Fuel Supply System for Internal Combustion Engines

The invention relates to a fuel supply system for internal combustion engines which includes a fuel container which is kept filled with fuel up to a constant level and which communicates through a tube with the induction tube of the engine. The amount of fuel which is metered out to the air aspirated by the engine is determined by the difference in the pressure of the fuel container and of the induction tube. The pressure in the fuel container can be altered by means of solenoid valves which function under the control of the intermittent sensor voltage of an oxygen sensor located in the exhaust line. These solenoid valves are disposed in the air conduits which lead to the air space above the fuel in the fuel container and permit connecting this air space with different portions of the induction tube in which different pressures prevail.

In order to meet the technical requirements of present day engines, fuel supply systems for internal combustion engines must automatically provide an appropriate fuel-air mixture under all operational conditions of the engine to permit complete combustion and to reduce, as much as possible, any toxic components in the exhaust gas while maintaining maximum power or least fuel consumption. For this purpose, the fuel quantity which is metered out to the engine has to be adapted extremely exactly to each and every operational state of the engine. Thus, the most favorable ratio of air to fuel must be changeable in dependence on motor variables, especially exhaust gas values, and in the fuel metering system described above, this change is effected by changing the pressure in the fuel chamber.

It is a principal object of the invention to provide a fuel metering system of the general type described above in which the change of the pressure in the fuel chamber may be made in a manner not requiring expensive structural elements and in a reliable manner.

This object is attained according to the invention by providing that the electromagnetic solenoid valves which are disposed in each air conduit which connects the air space above the fuel chamber with other portions of the engine are cycled in opposing phase.

A favorable feature of the invention is that the basic setting of the fuel-air mixture is made relatively rich and by providing that the fuel-air mixture ratio is controlled in such a manner that, when the sensor voltage of the oxygen sensor exceeds a certain predetermined threshold, the air space above the fuel chamber is connected through a first solenoid valve with the induction tube region in which a low pressure prevails, while, when the sensor voltage falls below the threshold, a second valve opens a conduit which connects the air space in the fuel chamber with the induction tube region in which a higher pressure prevails. It is another feature of the invention that the opening time of the solenoid valves is constant and that the valves are actuated by signals derived from the ignition system of the engine. Yet another feature of the invention is that the opening duration of each valve may be increased by a fixed factor if such a valve receives two sequential opening pulses from the ignition system.

In yet another favorable aspect of the invention, the valves are actuated cyclically and the duty cycle of each valve is made proportional to the output voltage of an integrating circuit which is part of an electronic controller. The input of the integrating circuit is provided in known manner with the oxygen sensor voltage and the output voltage of the integrator increases as long as the sensor voltage exceeds a predetermined threshold, while it decreases when the sensor voltage is smaller than the predetermined threshold. The output voltage of the integrator may be changed cyclically by providing that any change indicated by the sensor voltage is actuated by one of the ignition pulses and takes place over a predetermined length of time, whereas after that predetermined time, the output voltage of the integrator remains constant until the next trigger pulse from the ignition.

Another feature of the invention provides that the sum of the opening times of the cyclically controlled valves is constant and further that when the pressure sources of the engine pulsate, the sum of the opening times of the valves is made smaller than the pulse period and extends over a region which includes the highest pressure difference of the two sources of pressure for the air space in the fuel chamber.

The invention will be better understood as well as further objects and advantages thereof become more apparent from the ensuing detailed specification of four exemplary embodiments of the invention taken in conjunction with the drawing.

FIG. 1 is a diagram of a first exemplary embodiment of a fuel supply system according to the invention;

FIG. 2 is a second exemplary embodiment of the invention;

FIGS. 3-9 are diagrams which include timing information related to various possibilities for controlling the fuel supply system according to the invention;

FIG. 10 is a schematic diagram of a third exemplary embodiment of the fuel supply system of the invention; and

FIG. 11 is a schematic diagram of a fourth exemplary embodiment of the invention.

Turning now to FIG. 1, there is seen a portion of an induction tube 1 of an internal combustion engine, including an air flow control member 2 and an arbitrarily settable throttle valve 3. The air flow rate control element 2 has a needle-like extension 4 which terminates in a fuel metering location 5 of a conduit 6, thereby controlling the free aperture at the location 5. The conduit 6 extends into a fuel chamber 7 and its end remote from the air flow control element 2 extends below the level of the fuel in the chamber. The air space 8 above the fuel in the fuel chamber 7 communicates through a line 9 with two separate air conduits 12 and 13 which can be obturated by solenoid valves 10 and 11, respectively. The air line 12 leads to the region of the induction tube lying upstream of the air flow rate meter 2, while the air conduit 13 terminates at the narrowest part of the induction tube controlled by the air flow member 2 upstream of the throttle valve 3. In the normal case, when no air flows, the valves 10 and 11 are closed. An electronic controller 14 which includes an integrating operational amplifier exerts control over the solenoids valves and it acts in response to electrical variables which are transduced from operational variables 15 of the engine, for example, the engine rpm and a sensor voltage which is taken from an oxygen sensor 17 located in the exhaust line 16. A suitable valve controller is described in U.S. Pat. No. 3,874,171 whose descriptive portions are hereby incorporated by express reference.

The exemplary embodiment of the invention illustrated in FIG. 2 is substantially similar to that in FIG. 1 with the exception that the air line 13 terminates at the narrowest portion of the venturi cross section 18.

The oxygen sensor 17 disposed in the exhaust line 16 is a little tube closed on one side consisting of a solid electrolyte, for example, sintered zirconium dioxide. Both surfaces of the little tube are covered with evaporated microporous platinum layers which are provided with suitable electrical contacts on which an electrical potential may be impressed. One surface of the tube experiences atmospheric air while the other is exposed to the exhaust gases of the engine.

In known manner, the solid electrolyte becomes conducting for oxygen ions at elevated temperatures such as prevail in the exhaust gas. If the partial pressure of oxygen in the exhaust gas is different from the partial pressure of oxygen in the atmosphere, a potential difference occurs as between the two platinum layers, i.e., between the terminals on the tube, and this potential has a particular characteristic which corresponds to the air number .lambda. which is defined as proportional to the ratio of air to fuel. This potential difference across the two surfaces of the sensor is a logarithmic function of the quotient of the partial pressures of oxygen on the two sides of the solid electrolyte.

Thus the sensor voltage changes abruptly in the vicinity of the point when the air number .lambda. = 1. When .lambda. > 1 unused oxygen will suddenly appear in the exhaust gas. Because the output potential of the oxygen sensor 17 depends very heavily on the air number .lambda., this sensor is very suitable for controlling the above-mentioned solenoid valves 10 and 11. When the air number .lambda. < 1, the sensor potential is high, while it is low when .lambda. > 1.

FIGS. 3-9 are diagrams which are for illustration of the various control possibilities of valves 10 and 11. FIG. 3 is a diagram of a voltage U as a function of t. The upper curve in FIG. 3 indicates the sensor voltage U.sub.s as a function of time and is seen to fluctuate about a predetermined constant value U.sub.o indicated by a dash-dotted line. If the basic setting of the fuel-air ratio delivered by the fuel supply system is made rich and if the sensor voltage U.sub.s is larger than the threshold voltage U.sub.o, the fuel-air mixture is too rich and the valve 11 will be opened so that the air space in the fuel container 7 experiences a pressure decrease and a smaller quantity of fuel is aspirated at the metering aperture 5. If the sensor voltage U.sub.s drops below the threshold U.sub.o, the valve 11 is closed and the valve 10 is opened so that the air space in the fuel container is connected with that portion of the induction tube in which a higher pressure prevails so that, due to the greater pressure difference at the metering aperture 5, a larger amount of fuel is aspirated and the fuel-air mixture is thereby enriched. In this manner, the pressure in the air space 8 of the fuel container 7 is changed until the mixture is such that the air number .lambda. is approximately 1 and such a mixture has been shown to be particularly favorable and corresponds to a stoichiometric mixture of air and fuel.

The valve operating voltages U.sub.11 and U.sub.10 are shown in the two lower diagrams of FIG. 3. Another variant possibility of controlling the valves is indicated in FIG. 4, in which the opening pulses for the valves are the ignition pulses which may also be derived from rpm signals and wherein the opening time t.sub.o of the valves 10 and 11 is constant.

FIG. 5 shows yet another type of valve control in which the opening time t.sub.o of each valve 10, 11 is increased by a predetermined factor, for example, doubled in case this valve is opened consecutively by at least two sequential ignition pulses. Thus for example, if two opening pulses for the same valve occur in an arbitrarily settable time span t.sub.s, the opening time of that particular valve may be doubled, for example at the occurrence of a pulse after a time t.sub.s. Since the information about the magnitude of deviation from the air number .lambda. cannot be derived directly from the air sensor voltage, a repeated opening of a valve is assumed to imply a large deviation of the air number .lambda. from its nominal value and it is thus compensated for by a more rapid control due to a prolongation of the opening time of that valve.

In the variant control methods illustrated in FIGS. 6 to 9, the valves 10 and 11 are cycled in opposite phase. The duty cycles defined by T.sub.10 = t.sub.10 /(t.sub.10 + t.sub.11) and T.sub.11 = t.sub.11 /(t.sub.10 + t.sub.11) define the periods of time in which the air chamber 8 is connected to the higher or lower induction tube pressure respectively, and they thus create in the air chamber a pressure P.sub.1 whose average value corresponds to a value between the upper and lower induction tube pressures in proportion to the duty cycle ratio. This serves to create at the metering aperture 5 an effective pressure difference of such magnitude as to produce an air number .lambda. of approximately 1. The duty cycle ratios T.sub.10 and T.sub.11 are proportional to the output voltage of the integrator contained in the electronic controller 14 and the output voltage increases, for example, as long as the sensor voltage U.sub.s is greater than the threshold voltage U.sub.o and it decreases in the reverse case. An electric circuit which may be used for this type of control is described in the U.S. Pat. No. 3,874,171.

FIG. 6 shows the output voltage of the integrator U.sub.i as a function of time which, in turn, defines the duty cycle ratio of the valves 10 and 11, respectively, whereby the entire period t.sub.g = t.sub.10 + t.sub.11 is kept constant.

FIG. 7 illustrates a possibility of changing the output voltage U.sub.i of the integrator cyclically, i.e., any change induced by the sensor voltage is initiated by the ignition pulses, i.e., at a frequency f = 2n and then proceeds during a predetermined time period t.sub.i after which the output voltage U.sub.i until the next ignition pulse. This results in an average increase of the output voltage of the integrator proportional to the rpm. This method is described in U.S. Pat. No. 3,875,907.

Inasmuch as the oxygen sensor delivers its information at the operating frequency of the engine (for example in a four cylinder, four cycle engine, f = 2n), it could be useful to so control the change of the integrator output voltage that the same change of .lambda. takes place in any rpm-dependent cycle period T.sub.n = 1/2n.

The following relation holds: ##EQU1## when dT .about.du.sub.i, dp.sub.l .about. dT and d .lambda. .about. dp.sub.l, then ##EQU2## where T is the duty cycle (keying ratio) of the valve control pulses and is equal to the valve opening time divided by the engine period. In order to obtain the same response time for each cycle, the duty ratio T = T.sub.11 = t.sub.11 /T.sub.n may be generated by an output voltage of the integrator which changes in proportion to rpm. During the transition from one operational state of the engine to another having the same mixture ratio but different rpm, this duty cycle ratio must be maintained, i.e., t.sub.11 .about. 1/n. When du.sub.i /dt .about. n holds, .DELTA. .lambda. .about. n .multidot. 1/2n .multidot. const = const.

FIG. 8 shows the air chamber pressure p.sub.1 for valves 10 and 11 actuated at the ignition frequency and the duty cycle ratio is determined as discussed above by an rpm-proportional output voltage of the integrator.

FIG. 9 illustrates that it may be suitable to make the sum of the opening times of the valves 10 and 11 smaller than the pulse time t.sub.p when the pressure sources for the conduits 12 and 13 pulsate in the same phase, for example as do the pressures in the various regions of the induction tube of an engine. It is then suitable to place the operating domain of the valves in a region of maximum pressure difference between the two pressure sources. This brings the further advantage of preventing disturbances during the overlapping opening time of the valves.

A particularly advantageous possibility to actuate the valves is illustrated in the lower part of FIG. 9 which shows a curve illustrating the theoretical difference of the output times t.sub.11 - t.sub.10 determined by the output voltage from the integrator and in which only the valve 11 which has a theoretically longer opening time t.sub.11 is being opened during the difference t.sub.11 - t.sub.10.

This manner of construction avoids the situation where the pressure in the chamber 8 is lowered too far by the valve 11 and must then be built back up through the valve 10.

FIG. 10 illustrates a further embodiment of the invention which provides an increase of the pressure difference of the induction tube pressures used for controlling the air pressure p.sub.1. This is done by tapping off the larger pressure for the chamber 8 through a line 20 upstream of an air filter 21 in the induction tube. This construction provides a large pressure difference for controlling the fuel-air mixture due to the pressure drop across the air filter and, for example, the Venturi vacuum.

In all three embodiments of FIGS. 1, 2 and 10, it is generally required to make the basic setting of the fuel-air mixture rich. But when large amounts of mixture are flowing, (high Venturi vacuum) the air space 8 may experience a vacuum which would result in evaporation of the fuel components having a low boiling point and thus could produce disturbances in the pressure control. It may therefore be suitable, as illustrated in FIG. 11, to employ the pressure drop across the air filter 21 for controlling the pressure in the air chamber 8. For this purpose, the air chamber 8 may be connected via a line 9 with the air line 20 upstream of the filter 21 and secondly through a line 22 with the induction tube downstream of the air filter 21. The air lines 20 and 22 are controlled, respectively, by the solenoid valves 10 and 11. Since this type of mechanism can serve only to enrich the fuel-air mixture, the basic setting of the fuel supply system must therefore be made lean.

The solenoid valves 10 and 11 could also be operated in opposite phase by a common magnet as explained in the U.S. Pat. No. 3,974,813.

The foregoing description relates to preferred exemplary embodiments and other embodiments and variants of the invention are possible within the spirit and scope thereof, the latter being defined by the appended claims.

Details of the electronic controller 14 are known by one or more of the following U.S. Pat. Nos.:

3,874,171

3,782,347

3,759,232

3,745,768

3,483,851

and the allowed application Ser. No. 392,659, the descriptive portions of which are incorporated by express reference.