 1.
The input settings are made via 8. The output of the advance register 202 is input to a comparator 206,
Further, the output of the dwell register 204 is input to a comparator 208, and the comparator 206 is connected to the first
The comparator 208 generates an output pulse when the count value of the second counter register 210 reaches the set value of the advance register 202, and the comparator 208 generates an output pulse when the count value of the second counter register 212 reaches the set value of the dwell register 204. generates an output pulse. Comparator 206 outputs an ignition timing pulse, and comparator 208 outputs an energization start timing pulse. The RS flip-flop 214 repeats on and off outputs in response to the outputs of the respective comparators 206 and 208, and the ignition signal IGN is obtained at the output.

In order to execute the above ignition timing control, the first counter register 210 is designed to count the position pulse signal P1 input from the AND gate 216, and the AND gate 216 is set with the reference crank angle signal P2. It is set to open at the Q output of the RS flip-flop 218. The first counter register 210 is reset by the reference crank angle signal P2, and the RS flip-flop 218 is reset by the output of the comparator 206. That is, from the state in which the RS flip-flop 218 is reset by the comparator 206, the reference crank angle signal P2 is
When the RS flip-flop 218 enters the set state, the AND gate 216 is opened by the position pulse signal P1, and the first counter 210 is set to be able to count the position pulse signal P1 until the reference crank angle signal P2 is generated. There is.

The second counter register 212 also counts the position pulse signal P1 via an AND gate 220, but the conditions for opening the AND gate 220 are different from those of the AND gate 216. That is, when the Q output of the RS flip-flop 222 is set by the output of the comparator 206, which resets the RS flip-flop 218, the AND gate 220 is set.
is set to open, and the RS flip-flop 2
The reset of 22 is performed by the output of comparator 208. Therefore, after the comparator 206 generates an output, the second counter register 212 counts the position pulse signal P1 until the value set in the dwell register 204.

In addition, the output pulse of the knock detection device 30
KNCKP and OSMP are input to the counter register 234 via an AND gate 232, and a knock signal generated within the time of the time signal OSMP
KNCKP is counted. At the end of this count, the CPU 12 is interrupted by the status register 236, and at the same time the count value of the counter register 234 is fetched into the CPU 12 via the bus 18, the count value of the counter register 234 is cleared and the next count is started. Prepare for an outbreak. The number of pulses NP taken into the CPU 12 is data corresponding to the strength of the knock, and is used for calculating the ignition timing correction amount _{Δθ ADV1 =} f(NP).

The stator register 236 essentially consists of one flip-flop, and is set at the trailing edge of the time signal OSMP to serve as an interrupt processing signal, which is taken into the CPU 12. The counter 234 is reset by the output of a reset register 238.
A reset signal is input to the counter register 234 according to a command from the counter register 234.

The count value NP taken into the CPU 12 is data corresponding to the strength of the knock. That is, the frequency of the knock is empirically constant at approximately 7KHz, and the greater the intensity of the knock, the longer the duration of the knock. Therefore, the number of knock pulses KNCKP detected during a certain period of time started in synchronization with the start of the knock is proportional to the intensity of the knock. In the present invention, the correction amount Δθ _ADV1 for retarding the ignition timing according to the strength of the knock is determined as Δθ _ADV1 =f(NP).

In the above configuration, the results of the arithmetic processing performed by the CPU 12 according to the ignition timing control program are set in the respective registers 202 and 204, and the predetermined ignition signal IGN is output from the RS flip-flop 214.
obtained from. Also, when a knock occurs, an interrupt request is generated using the output pulse OSMP,
The count value of the pulse generated according to the intensity of the knock for a predetermined period of time is taken into the CPU 12, and a correction value is obtained.
Correct the ignition advance angle. After the correction, the ignition timing is controlled to increase stepwise in the advance direction in minimum correction units.

FIG. 3 is a timing chart showing the operation of the circuit shown in FIG. 2 described above. In the figure,
A is a reference crank angle signal P2, and B is a position pulse signal P1. C indicates the counting status of the first counter register 210, and C1 is the setting value of the advance register 202. D indicates an output signal of the comparator 206, indicating that an output is generated when the count value of the first counter register 210 reaches the set value of the advance register 202. E indicates the counting status of the second counter register 212, and E1 is the set value of the dwell register 204. F is the output of the comparator 208,
The operation is similar to that of comparator 206. G is an RS flip-flop 214 which responds to the outputs of comparators 206 and 208, that is, the pulses shown at D and F.
shows the output of H indicates the ignition coil current of the ignition coil 66 flowing in response to this output, and I indicates the ignition timing.

Next, in FIG. 4, the knock detection device 30 responds to the occurrence of a knock and generates a number of pulses corresponding to the intensity of the knock.
A knock detection circuit is shown that generates KNCKP. In the figure, a knock sensor 300 is constructed by winding a pickup coil around a magnetostrictive element, and converts knock vibration of an engine cylinder into an electrical signal. The output signal VIN of this knock sensor 300 is input to a band pass filter 302. This band pass filter 302 is provided to remove parasitic engine vibrations and effectively take out knock vibrations, and includes operational amplifiers 304, 306, capacitors 308, 310, 312, and resistors 314, 31.
6,318,320,322,324,326,
It is composed of 328, 330, and 332. This bandpass filter 302 has a center frequency of
7KHz, and the passband is set to 7±0.5KHz. A circuit including operational amplifier 304 constitutes a filter circuit, and operational amplifier 306 constitutes an amplifier circuit. Since these circuits are generally well-known circuits, detailed explanations will be omitted. The amplitude of the knock signal that has passed through the bandpass filter 302 changes depending on the strength of the knock state, that is, whether it is a light knock, a middle knock, or a heavy knock. The present invention performs control based on the occurrence of a light knock, and performs control so that a knock stronger than the light knock does not occur.

The knock signal that has passed through the bandpass filter 302 is input to an attenuator 340 in order to use it as an optimum value for comparison of sample and hold values. Attenuator 3
40 is an operational amplifier 342, a transistor 344,
Zener diode 346, diode 348,
350 and resistors 352, 354, 356, 35
It is composed of 8,360,362.

Sample and hold circuit 380 includes attenuator 340
A transistor 384, a diode 386, a capacitor 390, and resistors 392 and 39 are provided to sample and hold the peak value of the output signal of
It consists of 4,396. That is, the capacitor 390 is charged to the peak level of the noise that arrives during the period when the transistor 384 is in the on state, and this charged value and the connection voltage of the resistors 360 and 362 of the attenuator 340 are the same as that of the comparator circuit 700. It is compared by operational amplifier 702.

Here, as a second step of converting the knock vibration of the engine of the present invention into an analog electrical signal by the knock sensor 300, a band pass filter 3 is used.
02, attenuator 340 and sample hold circuit 3
80 falls under this category.

Further, by comparing the analog electric signal of the present invention with a predetermined knock judgment value, a third pulse signal having a pulse number corresponding to the frequency of the analog electric signal having a magnitude greater than the knock judgment value is converted.
The comparator 700 corresponds to this step.

Next, this connection point voltage is a voltage obtained by dividing the output voltage of the amplifier 342 by resistors 362, 360, 356, and 358, and is a voltage that is obtained by dividing the output voltage of the amplifier 342 by the resistors 362, 360, 356, and 358, and the output voltage of the amplifier 342 is
2, that is, only when the knock signal is detected, an output is generated in the comparator circuit 700. Therefore, the comparator circuit 700 generates a pulse corresponding to the knock when the intensity of the knock is greater than a predetermined value, that is, greater than the light knock.

A timing circuit 800 takes the operation timing of each of the circuits and uses unnecessary signals to control the comparison circuit 700.
The transistors 802, 804, 80 are provided to prevent the transistors from generating a pulse output.
6,808,810, diode 812,81
4,816,818,820,821, capacitor 822,824,826, resistor 828,83
0,832,834,836,838,840,
It consists of 842, 844, and 846. The transistor 802 of this timing circuit 800 has an ignition timing signal circuit 90 that generates an ignition timing signal.
An output of 0 is input.

A diode 9 is connected to the output side of the timing circuit 800 and the comparison circuit 700 along with a resistor 902.
04 constitutes a masking circuit, which turns on the transistor 802 for a predetermined time after the ignition timing, and grounds its connection point to make the output V _OUT zero, thereby preventing the output from being generated due to ignition noise. This noise is caused by the electromagnetic waves generated by the ignition signal reaching the sensor 300.
, and is superimposed on the output of the sensor 300. A mask is applied to prevent malfunctions caused by this noise.

Transistor 802 and transistor 804 constitute a first one-shot multivibrator, and transistor 802 constitutes an ignition timing signal circuit 90.
When the transistor 802 is turned on for a predetermined time with an ignition timing signal of 0, the transistor 802 is turned on in response to this.
04 turns off. Furthermore, transistor 802 and transistor 806 constitute a second one-shot multivibrator.

Transistor 808 constitutes a sample-and-hold discharge switch, and transistor 810 constitutes a sample-and-hold charge switch.
That is, when transistor 806 is off, transistor 808 is on, and when transistors 804 and 808 are both off, transistor 810 is on. Therefore, when transistor 810 is turned on, transistor 384 of sample and hold circuit 380 is turned on, and transistor 390 is turned on.
holds the sample. Further, when the transistor 808 is turned on, the capacitor 390 is discharged via the resistor 396 and the diode 821.

FIG. 5 shows the above operation timing, where A is the ignition timing signal output by the ignition timing signal circuit 900, B is the collector voltage of the transistor 802 which is turned on by the ignition timing signal, and C is the collector voltage of the transistor 804. It is voltage. D is the collector voltage of transistor 806. Transistor 806 and transistor 804 are turned off at the same time, but due to the subsequent voltage drop across resistor 832 due to the charging current to capacitor 824, transistor 8
06 turns on earlier. E is transistor 8
10 base voltage. This voltage becomes a high level only when the transistor 808 is off, that is, the waveform D is at a low level, and the collector voltage C of the transistor 804 is at a high level.

F indicates the output signal waveform of the knock sensor 300, F0 is a signal other than a knock signal, F1 is a late knock, and F2 is a middle knock. Ig is ignition signal noise. G is a bandpass filter 302
The output signal waveform of the knock sensor 300 that has passed through has an amplitude of a frequency of 7KHz.

H is the sample voltage in the sample and hold circuit 380, H1 is the discharge timing of the sample and hold corresponding to the high level period of waveform D, that is, the on period of transistor 808, and H2 is the sampled voltage of waveform E.
This is the sample and hold timing corresponding to the high level period of , that is, the period when the transistor 810 is on. The level of this waveform H corresponds to the level H in the waveform G, and the comparator circuit 700 outputs only the portion of the waveform G that exceeds this level H'.

I is the output V _OUT of the comparator 700 and is the knock signal waveform, I1 corresponds to the light knock, I2 corresponds to the middle knock, and it can be seen that the number of pulses increases according to the knock intensity.

Next, an ignition timing control program executed by the ignition timing control device described above will be explained based on FIG. 6 and subsequent figures.

FIG. 6 shows the positioning of ignition timing control and knock signal processing routines in various engine program controls.

In the figure, when an interrupt 500 occurs, the interrupt request is sent to the interrupt request NMI in the next step 502.
or another interrupt request IRQ. interrupt request
If it is an NMI, a knock signal processing routine is executed in step 504, while the interrupt request is an IRQ.
If it is determined that the interrupt is a timer interrupt (TIMER) or a reference interrupt (REF), the process moves to step 506, and the interrupt analysis process determines whether the interrupt is a timer interrupt (TIMER) or a reference interrupt (REF).

A timer interrupt is an interrupt that is used to start various tasks to be executed based on a program, starting from the task with the highest priority, at a fixed time period, preferably at a period of 10 msec. It is included. On the other hand, a reference interrupt is an interrupt for a task that is activated at every fixed rotation angle of the crankshaft.

If it is a timer interrupt, step 508
In the task scheduler, various tasks 510, 5
The execution order of each task is determined based on the priority determined in No. 12, 514, and 516. Task 51
0 is for various sensors, such as suction pipe negative pressure analog
Digital conversion A/D and rotation speed are taken in, task 512 is ignition timing (ADV) and ignition coil energization start timing (DWL) control, task 514 is starter switch and idle switch monitor control, task 516 is correction processing. . And 1
As a result of the execution of the two tasks, a task completion report is made to the task scheduler 508 in step 518. Then, task scheduler 50
8 instructs execution of the next priority task.

If the result of the interrupt request analysis process in step 506 is that it is a reference interrupt, the process proceeds to step 520.
Then, rotation interrupt processing is performed. Note that after completing steps 504 and 520, step RTI
The state is returned to the state before the interrupt.

FIG. 7 shows the knock signal processing routine 504.
It is shown. In the figure, task launch 522
Accordingly, in step 524, the stator register 236, which was set when the counter register 234 ended, is reset. Step 52
6, the contents of the counter register 234 NP
is captured, and after the capture, the counter register is cleared in step 528. The number of pulses taken in NP is a value corresponding to the intensity of the knock, and from this value, in the next step 530, the ignition timing correction amount Δθ _ADV1 is calculated from Δθ _ADV1 = f(NP). However, in general, Δθ _ADV1 is determined by table search instead of calculation. This table is ROM14
It is stored in , and has the following value.

Table 1 NP 1 2 3 4 5 or more Δθ _ADV1 0゜ 1゜ 1゜ 1゜ 2゜ The correction amount Δθ _ADV1 searched in step 530 is
is stored in RAM 16, and then in step 532
At step 534, a flag instructing execution of the ignition timing correction execution report, that is, execution of the ignition timing correction operation, is set, and at step 534, the task is completed and the state before the interruption is returned.

Here, as a fourth step of counting the pulse signal for each ignition cycle of the present invention and detecting the counted value (NP) as the knock intensity, step 5 is performed.
24,526,528 are applicable.

In addition, a fifth part determines a retard correction value (ΔθADV1) for retarding the ignition timing based on the count value (NP) representing the knock strength of the present invention.
Step 530 corresponds to this step.

In the above explanation, the number of knock pulses NP is detected by the circuits 232, 234, 2 in the interface 20.
The case where this is done in terms of hardware using 36 and 238 has been described. A method of detecting the number of knock pulses NP using only software instead of such a circuit will be described next.

Knock pulse from knock detection device 30
KNCKP passes through the track indicated by the broken line in Figure 1.
It is supplied to the CPU 12. The CPU 12 generates a non-maskable highest priority interrupt request NMI in synchronization with the first pulse of the knock pulse KNCKP, and the RAM 1
The software timer (hereinafter referred to as S timer) provided in 6 is started, the number of pulses of the pulse train signal KNCKP received during the timing period of the S timer is counted, and the optimum ignition timing is determined based on this counted value. Performs calculation processing for control.

A knock signal processing routine 504 is shown in FIG. In the figure, when an NMI interrupt occurs in synchronization with the first pulse of the knock signal KNCKP from the knock detection device 30 in step 521A,
In step 522A, the counter CNTR (not shown) provided in the RAM 16 starts counting, and the counter CNTR starts counting every time a pulse of the knock signal is input.
The contents of CNTR are incremented by +1 and updated. Then, in the next step 523A, it is determined whether the S timer is set or not based on a flag to be described later. If the S timer has already been set, jump to step 531A;
The program returns to the program execution state immediately before the interrupt at step 500 in FIG. However, since the S timer should not have been set when no knock has occurred, when the first pulse of the knock signal KNCKP is taken into the counter CNTR, the process moves to step 524A, and the S timer set execution report is sent. The S timer flag is set. Then, in step 525A, the S timer is activated by the first pulse of the knock signal KNCKP, and counting of the S timer is started. Next, in step 526A, it is determined whether the contents of the S timer have reached the specified value. This specified value corresponds to the general maximum duration of a write knock, is determined empirically, and is stored in the ROM 14.

If the updated value of the S timer has not reached the specified value, the process returns to step 525A and the S timer is updated.
The S timer is updated by 1, and while the S timer is activated, the S timer is updated in a loop formed by steps 525A and 526A. During the operation of this S timer, a number of pulse train signals KNCKP corresponding to the strength of the knock state are taken into the counter CNTR and counted.

When the contents of the S timer reach the specified value in step 526A, the process moves to step 527A to report the completion of timer execution, ie, reset the flag set in step 524A. Next step 5
At 28A, the ignition timing correction amount Δθ _ADV1 is Δθ _ADV1 = f
Calculated from (NP). Here, the ignition timing correction amount Δθ _ADV1 is a function of the count value NP of knock pulses KNCKP received during activation of the S timer in the counter CNTR. Generally, the correction amount Δθ _ADV1 is found by searching the table.

Furthermore, in step 529A, the counter
Clears the contents of CNTR and S timer, prepares for the next knock, and executes the next step 530.
At A, the ignition timing correction execution report is set, that is, the flag for ignition timing correction execution is set. If the ignition timing correction execution report is made, the process returns to the state before the interruption from step 531A.

Here, as a fourth step of counting the pulse signal for each ignition cycle of the present invention and detecting the counted value (NP) as the knock intensity, step 5 is performed.
22A, 523A, 524A, 525A, 526
A falls under this category.

In addition, a fifth part determines a retard correction value (ΔθADV1) for retarding the ignition timing based on the count value (NP) representing the knock strength of the present invention.
Step 528A corresponds to this step.

The above description will further explain the timer update in steps 525A and 526A and the counter update in step 522A. As mentioned above,
When a knock occurs, an interrupt is generated based on a pulse from the knock detection circuit. This interrupt causes the counter CNTR to be updated in step 522A. Since it is necessary to start the S timer at the beginning of the pulse representing a series of knock conditions from the knock detection circuit,
Initially, the process proceeds to steps 523A and 524A. By setting a flag in step 524A, it is known that the timer is running, and from the next interrupt, the flag is determined in step 523A and the process advances to return-to-interrupt 531A.

In step 525A, the timer counter (STIME) is updated by creating a software reference time elapsed, and in step 526A it is determined whether a certain time has elapsed. In other words, the passage of a certain period of time can be detected in steps 525 and 526A.

During time measurement in steps 525A and 526A, interrupts (NMI) are not inhibited, so subsequent knock pulses from the knock detection device
When an interrupt occurs based on KNCKP, step 5
25A and 526A are interrupted and an interrupt is made to 521A, which is the start of KIR processing. Then, in step 522A, the counter CNTR is updated again.
In this case, since the flag was previously set in step 524A, steps 523A to 531A are executed.
The point where it was jumped to and interrupted by an interrupt, that is,
Return again to steps 525A and 526A. Therefore, YES in step 523A is step 5.
This is a case where an NMI interrupt occurs during execution of 25A and 526A. When a certain period of time has elapsed, the process advances to step 527A as described above.

FIG. 9 shows the ignition timing control routine 51 of FIG.
2, and this ignition timing control routine is configured to correct the basic ignition advance angle θ _ADV (BASE) by the ignition timing correction amount Δθ _ADV1 and the normal correction amount Δθ _ADV2 . In the figure, from task start 540 to step 542, the basic ignition timing θ _ADV (BASE) is set.
is calculated from θ _ADV (BASE) = f (N, L).
Here, N is the engine speed and L is the suction pipe negative pressure. This calculation has been performed by a method that has been generally used in practice, and is the first step of the present invention that takes in engine rotation information and load information and calculates the engine ignition timing data (θADV (BASE)). It corresponds to the step of

After the calculation, in the next step 544, it is determined whether or not the modification execution by knocking has been completed (Yes, No) based on the flag set in step 532 or 530A. If a knock has occurred and the correction execution has not yet been completed, the flag is set and the result is NO, and the process proceeds to step 546. In step 546, the ignition timing correction amount Δθ _ADV (t) is determined by the following equation.

Δθ _ADV (t) = Δθ _ADV (t-1) − Δθ _ADV1 ...(1) In equation (1), Δθ _ADV1 is the current correction amount and should be added to the basic ignition advance angle θ _ADV (BASE). It's the amount. Δθ _ADV (t-1) is the previous correction amount immediately before correction. Since Δθ _ADV1 is the amount to be retarded, it is subtracted from Δθ _ADV (t-1).

Ignition timing correction amount Δθ _ADV found in step 546
(t) in step 548.
It is added to the basic ignition advance angle θ _ADV (BASE) obtained in step 42 as shown in the following formula.

θ _ADV (t) = θ _ADV (BASE) + Δθ _ADV (t) ...(2) As a result of the correction in step 548, the next step 5
At step 50, report the completion of modification execution, that is, at step 532.
Alternatively, the flag set at 530A is reset.

In step 552, ignition timing data θ _ADV is obtained from the following equation and is set in the advance register 202.

θ _ADV = θ _REF - θ _ADV (t) ...(3) Here, θ _REF is the crankshaft angle from the reference crank angle signal P2 to the engine top dead center TDC, as shown in FIG. It is a fixed value.

Here, as a sixth step of retarding the ignition timing data (θADV(BASE)) based on the retardation correction value (ΔθADV1) of the present invention, step 54 is performed.
6,548 fall under this category.

Furthermore, step 552 corresponds to the seventh step of setting the ignition timing based on the retarded ignition timing data (θADV).

After the ignition timing data θ _ADV is set in the advance register 202 as described above, step 5
At step 54, an energization start timing control routine for setting energization start timing data θ _DWL in the dwell register 204 is executed, and the task ends at step 518.
Go to. Since the method of calculating θ _DWL is already known, it will not be explained here.

On the other hand, if it is determined in step 544 that the correction execution by knocking has been completed, the process proceeds to step 55.
In step 7, the correction amount Δθ _ADV2 that depends on the engine speed N is determined. This correction amount is determined at a predetermined period, e.g.
This represents the amount by which the ignition advance angle is advanced every 20 msec, and is the smallest unit of advance angle correction amount. This correction amount Δθ _ADV2 is set to be 0.4°/sec when the rotational speed N is N1000 rpm. In the preferred embodiment, the advance period is 20 msec, so the correction amount Δθ _ADV2 is 0.008°. Also, the rotation speed N is N<
At 1000 rpm, Δθ _ADV2 is set to 1°/sec, that is, 0.02°. The reason why the advance angle correction speed is reduced in the high-speed rotation range is that the higher the rotation speed, the greater the background noise that enters the knock detection circuit, making knock detection that much more difficult.

In step 558, the ignition timing correction amount Δθ _ADV
(t) is obtained from the following equation.

Δθ _ADV (t)=Δθ _ADV (t-1)+Δθ _ADV2 (4) In equation (4), Δθ _ADV (t-1) is the amount of correction immediately before the correction. The ignition timing correction Δθ _ADV (t) determined at step 558 is added to the base ignition timing θ _ADV (BASE) determined at step 542.
That is, the ignition timing θ _ADV (t) is obtained from θ _ADV (t)=θ _ADV (BASE)+Δθ _ADV (t) (5).

As is clear from this equation (5), after the knock disappears, the ignition timing is retarded according to the strength of the knock due to the occurrence of the knock, and the ignition timing is retarded in the minimum correction unit every time task 512 is executed. The ignition timing is advanced in steps of Δθ _ADV2 .

Ignition timing θ _ADV (t) determined in step 560
is set in the advance register 202 in step 552, and then the energization start timing control routine is executed in step 554 as described above, and the process proceeds to task end step 518.

As a result of the ignition timing being controlled as explained above, the knock extinguishing process is carried out in accordance with the strength of the knock at the same time as the knock occurs, and after the knock extinguishing process, the ignition timing is increased in the advance direction by the minimum correction unit. The ignition timing is always set to the optimum state with high precision just before the engine starts, improving driving performance and maintaining maximum engine output, which is extremely beneficial in terms of improving engine efficiency and reducing exhaust emissions. . Particularly, it is extremely effective for engines equipped with a turbocharger.

In the embodiment configured to perform time management using software, such as setting the acquisition time of the number of pulse train signals corresponding to the intensity of the knock state output from the knock detection device, no hardware part for time management is required. Therefore, the number of parts is reduced and costs can be reduced. Furthermore, in this embodiment, the knock signal is directly input to the CPU without going through the input/output interface circuit, and an interrupt is generated. Accordingly, according to the present invention, the ignition timing can be corrected after the knock occurs. The processing time until control starts is shortened, and controllability can be improved.

Steps 557 to 560 in FIG. 9 show an example in which the ignition timing is advanced in steps at a predetermined time period, but instead of this, in the rotation interrupt processing routine 520 in FIG. A similar effect can be performed at every predetermined angle of rotation of the engine, preferably during the ignition cycle. Routine 5 in this case
Some contents of 20 are shown in FIG. Each operation of steps 557A, 558A, 560A, and 552A in the figure is equivalent to step 55 in FIG.
7,558, 560 and 552, so a duplicate explanation will not be given here. In this embodiment, steps 557 and 55 in FIG.
8, 560, and 552 do not exist, and therefore, when it is determined ``YES'' at step 544, as shown by the broken line, the process jumps to step 554.

FIG. 11 shows another embodiment of the ignition timing control routine 512. This ignition timing control routine is the same as the ignition timing control routine shown in FIG. 9, with a control step based on the engine rotation speed added thereto to eliminate the adverse effects of background noise in the cruising rotation range. That is, after step 542, step 600 is provided to set the engine rotation speed N to a preset value NK (NK>
1000rpm) or higher (Yes, No),
If it is less than the preset value NK (No), the ignition timing control routine is the same as the one shown in FIG. 8, but if it is greater than the preset value NK (Yes), the ignition timing correction amount Δθ is _ADV (t)=0
and stops the advance angle operation. As a result, in step 604, the ignition timing θ _ADV (t) is changed to the basic ignition timing θ _ADV
(BASE) only, and from this value step 55
In step 2, ignition timing data θ _ADV is determined and set in the advance register. The other steps are the same as the routine shown in FIG. 9, so their explanation will be omitted.

Next, the results of the ignition timing control of this embodiment will be explained with reference to FIG. In the figure, A indicates the change in engine rotational speed N (rpm) from starting the engine to starting the vehicle and driving the vehicle, and B indicates the vehicle speed. C indicates the advance angle (°) in the ignition timing, and the vertical axis indicates the angle from top dead center. The change in the ignition advance angle indicates the execution of knock control.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing an embodiment of the engine ignition timing control system according to the present invention, FIG. 2 is a block diagram showing the interface circuit configuration in FIG. 1, and FIG. 3 is the circuit shown in FIG. 2. 4 is a circuit diagram showing a specific circuit configuration of the knock detection device shown in FIG. 1, and FIG. 5 is a waveform diagram of the main part to explain the operation of the circuit shown in FIG. 4. , FIG. 6 is a flowchart showing the overall procedure of the engine control system according to the present invention, and FIGS.
9 is a flowchart showing the knock signal processing routine shown in FIG. 6, the former using a hardware circuit, the latter using only software, and FIG. FIG. 10 is a flowchart showing the ignition timing control routine of
FIG. 11 is a flowchart showing an alternative example to the example shown in FIG. 9; FIG. 11 is a flowchart showing another embodiment of the ignition timing control routine shown in FIG. 6;
FIG. 2 is a diagram showing the measurement results of the apparatus implementing the present invention. 12...CPU, 16...RAM, 14...ROM, 3
0: Knock detection device.

Claims (1)

[Scope of Claims] 1. Engine rotation information and load information are taken in and ignition timing data (θADV
(BASE)); the knock vibration of the engine is detected by the knock sensor 30;
A second step 302, 340, 380 of converting the analog electrical signal into an analog electrical signal by 0; A third step 700 of converting the pulse signal into a pulse signal having a number of pulses corresponding to the vibration frequency; a fourth step of counting the pulse signal for each ignition cycle and detecting the count value (NP) as the knock intensity.
Steps 524, 526, 528, 522A,
523A, 524A, 525A, 526A; a fifth step 530 of determining a retardation correction value (ΔθADV1) for retarding the ignition timing based on the count value (NP) representing the knock strength;
528A; Sixth step 546, 548 of retarding the ignition timing data (θADV(BASE)) based on the retardation correction value (ΔθADV1); retarding the ignition timing data (θADV);
A seventh step 55 of setting the ignition timing based on
2. An engine ignition timing control method comprising: