EP0187738A1 - Electric circuit for electroerosion machining - Google Patents

Abstract

The electrical circuit for an electro-erosion machining machine comprises means for delivering erosive discharges (1), and pre-ignition means (6-12). The pre-ignition means, with transformer (6) and Zener diode (3), comprise a clipping branch (11, 12) also with Zener diode. In addition, energy regulating means (14, 15, 22, 23, 24) are connected to limit the current of the preignition pulses as a function of the clipping current, measured by the voltage drop across a terminal. resistor (12) in series with the Zener clipping diode (11). The arrangement of the pre-ignition means mainly relates to the control of the electrical and geometric parameters relating to the pre-ignition and to the general operation of the machining machine.

Description

Electrical circuit for machining by electro-erosion

The present invention relates to an electrical circuit for an EDM machine working a part by electro-erosive discharges in a dielectric medium, in a machining space between a tool electrode and said part which also constitutes itself a electrode, comprising means for delivering erosive discharges, having output terminals which respectively supply said electrodes with pulses of erosive electrical energy, and means of pre-ignition to supply, each time in proportion to a so-called erosive pulse, pre-ignition pulses which have a higher voltage and a lower energy than said erosive pulses, and which ignite, that is to say ionize, in the machining space within dielectric fluid, each time a channel which is then traversed by the erosive discharge.

A circuit conforming to the generic definition set out above is already known, in particular by the disclosure of patent US 3,893,013, in the name of the same inventor and applicant. In general, the present invention aims to improve the aforementioned known system, in particular from the points of view which emerge from the considerations which follow.

In the technique of EDM machining, a key point consists in an adequate adaptation between the parameters of the applied electrical pulses and the machining distance, that is to say the distance between the electrode- tool of the part to be machined. We note that, in the terminology of tradespeople, this machining distance is, in all languages, generally designated as "GAP",

WIPO or more precisely "machining GAP", and it is this term which will be used later.

The adaptation between the machining GAP and the electrical parameters can be carried out either by acting on these electrical parameters, or by acting on the dimension of the GAP, using means, generally motorized, in advance. and retraction of the electrode-tool, which are present on practically all existing electro-erosion machining machines. A first difficulty consists in knowing exactly the dimension of the machining GAP, taking into account the fact that EDM itself digs the workpiece and that, if the machining electrode was not moved, the GAP would increase by itself to finally settle at a value large enough to interrupt the process of digging by EDM. In such a case, however, the power of the machining would begin to decrease progressively, and this is not what we want. It is indeed important to maintain the best possible working conditions, which vary moreover depending on whether one wishes a fine, finishing or superfinishing machining, which must be carried out in a regime in which the electro-erosive pulses are relatively weak. , or rapid machining, for example of roughing, which is most advantageously carried out in a regime comprising relatively powerful electroerosive pulses.

One might think that, taking into account the phenomenon mentioned above, it suffices to set a certain standard of power of the pulses and to advance the tool electrode at a speed such as the parameters of the erosive pulses, taking into account the par¬ peculiarities of the circuit which supplies them, remain substantially constant, in order to obtain a machining regime adequate, the value of the GAP adjusting itself somehow automatically. However, this procedure only leads to mediocre results, and it is hardly applicable in the case of machines using the principle of pre-ignition. The application of this principle nevertheless makes it possible to obtain very advantageous results which cannot be read without it, and it would be regrettable to give it up simply because it is difficult to grasp the parameters - electrical and material - which condition its proper use.

Before considering the very principle of pre-ignition, the fundamental problem of igniting a conductive channel between the tool electrode and the workpiece should be considered. space of the GAP, within a dielectric liquid in which the electrode-tool and the part bathe. (In the context of a general definition, it is better to speak of a "dielectric fluid", taking into account the phenomena of almost instantaneous vaporization which occur, and also taking into account the fact that one must not exclude any possibility of using a fluid other than a liquid, for example a "mist".)

In studying the phenomena involved in a simple EDM machining process, in which an electrode assumed to be cylindrical, for example, "gradually sinks" into a workpiece which it digs by EDM, distinguish between two different notions of "GAP". First of all, there is a distinction between the concept of "frontal GAP", which corresponds to the distance over which the erosive pulses occur, passing through channels of high electrical conductivity. It can also be called "effective machining GAP", it is indeed the GAP in relation to which we strive to obtain the maximum effectiveness of the erosive effect of the pulses at the same time as good stability of the machining process. But it is also necessary to consider the notion of "lateral GAP" which corresponds to the lateral distance between the peripheral dimensions of the electrode-tool and the dimensions "in width" (if the electrode-tool is cylindrical, the diameter) of the "hole" dug in the workpiece by the tool electrode. It is clear that, as long as the lateral gap between the tool electrode and the workpiece has not reached a limit value, electro-erosive pulses will also occur laterally, further widening this lateral gap. The concept of "lateral GAP" is therefore that of a "limit GAP" which corresponds to the distance beyond which conductive channels can no longer be created taking into account the electrical parameters applied. This "lateral GAP", necessarily significantly larger than the "machining GAP" determines the enlargement of the machined area with respect to the dimensions of the electrode tool, that is to say the sub -sizing that the tool electrodes must have in order to obtain precise machining dimensions.

The "lateral GAP" might be easily measurable, but the trouble is that it does not correspond to the "frontal GAP" which primarily concerns the machining process, and that the relations between the two dimensions are essentially variable according to the machining regime, although the two depend in particular on the different parameters that are the energy of the pulses, their voltage, the intensity of the pulse currents, as well as the characteristics of the medium or dielectric liquid.

The value of these two "GAPS" results from the physico-chemical process which determines the breakdown conditions of the dielectric and the formation of channels for the current of the erosive pulses. We usually talk of "ionization of the channel" but in reality this phenomenon has a double aspect: First, we have a polarization of the dielectric in the zone concerned, then then we have the ionization proper which makes the channel highly conductive.

The polarization is a function of the electric field which is established between the electrode and the part. To designate the degree of polarization, we use the term of vector of polarization P which is equal to the sum of the moments of the molecules in a volume _4V.

- ^r "ΔV

Since the degree of polarization is proportional to the intensity of the electric field, the polarization vector can also be expressed by:

P = x £ where _ £ is the coefficient of polarization (or the dielectric susceptibility of the medium).

On the other hand, the intensity of the electric field can be expressed as a function of the charge on the surfaces of the electrodes:

where £ is the dielectric constant and S - the surface of the electrodes.

But the intensity of the field is also linked to the voltage applied to the electrodes by the relation:

So the polarization vector can be expressed by

This shows that the degree of polarization is proportional to the load and the applied voltage, assuming that the surface is constant, and inversely proportional to the surface of the machining area, if the load is admitted constant, and to the distance from the workpiece, if the voltage is assumed to be constant. When the intensity of the field reaches a sufficient degree at any point in the machining area (for example between the roughest edges of the surfaces of the electrode and of the workpiece), ionization takes place by the emission of electrons by the cathode. The minimum difference in potential for producing this ionization is proportional to the "ionization potential of atoms or molecules", that is to say to a physical quantity depending on the dielectric medium, as well as to the number of atoms or molecules to be ionized "in series", that is to say at the dimension (d) of GAP, again taking into account the properties of the dielectric medium. As for the energy required, it is proportional to the minimum difference in potential (or voltage) mentioned above, and at the same time proportional to the number of electrons involved, that is to say in principle at number of atoms or molecules concerned by ionization, that is to say ultimately by the section (area) of the channel which is ionized.

The two polarization and ionization processes are therefore conditioned by geometrical parameters relating to the channel, as well as electrical parameters. We have seen that the geometric parameters tend in the long term to establish themselves as a function of the electrical parameters, although the instantaneous value the length (d) of the GAP can be adjusted. But it is above all on the electrical parameters that it is preferable to act, with a view to obtaining the ignition of a channel suitable for the desired machining. Indeed, if one can determine in a relatively precise way the energy con¬ sacred to the lighting of the channel, one can thereby determine the section of this channel, taking into account its length. The voltage, or potential difference, applied for the lighting of the channel is either sufficient or insufficient to allow ignition. However, an insufficient tension for a certain length of GAP will become sufficient for a shorter length, and vice versa. It is therefore necessary to firstly set the voltage of the pulses used for ignition (assuming that the length of the GAP will be adapted to this voltage), and then to determine as precisely as possible the energy that will be devoted to the ignition of the channel. . If too much energy is devoted to this ignition, the channel becomes very wide and the machining is then of poor quality. On the other hand, a too low amount of energy would theoretically suffice to light a very weak section. However, it has been observed that, below a certain energy threshold, the ignition of a channel of very small section does not occur. In a way, there are only "aborted attempts at ionization".

Once the channel is turned on, the energy is used for erosion and no longer for ionization of the channel. The quality of the erosion work (fine machining, rough machining) depends on the energy of the erosive discharges, but also depends on the relationship between this energy and the parameters of the channel which is on for its passage. When, as is the case in many machines of the prior art, it is left to ignite the channel at the initial portion of the erosive pulse, the parameters of the channel can hardly be determined, that is to say controlled, in a precise manner. On the other hand, this is possible if we separate the two functions on the one hand from ignition and on the other hand from erosion. This is what we do in machines of the "pre-lighting" type, and we then obtain extremely advantageous results, provided, however, that we are able to control the various parameters involved in the ignition process. It should not be forgotten that this process also depends on technological factors, which may vary, and relating mainly to the dielectric liquid. Thus, for example, the pollution of the dielectric liquid, the speed of circulation of this liquid in the machining area, etc., play a non-negligible role. It is necessary that the control of the parameters can be preserved in spite of disturbances of this kind. The invention will propose measures capable of greatly improving this control of the pre-ignition parameters.

It should also be recalled here that the tests which led to proposing the technical measures in accordance with the invention were carried out mainly on machines of the type proposed for example by the disclosure of US Patent No. 3,893,013, previously cited. These machines work on the principle of relaxation, that is to say that a machining capacitor takes charge of a certain amount of energy, then discharges by providing the electro-erosive pulse. in the GAP. In order for this discharge to take place, the machines in question generate pre-ignition pulses, when the voltage available for the erosive pulse is sufficient. The generation of the pre-ignition pulse is typically established using a step-up transformer. tension. However, for the study of the pre-ignition (or generally ignition) phenomena, circuits similar to those which provide the erosive pulses have been used, but working at much higher voltage levels and with much lower energies. Furthermore, instead of determining the necessary voltage as a function of the distance from GAP, it has been preferred to determine the distance from GAP for which a given ignition voltage is sufficient. This corresponds to practical reality, since the machines practically all comprise an electrode displacement motor which, gradually, brings the electrode to the vicinity of the workpiece. The tests carried out have shown that, for ignition (or pre-lighting) by varying the energy available, for example by connecting the electrodes to capacitors of different capacities having been charged at the same voltage, or by varying the charging voltage, for the same capacity, the ignition distance varies, in both cases, in the same direction as this voltage or this capacity, at least in the field where the distances and the energies involved can vary. Experience thus confirms what has been established previously in a more theoretical manner.

It has been said that a defect encumbering EDM machines without pre-ignition is that the ionization of the channel is obtained by the same voltage which supplies the erosive pulses, which prevents the parameters of the 'ignition. In this regard, we note that a good ionization of the channel depends mainly on the voltage applied for lighting, in connection with the energy available for this purpose, while the erosive efficiency of the pulse is essential. partly a function of the current of the erosive pulse. The tension required to maintain the erosive discharge, once the channel is lit, varies only very little with the distance from GAP (d). On the other hand, the voltage required for lighting the channel is practically - all other things being equal - proportional to this distance from GAP. If the tension

(more precisely the EMF) causing the erosive discharges is too high, there is, as the tests show, in particular excessive wear of the electrode-tool, and moreover, the risk of creating an arc is increased electric, whose permanent nature blocks electro-erosive operation. If, on the other hand, you want to be able to switch on the channel using a relatively low voltage, you must have an extremely small GAP distance, which results in many troubles, such as for example the risk of short-circuiting by particles of matter, in addition to the fact that the parameters of the ignition are not controlled. In fact, in machines which do not use pre-ignition, the compromise necessary between the two aforementioned requirements (ignition, erosion) limits the possibilities of adaptation of the erosive impulses to the technological conditions of the machining, and it also limits the choice of the parameters relating to ignition, which themselves condition the characteristics of the ionized channel which is established. When there is applied, on the GAP electrodes, a voltage (capable of delivering a large current pulse) sufficient to maintain an erosive discharge but insufficient to ignite a channel, taking into account a relatively large distance from GAP , no discharge occurs due to the application of this single voltage. If, under these conditions, a short pulse is applied, at a significantly higher voltage, a channel is lit and, from that moment, the lower voltage can deliver the erosive pulse at strong current.

The principle of pre-ignition thus achieves a separation of functions, and therefore of the parameters which direct them. Leaving aside the questions of the erosive pulse itself, we focus, in the context of the present invention, on the parameters of the pre-ignition pulse. We have seen that a good quality pre-ignition requires control of the pre-ignition voltage, and, if possible, of its energy.

Note that if an adaptation must be made between the voltage of the pre-ignition pulse and the GAP distance (d), this adaptation does not necessarily relate to the electrical voltage parameter, but it can also - and advantageously - be fai¬ you by acting on the distance of GAP which, in any case, is most often controlled by a device for automatic positioning of the tool electrode. Under these conditions, it is possible to admit a practically fixed pre-ignition voltage and wait for the electrode positioning device to establish, in relation to this voltage, the appropriate GAP distance. It is also necessary that the pre-ignition voltage has indeed a good stability because, as the pre-ignition pulse is extremely fast, the slightest disturbance of an electrical or material order can significantly modify the peak voltage of the pulse. (the duration of which is most often significantly less than a microsecond). In known devices using the pre-ignition principle, tests have shown (with an oscilloscope) that even the voltage of the pre-ignition pulses in relatively stable mode were subject to almost random fluctuations. Other tests, which could only take place within the framework of the proposed measures

_ _ - ,. p_ by the present invention, have shown that there were also significant, almost random, fluctuations in the energy of the pre-ignition pulses.

The advantages of the pre-ignition system can be listed as follows: 1) With a relatively high ignition voltage, the operation of the circuit will intervene for dis-. Relatively large GAP tances, in the sense that, with this distance, the voltage used only to supply the erosive pulses will not be able to ignite the channel. The difference between the two voltages represents security with regard to short circuits and arcs, it makes the machining process particularly stable and safe, it allows control of electrical parameters by separating the functions of 'ignition and erosion.

2) The ionization of the channel being independent of the electrical parameters of the erosive pulses, it becomes possible to vary the latter in extremely large proportions by easily adapting them to the best technological conditions while maintaining perfect stability and efficiency. of usi¬ swimming.

3) It is known in the EDM technique that the wear of the tool electrode is strongly influenced by the importance of the machining GAP. This wear increases notably when this distance of GAP decreases. This explains the difficulty of obtaining, without pre-ignition, low wear of the electrode in the regimes where the GAP must be particularly low, that is to say in the finishing regimes, having to be done at pulse energy. relatively low, that is to say with a short distance from GAP, if we stick to the explanations previously stated.

^' . ,. <_ _ ^"' ~ The application of the pre-ignition makes it possible to have a significantly higher GAP distance regardless of the parameters of the erosive pulses. It thus becomes possible to reduce the wear of the electrode-tool in all regimes, and in particular in the finishing regimes where this wear, at equal value, would be most regrettable, but where precisely this wear can be reduced to a value close to zero by application of the pre-ignition principle, at least under the particularly advantageous conditions established in the context of the present invention. 4) Frequent disturbances of the conditions in the machining area which destabilize the working process by causing short circuits, continuous arcs and other harmful phenomena, require the application of "adaptive control""parameters of the erosive pulses as well as the position of the electrode-tool. The complex nature of these disturbances makes it necessary to take into account a large number of factors, hence the use in very conventional devices without pre-ignition, of very complicated systems for carrying out a valid adaptive control. However, experience shows that whatever the nature of these disturbances, they are always reflected in the conditions of the ionization of the channel. The application of the pre-ignition with an automatic adaptation of the energy of the pre-ignition pulses to the single ionization factor makes it possible to obtain perfect machining stability without any modification of the parameters of the erosive pulses and without sudden changes in the position. of the tool electrode. 5) The stability and safety of the machining process make it possible to use the simplest erosive pulse generator circuits such as, for example

O PI example, of the capacitive type (LC or similar). These circuits then give comparable and even superior technological results to the most sophisticated circuits used today. 6) The channel pre-ignition system can be applied to all systems for generating erosive pulses, giving them all the advantages mentioned above. In view of the above considerations, an object of the present invention is to provide an electrical circuit for an electro-erosion machining machine working according to the principle of pre-ignition, in which the drawbacks of the prior art are avoided, and in particular in which the parameters which condition the quality of the ignition of the channels and of the electro-erosive machining are adequately controlled.

In accordance with the invention, the electrical circuit for an EDM machining machine, in accordance with the generic definition previously stated, achieves the targeted performances due to the presence of the characters stated in one or other of the claims independent. It is noted that the characters, either of one or of the other, independent claims intervene respectively for different aspects of the problem of stabilization of the pre-ignition. The dependent claims define embodiments of the subject of the invention which are particularly advantageous, in particular from the points of view of the voltage stability, of the detection of tendencies in the increase of pre-ignition energy, of the constitution of circuits allowing a stabilization of the energy, manually or automatically, of the simplicity or the convenience of use of the circuits generating the pre-ignition pulses, etc.

Regarding the features of the invention,

WIPO it should be noted that the clipping means which are mentioned can be either with voltage limiting characteristic, but not voltage stabilizer (pe) Zener diode in series with significant resistance) or with truly current stabilizing characteristic (Pe diode Zener without notable ohmic resistance in series).

As the machining GAP is presumed to have an electro-disruptive characteristic, the connection in parallel with it of a clipping circuit also presumed to be electro-disruptive may seem para¬ doxal. In fact, all things considered, this paradox does not intervene, given that on the one hand the machining GAP is not the equivalent of a conventional electro-disruptive element, and that, on the other on the other hand, there are means which control the distance of GAP to the parameter of the pre-ignition pulses, that is to say which maintain substantially equal the two voltage values "apparently electro-disruptive" (but which in reality at at least one admits a certain variation allowing them to be matched to one another).

With regard to the aforementioned means for servo-servicing the distance from the GAP, that is to say, the means, generally motorized, for repositioning the electro-tool relative to the workpiece, it should be noted that , in certain particular embodiments, such means could be dispensed with. This would be the case in particular with an EDM machining machine in which the electrode is intended solely for printing a shallow bas-relief in the workpiece. The latter could be put in place, the electric circuit being stopped, with the tool electrode pre-positioned just at the desired distance, after which the electrical means would be put into operation to print the bas-relief in question, without displacement of the tool electrode, whereupon the machining would be finished, the machined part being removed to make room for the next part to be machined. For this reason, it does not appear appropriate to include the mention of the electrode positioning means in question in the most general definition of the subject of the invention.

The accompanying drawing illustrates, by way of example and starting with presenting an embodiment known from the prior art, embodiments of the subject of the invention; in this drawing: fig. 1 shows a circuit for machining machine for electro-erosion, using the pre-ignition system, known from the prior art, in particular by US 3 893 013, FIG. 2 shows the diagram of an electrical circuit for an EDM machine according to the invention, this FIG. 2 showing in more detail the part used for the generation and application of the pre-ignition pulses, FIG. 3 is a diagram of a form of execution of a circuit according to the invention, similar to FIG. 2, but further comprising signaling means providing a perceptible indication of an increased state of pre-ignition energy, as well as manual control means making it possible to reduce this energy, in particular on the basis of said indication, fig. 4 is a diagram of a form of execution of a circuit according to the invention, similar to that of FIG. 2, but also having automatic means for limiting the pre-ignition energy, FIG. 5 is a partial diagram of a variant which can advantageously in certain cases be used to replace the circuit part lying about the primary winding of a transformer for generating the pre-ignition pulses, fig. 6 is a block diagram, accompanied by a diagram representing the relative evolution of the voltage and the current, which illustrates the operating principle of the entire circuit in the case where the means of Erosion discharge delivery works on the relaxation principle, with a machining capacitor charging and discharging periodically, and ^* fig. 7 is a block diagram, accompanied by a diagram representing the relative evolution of the voltage and the current therein, which illustrates the operation of the electric circuit in the case where the means for delivering erosive discharges comprise organs switching electronics delivering a waveform with rectangular slots.

In the circuit for an EDM machine shown in FIG. 1, previously known from the disclosure of US Pat. No. 3,893,013, there is a machining capacitor C, supplied by a direct current source via impedances R. And L, _, ch ch which charges up to a certain voltage to discharge periodically, in the form of electro-erosive pulses, in a machining GAP d. An adjustable inductance L, makes it possible to adjust the duration of the pulses and softens the jumps of current, in particular at the moment when the electro-erosive discharge begins. D and D diodes make the erosive pulses unipolar, preventing the appearance of oscillations during discharge. An inverter I reverses the polarity between the electrode and the workpiece.

The principle of generation of electro-erosive pulses in the circuit of fig. 1 is analog see the one illustrated in fig. 6. However, the circuit of fig. 1 comprises, according to the prior art, still little perfected pre-ignition means, whereas FIG. 6 illustrates the very principle of the generation of electro-erosive pulses as one of those to which the invention can be applied (although the elements which characterize the subject of the invention, namely those which relate to pre-ignition, are not shown in this fig. 6). Returning to fig. 1, it can be seen that the circuit cooked according to this figure comprises pre-lighting means including a voltage step-up transformer T, a Zener diode DZ, a variable resistance R .., a variable resistance R_ and a diode D_. It is well understood that, when the charge voltage of the capacitor

C reaches a value determined by the voltage of the diode -Zener DZ, somewhat lower than the supply voltage, a current begins to flow in the primary winding of the transformer T, through the resistor R,. The increase in this current induces a pulse, of higher voltage, in the secondary winding of the transformer, which is connected in parallel, on the GAP d, through the diode D and the (adjustable) resistor R_. The energy that this pre-ignition pulse can deliver depends on the parameters of the components of the pre-ignition circuit, the pre-ignition voltage falling naturally when the pre-ignition current occurs, but in a way that is difficult to determine. In this device, neither the pre-ignition voltage nor the pre-ignition energy can be adequately controlled.

Experience has shown that the circuit according to the prior art shown in FIG. 1 could only be used in a very restrictive manner in 19 real industrial conditions. The difficulties arose in particular from the following reasons:

- Due to the mode of connection of the windings of the transformer T, it had to present a transformation ratio having a high value of voltage rise, which caused a significant consumption of the current coming from the capacitor C. In addition, this situation had the effect that the transformer T had inductive characteristics which were not very favorable to pre-ignition and in particular to the control of its parameters.

- After the current increase phase, a large direct current passed through the primary winding of the transformer, the Zener diode, and the resistor in series with it, a large part of the energy being thus lost (and causing as the case may be excessive heating). This situation also harmed the power of the machining.

- In the cases where the charge of the machining capacitor was characterized by very steep curves, as in the case where one would have liked to use erosive pulses of the rectangular type generated electronically, the transformation of a part of the energy in pre-ignition pulses was practically no longer possible as a result of insufficient efficiency.

- The pre-ignition pulses being entirely dependent, as regards their voltage and current, that is to say their energy, the machining power and the characteristics of the erosive pulses, the parameters of the pre-ignition pulses do not could hardly be controlled, even getting out of hand. Significant variations in the erosive pulses produced very harmful parasitic effects. We

O often encountered a spontaneous increase in the energy, and if necessary in the voltage, of pre-lighting, which caused excessive ionization of the dielectric medium, favoring short circuits and continuous arcs.

- In addition, parasitic phenomena (such as those mentioned above or others still) provided the servo-control system with automatic advance (or automatic positioning) of the tool electrode with often incoherent information from so that the servo motor was unable to properly maintain the machining GAP value, making the entire machining process unstable.

- In short, the use of this circuit was practicable only under very limited, special conditions of voltage, current and usi¬ ing capacity, which could exist for example for a usi¬ weak power. In addition, pre-lighting means did not allow the principle to be applied to different systems for generating erosive pulses.

Thus, it can be seen that the principle of pre-lighting can only be correctly applied, so as to provide all the benefits which may arise therefrom, if certain conditions are fulfilled. A performance targeted by the invention consists precisely in satisfying these conditions, which are as follows: 1) The electrical parameters of the pre-ignition pulses must, to a large extent, be set only as a function of their ionizing effect (polarization + ionization), remaining as much as possible independent of the speed and of the machining power, of the current waveform of the erosive pulses, and of other characteristics of the generator

O PI of pulses. 2) The voltage of the pre-ignition pulses must be limited, if possible strictly, to a defined and constant value, because it is this voltage which, through the play of electrode positioning means, determines the value of the GAP, just as it determines the value of the "cutting gap" (that is to say the difference between the dimensions of the tool electrode and those of the hollow machined in the part). 3) The energy of the pre-ignition pulses must be only a small fraction of the energy of the erosive pulses and must only be used for ionizing the channel; there should be as little mutual influence as possible between the parameters of the pre-ignition pulses and those of the erosive pulses.

4) To ensure regular ionization of the channel under the different operating conditions and machining power, it is good for the energy of the pre-ignition pulses to be adapted to these conditions, with at least one possibility of manual adjustment, and if possible a possibility of automatic adjustment.

5) The instant of application of the pre-ignition pulses must be adequately determined as a function of the electrical potential available for the erosive pulses.

6) The duration of the pre-ignition pulse, that is to say of the ionizing effect, must be as short as possible; the pre-ignition pulse should drop as soon as the channel is ionized and the erosive pulse current begins to flow.

Experience shows that the best technological results, i.e. the greatest removal of the machined material and the lowest wear of the tool electrode are obtained when the conditions 22 of the above are met. It is clear that certain conditions may be more imperative than others, the first rank returning to the regularity of the pre-ignition voltage, followed very closely by the conditions relating to the pre-ignition energy.

The object of the invention, examples of which will now be described in conjunction with FIGS. 2 and following, achieves a very complete mastery of the parameters of the pre-ignition pulses, making their ionizing effect adequate, whatever the operating speed and power conditions, and without excessive dependence either on other technological conditions. Another object of the invention is to provide this control automatically. In the circuit according to the invention which is schematically represented in FIG. 2, the means for delivering erosive discharges 1 may be similar to all of the capacitor C and of the peripheral components shown in FIG. 1. However, the erosive discharge supply circuit 1 of FIG. 2, can also be constituted in another way, for example using electronic communications means. The upper parts of figs. 6 and 7 illustrate, moreover, two different ways in which the supply circuit 1 for erosive discharges can be formed and can operate.

In fig. 2, the material device 2 for machining by electro-erosion comprises a tank filled with dielectric liquid and in which an electrode-tool El is located up to the top of a workpiece Pi. The circuit 1 for delivering electro-erosive pulses is connected to the terminals of the two electrodes that constitute the electrode-tool on the one hand and the workpiece on the other hand, only by means of 23 diary of a diode 7 necessary for the pre-ignition pulses to be applied. A diode 10 prevents any reversal of voltage polarity across the terminals of the GAP, that is to say on the electrodes Ei and Pi.

For the rest, the diagram in fig. 2 il¬ only shows the pre-ignition circuit in detail. It can be seen that it comprises a transformer 6, voltage booster, in which the pre-ignition pulses are generated. An adjustable resistor 5, a capacitor 4 and a Zener diode 3 are connected in series with the primary winding of the transformer 6, all of this series branch being connected to the output terminals of the supply circuit 1. Furthermore, the secondary winding of the transformer 6 is connected, via a limiting resistor 8 and a non-return diode 9, to the terminals of the diode 1, previously mentioned.

It is well understood that, before the start of the erosive pulse, the supply voltage delivered by the circuit 1, which had been made zero at the end of the previous erosive pulse, begins to increase and comes to exceed the voltage characteristic of the Zener diode 3. The capacitor 4 then begins to charge through the resistor 5, the Zener diode

3 and the primary winding of the step-up transformer 6. This current, first limited by the transformer self-inductance, then increases rapidly, as long as the supply voltage is in an increasing situation. The voltage induced by the increase in current in the primary winding of the transformer is also present in the secondary winding of the latter, in the form of a pulse of a significantly higher voltage, capable only of translating into a 24 relatively low current, which voltage is applied to the terminals of the diode 7, then polarized in the blocking direction. Thus, the voltage across the terminals of the GAP d is equal to the sum of the supply voltage in erosive pulses (output terminals of circuit 1), and of the pre-ignition pulse (secondary winding of the transformer 6).

Note that the capacitor 4 serves to prevent the current in the primary winding of the transformer from being determined exclusively by the characteristics of the Zener diode, the ohmic resistance and the self-inductance of the transformer. Experience has shown that, without a capacitor, the pre-ignition pulse is lost, probably due to the fact that it comes with a flank that is too stiff or burdened with too high series impedance. On the contrary, with the capacitor 4 in series, one obtains, as tests have shown, an adequate and effective pre-ignition pulse. It is likely that reverse polarity charging and discharging processes take place in capacitor 4. Without going into an in-depth study on this subject, we note in any case that the presence of capacitor 4 has a beneficial effect on the pulses of pre-ignition. In short, it is noted that the presence of the capacitor 4 has the following advantages:

1) Elimination of direct current circulation in the primary winding of the transformer

2) Better determination of the time of the increase of the current in the primary winding of the transformer, making it possible to ensure the generation of the quantity of energy necessary for the pre-ignition and allowing to control the instant of application of the pre-ignition.

3) Maintaining the pre-ignition voltage at least for for the time necessary for correct ionization of a channel in the GAP. 4) Return to the machining circuit of the amount of energy drawn in excess. 5) Possibility, by adequate choice of the capacitor 4 and the value of the resistor 5 _t, to ensure the production of the pre-ignition pulse for a short time, just sufficient to effectively ignite the channel but not likely to cause an excessive prolongation of the ionizing effect.

The effect of the series connection of the condenser 4 is thus advantageous in that it provides a pre-ignition pulse of adequate power. However, there is something else which contributes to the rational application of the pre-ignition principle. In fact, despite the fact that a regular generation of the pre-lighting pulses is obtained in the above-mentioned manner, the variations in the conditions in the machining zone itself (that is to say in GAP), due to fluctuations in the characteristics of the dielectric medium - such as an accumulation of machining residue in the form of conductive particles, gas formation, carbon deposition, etc. - can cause variations in the resistance of this medium, which makes it difficult to control fluctuations in the voltage of the pre-ignition pulses.

This is the reason why there is provided in the circuit according to the invention a branch for clipping the pre-ignition voltage, composed in this case of a Zener diode 11 (possibly several Zener diodes in series) and a resistor 12 in series with this Zener diode, all of these elements being connected directly to the terminals of the GAP.

WIPO 26

This stabilizes the voltage of the pre-ignition pulse.

In the comments set out prior to this description, it has already been indicated that 5 the parallel connection of two elements having a characteristic of electro-disruptive type was not paradoxical, contrary to appearances. This is due on the one hand to the fact that the GAP does not have a conventional electro-disruptive characteristic and this is

10 also due to the fact that there is most of the time, in a manner not shown in the drawing, a device for controlling the position of the electrode, or for automatic electrode displacement, which, when pre ¬ feels a given voltage to pre-light a channel,

15 substantially automatically establishes the distance from GAP to the desired value. In this case, the characteristic of the series assembly 11, 12 could be not only a throttle, but a voltage stabilizer, that is to say that the maximum voltage drop envisaged at

20 terminals of the resistor 12 will be small compared to the characteristic voltage of the Zener diode 11. The presence of the resistor 12 is in fact explained by the use that will be made of it in improvements which will be explained later in conjunction with

25 fig. 3 and 4. It is noted that, in the absence - always pos¬ sible - of an automatic electrode positioning device, the resistor 12 would advantageously be made larger, so as to establish an ohmic voltage drop suitable for sharing the current between the

3 ₀ GAP and the Zener diode 11 even in case the distance of GAP d is not entirely adequate.

With the clamping branch 11, 12, the cooked circuit of FIG. 2 provides stabilization of the voltage of the pre-ignition pulses. However, we have seen that,

. 1 ^~ . ~ ('^' _ ~ '" ^{fpI ""}'^' IFO to obtain a very good pre-ignition pulse, it was also important to control the energy parameter of the pre-ignition pulse. The other embodiments which will now be described respectively in conjunction with FIG. 3 and with fig. 4, moreover provide the means of controlling the energy parameter of the pre-ignition pulses.

It can be seen that the embodiment according to FIG. 3 includes all the elements which were in FIG. 2, with in addition a transistor 14 for limiting the pre-ignition current, an auxiliary voltage source 15, a variable resistor 16, a non-return diode 18 and a signaling device 17. The emitter-collector section of the transistor 14 is connected in series with the resistor 8 and the diode 9 already considered in connection with FIG. 2. This emitter-collector section is controlled by the current in the emitter-base section of this transistor 14. This current, coming from the auxiliary voltage source 15, is adjustable by means of the variable resistor 16. More precisely, this resistor variable 16 makes it possible to establish the maximum value of current which can pass through the emitter-collector section of transistor 14, during the pre-ignition pulse. The signaling circuit 17 is also supplied by the auxiliary voltage source 15 and it includes an input connected to the connection point between the diode 11 and the resistor 12, in the clamping branch already considered in connection with its fig. 2. The signaling circuit 17 can be designed in several different ways. It may for example be an electronic voltmeter memorizing the peak voltage drop values in the resistor 12, which are representative of the values of the poin- te of clipping current. The signaling circuit 17 could also be an oscilloscope showing the evolution of the clipping current, it could also be an electronic circuit detecting only the fact that the voltage drop on the resistor 12 (increased by the drop in almost constant voltage on the diode 18 in the passing direction) has poin¬ your which exceed or do not exceed a given threshold value. Depending on the case, an indicator 17a will provide an indication allowing an operator to know the desired information as for the voltage drop in the resistor 12. This indicator 17a could be for example a lamp or an optical device, indicating that a threshold is exceeded or not by the tips of the clipping current, it could also be a digital indicator, providing the very value of this current, or an analog needle indicator, etc.

With this embodiment according to FIG. 3, an operator perceives the indication provided by the member 17a of the signaling circuit 17 and he can adjust the variable resistance 16 so as to decrease (or increase) the maximum value of the current which can pass through the emitter-collector section of the transistor 14. i.e. the maximum value that the current available for pre-ignition can have.

The current which the transistor 14 allows to pass, however, is not, in all cases in the general case, the total current which passes through the GAP, since a part of this current passes through the clipping branch, 11 , 12. The distribution of the pre-ignition current between the clipping branch and the GAP is conditioned by the distance d of the GAP, which must be adequately established. In most of the cases,

WIPO this distance of GAP is established by a device for controlling the position of the electrode which does not automatically give the desired value at this distance d. We will see later how such a device can, for example, be automatically controlled. In principle, the conditions are such that the current in the clamping branch 11, 12 is approximately of the same order of magnitude as the pre-ignition current in the GAP. The embodiment according to fig. 3 has the advantage of having a "limit value of the pre-ignition current" which is normally established in a fixed manner, according to the adjustment of the variable resistor 16. This limit value of the current available for pre-ignition is therefore sort of memorized by the value given to the adjustable resistor 16, taking into account the voltage of the auxiliary power source 15.

The adjustment or adjustment of the available pre-ignition energy (maximum value of the current in the emitter-collector section of the transistor 14) is manually adjusted, in a manner thus remaining memorized, in the embodiment according to FIG. 3. On the other hand, in the embodiment according to FIG. 4, this adjustment occurs automatically, a limitation of the current in the emitter-collector section of the transistor 14 intervening only when the current in the resistor 12, that is to say the clipping current, exceeds a certain limit value . For this, in the form of execution according to FIG. 4, there is again the transistor 14, the emitter-collector section of which is capable of limiting the value of the current available for pre-ignition (that is to say the sum of the clipping current and the current of pre-ignition). On the other hand, in this form of execution according to FIG. 4, transistor 14 is controlled using a circuit comprising a transistor 22, pre-ignition current control command, an auxiliary voltage source 15, a fixed resistor 23 and an adjustable resistor 24. A connection of the resistor 23 is connected to the pole negative of the auxiliary voltage source 15, and is connected at the same time to the connection point between the Zener diode 11 and the resistor 12, in the clipping branch. The resistor 23 is above all a resistor for neutralizing the leakage current of the transistors 14 and 22. When no current crosses the resistor 12, the negative pole of the auxiliary power source 15 is at the same potential as the electrode-tool El, that is to say that almost all of the voltage of the auxiliary source 15 is applied to the variable resistor 24 The transistor 22 is mounted, as seen in "emitter-follower" and the current which is delivered by the emitter of transistor 22 is approximately equal (as long as one can consider in resistance 23 as practically negligible) to the quotient of the voltage of the auxiliary source 15 by the value of the variable resistance 24 , multiplied by the amplification coeffi¬ cient of transistor 22. It is this current, coming from the emitter of transistor 22, which controls transistor 14 by its emitter-base section. When a voltage drop occurs on the resistor 12, the negative pole of the auxiliary voltage source 15 goes to a lower potential compared to that of the tool electrode And, and the voltage which remains applied to the variable resistance 24 decreases correspondingly, automatically causing a reduction in the current which controls the transistor 14. If the voltage drop in the resis¬ tance 12 becomes large, the transistor 22 could even be blocked completely, also blocking the transis- tor 14, but precisely at this time the current in the resistor 12 would necessarily decrease, so that the blocking state does not happen to occur and that one tends towards a situation "stabilized by effect c of reaction". The value of the current in the resistor 12 for which this stabilized situation occurs can, as is easily understood by applying the rules of the electronic circuits, be adjusted using the variable resistor 24. 0 In operation the maximum value of the authorized current to cross the emitter-collector section of the transistor 14 is not fixedly established, as in the case of the embodiment according to FIG. 3, but it is, a priori, very high. It is only at 2_5 when a large clipping current occurs that this current limit value in transistor 14 is reduced, by the effect of transistor 22 and the elements associated with it.

The advantage of the circuit according to fig. 4 ₀ 2 is that if the current clipping tends to increase the current limit through the transistor 14 automatically becomes more severe, so that the system automatically adapts to different conditions likely to occur (eg in the case of "hesitant" ionization, at the moment when the tool electrode is distant from the workpiece) • An advantageous effect of the circuit for automatic limitation of the current available for pre-ignition, according to FIG. 4, you consis¬ in that, when the tool electrode is 3 to ₀ for one reason or another, remote from the workpiece, the interruption of the preignition pulses is free, there is no a period where pre-ignition occurs "at the limit of the possible", which opens only an ionized channel of poor quality.

In fig. 4, we also planned, but in a manner which would not be compulsory, a first semi-clipping branch 19, 20, 21, which acts upstream of the transistor 14, but downstream of the resistor 8 and the diode 9. It has been found that the effect of the automatic energy limitation controlled by the transistor 22 was more effective if the available voltage was already clipped, and this is the role of the prior clipping branch 19, 20. It is however very clear that the total voltage drop in the Zener diode 19 and the resistor 20 must forcé¬ be greater than the total voltage drop in the diode 11 and the resistor 12, otherwise the effect of the clamping branch itself , 11, 12, would be zero. Anyway, the diode 21, provided for all practical purposes, is not essential, even in the case where the prior clipping (or semi-clipping) branch comprising the Zener diode 19 is provided. and resistance 20.

The circuit according to fig. 4 ensures complete and automatic control of the electrical parameters of the pre-ignition pulse, namely the peak voltage of the latter and of the energy available for pre-ignition.

Fig. 5 shows a possible modification, and at least in some cases advantageous, according to which the elements which control the flow of current can be arranged in the primary winding of the transformer which supplies the pre-ignition pulses. In the variant according to FIG. 5, we find, in 3 ', 4', 5 'and 6 *, the homologous elements 3-6 of FIGS. 2, 3 and 4, but according to a different provision.

In fig. 5, a transistor 25 is provided, provided with a resistor for neutralizing the leakage current 26, which is controlled by its base by the branch- ment-series of a Zener diode 3 ', corresponding to diode 3 of fig. 2 to 4, and a capacitor 4 ', corresponding to the capacitor 4 of FIGS. 3 to 4. By the way, in series on the emitter-collector section of this transistor 25, there is disposed the primary winding of the transformer, here called 6 ', itself in series with a 5' adjustment resistor. , corresponding to the resistance 5 of FIGS. 3 and 4. A resistor 27 is connected to the terminals of the series assembly comprising the primary winding of the transformer 6 'and the adjustable resis¬ tance 5', so as to form a path allowing, if necessary, the passage of a demagnetization current of the transformer 6 '. The variant diagram according to FIG. 5 presents, similarly to the diagrams according to FIGS. 2 to 4, the property according to which, due to the presence of an element avoiding a direct galvanic connection (in this case the capacitor 4 'corresponding to the capacitor 4 of FIGS. 2 and 4), no current flows from a permanent or permanent way in the pre-ignition means, these being the seat of no current which would be foreign to the function of generation of pre-ignition pulses. On the other hand, the variant diagram according to FIG. 5 allows the application to the terminals of the primary winding of the transformer of a voltage practically equal to the voltage delivered by the circuit 1 which supplies it, this voltage applied to the primary winding being not, as in the case of figs. 2 to 4, reduced at least by the Zener voltage of the 3 or 3 'Zener diode. This allows the use of a transformer with a lower transformation ratio. Ultimately, taking into account that the pre-ignition pulse is added to the voltage supplied by the circuit 1 of erosive voltage supply 1, the transformer 6 'could

OMP not even be a voltage booster, since, in principle, it suffices that the pre-ignition voltage is at least notably greater than the voltage of the electro-erosive pulses. In practice, since the pre-ignition voltage is preferably greater than twice the voltage of the erosive pulses, the transistor 6 ′ will nevertheless be at least slightly voltage-boosting, but this is not an absolute necessity.

In fig. 5, there is also shown, in dotted lines, a possibility of complement which may be interesting. This is the circuit part which includes a resistor 28 in series with the resis¬ tance 27, a transistor 29 whose emitter-base section is connected in series on the resistor 28 and a resistor 30 constituting a resistor of load connected to the collector of transistor 29. A coupling capacitor 31 also connects the collector of the reaction transistor 29 with the base electrode of the ignition control transistor 25. Because of this arrangement, when the transistor 25 has become conductive, the transis¬ tor 29 is also conductive and, via the condenser 31, this conduction state is maintained at least during the charging time of this capacitor 21, even if the current through the Zener diode 3 'and the capacitor 4' was no longer sufficient to control the conduction of the pre-ignition control transistor 25 on its own.

Note that the circuit of fig. 5, both with the addition drawn in dotted lines and without it, is a circuit of the type which does not add any current at rest.

In fig. 5, there is also shown, at 32, the possibility, likely to be very interesting, of providing an additional input connection

WIPO causing, on request, the emission of an additional pre-ignition pulse. In fig. 5, the possibility has been shown of bringing the additional pre-ignition pulse control 32 to the base electrode of the reaction coupling transistor 29, but it is clear that this additional control pulse could be brought at any other point on the circuit.

Note that, with the circuit components drawn in dotted lines, the diagram in fig.

5 gives the pre-ignition control means the characteristic function of a monostable rocker, which can make it possible, for example, to calibrate the pre-ignition pulse in time independently of the control supplied by the diode Zener 3, 3 'and capacitor 4, 4', control which, in turn, is dependent on the evolution of the voltage delivered to the output terminals of circuit 1, which will provide the erosive pulse . It is also necessary to consider figs.

6 and 7. These two figures each consist of a schematic diagram (upper part) and a diagram of the evolution of voltages and currents (lower part). Fig. 6 illustrates the fact that, as a circuit for delivering erosive discharges, a circuit of the "relaxation" type can be used, such as, without regard to pre-ignition questions, the prior art knew of it, in particular by the description of the previously cited USA patent. At the terminals a, b, a ', b ¹ of fig. 6, the pre-ignition arrangement is connected, as shown in FIGS. 2 to 4.

The lower part of fig. 6 shows how, on the one hand, the voltage and, on the other hand, the current in the GAP evolve. The tension, established at

OM? I terminals of the main pulse delivery capacitor (or machining capacitor) first increases to a value U_, at which the pre-ignition process is triggered. The voltage then passes through the value U which constitutes the pre-ignition peak voltage, then, with the channel on, this voltage drops to a relatively low value, of the order of 30 to 40 V, where it is maintained, with an evolution substantially flat, until the end of the pulse. In addition, the current begins to gradually increase when the channel has been turned on by the voltage spike.

U, and it follows an evolution corresponding substantially to an alternation of a sinusoid, passing through a maximum and then returning to zero. When this current returns to zero, the channel deionizes and it cannot flow in the opposite direction (this would in any case be prevented by diodes 7 and 10).

In the upper part of fig. 6, we also note that there is, in series with the main machining condenser, an adjustable induction coil which is responsible for supporting the voltage jolt which occurs when the channel has just been turned on. This inductive impedance thus contributes to "softening" the current jumps which would otherwise be excessively large.

Fig. 7 illustrates, in a manner similar to FIG. 6, the possibility of having the electro-erosive discharges delivered not by a capacitor circuit, that is to say mainly of the relaxation type, but by an electronic circuit comprising "en" switches and "off", applied to the output of an energy source capable of delivering an adequate voltage (voltage U) to control the pre-ignition, and

_ ^ _ - ' a current high enough to constitute the electro-erosive pulse, the voltage then being of course dropped to correspond to the maintenance voltage of the current of erosive pulses. The lower part of fig. 7 shows that, with the use of such an electronic circuit for delivering electro-erosive discharges, the function of the pre-ignition circuit is practically not modified. On the other hand, the erosive current pulses are, as we can see, much more "square".

The connection points a, b, a 'and b' of fig. 7 are, like the homologous points in FIG. 6, intended for the insertion of the circuit part concerning the pre-ignition. We have spoken, on several occasions, of the means of positioning the electrode-tool, which exist in almost all of the machines for machining by electro-erosion, and which, naturally, are normally also present in the machine controlled by the circuit. according to the invention. It has even been said that it is the presence of these electrode-tool position control means which allow the clipping system to develop all of its advantageous effects, in particular in the case where the resistance 12 is of low value and where the clipping branch, formed essentially from the Zener diode 11, not only has a voltage reducing characteristic, but a voltage stabilizing characteristic.

In general, the means for positioning the tool electrode include a motor which is controlled to bring the electrode closer to the part or to move it away from it. This engine is controlled by conventional means; in order to determine its functioning, we generally have means which react

WIPO with different parameters, electrical, even mechanical, which can be understood on the circuit or the machine. Substantially, there are three situations, that where an approximation of the electrode is necessary (the motor turns in one direction), that where no movement of the electrode is necessary (the motor is stopped), and that where a distance from the electrode is necessary (the motor turns in the other direction). Hereinafter, we will mention some of the numerous methods according to which different parameters can be treated as criteria for determining the control of the electrode positioning motor. A first method which will be described is suitable for machines according to the prior art as well as for machines whose circuit corresponds to the invention; another method, possible for the prior art, is not applicable for a machine having a circuit such as that which the invention proposes, but, with modifications, this second method can be applied to the circuit according to the invention, and then it constitutes a particularly advantageous method for controlling the electrode positioning motor. In the drawing, we have marked, in different places, indications x, y, z, m, q and w. It is to the voltages which occur respectively between some of these points that one can refer, as a criterion for controlling the positioning of the electrode. It will be noted that by designating a voltage by two of the aforementioned points, the voltage of the first of these points relative to the second of these points will be designated (for example, the voltage xy and the voltage on the resistor 12, x being positive with respect to to y).

The first method for determining

^" JTJRE WIPO necessary commands of the electrode positioning motor consists in first of all entering as information the signal qx (fig. 2 to 4) or mw (fig. 1). If this signal is constantly positive, it means that there is very probably a short circuit or an arc in the GAP, and it is necessary to command a distance of the electrode. If this signal is constantly zero or negative, this means that there is never a significant current in the GAP and it is advisable to order an approximation of the electrode. Finally, if there are positive voltage pulses, separated by time intervals with zero or negative voltage, it is advisable to consider, as a second criterion, the signal xm (fig. 1-4). If this signal has pulses whose tip reaches a value close to a setpoint (designated as V), this means that the distance from GAP is suitable and the motor must remain stationary. If this signal xm has voltage spikes which nevertheless remain significantly lower than the said setpoint value V, this means that the distance of GAP is too small and that it is necessary to actuate the motor to move the electrode away. On the contrary, if the voltage peaks of the signal xm are notably higher than the said setpoint value V, it is because the motor must be controlled so that it causes a reduction in the distance of GAP.

It is noted that the possibility of an absence of impulse in the signal xm would only exist in the two cases where the situation is already determined by the signal qx (or mw), and where the signalxm does not intervene.

Another control method, practicable according to the prior art, consists in using only the signal xm and in carrying out only the determination that the first method made in the alternative, in

i RE WIPO 40 function of this signal xm. It is noted however that one can then have cases where there appears no impulse on the signal xm. This can just as well correspond to a short-circuit in the GAP, requiring a distance from the electrode, as to a supply voltage too low, requiring a rapprochement of the electrode. To get out of this difficulty, it was planned to carry out in this case first of all a separation from the electrode which, in any case, eliminates the short-circuits and suppresses the discharges in the gap, followed by a gradual approximation of the electrode, until the pulses resume normally. Note however that this second method, applicable according to the prior art, is not applicable as presented with a circuit according to the invention (fig. 2 to 4). Indeed, in this case, the setpoint voltage Ves will naturally be the voltage

Zener of diode 11, and the xm signal can never significantly exceed this setpoint. However, this second method had the advantage of not having to measure a voltage drop on a diode such as diode 7 (fig. 2 to 4) or diode D_ (fig. 1).

The circuit according to the invention nevertheless provides a very advantageous possibility of practicing a method close to the second method previously considered, but using another criterion to know the situations where the electrode must be distant. In fig. 2 to 4, we see that we have drawn, in dotted lines, a resistor 12a which is in series on the supply of the pre-ignition voltage and current. This additional resistance has approximately the same value as resistance 12. We can then measure the two voltages zx and xy. If these two measurements give at least approximately the same value of current, this means that all the current delivered by the secondary winding of the transformer 6 is directed to the clipping branch 11, 12, that is to say that the GAP is not subject to an active pre-ignition pulse (the voltage may appear but it is insufficient to establish pre-ignition). It is deduced therefrom that the motor must be controlled so that it brings the tool electrode closer together. In addition, concerning the indeterminacy which existed in the prior art and required a "back and forth" of the tool electrode, it is noted that, in the circuit according to the invention, the case in which no pulse occurs on the signal xm can be treated first, assuming that there is the variant according to fig. 5 including what is drawn in dotted lines, by sending a pulse to the connection 32, so that the circuit attempts to cause the application of an additional pre-ignition pulse. If this pulse occurs and it actually causes the pre-ignition, the machining process will continue normally, there will be at most a correction of the distance of the electrode, if the pulse xm is too low. If no pulse occurs after the application of a control pulse on the input connection 32 (fig. 5), it is likely that there is a short- circuit and it is then advisable to cause a distance from the electrode. Finally, if the pre-ignition pulse occurs but it does not cause any effective lighting of the channel, then we will certainly have the detection, by means of the two signals zx and xy, of the previously mentioned situation in which there is reason to bring the electrode closer. Thus, with the circuit according to the present invention, using only- ment voltages on points x, y, z, m, it is possible to realize an automatic control at the same time very simple and very advantageous of the device (with motor) of positioning of the electrode-tool (EÊ) compared to the part to be machined (Pi). In the case where the possibility of controlling an attempt to send pre-ignition pulses does not exist, the case of the absence of pulses in the signal xm can naturally be treated as it was according to the prior art. However, it should be noted that, with the circuit according to the invention, the risk that this difficult situation arises is extremely low.

Claims

CLAIMS:

1. Electrical circuit for an EDM machining machine working a part (Pi) by electro-erosive charges in a dielectric medium, in a machining space (GAP, d) between a tool electrode (And) and said part (Pi) which also itself constitutes an electrode, comprising means for delivering erosive discharges (1), having output terminals which respectively provide on said electrodes (And, Pi) erosive electrical energy pulses, and pre-ignition means (6, 3, 9, ...; 6 ', 3', ...) to supply, each time in prologue to a said erosive pulse, pre-ignition pulses which have a higher voltage and a lower energy than the said erosive pulses, and which ignite, that is to say ionize, in the machining space (GAP, d) within the dielectric fluid , each time a channel which is then run through by the erosive discharge, characterized in that it comprises means clipping (11, 12) arranged to reduce the voltage of said pre-ignition pulses on said electrodes (Et, Pi) when it exceeds a set value.

2. Circuit according to claim 1, for a machine for machining by electro-erosion provided with means which control the positioning of the electrode-tool (EL.) To adequately establish the machining distance (d) that is i.e. the dimension (d) of said machining space (GAP) between the two electrodes (Et, Pi), characterized in that said clipping means (11, 12) comprise at least one component (11) with voltage limiting characteristic, the limit voltage value of which corresponds substantially to the said setpoint, this or these components being connected to the said electrodes (Et, Pi) and / or to the output connections of the said means pre-ignition (6, 3, 9, ...) by conduc¬ tion paths substantially free of obstacle dropping the voltage, that is to say comprising at most elements (12, ...) n ' involving no significant voltage drop relative to said limit voltage value.

3. Circuit according to claim 1, carac¬ terized in that the said pre-ignition means (6, 3,

9) are also provided with limiting means (14) allowing both the energy available for each pre-ignition pulse to be reduced.

4. The circuit as claimed in claim 1 or claim 2, in which the said pre-lighting means comprise an arrangement (3-6) which generates, and delivers at its output, each pre-ignition pulse with finite energy, characterized in that that said pre-ignition means further comprise pre-ignition energy limiting means, which comprise, in series with the output of said arrangement, a controlled element (14) of the transistor type, which reduces the energy available in the pre-ignition pulse, which energy also depends on the time and voltage parameters of the pulse, that of the voltage also remaining subject to the action of said clipping means (11, 12).

5. Circuit according to claim 4, characterized in that it comprises a signaling circuit

(17) connected to said clipping means (11, 12) and providing a perceptible indication relating to the current passing through them, and a manually adjustable component (16) which acts on said controlled element (14) and allows to establish the conduction limit of the latter in dependence on said perceptible indication provided by said signaling circuit (17).

6. Circuit according to claim 4, carac¬ terized in that it comprises means of limitation au- pre-ignition energy tomatoes (12, 15, 22, 23, 24) arranged to control said controlled element (14) automatically so that it reduces its conduction when the current in said clipping means exceeds one given value.

7. Circuit according to claim 6, charac¬ terized in that the said means for automatically limiting pre-ignition energy (12, 15, 22, 23, 24) comprise a manually adjustable component (24) which makes it possible to adjust at will the said given value of the current in the clipping means, this value being supplied by a resistor (12) connected in series with a Zener diode (11) to constitute the said clipping means.

8. Circuit according to claim 1, charac¬ terized in that the said pre-ignition means (6, 3, 9; 6 ', 3', ...) are connected to the said means for delivering erosive discharges (1 ), so that a constituent with a voltage threshold (3; 3 ′) of the pre-lighting means recognizes the instant when a certain voltage value is present at the output terminals of said means of erosive discharges (1 ), and that the pre-ignition pulse occurs at this time.

9. The circuit of claim 8, characterized in that said pre-ignition means are not only controlled but also supplied by the output terminals of said means for delivering erosive discharges (1), these pre-ignition means comprising a transformer - voltage booster (6), and, in series with the primary winding of this transformer (6), all connected to said output terminals of said means for delivering erosive discharges (1), a circuit part pre-ignition control (3, 4, 5; 3 ', 4', 26, 25, 27, 5 ') comprising the said component with voltage threshold sion (3; 3 ') and a capacitor (4; 4'), this condensa¬ having the effect of preventing the passage of a foreign current to the generation of pre-ignition pulses in said primary winding.

10. The circuit of claim 9, ca¬ characterized in that said part of the pre-ignition control circuit consists of a circuit branch which is in series with said primary winding of said transformer (6), and which comprises in series at least said capacitor (4) and said component with voltage threshold (3), the latter being formed of at least one Zener diode.

11. The circuit of claim 9, charac¬ terized in that said pre-ignition control circuit part comprises a pre-ignition control transistor (25) whose base is connected, for controlling the conduction of its section transmitter-collector, to a circuit branch comprising in series the said capacitor (4 ') and the said threshold voltage element (3'), the said primary winding of the transformer (6 *) being connected in series with the emitter-collector section of said transistor (25) to receive current pulses therefrom, under said command.

12. The circuit as claimed in claim 11, ca¬ characterized in that said part of the pre-ignition control circuit further comprises an active pulse reaction loop (27-31) comprising a reaction amplification tran¬ ( 29) capable of applying, by means of a reaction capacitor (31), a reactive looping pulse on the basic connection of the pre-ignition control transistor (25), in order to give said part of the control circuit prior to lighting the operating characteristic of a monostable rocker.

O PI

13. The circuit as claimed in claim 12, characterized in that it comprises, connected to one of the transistors of said part of the pre-ignition control circuit, an input (32) to which a signal can be applied which, independently of the evolution of the voltage on said output terminals of said means for delivering erosive discharges, tends to cause an additional double tilting of these transistors, in monostable rocker operation, to generate an additional pre-ignition pulse, the said signal being applied upon observation of an unwanted absence of pre-ignition.

14. Electric circuit for machine for machining by electro-erosion working a part (Pi) by 5 electro-erosive discharges in dielectric medium, in a machining space (GAP, d) between a tool electrode (Ξi ^" ) and said part (Pi) which also itself constitutes an electrode, comprising means for delivering erosive discharges (1), having output terminals which supply pulses respectively on said electrodes (E_t-, ° pi) erosive electrical energy, and pre-ignition means (6, 3, 9, ...; 6 '. 3', ...) to supply, each time in prologue to a said erosive pulse, pre-ignition pulses which have a higher voltage and a lower energy than the so-called 5 erosive pulses, which ignite, that is to say ionically, in the machining space (GAP, d) within the fluid dielectric, each time a channel which is then par¬ run by the erosive discharge, characterized in that the said means preignition (6, 3, 9; 6 ', 3', ...) are con- nected ⁰ to said output means discharges érosi¬ ves (1), so that a voltage threshold component (3 ; 3 ′) pre-ignition means recognize the instant when a certain voltage value is present at the output terminals of said erosive discharge means (1),

o .. < and that the pre-ignition pulse occurs at this time.

15. The circuit of claim 14, ca¬ characterized in that said pre-ignition means are not only controlled but also supplied by the output terminals of said means for delivering erosive charges (1), these pre-ignition means include ¬ from a voltage step-up transformer (6), and, in series with the primary winding of this transformer (6), all connected to said output terminals of said means for delivering erosive discharges (1), one by ¬ pre-ignition control circuit tie (3, 4, 5; 3 *, 4 ', 26, 25, 27, 5') comprising the said threshold voltage component (3; 3 ') and a transmission without direct contact (4, 4 ') capable of preventing the passage in said primary winding, of a current foreign to the generation of the pre-ignition pulses.

16. The circuit of claim 15, ca¬ acterized in that said transmission member without direct contact is a capacitor (4; 4 ').

17. The circuit of claim 16, ca¬ characterized in that said part of the pre-ignition control circuit consists of a circuit branch which is in series with said primary winding of said transformer (6), and which comprises in series at least said capacitor (4) and said component with voltage threshold (3), the latter being formed of at least one Zener diode.

18. The circuit as claimed in claim 17, characterized in that said series branch further comprises an adjustable resistor (5) by which the evolution of the current in said primary winding can be modified during the pre-ignition pulse.

19. The circuit of claim 16, ca¬ characterized in that said part of pre-ignition control circuit comprises a transistor (25) of pre-ignition control whose base is connected, for controlling the conduction of its emitter-collector section, to a circuit branch comprising in series said capacitor (4 *) and said threshold voltage element (3 '), said primary winding of transformer (6') being connected in series with the emitter-collector section of said transistor (25) to receive current pulses therefrom, under said command.

20. Circuit according to claim 19, ca¬ characterized in that said part of the pre-ignition control circuit further comprises an active impulse reaction loop (27-31) comprising a reaction amplification tran¬ ( 29) capable of applying, by means of a reaction capacitor (31), a reactive looping pulse on the basic connection of the pre-ignition control transistor (25), in order to give said part of the control circuit prior to lighting the operating characteristic of a monostable rocker.

21. Circuit according to claim 20, characterized in that it comprises, connected to one of the transistors of said part of the pre-ignition control circuit, an input (32) on which a signal can be applied which, independently of the evolution of the voltage on said output terminals of said means for delivering erosive discharges, tends to cause an additional double tilting of these tran¬ sistors, in monostable rocker operation, to generate an additional pre-ignition pulse, said signal being applied upon observation of an unwanted absence of pre-ignition.

22. Circuit according to one of claims 1, 8, 9, 10, 11, 14, 15, 17, 19, characterized in that the said means for delivering erosive discharges

(1) comprise a machining capacitor which is charged and discharged periodically according to a relaxation mode of operation, voltage drop means being provided to soften the current jumps (fig. 6).

23. Circuit according to one of claims 1, 8, 9, 10, 11, 14, 15, 17, 19, characterized in that the said means for delivering erosive discharges

(1) include electronic switching means which deliver, with the energy reserve necessary to provide the erosive discharges, a potential waveform with rectangular slots, the high level parts of said waveform providing a potential of a value suitable for maintaining the pre-ignited erosive discharges, while the low level parts of this waveform provide a substantially zero potential suitable for interrupting any erosive discharge (FIG. 7).

_ Oiv. ^' PI, _. IPO