Electric Discharge Machining Apparatus With a Wire-Shaped Electrode

This invention relates to electric discharge machining apparatuses with a wire-shaped electrode for machining a workpiece by the use of electric discharge, and more particularly to a control device for use in the electric discharge machining apparatus, which is capable of optionally controlling machining electrical conditions according to variations in thickness of the workpiece and to the quality thereof.

In an electric discharge machining apparatus of this type, a metal wire about 0.05 to 0.3 mm in diameter is employed as its one electrode. Electrical discharge is caused between the electrode and a workpiece to be machined to melt the corresponding portion of the workpiece. The electrode is displaced relative to the workpiece in the X-Y plane to thereby electrically cut or cut a way the workpiece in a desired configuration. In this case, in general, the electrical discharge is carried out with a constant speed feed of 1 pulse/.mu.m; that is, the discharge is effected at a constant voltage in the machining gap without controlling the discharge energy.

This constant speed feed method can satisfactorily machine a workpiece if the workpiece is constant in thickness. However, in machining a workpiece which is variable in thickness, initially it is necessary to set the machining feed speed to a value for the maximum thickness (that is, the maximum machining area) of the workpiece, so that no short-circuiting occurs and the wire-shaped electrode may not be broken. Accordingly, in the constant speed feed method, the machining feed speed is relatively low. Thus, even in machining a portion of the workpiece which is smaller in thickness than the other portions, the machining feed speed is still maintained low. Therefore, the constant speed feed method is considerably low in machining efficiency. On the other hand, it is well known in the art that even if a workpiece is constant in thickness, in order to shape a workpiece to a configuration having an angular corner it is preferable to increase the machining feed speed or to decrease the discharge energy. Because the workpiece is machined sharply, i.e., the resultant corner is not round. Thus, the constant speed feed method in which the discharge energy is maintained unchanged still involves problems to be solved.

An electrical discharge machining apparatus has been proposed in the art in order to eliminate the above-described difficulties accompanying the constant speed machining method. In the conventional apparatus, a voltage across a machining gap, i.e., a machining voltage, is detected and a machining feed speed is controlled so that the voltage thus detected is constant. The arrangement and operation of the conventional apparatus will be described with reference to FIG. 1.

A machining electric source (3) supplies a machining current to a wire electrode (1) and a workpiece (2) to be machined. The average value Eg of the machining voltage and a reference voltage Eo are applied to an error voltage amplifier (4), which determines a machining feed speed F proportional to an error voltage which is the difference value between the machining voltage Eg and the reference voltage Eo. The machining feed speed F thus determined is distributed as an X-axis component Fx and a Y-axis component Fy by a speed distributor (5), which are adapted to drive an X-axis motor (6) and a Y-axis motor (7), respectively. In this connection, the following relation is established between the machining feed speed F provided by the amplifier (4) and the outputs Fx and Fy of the speed distributor (5):

In the above-described arrangement, when the gap between the wire-shaped electrode (1) and the workpiece (2) becomes smaller and accordingly the machining voltage Eg becomes lower than the reference voltage Eo, the machining feed speed F is reduced and the gap is widened so that the machining voltage Eg approaches the reference voltage Eo. In contrast, when the machining voltage Eg becomes higher than the reference voltage Eo, the machining feed speed F is increased to thereby cause the machining voltage Eg to approach the reference voltage Eo. This is a system in which a machining voltage is fed back to vary the machining feed speed. In this system, the machining feed speed is increased when a relatively thin portion of the workpiece is being machined, whereas it is decreased when a relatively thick portion of the workpiece is being machined.

The employment of the above-described machining feed speed control in which the machining voltage is maintained unchanged makes it possible to eliminate to some extent the loss in machining feed speed which is involved in constant speed feed.

FIG. 2 shows the waveform of current for charging a charge-discharge capacitor. In FIG. 2, Ip is the peak value of the charging rest current, .tau.p is the pulse width, and .tau.r is the pause period of time. FIG. 3 is a circuit diagram showing the machining electric source (3) in FIG. 1. The electric source (3), as shown in FIG. 3, comprises a capacitor (8) which affects a machined surface roughness, a current limiting resistor (9) determining the peak value Ip of the charging current, a switching transistor (10), an oscillator (11) for determining the pulse width .tau.p and the rest period of time .tau.r of the machining current, and an internal DC source (12) providing a no-load voltage between the electrodes.

Depending on these electrical conditions, discharge machining energy is variable even with the average machining voltage Eg maintained unchanged. In general, in machining a workpiece relatively small in thickness the discharge machining energy is liable to concentrate at a point, and therefore it is necessary to lower the above-described electrical conditions to decrease the discharge machining energy, otherwise the wire electrode would be broken.

The machining feed speed F is so controlled that the machining voltage Eg is maintained unchanged and in machining a workpiece variable in thickness the electrical conditions are so set that the wire electrode (1) is not broken at the portion of the workpiece which is the smallest in thickness. Therefore, at the portion of the workpiece which is larger in thickness, the electrical conditions are not sufficient and accordingly the machining speed is decreased. In addition, it is known that the machining accuracy is improved by increasing the electrical conditions at the relatively large portion of the workpiece in thickness.

When the machining direction is changed as in the corner of a machining configuration, the discharging area is decreased, i.e., equivalently the thickness is decreased. Therefore, with the electrical conditions maintained unchanged, over-cutting occurs, as a result of which the accuracy of the machined corner is lowered. Accordingly, in this case, it is necessary to decrease the electrical conditions in order to machine the workpiece sharply at the corner with high accuracy.

As is apparent from the above description, the conventional constant speed feed method and the conventional machining speed control maintaining a machining voltage unchanged are disadvantageous in that in machining a workpiece variable in of variable discharge machining area, i.e., in thickness or in machining a workpiece for providing corners, the machining feed speed and the machining accuracy are unsatisfactory. Furthermore, the conventional methods are low in reliability because the electrical conditions are manually set, with the result that the machining operation depends greatly on the operator's experience and because the setting of the electrical conditions is rather difficult, with the result that the wire-shaped electrode is often broken.

Accordingly, an object of this invention is to provide electric discharge machining apparatuses with a wire-shaped electrode in which all of the above-described difficulties accompanying conventional apparatuses have been eliminated, and in which operational amplifier means is employed to automatically set the electrical conditions, namely the charge peak current Ip, pulse width .tau.p, rest period of time .tau.r and capacitance C to optimum values according to variations in the discharge machining area of a workpiece.

In the apparatus according to the invention, a machining speed is detected, the machining speed thus detected in compared with a reference machining speed in an operational amplifier to thereby detect variations in the machining area of a workpiece, and in response to the thus detected variations in machining area, optimum electrical conditions in electrical discharge machining are automatically set in accordance with the variations in thickness of the workpiece.

According to the invention, the electrical conditions which have been manually set are completely automatically set, and accordingly the wire-shaped electrode will never be broken carelessly. Furthermore, in machining a workpiece whose thickness is variable, optimum electrical conditions are automatically set for the variations of the thickness. Accordingly, in the invention, unlike the prior art, the method of machining a workpiece under decreased electrical conditions corresponding to the smallest thickness of the workpiece is not employed, and therefore the machining feed speed can be greatly increased and the machining accuracy is also considerably improved.

Another object of the invention is to provide electric discharge machining apparatuses with a wire-shaped electrode in which in response to variations in discharge machining area of a workpiece, among the electrical conditions, the charge peak current Ip, pulse width .tau.p, rest period of time .tau.r, capacitance C and no-load voltage are completely automatically controlled to optimum values individually or in combination.

In the apparatus of the invention, the electrical conditions different in energy which have been prepared in the form of a data table with respect to the variations in thickness of a workpiece and the variations in machining area of the workpiece in an equivalent sense as in machining corners are successively read out to be subjected to comparison to obtain the optimum electrical conditions at all times. In order to detect the variations in discharge machining area, an averaged machining feed speed (for a predetermined time) is employed, and the electrical conditions are changed every predetermined time. Thus, the electrical conditions are controlled exactly in conformance to the variations in discharge machining area.

Accordingly, the electrical condition switching control of the invention, unlike the prior art, is free from the trouble that it is difficult to suitably set the electrical condition according to the variations in thickness of a workpiece. Thus, the apparatus of the invention is improved in operability and is free from the problem that the wire electrode is broken by erroneously setting the electrical conditions. The most significant merit of the invention resides in that the drawbacks accompanying the prior art apparatus such as the lowering in of machining speed can be remarkably improved. In addition, since the electrical conditions are decreased in forming a corner by machining a workpiece, the resultant corner is much sharper than that formed by conventional apparatuses. Thus, the apparatus of the invention is remarkably high in reproducibility, reliability and machining characteristics in machining workpiece of variable thickness.

In the accompanying drawings:

FIG. 1 is a schematic diagram showing the arrangement of a conventional apparatus;

FIG. 2 is an explanatory diagram showing the waveform of a pulse current supplied to the electrodes of the apparatus;

FIG. 3 is a circuit diagram showing a machining electric source;

FIG. 4a is a diagram showing one example of an electric discharge machining apparatus with a wire-shaped electrode according to this invention;

FIG. 4b is an explanatory diagram for a description of a method of changing electrical conditions according to a machining feed speed signal;

FIG. 5 is a graphical representation indicating thickness t with machining feed speed F in the conventional discharge machining apparatus under the conditions that the machining voltage is constant and the electrical condition is maintained constant;

FIG. 6 is a graphical representation indicating electrical conditions and machining feed speeds with respect to variations in thickness of a workpiece under the conditions that optimum conditions are set according to the invention;

FIG. 7 is an explanatory diagram showing another example of the apparatus of the invention, in which the machining feed speed is averaged by a low-pass filter and electrical conditions are set;

FIG. 8 is an explanatory diagram showing a further example of the apparatus according to the invention, which is modified by adding to the apparatus in FIG. 4a a circuit for sampling the feed speed at predetermined time intervals;

FIG. 9 is an explanatory diagram showing a still further example of the apparatus of the invention, which has the functions of the low-pass filter in FIG. 7 and of the feed speed sampling circuit in FIG. 8;

FIG. 10 is a graphical representation indicating variations in discharge machining area of a workpiece with respect to variations in thickness of the workpiece;

FIG. 11 is a schematic diagram showing an arrangement of the apparatus according to the invention;

FIG. 12 is a diagram showing one example of an electrical condition table;

FIG. 13 is a explanatory diagram for a description of an automatic electrical condition setting system according to the invention;

FIG. 14 is a block diagram showing an operational amplifier circuit and an electrical condition control device;

FIG. 15 is a diagram showing the waveform of a clock pulse signal outputted by an oscillator; and

FIG. 16 is a block diagram showing a part of the electrical condition control device.

FIG. 4 shows one example of an electric discharge machining apparatus with a wire-shaped electrode according to the present invention which is obtained by adding operational amplifiers (13) to the apparatus shown in FIG. 1. Data representative of a machining feed speed F and a reference feed speed Fo are applied to the operational amplifiers (13), so that output voltages Ki(Fo-F) proportional to the difference between the speeds F and Fo are applied to the pulse source (3) and machining energy is proportional to the difference between the speeds F and Fo, i.e., an electrical condition is set. In the above-described output voltages Ki(Fo-F), Ki is the proportional constant (where i=1 to 4); that is, K.sub.1 through K.sub.4 are coefficients provided respectively for the electrical data Ip, .tau.p, .tau.r and C.

The operation of the apparatus shown in FIG. 4 will be described. Similarly as in the conventional apparatus in FIG. 1, the machining voltage Eg is compared with the reference voltage Eo, and the machining feed speed F is so controlled that the machining voltage Eg approaches the reference voltage Eo at all times.

If control is attained so that, with respect to the variations in thickness of the workpiece (2), the gap voltage between the wire-shaped electrode (1) and the workpiece (2) is maintained constant and the electrical conditions are the same, then the machining energies are substantially constant and the machining area speed is also substantially constant, and therefore the machining feed speed F is in inverse proportion to the thickness of the workpiece.

Under this condition, the machining feed speed F is applied to the operational amplifiers (13). As the machining feed speed F is decreased, the output voltages Ki(Fo-F) of the operational amplifiers (13) are increased; that is, the four electrical data Ip, .tau.p, .tau.r, and C are increased, so that the machining feed speed F is not so greatly decreased.

FIG. 5 is a graphical representation indicating machining feed speed F with thickness t under the condition that, in the conventional apparatus, machining energy EC is so set that the wire-shaped electrode (1) is not broken in machining the portion of the workpiece which is smallest in thickness. In this case, the electrical conditions EC are maintained unchanged, and therefore the machining feed speed is decreased substantially in inverse proportion to the thickness t.

FIG. 6 is also a graphical representation indicating machining feed speed F with thickness t under the condition that, in the apparatus according to the invention, the optimum electrical condition EC is automatically set. In this case, as the thickness t increases, the electrical condition EC increases, that is, the electrical data Ip, .tau.p and C increase while the electrical data .tau.r decreases. Accordingly, the machining feed speed F is not so greatly decreased with respect to the reference feed speed Fo; however, it has a negative inclination because, as the pause period of time .tau.r decreases, the machining energy increases.

In general, during a machining operation, it is difficult to detect the thickness t, and therefore the electrical condition EC can be determined by utilizing the fact that the thickness t is substantially in proportion to the machining feed speed F. If the machining feed speed F and electrical condition EC indicated in FIG. 6 are approximated with straight lines, then

where a and b.sub.1, b.sub.2, b.sub.3 and b.sub.4 (represented by "b.sub.i ") are the constants representing the inclinations of the straight lines F, Ip, .tau.p, .tau.r and C in FIG. 6, respectively, and t is the thickness.

Eliminating the thickness t from the above equations, the following equation (3) is obtained:

where Ki is the constant, and Ki=b.sub.i /a

This means that, if the output voltages Ki(Fo-F) are provided by the operational amplifiers (13) and electrical data proportional to the output voltages are applied to the pulse source (3), then the electrical condition EC is automatically set with respect to the thickness t.

This will be described in more detail. FIG. 4b shows a technical concept of how the peak value Ip of the pulse current is varied. It should be noted that the technical concept can be similarly applicable to variations of the other electrical conditions such as pulse width .tau.p, rest period of time .tau.r and capacitance C.

If, during a machining operation, the thickness t increases, then Eg<Eo, and therefore E(F)'=K(Eg-Eo)<0 and .vertline.E(F).vertline.<.vertline.E(F)'.vertline., where K is the constant, E(F) is the voltage value corresponding to the machining feed speed F before the thickness increases, and E(F)' is the voltage value corresponding to the machining feed speed after the thickness increases.

The voltage value E(F)' is added to the voltage E(Fo) corresponding to the reference feed speed. In this case, E(Fo)-E(F)<E(Fo)-E(F)' where E(F) is the voltage value corresponding to the machining feed speed before the thickness increases. The increased E(Fo)-E(F)' is applied to the operational amplifier (13), so that a voltage value E(Ip)' corresponding to the amount of increase (being a voltage value corresponding to the peak current value) is outputted by the amplifier (13). A value E(Ipo) (which is the value of the EC axis segment in FIG. 6) is added to the value E(Ip), and the resultant value is applied to the machining source (3), and Ip corresponding to E(Ip')+E(Ipo) is selected. The analog data E(Ip)'+E(Ipo) is converted into a digital data by an analog-to-digital (A/D) converter before applied to the machining source (3), if necessary.

In the case where the thickness of the workpiece decreases, Eg>Eo, E(F)">0 and .vertline.E(F).vertline.<.vertline.E(F)".vertline.. As a result, E(Fo)-E(F)>E(Fo)>E(Fo)-E(F)". The E(F)" is the voltage value corresponding to the machining feed speed when the thickness decreases.

The electrical data can be changed as described above.

Since the electrical condition EC has the parameters, namely, the charge peak current Ip, the rest period of time .tau.r, the pulse width .tau.p the capacitance C and the like, there may be a variety of electrical conditions with the same machining energy. However, in general, the most stable parameters should be selected through experiment.

In the above-described examples, the operational amplifier is employed so that the resistance is varied in proportion to the output of the operational amplifier to set the data Ip, and the oscillator 11 is controlled to vary the pulse with .tau.p and the rest period of time .tau.r, while the capacitance is automatically set. However, instead of the operational amplifiers 13, a number of comparators may be connected in such a manner that various data are applied to first input terminals of the comparators. In this case, the comparators provide digital outputs. Therefore, if the digital outputs are utilized to operate (on and off) switches provided for the relevant electrical data, the same effects described above can be obtained.

In the circuit shown in FIG. 4, although the electrical data are set immediately in response to the feed speed F, it should be noted that, since the feed speed F is somewhat varied during machining, the electrical data are also varied. The variations of the electrical data result in machining energy variations, which affects the feed speed F. Thus, the control system becomes unstable as a whole, as a result of which the electrical data are greatly varied and the wire-shaped electrode may be broken by momentary excessive energy.

This difficulty is eliminated by another example of the apparatus according to the invention shown in FIG. 7. In this example, a low-pass filter consisting of a resistor (14) and a capacitor (15) is employed to apply the average value F of the feed speed F to the amplifiers (13), so that a constant voltage is applied to the amplifiers (13), and the electrical data provided for the pulse source (3) are made stable.

A circuit shown in FIG. 8 is also practical to stabilize the control system. The circuit comprises a sample-and-hold switch (16), a sample-and-hold capacitor (17) and a timer (18) adapted to close the switch (16) momentarily at predetermined time intervals. That is, a feed speed at a time instant is stored in the capacitor (17), and the capacitor (17) continuously holds a constant value until the switch (16) is closed by the timer (18) again. In other words, the voltage applied to the operational amplifiers (13) is maintained held for a period of time, during which the electrical condition is maintained unchanged.

A circuit shown in FIG. 9 is obtained by adding a sample-and-hold switch (16), a sample-and-hold capacitor (17) and a timer (18) to the low-pass filter shown in FIG. 6. In this case, a considerably stable control system can be provided, because even when the switch (16) is closed by the timer (18), the voltage across the capacitor (17) is not greatly changed, and accordingly the electrical data are slightly changed every predetermined time.

In the above-described examples, the switch (16) is closed momentarily by the timer every predetermined time. However, the same effects described above can be obtained by utilizing a machined distance detector instead of the timer in such a manner that the switch (16) is closed whenever the workpiece is machined for a predetermined distance.

In the above-described examples, the electrical condition and the machining feed speed are approximated with straight lines with respect to the thickness t of a workpiece. However, more accurate control can be carried out by performing approximation as shown in FIG. 6 with operational amplifiers having non-linear circuits.

This invention is further intended to control the electrical data completely in accordance with a discharge area. As conductive to a full understanding of this, variations of the discharge area with respect to variations of the thickness of the workpiece will be described. In FIG. 10, the horizontal axis indicates the positions of the wire-shaped electrode (1) in machining the workpiece (2) which is variable from t.sub.1 to t.sub.5 in thickness whereas the vertical axis indicates the values S.sub.1 -S.sub.5 of the discharge area S which corresponds to t.sub.1 -t.sub.5, respectively. In the figure, the machining direction is designated by an arrow Q.

As is apparent from FIG. 10, in the case where the thickness of the workpiece is increased from t.sub.1 to t.sub.5, when the wire-shaped electrode (1) reaches the position A, electrical discharge occurs between the electrode (1) and the end face of the portion of the workpiece, which is t.sub.5 in thickness. In other words, in the case where the thickness is abruptly increased from t.sub.1 to t.sub.5, the discharge area is not abruptly increased from S.sub.1 to S.sub.5 ; that is, the discharge area is gradually increased from S.sub.1 through S.sub.2, S.sub.3 and S.sub.4 to S.sub.5. In FIG. 10, the distance AB is theoretically the sum of the radius of the wire-shaped electrode (1) and the discharge gap.

In the case where the thickness is decreased from t.sub.5 to t.sub.1, similarly as in the above-described case, the discharge area is not abruptly decreased from S.sub.5 to S.sub.1 at the position D; that is, it begins decreasing at the position C, then decreases gradually and finally decreases to S.sub.1 at the position D.

It goes without saying that, if the discharge area based on the wire electrode (1) and the discharge gap are computed and graphed under the condition that theoretically the discharge gap is maintained unchanged even when the thickness changes, as shown in FIG. 10, then AB=CD.

In the case where the thickness is increased from t.sub.1 and t.sub.5 or decreased from t.sub.5 to t.sub.1 (especially where the thickness is changed abruptly as in a configuration having right angles), practically the intermediate thickness t.sub.2, t.sub.3 and t.sub.4 are included, and therefore it is necessary to vary the electrical data in accordance with these different thicknesses.

However, if the electrical data are abruptly changed because of the abrupt change in thickness at the positions B and D corresponding to both end faces of the workpiece's portion t.sub.5 in thickness, then the following problems will occur. It is assumed that the thickness is increased from t.sub.1 to t.sub.5. If the electrical data for the thickness t.sub.1 is changed to the data for the thickness t.sub.5 at the position B, the portion A-B of the workpiece is machined with the decreased electrical data for the thickness t.sub.1. Therefore, as the discharge area increases gradually, the machining speed is decreased resulting in a lot of loss. In contrast, it is assumed that the thickness is decreased from t.sub.5 to t.sub.1. If the electrical data for the thickness t.sub.5 is changed to the data for the thickness t.sub.1 at the position D, then the machining is carried out with the electrical data for the thickness t.sub.5 although the thickness of the portion C-D is smaller than the thickness t.sub.5. Accordingly, in this case, the current density is increased, or the discharge energy is concentrated at one point, and therefore the wire-shaped electrode is liable to be broken. Experimentally, the value AB (=CD) is, in general, 0.15 to 0.2 mm (in the case of a wire-shaped electrode 0.2 mm in diameter). This value is a distance long enough to break the wire electrode.

FIG. 11 shows another example of the apparatus according to the invention. In the apparatus in FIG. 11, similarly as in the apparatus shown in FIG. 1, the machining voltage Eg is compared with the reference voltage Eo, and the machining feed speed F is so controlled that the machining voltage Eg approaches the reference voltage Eo at all times. As described with reference to FIG. 1, the machining feed speed F proportional to the error voltage is averaged in an arithmetic circuit 140, so as to be outputted as a machining feed speed average value F every predetermined time.

The average value F is applied to an electrical condition control device 150, in which, for the inputted average value F, an electrical condition Ec most suitable for a thickness being machined is selected from a data table which provides electrical conditions Ec for machining feed speed average values F, respectively, and the electrical condition Ec thus selected is applied to the machining electric source (3).

FIG. 12 shows one example of the data table which is stored in memory means in the electrical condition control device 150. In the data table, as shown in FIG. 12, for various thicknesses of a workpiece there are provided electrical conditions and the upper limit values Fu and lower limit values Fd of machining feed speed average values F, respectively.

FIG. 13 is a graphical representation for a description of an automatic electrical condition setting system according to the invention. In FIG. 13, the horizontal axis indicates thickness t (t.sub.0 <t.sub.1 < . . . <t.sub.4 < . . . ), whereas the vertical axis indicates machining feed speed average value F and electrical condition Ec. As is apparent from FIG. 13, for a thickness between 0 and t.sub.0 an electrical condition Ec.sub.0 is provided and a machining feed speed average value F is between F.sub.0 and F.sub.0 '; and for a thickness between t.sub.0 t.sub.1 an electrical condition Ec.sub.1 is provided and a machining feed speed average value f is between F.sub.1 and F.sub.1 '.

The principle of operation will be described with reference to FIGS. 12 and 13 in detail. It is assumed that a workpiece's portion whose thickness t is between t.sub.3 and t.sub.4 (t.sub.3 <t<t.sub.4) is machined under the electrical condition Ec.sub.4. In this case, the machining feed speed average value F is between F.sub.4 and F.sub.4 '. Furthermore it is assumed that under this condition, the machining operation is advanced to another portion whose thickness t is between t.sub.1 and t.sub.2 (t.sub.1 <t<t.sub.2). In this case, the electrical condition Ec.sub.4 is still maintained, and therefore the machining feed speed average value F is increased to be F>F.sub.4 ; that is, it exceeds the upper limit value F.sub.4 of the electrical condition Ec.sub.4. Accordingly, the electrical condition control device 150 decreases the electrical condition by one step, i.e., it applies the electrical condition Ec.sub.3 to the machining electric source 3. The upper limit value Fu corresponding to the electrical condition Ec.sub.3 is F.sub.3 . However, the actual thickness t is between t.sub.1 and t.sub.2 (t.sub.1 <t<t.sub.2) as described above, i.e., it is smaller than the thickness (t.sub.2 <t<t.sub.3) corresponding to the electrical condition Ec.sub.3, and therefore the machining feed speed average value F exceeds the upper limit value F.sub.3.

Accordingly, in the electrical condition control device 150, the electrical condition Ec.sub.3 is decreased by one step according to the data table; that is, the electrical condition Ec.sub.2 is outputted. Since the thickness of the workpiece's portion is between t.sub.1 and t.sub.2, the machining feed speed average value F is between F.sub.2 and F.sub.2 ", and the electrical condition control device 150 outputs this electrical condition Ec.sub.2 continuously.

In the case also where the thickness increases, similarly as in the above described case, a suitable electrical condition is selected so that the present machining feed speed average value F is between the upper limit value Fu and the lower limit value Fd.

By way of example, the selection of a suitable electrical condition will be described with respect to the peak current Ip.

In a circuit shown in FIG. 14, a voltage E(F) (which is a voltage value corresponding to a machining feed speed F) is added to a reference voltage EO(+). The resultant value E(F)+Eo is averaged into a value E(F)+Eo by a low-pass filter (141) and is then converted into a digital data by an analog-to-digital converter (142). The digital data is applied to a latch circuit (143). The latch circuit (143) stores and outputs the digital data with the timing of the rise of a clock pulse generated by an oscillator (144); that is, the latch circuit (143) holds the digital data until the rise of the following clock pulse. The clock pulse is outputted by the oscillator (144) every 0.5 to 1 sec. as shown in FIG. 15.

The output of the latch circuit (143) is applied to a decoder (151), where the output bits corresponding to the data are raised from "0" to "1". In FIG. 14, the A/D converter (142) is a 4-bit A/D converter, and therefore the decoder (151) can provide sixteen different outputs. The 2-bit data of peak currents Ip have been stored in a memory (152) in advance. Therefore, the memory (152) can provided four different peak current data Ip. When the output of the decoder (151) is supplied to a line Ip.sub.1, a 2-bit data corresponding to the line Ip.sub.1 is applied through an AND circuit (153), as a binary coded decimal number (BCD), to the machining electric source (3). In response to the binary coded decimal number, the machining electric source (3) causes a relay or the like (not shown) to operate to control the resistor (9) in FIG. 3.

The selection of a suitable electrical condition has been described with respect to Ip; however, this technical concept can be similarly applied to the selection of the electrical data .tau.p and .tau.r. If, as shown in FIG. 16, the output bit data (Ip.sub.1) of the decoder (151) is connected to other AND circuits (153) provided for the electrical data .tau.p.sub.1 and .tau.r.sub.1, then the electrical data Ip.sub.1, .tau.p.sub.1 and .tau.r.sub.1 can be applied, in combination, to the machining electric source (3). In FIG. 16, the other data (sixteen different data) similar to the electrical data Ip.sub.1, .tau.p.sub.1 and .tau.r.sub.1 are applied to lines Ip.sub.n, .tau.p.sub.n and .tau.r.sub.n. These data are selected in response to the other output bit data of the decoder (151).

For instance, when the thickness increases, in general Eg<Eo and therefore the value Eo+E(F) is decreased. Accordingly, the output bit data of the A/D converter (142) is lower in weight. Thus, in this case, the output of the decoder (151) should be coupled to the memory (152) so that a high energy electrical condition is selected from the memory (142). It goes without saying that when the thickness decreases, a method opposite to the above-described one should be employed.

As described above, a variety of data can be made available by changing the connection between the output of the decoder and the memory or the contents of the memory.

In the apparatus of the invention, for example as shown in FIG. 11, the machining feed speed F determined by the error amplifier (4) is applied to the averaging circuit 140, where it is averaged every predetermined time. The reason for this will be described.

In an actual machining operation, control is made with a machining feed speed F so as to cause the average machining voltage Eg to approach the reference voltage Eo, and therefore the machining feed speed F is variable. In addition, the machining feed speed F sometimes greatly varies when the interelectrode discharge changes, or when the wire-shaped electrode changes in tension, or because the wire-shaped electrode has a non-uniform diameter, or because of external effects. If, in this case, the electrical condition is changed in accordance with the machining feed speed F which varies with time as described above, then the electrical condition is unstable while the thickness is constant. Further, if the electrical condition is changed to be higher in electrical energy, then at worst the wire-shaped electrode is broken. In machining a workpiece whose thickness is variable or in machining a workpiece to provide a corner, it is difficult to obtain electrical conditions exactly according to the discharge area, and the wire-shaped electrode is therefore broken or a loss is involved in machining feed speed. Especially in the machining of a corner, the machining direction is often abruptly changed, and therefore in the initial machining the wire-shaped electrode is liable to bite the end face of the corresponding portion of the workpiece. This will greatly change the machining feed speed F, as a result of which the electrical condition is also greatly changed resulting in a so-called "hunting state", or the wire electrode may be broken.

Because of the above-described various factors, according to the invention, instead of the machining feed speed F which varies with time, the machining feed speed F which is averaged to a certain extent is employed. This machining feed speed average value F does not respond to momentary external effects and other variations. Therefore, with the machining feed speed average value F, portions of a workpiece where its thickness changes or corners should be provided can be accurately machined. Furthermore, according to the invention the average value is obtained every predetermined time, and therefore an actual machining operation can be carried out separately according to a portion of a workpiece for which the machining feed speed average value F should be changed and a portion of the workpiece for which the average value should be maintained unchanged. According to the experiments made by the inventor, the above-described predetermined time is preferably within a range of 0.5 sec to 1 sec. With this time range, not only is an ordinary machining operation carried out, but also a portion of a workpiece where the thickness changes or where a corner should be provided can be satisfactorily machined; that is, the apparatus of the invention can satisfactorily respond to transient phenomena which may be involved in a workpiece machining operation.

In the apparatus according to the invention, the time intervals for changing the electrical condition are of the same as the above-described time intervals; that is, the time intervals are also set to 0.5 to 1 sec by sufficiently taking into consideration the response of machining feed speed due to the change of the electrical condition. It goes without saying that, if the time intervals are set to an excessively large value, then a trouble occurs that, although a thickness change has been detected, the electrical condition is not changed. If this trouble occurs when the thickness decreases, then the wire electrode may be broken.

In the above-described apparatus of the invention, the arithmetic circuit or averaging circuit and the electrical condition control device and the like may be replaced by a computer so that a variety of data in a wider range can be stored and processed. In this case, the application range of the apparatus of the invention is further increased. The electrical data, namely, the charge peak current Ip, the pulse width .tau.p, the rest period of time .tau.r, the capacitance and the no-load voltage can be controlled individually or in combination. In this connection, if the data table is experimentally prepared in advance depending on the most important one of the machining feed speed, machining accuracy and machined surface roughness the data table is prepared through experiment in advance, then a variety of machining operations can be achieved with the apparatus of the invention.