JPS5967407A - Processing device for signal of engine rotational angle sensor - Google Patents

Abstract

PURPOSE:To enable the prevention of a false operation even when an interrupt process is made to stand by a method wherein a timing at which a reference position signal is latched and delivered to CPU is determined according to a rotational angle signal delivered after a time when the reference position signal is actually generated. CONSTITUTION:The titled device comprises D type FF81 for memorizing the generation of a reference position (REF) signal, FF82 taking in an output Q1 of said FF81 at the timing of the subsequent rotational angle (CA) signal, and a NAND gate 83 clearing FF81 at an input stage when an output Q2 of FF82 reaches an H level while the CA signal continues. Since CPU takes only the Q2 output of a latch 8 into account, CPU10 generates no REF signal practically until a 0-th CA signal is generated. Therefore, even when on n+1-th CA interrupt is made to stand by a time passing over the timing whereat the REF signal is generated, said interrupt is never misread as the 0-th CA signal. As the result, a false operation of misreading the CA signal as the REF signal can be prevented even when the interrupt process of the CA signal is made to stand by.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a sensor signal processing device that stably detects a reference position of a rotation angle.

An engine rotation angle sensor (CA sensor) is essential in an engine control system using a microprocessor. This sensor generates a rotation angle signal (CA double signal) indicating the engine speed and rotational position, and a reference position signal (REF signal) indicating the control start position.There are various methods for CA sensor, but A magnetic detection method that utilizes changes in magnetic flux as the engine rotates is common.Figure 1 is an explanatory diagram of this system, where ■ is a crankshaft, 2 and 3 are toothed rotors attached to the shaft, and 4 and 5 are detection coils that convert magnetic flux changes accompanying the rotation of rotors 2 and 3 into voltage changes.These rotors and coils constitute a sensor,
The output of coil 4 is the REF signal shown in Fig. 2 (al), and the output of coil 5 is the CA multiplied signal shown in fbl.Actually, (
δ1. (tc+ which shaped the waveform of bl, (di pulse is used as REF signal and CA multiplied signal.

When the rotor 2 is provided with teeth 2a at only two locations, the REF signal is generated once every 360 degrees of crank angle.

On the other hand, the rotor 3 has a plurality of teeth 3a arranged at equal angles LJ.
Therefore, if 24 i'Ji3a are installed, CA
The double signal occurs once every 30 degrees. Generally REF
The signal is not used as it is as a reference position. This is because the relative position accuracy when attaching the rotors 2 and 3 to the shaft I is not good. Instead, the current is applied to the CA double signal reference position immediately after the REF signal. In a vehicle, the sensor is mounted so that this reference position is at a constant angle with respect to the dead center position of the specific cylinder. Even if the REF signal is not used directly as a reference position, the current energization method is suitable for rotor 2.
The tooth 2a of the rotor 3 is set to be located between the adjacent teeth 3a, 3a of the rotor 3. However, in this case as well, a certain tolerance λj is allowed (for example, ±8°), so υi
The angle between 2a and 3a is not necessarily 15°.

Fig. 3 shows a conventional signal processing device that calculates a reference position from the RE F signal and the CA multiplied signal.
A comparator for shaping with ref2, 8 a latch set by the output REF of the comparator 6, 9 an interrupt flag set by the output CA of the comparator 7, 10 a CPU that calculates the reference position from the contents of the latch 8 and flag 9.
(Central processing unit). Figure 4 shows the timer 1, which shows the operation, and the number of the CA double signal is ``n'',
n4-1.0.1°2... is the number (CA-No) of each CA with the reference position being 0. This CA-N
o can be initialized using a soft counter in CPU10. cpui.

When the interrupt flag 9 is set, it is cleared at the timing of ■. Further, the contents of the latch 8 are monitored by the above-mentioned interrupt processing program, and if it is set to 1-, it is cleared at the timing (2). FIG. 5 is a processing flow of CA interrupt.

In such a signal processing device, (1) the relative accuracy of the REF position with respect to the CA position is poor; (2) even if a CA interrupt occurs during processing of another interrupt, it will wait until the processing of the former is completed;
■ Malfunctions may occur under adverse conditions such as when the engine is running at high speeds. This will be explained with reference to FIG. The figure shows (n+
1) The manufacturing tolerance (±8
)'s J & Bad Mf Te] The case where the angle is narrowed to 5° - 8° - 7° is taken as an example. CA multiplication signal time 1ζ with angle difference of 30°? A varies depending on the rotation speed, and is shortened to 833 μs at 0000 rpm, so 7° corresponds to 190 μs. If CPtJIO generates the (n+1)th CA multiplication signal while processing another interrupt, the CPU 10 can see the flag only after completing the other interrupt processing. For this reason, if processing other interrupts takes 200 μs, for example, the REF signal is already generated before processing the CA interrupt, and r2F, F8 is senored. In 1 L, CPU 10 is the (n→-1)th C
It will be regarded as the signal to C after the A-fold signal and the REF signal are generated (if it is positive, it will be the O-th signal). This malfunction causes a 30° error in angle control.

The present invention improves the launch circuit for the REF signal.
This is intended to prevent double malfunctions, and its characteristics are that the rotation angle signal obtained from the engine rotation angle sensor is used to interrupt the central processing unit, and the reference position signal obtained from the sensor is used to interrupt the central processing unit. Separate the circuit L-L
, and when it is determined that a launch output of the reference position signal is generated from the latch circuit during interrupt processing of the rotation angle signal by the central processing unit, the rotation angle signal position to which the separation is applied is set to the engine rotation. In a signal processing device that is regarded as a reference position of a corner, the launch circuit includes a first circuit that stores the generation of a position signal for the nucleus /-(; The present invention also includes a second circuit that outputs an output to the central processing unit in synchronization with a predetermined rotation angle signal.

FIG. 7 is a diagram explaining the principle of the present invention. In the present invention, REF
The launch timing of the signal is delayed not to the actual generation time of the REF signal but to the generation time of the CA multiplied signal immediately thereafter.

(C1 in the figure is a conventional method of latching at the time of generation of the REF signal, but in the present invention, it is changed to (di). In this way, the (n+1)th CA interrupt is processed by another interrupt process. Even if you are kept on standby (indicated by the dashed line)
, the (n+1)th OCA interrupt has been processed by the time the Oth CA multiplied signal is generated and the REF signal is latched, so the CA multiplied signal immediately after the REF signal is latched is always the zeroth CA multiplied signal.

Next, an embodiment of the present invention will be described with reference to FIG. This figure shows the 8 portions of the lunch that are improved in Fig. 3. The latch 8 in this example is a D-type flip-flop (the first
za) 81 and a D-type flip-flop (second circuit) 82 that takes in the output Q+ of the flip-flop 81 at the timing of the next CA Shingetsu (here, consider the 0th circuit).
, the flip-flop 82 is activated during the period when the CΔ other signal remains
It consists of a Nant gate 83 that clears (CLR+) the flip-flop 81 at the input stage when the output Q2 (latch output) becomes )l level. FIG. 9 is a time chart. Since CPLIIO sees 4J, which is the Q2 output of latch 8 (it does not see Q+ small output), it is equivalent to the REF signal not being generated for CPU10 until the 0th cA (g) is generated. Therefore, as shown in FIG. Actually R as in the example
Until the time exceeding the timing at which the EF signal is generated (n+
1) Even if the th CA interrupt is put on standby, there is no need to worry as it is treated as the 0th CA multiplication signal.

Note that the output Q 2 =H of the flip-flop 82 does not remain even when the first and subsequent C persons are included (malfunction occurs). In the example of FIG. 8, 02 receives the output Q + = L from the previous stage and becomes Q 2 = L at the timing of the first OCA signal. Therefore, if it is assumed that Q 2 =H during the first CA multiplication signal interrupt processing, a malfunction will occur. However, the Q2 output of flip-flop 82 is
(Switching from to L is fast (on the order of several tens of ns),
On the other hand, it takes about 10 μS of power supply for CPU10 to accept the first CA interrupt and actually start processing, so by then Q 2 = L.
No problem.

The flip-flops 81 and 82 can also be cleared from the CPU 10. In this case, it may be performed immediately after processing the CA interrupt, so it is not necessarily necessary to wait until the first CA multiplication signal timing. However, if the gate 83 is used as shown in FIG. 8, the burden on the CPU 10 will be lightened, and there is an advantage that one i/stone boat for clearing can be omitted. Furthermore, if the reference position is set to the first CA multiplied signal, the flip-flops 82 may be connected in (i + 1) stages.

FIG. 1O is an explanatory diagram of the maximum rotational speed that can be achieved when R=I by the signal processing device of this example. The same figure shows CA/rib-23 C
It is assumed that the processing time for other interrupt processing that started immediately before the generation of the signal is 200 μs, and that the processing time for subsequent (CA-No = 23 C person separate processing) also takes 200 μs. .The total h1 value of both processing times is 4.
Since it is 00 μs, CA・Nδ−11 and CA・N′i
5 = It cannot cope with the rotation speed where the interval between O is less than 40Otts. However, this rotation speed N is 8000 rpm, which is considered to be the mechanical limit for energization.
m is a value that can be sufficiently covered. surface, vehicle C
Other interrupt processing that causes the A interrupt to wait is processing that receives the completion of an A/D converter that converts the sensor output of another channel into a digital value, captures that data, and further switches the channel.

As described above, according to the present invention, the reference position signal REF and rotation angle signal %cA from the engine rotation angle sensor are transmitted to the CPU.
In a signal processing device that determines the reference position of the engine rotation angle based on the reference position signal REF, the timing at which the reference position signal REF is latched and output to the CPU is set based on the rotation angle signal CA after the time when the reference position signal REF is actually generated. Since the reference position signal REF is determined by
Interrupt processing of rotation angle signal CA immediately before L is on standby.-d
Even in the case where the rotation angle signal CA is changed, it is possible to prevent an erroneous operation in which the immediately preceding rotation angle signal CA is mistaken for the reference position.

[Brief explanation of drawings]

Figure 1 is a configuration diagram of a magnetic detection type engine rotation angle sensor, Figure 2 is its output waveform diagram, Figure 3 is a configuration diagram of a conventional sensor signal processing device, Figure 4 is its operating waveform diagram, and Figure 5 The figure is a flowchart showing the rotation angle signal interrupt processing, Fig. 6 is a time chart 1 of malfunction due to other interrupt processing, Fig. 7 is a time chart I showing the principle of the present invention, and Fig. 8 is a time chart of the present invention. FIG. 9 and FIG. 10 are time charts showing the main g1 configuration diagram and FIG. 10 showing one embodiment. In the figure, 2 to 5 are engine rotation angle sensors, 8 is a reference position signal launch circuit, 9 is a rotation angle signal interrupt ij flag, 1
0 is a central processing unit, 8I is a first circuit, 82 is a second circuit, and 83 is a clearing circuit. Applicant Fujitsu Ten Ltd. (and 1 other person) Representative Patent Attorney Minoru Aoyagi Figure 1~”! i Reference A 11 Reference position 1 Fig. 3 Fig. 4 Reference position 1 Fig. 5 = 'H・9 Reference A position 1 ゛4 para/+・7 So 7 Fig. 6 Near°'. Clear■ ---111---t----
-----1111---Figure 7 CA control (e) 7ri・---
Figure 1, Figure 8, Figure 9, Figure 10

Claims (1)

[Claims] The rotation angle signal received from the engine rotation angle sensor is used to interrupt the central processing unit, and the reference position signal obtained from the sensor is used to control the launch circuit. When it is determined that a latch output of the reference position signal is generated from the launch circuit during interrupt processing, the signal processing device considers the rotation angle signal position at which the interrupt occurs as the reference position of the engine rotation angle. In,
The launch circuit includes a first circuit that stores the occurrence of the reference position signal.
and a second circuit that outputs the output of the first circuit to the central processing unit in synchronization with a predetermined rotation angle signal after the reference position signal is generated. Signal processing device for engine rotation angle sensor.