FR2698218A1 - Process for inserting components at heart of energy converter structure - using parallel controlled switching arrays, with associated capacitors and diode, connected between side tracks on either side of board - Google Patents

Abstract

The process involves using switching elements, each of which comprises a diode (Da,Db,Dc) in series with a controlled switch (74a,74b,74c) and a capacitor (Ca,Cb,Cc) in parallel to them. The switching elements are connected to a power source and to an output (73a,73b,73c). An insulating board supports conducting tracks (71,72) to power lines with side tracks (76a,76b,76c,79a,79b,79c) extending from them. The second side tracks (79) are shorter and have a separator extension (75a,75b,75c). The switch is connected between the first side track (76) and the separated extension. The diode joins the second side track and the extension through insulated holes in the first side track. The capacitor lies between the first power track (71) and the second side track through an insulated hole in the first side track. USE/ADVANTAGE - Process for inserting components within energy converter such as electronic power supply using choppers. Improved separation of high frequency currents, and improved electromagnetic compatibility and heat dissipation.

Description

 The present invention relates to a method for implanting components within energy conversion structures.

 It also targets applications of this process for the production of switching power supplies and choppers.

 In many sectors involving conversion of electrical energy, and particularly in the automobile sector, the industrialization of a static converter imposes a particularly strong constraint on costs. These are both the direct costs of passive and active components and their implementation (mechanical assembly, thermal aspects, etc.). Putting passive and semiconductor components in parallel is one solution to reduce these costs. In fact, the prices of component suppliers show that, for a given performance switch, it is most often advantageous to choose several widely distributed components than one single component in the best of current technology. This choice also improves heat dissipation. This is also true for passive components, and the cost savings of the components should not be compromised by delicate assembly.

 There is an abundant literature on the paralleling of components. the parallelization techniques currently proposed are industrially costly. Mention may be made of 1 balancing by a circular symmetry technique, by the use of additional passive components in series with a power or control electrode, for example coupled inductors as proposed in the communication "Paralleling of high bipolar power Darlington "(paralleling of high voltage power bipolar Darlington) published in the document" Proc. of EPE ", 1991, ppl.7 to 1.13, or by pairing of transistors, as described by JT HUTCHINSON in the communication entitled" Paralleling Switching bipolar power transistors "(paralleling bipolar power transistors), PCIM September 1985, pp 40 to 44.

 We can also cite the paralleling of complete converters and the addition of currents by inductance.

 The paralleling of components must respect a certain number of rules. To avoid overvoltages, it is first of all necessary to reduce parasitic cabling elements (connection chokes). However, the induced compactness leads to a concentration of heat sources which is unfavorable from a thermal point of view.

 Beyond two components, perfectly symmetrical wiring quickly opposes industrial economic imperatives (circular symmetry, supply of power and control electrodes, etc.).

 European patent EP 215 325 filed on March 25, 1987 in the name of the company CURTIS INSTRUMENTS discloses a semiconductor control system for a DC motor comprising power transistors connected in parallel and freewheeling diodes also connected in parallel, these diodes and transistors being mechanically and electrically connected to a common radiator and arranged so as to facilitate the transfer of current between a diode and the associated transistor. However, with the arrangement proposed in this patent, a regular distribution of the current between the various switch components seems difficult to obtain because these switches are all connected in parallel and placed on a massive and unique radiator.

 The object of the present invention is to remedy these drawbacks by proposing a method for implanting components which leads to an improvement in the distribution of high frequency currents, in electromagnetic compatibility and in heat dissipation.

 According to the invention, the method for implanting components within an energy conversion structure comprising an input stage connected to a power source, an output stage and a predetermined number of switching arms connected in parallel on said power source and each comprising several switching cells connected in parallel and each comprising a diode and a controlled switch connected in series, each switching cell being further provided with a capacitor connected in parallel on said cell, is characterized in that each switching cell is connected to said output stage from a conductive link element connecting each controlled switch to the diode associated with it and independent link means for connecting said conductive link element to said stage exit.

 Thus, with the method according to the invention, it is not the controlled switches, in this case the transistors, but the switching cells which are in parallel, and the currents are added as in the paralleling of converters. but with the invention, this paralleling is purely internal and does not require any additional component since it uses wiring inductances or else leakage inductances of a transformer with multiple primaries. Thus, a better balancing of the currents in the components is obtained and a reduction in the stresses on them since the inductances of the cables which constitute the means of connection to the different switching cells are thus in parallel, the overall inductance then being divided by the number of switching cells.

According to a first advantageous embodiment of the method according to the invention, it comprises the following stages of implementation
- on a first face of a material support
substantially flat insulation, a first track
conductive comprising one of said lines
feed and first elements
conductors forming fingers equal in number to that
switching cells
- on the other side of said support
* a second conductive track comprising
the other supply line and second
conductive elements forming fingers
respectively opposite these first
conductive elements and significantly more
shorter than these; and
* conductive connecting elements in number
equal to the number of switching cells,
extension of said second elements
conductors and opposite the ends
respective of said first elements
conductors, and the method further comprises,
for each switching cell, the
following locations
. the controlled switch is connected between a
first conductive element and an element
opposite conductor
. the diode is connected between the second
conductive element opposite said first
conductive element and the conductive element
of connection which is in its extension; and
. the capacitor is connected between the
first supply line and said second
conductive element, a recess in the
first conductive element and in the support
insulation being provided to allow the
connection of a terminal of said capacitor with
said second conductive element.

According to a second advantageous embodiment of the method according to the invention, it comprises the following stages of implementation
- on a first face of a material support
substantially flat insulation, a first track
conductor comprising a first line
feed and first elements
conductors forming fingers equal in number to that
switching cells
- on the other face of said support:
* a second conductive track comprising
the other supply line and second
conductive elements forming fingers
respectively opposite these first
conductive elements and significantly more
shorter than these; and
* conductive connecting elements in number
equal to the number of switching cells,
extension of said second elements
conductors and opposite the ends
respective of said first elements
conductors, and the method further comprises,
for each switching cell, the
following locations
. the controlled switch is connected between a
first conductive element and an element
opposite conductor
. the diode is placed on the second face of the
insulating support in a recess made
in this support so that said diode is
finds substantially aligned with the second
conductive element and the conductive element of
liaison; and
. the capacitor is placed on the second
face, a recess being provided to allow
the connection of a terminal of said capacitor
with the first conductive element.

According to yet another advantageous embodiment of the method according to the invention, it comprises the following stages of implementation
- on a first face of a material support
substantially flat insulation, a first track
substantially rectangular conductor, including
a first supply line; and
- on the other face of said insulating support, a
second conductive track comprising the second
power line, and multiple strips
substantially rectangular conductors,
electrically isolated one from
each other and compared to the second track
conductive, and each corresponding to a
switching cell, recesses being
provided in said first conductive track and
in said insulating support to allow, for
each switching cell, a connection of
electrodes of each diode with respectively the
second conductive track and the conductive strip
associated with said cell, and a connection of a
terminal of the capacitor associated with said cell
with said second conductive track, each
controlled switch being connected between said
first conductive track substantially
rectangular and its conductive strip
associated.

According to another embodiment of the method according to the invention, corresponding to a bilateral implantation mode suitable for diodes requiring a radiator, comprises the following implementation steps
- on a first face of a material support
insulator, a first conductive comb comprising
a first supply line and several
first bands each associated with a cell
switch, and a second conductive comb
comprising a second supply line and
several second bands substantially aligned
with said first teeth, the first and
second supply lines being parallel
placed side by side
- on the other face of said support, several strips
electrically conductive
independent of each other, associated
each to a switching cell, and placed
opposite a first and a second
strips, and the method further comprises, for
each switching cell, the locations
following:
- the controlled switch is connected between the
first strip and the connecting strip,
- the diode is connected between the second strip and
the connecting strip; and
- the capacitor is connected between the first
strip and the second strip.

 According to another aspect of the invention, the method according to the invention can be applied for the production of a switching power supply including an asymmetrical bridge comprising a set of switching cells placed in parallel. The switching cells belonging respectively to each switching arm are arranged symmetrically on either side of the power source.

 The power source can also be transferred outside of said converter and the switching cells are then installed in line on the same support and constitute an in line structure further comprising two power lines to which the power source is connected. food.

 Preferably, the output stage comprises a transformer comprising as many primary windings as pairs of switching cells, each pair of switching cells consisting of a switching cell belonging to said first switching arm and a cell switching elements belonging to said second switching arm which are respectively connected to the two terminals of one of said primary windings associated with said pair of cells.

 It should be noted that the connections of the primary windings are preferably alternated to ensure symmetry of the currents since all the paths of passage of the current are thus approximately the same length from the source to the other primary windings during the conduction period of the transistors. , or through the diodes during freewheeling periods.

 All that has just been said for the primary windings can also be applied to the secondary windings of the transformer by dividing them into multiple secondary windings delivering on diode cells in parallel.

 The method according to the invention can also be applied for producing a chopper including at least one switching arm comprising a set of switching cells placed in parallel.

 This process thus allows easier industrialization of converters and switching power supplies. In addition, it provides redundancy of operation since several independent circuits are placed in parallel.

Other features and advantages of the invention will appear in the description below. In the appended drawings given by way of nonlimiting examples:
- Figures 1A and 1B show examples of
classic structures of a chopper and a
switching power supply Figure 2 illustrates the area of loops to
minimize in an example of a converter figure 3 represents a decoupling mode
according to the invention; FIG. 4 represents a paralleling of
switching cells according to the invention for
a switching power supply structure FIG. 5 illustrates a transfer of the source
of energy outside in a structure of the
type of that shown in Figure 4 Figure 6 is a perspective view of a
example of the use of three windings
primary in a feeding structure at
Figure 7 illustrates a first example of
geometric arrangement of the cells of
switching according to the invention Figure 8 is a sectional view of a first
switching cell Figure 9 is a sectional view of a second
cell with minimization of inductances
parasites;; Figure 10 is a sectional view of a third
cell with minimization of inductances
parasites; Figure 11 is a perspective view
illustrating a filling of a comb
Figure 12 is a perspective view
representing a bilateral implantation for
diodes on a radiator; Figure 13 is a perspective view
illustrating alternate connections of
capacitors; Figure 14 shows connections of
capacitors for bilateral installation
- Figure 15 is a perspective view of lines
of command in unilateral implantation; and
- Figure 16 is a perspective view of lines
of command in bilateral implantation.

 We will now describe several particular embodiments of the invention with reference to the above figures.

 To better understand the invention, we consider two concrete examples intended for the electric vehicle market. The first is a battery charger, with reference to FIG. 1A, the second is an auxiliary converter, with reference to FIG. 1B, the function of which consists in supplying the DC voltage, in this case 12 V, necessary for the equipment of onboard (headlights, indicators, stop lights, ...) from a main battery called HT 148 V).

 In the two examples chosen, the power supply is 48 V. The battery charger essentially consists of a step-down chopper. Its power is 900 W. The auxiliary converter has a nominal power of 350 W. Its power structure uses an asymmetrical bridge, with reference to FIG. 1B.

In both cases, the switching frequency is equal to 100 kHz. The aforementioned electrical quantities are given only by way of examples and the method according to the invention can be applied to much wider power and frequency ranges.

 By way of example, the battery charger 1, the structure of which is shown in FIG. 1A, comprises a high frequency stage comprising a controlled switch 3, a diode 5, and a smoothing inductor 4, supplied by the rectified sector 2 and connected output to a battery 6.

 The auxiliary converter 10, the structure of which is shown in FIG. 1B, comprises an asymmetrical bridge supplied by a main battery 7 called high voltage (greater than or equal to 48 V), consisting of two asymmetrical switching arms each comprising a diode 9, 11 and a controlled switch 8, 12 and at the midpoints of which is connected an output stage comprising a transformer 13 and a rectifier 14 for delivering a substantially continuous voltage V.

 The switches shown in Figures 1A and 1B are MOS transistors. Depending on the applications, they can be replaced by other semiconductor switches such as bipolar transistors, IGBT, or equivalent means.

 In the case of switching power supplies, the diodes are used only to demagnetize the transformer. They therefore have a much smaller caliber than transistors. Generally, they do not need to be placed on a radiator. In contrast, in a chopper application, the diodes have substantially the same caliber as the transistors.

These two cases require considering two different locations which will be described in the following.

 The asymmetrical bridge of FIG. 1B can be redrawn by showing the areas to be minimized in order to reduce the stresses on the switching cell, as illustrated in FIG. 2 which represents a conversion structure 20 comprising an asymmetrical bridge comprising a first arm switching consisting of a controlled switch 21 and a diode 22, and a second switching arm consisting of a diode 23 and a controlled switch 24, in which the voltage source 25 is connected between the two switching arms . The connection points of each switching arm are connected to an output stage comprising a transformer 28 and a rectifier 29.

 There appear two loops 26, 27 whose surface is to be minimized. The voltage source being distant from the auxiliary converter, it is conventional to use one (or more) capacitor (s) wired as close as possible to the switching cells, with reference to FIG. 3 which represents the insertion of two capacitors 31, 32 which makes it possible to significantly reduce the area of the loops 33, 34.

 For economic reasons, leading for example to the use of standard boxes, for example, of the TO220 type of current diffusion, each transistor is actually made up of several components in parallel (three in the case developed here). Rather than putting these components directly in parallel, the switching cells, for example 2 * 3 diodes and 2 * 3 capacitors, are put in parallel, with reference to FIGS. 4 to 6.

Thus, in a first example 40 of structure implementing the method according to the invention, with reference to FIG. 4, each switching arm comprises three parallel switching cells 41a, 42a, 43a; 41b, 42b, 43b, a power source 25 arranged for example in the middle of the structure and an output stage 28, 29. Each switching cell comprises a controlled switch Tla-T3a; Tlb-T3b, a Dla-D3a diode; Dlb-D3b and a decoupling capacitor
Cla-C3a; Clb-C3b.

 In another embodiment, the structure 50 has an in-line structure comprising two groups of three switching cells 51a-53a; 51b53b and the power source 25 is carried outside the structure.

 This adopted solution (online installation) is already interesting from an industrial point of view because it facilitates assembly (radiator printed circuit interface). It does not yet have a sufficient character of symmetry. However, since the high switching frequency generally requires the use of divided wire for the winding of the transformer, it is possible to envisage winding three identical primaries in parallel which are connected as shown in FIG. 6.

 In the structure 60 shown in FIG. 6, a first switching cell 61a of the first switching arm is associated with a first switching cell 61b of the second switching arm by having their respective connection points connected to a first primary winding 61 of an output transformer 64. It is the same for the second 62a, 62b and third 63a, 63b respective cells of each switching arm connected respectively to a second and to a third primary windings 62, 63 of the transformer 64 which also includes a secondary winding 65 connected at the input of a rectifier stage 29. It should be noted that the connections of the primary windings 61, 62, 63 are alternated to ensure symmetry of the currents because all the current passage paths are thus of approximately length identical.

 Thanks to this arrangement, the geometry on the one hand and the leakage inductors on the other hand, favor the balancing of the currents. The power source connections can be placed either in the center (maximum symmetry) or on the side (ease of wiring).

 It is necessary to note that the connections between the various switching cells have an inductance, admittedly low, but which, because of the perfect non-simultaneity of the priming and blocking of the different components will create, with the decoupling capacitors, oscillations parasites. Unlike electrochemical components, the technology used here (plastic film capacitors, possibly ceramic) does not dampen these oscillating circuits. It is therefore necessary to minimize the length of the connections.

 The need to pick up capacitors rather than semiconductors requires three-dimensional wiring. The semiconductors are implanted on the base surface. The capacitors are arranged as close as possible to the two supply lines.

 The diagram in FIG. 7 thus illustrates a proposed topology for implantation of the switching cells according to the invention. We will now specify the possibilities of geometric layout.

If the diodes do not need to be mounted on a radiator, diodes and transistors will be placed on the same side of the supply lines. If the diodes require a radiator, the diodes and transistors will be installed on either side of the supply lines.

FIG. 7 is a perspective view of the mounting of the components when the diodes do not need a radiator. In this structure 70, a first conductive track 71, placed on a first face of a support made of insulating material (not shown), comprises a supply line 77 and three first conductive elements 76a, 76b, 76c. A second conductive track 72 comprises a second supply line 78 and second conductive elements 79a, 79b, 79c arranged opposite the first conductive elements but substantially shorter. Connecting conductive elements 75a, 75b, 75c are also placed on the second face of the support in the extension of the first conductive elements but electrically isolated from the latter. Each switching cell 69a, 69b, 69c comprises a controlled switch 74a, 74b, 74c connected between the first conductive element 76a, 76b, 76c and the connection element 75a, 75b, 75c, a diode Da, Db, Dc connected between the second conductive element 79a, 79b, 79c and the connection element 75a, 75b, 75c, and a capacitor
Ca, Cb, Cc connected between the first supply line 77 and the second conductive element 79a, 79b, 79c. To each connecting element 75a, 75b, 75c is connected a conductor 73a, 73b, 73c for connection to the output stage (not shown).

 Figure 8 is a section of the same layout where the diodes are mounted on the component side. Thus, in this first form of implantation, for each switching cell, the diode 84 is placed on the same face of the insulating support 82 as the capacitor 83 and the controlled switch, placed at the end of the support 82, is connected between the first conductive element 87 and the connecting conductive element 88a to which the conductor 89 for connection to the output stage is connected. Two recesses E2, E3 are provided in the first conductive element 87 and in the insulating support 82 to allow the connection respectively of the anode 84b and the cathode 84a to the second conducting element 88B and to the connecting element 88a and another recess E1 is provided to allow the connection of a terminal 83b of the capacitor 83 to the second conducting element 88b , the other terminal of this capacitor 83 being connected to the first conductive element 87.

 To minimize the line inductance, it is better to align the diodes with the connecting conductive elements. Figures 9 and 10 show such an arrangement. In the second example 90 of implantation shown in FIG. 9, the diode 94 is placed on the second face of an insulating support 92 and is connected between the connecting element 98a and the second conducting element 98b, while the capacitor 83 placed on the side of the first face of the insulating support 92 has a first terminal 83B connected to the second conductive element through a recess made in the first conductive element 97 and in the support 92 and a second terminal 83a connected to the first conductive element 97.

 FIG. 10 illustrates an embodiment in which the diode is aligned by making a cut in the printed circuit. In this structure 100, the diode 104 and the capacitor 83 are placed on the same side of the insulating support 102 in which a recess 102a is formed with a dimension sufficient to at least partially accommodate the diode 104 which is then substantially in alignment with the second conductive element 109 and the connecting element 107. The capacitor 83 has a first terminal 103b connected to the second conductive element 109 while its second terminal is connected to the first conductive element 108 by a recess 102b formed in the insulating support 102.

To reduce the inductance, the distance between the two supply tracks must be as small as possible. It is determined by the insulation to be provided, provided for example by the thickness of a printed circuit as described by L.PERIER & J.BARRET in an application note entitled "An innovative high frequency high current transistor chopper nouveau chopper high current and high frequency transistor) AN366 / 06.89 published in the power product designer's guide, 2nd edition SGS
Thomson. June 1992, or better still, by the thickness of a polyimide film (For example: KaptonGR).

 The resulting increase in capacity is not significant in this application. Similarly, to further decrease the inductance, it is advantageous to fill the combs consisting of the first and second conductive tracks, as illustrated in FIG. 11, in which the structure 110 comprises a first conductive track 111 substantially rectangular, a second track conductive 115 also substantially rectangular and separated from the first track by an insulating support (not shown, and three conductive connecting strips 114a, 114b, 114c each corresponding to a switching cell comprising a controlled switch 113a, 113b, 113c connected between the first track 111 and a connecting strip 114a, 114b, 114c, a diode 116a, 116b, 116c connected between the second track 115 and the connecting strip 114a, 114b, 114c and a capacitor 117a, 117b, 117c connected between the first track 111 and the second track 115. To each connecting strip 114a, 114bu 114c is connected a conductor 112a, 112b, 112c connecting to the output stage (not shown).

 It should be noted that to minimize the inductance, the two supply tracks must always be superimposed as shown in the application note entitled "Environment design rules of MOSFET in medium power application" (rules for environment design in of medium power applications) AN358 / 06.89 published in the application manual for power transistors, 1st edition SGS-Thomson. Nov. 1989, pp. 63 - 71.

By neglecting the edge effects, the capacity resulting from two superimposed tracks is written C = eS / e where
S represents the facing surfaces and e is the thickness of the dielectric. For 1.6mm thick epoxy, the surface capacity is 3pF / cm2.

It is practically negligible for commonly used surfaces. Still neglecting the edge effects, the linear inductance has the expression L / k =, uoe / w where W represents the width of the track. For a track 10cm long and 2cm wide, the capacity is 60pF and 1 the inductance 10nu. This inductance value decreases with the thickness of the insulation. For Kapton with a thickness of 40ut, the surface capacity is of the order of 120pF / cm2, that is, for the preceding track, a capacity of 2.4nF and an inductance of 0.25nH.

 It should be noted that the output wires to the transformer can be placed anywhere on the diode-transistor connection strips and, in particular, at the ends.

 If the diodes must be installed on a radiator as is the case in choppers, one can keep a linear arrangement of the power components by adopting the arrangement in FIG. 12 illustrating a bilateral layout. The structure 120 comprises in a first plane, a first conductive comb 122 and a second conductive comb 121 each comprising a supply line 128, 129, these supply lines being placed side by side and parallel.

A second plane parallel to the first includes three connecting strips 123a, 123b, 123c placed opposite the teeth of the first and second combs 122, 121, and to which are connected conductors 12 Sa, 125b, 125c for connection to the floor Release. For each switching cell, a controlled switch 124a, 124b, 124c is connected between a tooth of the first conductive comb 122 and the connecting strip 123a, 123b, 123c which is opposite it, a diode 127a, 127b, 127c connected between the corresponding tooth of the second conductive comb 121 and the connecting strip 123a, 123b, 123c, and a capacitor 126a, 126b, 126c connected between the two supply lines 128, 129.

 Here again, the distance between the two supply tracks is determined by the insulation to be provided. The distance between the supply tracks and the output tracks can be the thickness of a printed circuit or, better still, the thickness of a polyimide film.

 It can be noted that this arrangement has the same effect on the reduction in the inductance of the switching cell as the above-mentioned arrangement providing for implantation with diodes without a radiator. the inductance of the supply tracks is also reduced if bar-buses are used8. Filling the "comb" reduces the inductance of the switching loop and the resistance of the conductors. Note that the connections between diode and transistor cross the supply lines providing a minimum capacity.

 Even more than in the previous case, the output wires to the primaries of the transformer can be placed anywhere on the diode-transistor connection strips and, in particular, at the ends.

 To minimize the magnetic fields radiated by the currents flowing in the capacitors, it is advantageous to divide each of these components into two components and to have them traversed by opposite currents. This is all the more important as the capacitors used have a large "emission surface".

 In the case of diodes without a radiator, the solution is simple. The connections of the two capacitors to the supply tracks are alternated, with reference to FIG. 13 which represents a particular arrangement of a switching cell 130 which comprises, in a first plane, a first conductive track 138 on which a diode is placed 135 and two capacitors 137, 136, and in a second plane, a second conductive track 133 and a connecting strip 134 to which a conductor 131 is connected. A controlled switch 132 is connected between the first track 138 and the connecting strip 134 , the diode 135 is connected between the connecting strip 134 and the second conductive track 133 by recesses made in the first conductive track 138 and the insulating support (not shown). A first capacitor has a first terminal connected to the first conductive track 138 and a second terminal connected to the second conductive track 133, while the second capacitor 136, placed upside down opposite the first capacitor 137, has a first terminal connected to the second conductive track 133 and a second terminal connected to the first conductive track 138.

 In the case of diodes with radiators in a bilateral installation, the connections can be made according to the diagram in FIG. 14. The switching cell 140 comprises in a first plane, a first supply line 141 and a first conductive element 148, a second supply line 142 and a second conducting element 147, and in a second parallel plane, a connecting element 149. It also comprises a controlled switch T connected between the first conducting element 148 and the connecting element 149, a diode D connected between the second conductive element 147 and the connecting element 149, a first capacitor C1 connected between the first supply line 141 and the second supply line 142 via respectively a first and a second conductive elements 145, 146 respectively connected to the first and second supply lines 141, 142, and a second capacitor C2 also connected between the two supply lines but in the opposite direction to that of the first capacitor C1, via first and second conductive elements 143, 144.

 The implementation of the control of the conversion structure can of course be carried out within the framework of the method according to the invention. To keep an economical printed circuit, we limit ourselves to two layers of conductors. A possible inexpensive third layer consists of a component or a strap.

Here, it is either the grid resistance which is used in unilateral implantation, with reference to FIG. 15, or a grid resistance and a source resistance which are used in bilateral implantation, with reference to FIG. 16.

 In the unilateral implantation mode 150 shown in FIG. 15, a first conductive track 152 and a second conductive track 153 are located in two parallel planes, in a configuration shown in FIG. 7. Three connecting strips 155a, 155b, 115c are placed in the same plane as the second conductive track 153 and opposite each conductive element 159a, 159b, 159c of the first conductive track 152. Each switching cell comprises a controlled switch 154a, 154b, 154c, for example a MOS transistor , whose source electrode is connected to the first conductive element 159a, 159b, 159c, the drain electrode is connected to the connecting strip 155a, 155b, 155c, and the gate electrode is connected, via a resistor gate 156a, 156b, 156c to a first control line 157. The source electrode of each transistor 154a, 154b, 154c is connected by a conductor 158a, 158b, 158c to a second line d e control 156. These two control lines 156, 157 can be located in a third plane parallel to the other two planes containing the conductive tracks.

 In the case of a bilateral installation, with reference to FIG. 16, two control lines 168, 169 are placed between the two supply lines 172, 173. For each switching cell, the controlled switch 167a, 167b, 167c, for example a MOS transistor, has its drain electrode connected to a connecting strip 165a, 165b, 165c, its source electrode connected to a first conductive element 174a, 174b, 174c and its gate electrode connected to a first line control element 168 via a gate resistor 171a, 171b, 171c. A second control line 169 is connected to each first conductive element 174a, 74b, 174c via a source resistor 170a, 170b, 17 Oc. Furthermore, each switching cell further comprises a diode connected between a second conductive element 175a, 175b, 175c belonging to the second conductive track 161 and the connecting strip 165a, 165b, 165c, and a connected capacitor 164a, 164b, 164c. between supply lines 168, 169.

 The insertion of a source resistor firstly simplifies the wiring but also makes it possible not to make a connection between the sources (emitters) of the transistors when the latter are on the positive supply line.

 This leads to a reduction in the stresses on the components. in fact, as the inductors for connection to the various transistors are in parallel, the overall inductance is divided by the number of transistors. Operating redundancy is also obtained since several independent circuits are placed in parallel.

 Of course, the invention is not limited to the examples which have just been described and numerous modifications can be made to these examples without departing from the scope of the invention.

Claims (16)

 1. Method for implanting components within an energy conversion structure (40, 50, 60) comprising an input stage connected to a power source (25), an output stage (28, 29 , 64) and a predetermined number of switching arms connected in parallel to said power source and each comprising several switching cells (41a-43a, 41b-43b; 61a-63a, 6lb-63b; 69a-c) connected in parallel and each comprising a diode (Dla-Dlb, Dlb-D3b) and a controlled switch (Tla-T3a, Tlb-T3b) connected in series, each switching cell (41a43a, 41b-43b; 61a-63a, 61b-63b) being additionally provided with a capacitor (Cla -C3a, Clb-C3b) connected in parallel on said cell, characterized in that each switching cell (41a-43a, 41b-43b; 61a63a, 61b-63b) is connected to said output stage (28, 29, 64) from a connecting conductive element (75a-c, 88a, 98a, 107, 114a-c, 123a-c, 134, 149, 155a-c) connecting each controlled switch (Tla-T3a, Tlb-T3b) to the diode (Dla-Dlb, Dlb-D3b) associated therewith and independent connecting means (73a-c, 89, 112a-c, 125a-c, 131, 151a-c, 166a-c) to connect said element connecting conductor to said output stage.  2. Method according to claim 1, characterized in that it comprises the following production steps:  - on a first face of a material support  substantially flat insulation, a first track  conductor (71) comprising a first line  supply (77) and first elements  conductors (76a, 76b, 76c) forming fingers in  number equal to that of the switching cells,  - on the other side of said support  * a second conductive track (72) comprising  a second supply line (78) and  second conductive elements (79a, 79b, 79c)  forming fingers, respectively opposite  of said first conductive elements (76a,  76b, 76c) and significantly shorter than these  last ; and  * connecting conductive elements (75a, 75b,  75c) in number equal to the number of cells of  switching, as an extension of said seconds  conductive elements (79a, 79b, 79c) and in screws  with respect to the respective ends of said  first conductive elements (76a, 76b, 76c),  and in that it further comprises, for each  switching cell, locations  following  the controlled switch (74a, 74b, 74c) is  connected between a first conductive element  (76a, 76b, 76c) and a conductive element of  connection (75a, 75b, 75c) opposite;  the diode (Da, Db, Dc) is connected between the  second conductive element (79a, 79b, 79c) in  with respect to said first conductive element  (76a, 76b, 76c) and the conductive element of  link (75a, 75b, 75c) which is in its  extension ; and  the capacitor (Ca, Cb, Cc) is connected  between the first supply line (77)  and said second conductive element (79a, 79b,  79c), a recess in the first element  conductor (76a, 76b, 76c) and in the support  insulation being provided to allow the  connection of a terminal of said capacitor (Ca,  Cb, Cc) with said second conductive element  (79a, 79b, 79c)  3. Method according to claim 2, characterized in that the diode (84) of a switching cell is located on the first face of the insulating support (82), recesses (El, E2, E3) being provided in the first conductive element (87) and in the insulating support (82) for connecting the cathode (84a) and the anode (84b) of said diode (84) respectively to the conductive connecting element (88a) and to the second conductive element ( 88b).  4. Method according to claim 2, characterized in that the diode (94) of a switching cell is located on the second face of the insulating support (92), the cathode and the anode of said diode (94) being connected respectively to the connecting conductive element (98a) and to the second conductive element (98b).  5. Method according to claim 1, characterized in that it comprises the following production steps  - On a first face of a support (102) in  substantially planar insulating material, a first  conductive track including a first line  feed and first elements  conductors (108) forming fingers equal in number to  that of switching cells;  - on the other face of said support (102)  * a second conductive track comprising  the other supply line and second  conductive elements (109) forming fingers  respectively opposite these first  conductive elements (108) and substantially more  shorter than these; and  * connecting conductive elements (107) in  number equal to the number of cells in  switching, as an extension of said seconds  conductive elements (109) and opposite  respective ends of said first  conductive elements (108), and in that it  further includes, for each cell of  switching, the following locations  the controlled switch (81) is connected  between a first conductive element (108) and  a screw conducting element (107)  with respect to said first conductive element (108)  the diode (104) is placed on the second face  of the insulating support (102) in a recess  (102a) produced in this support (102) so  that said diode (104) is located substantially  aligned with the second conductive element  (109) and the connecting conductive element  (107); and  the capacitor (83) is placed on the second  face of the insulating support (102), a recess  (102b) being provided to allow the  connection of a terminal (103a) of said  capacitor (83) with the first element  conductor (108).  6. Method according to claim 1, characterized in that it comprises the following production steps:  - on a first face of a material support  insulator, a first conductive track (111)  substantially rectangular, including a first  feeder ; and  - on the other face of said insulating support, a  second conductive track (115) comprising a  second supply line, and several strips  substantially bonding conductors  rectangular (114a, 114b, 114c) electrically  isolated from each other and by  relative to the second conductive track (115), and  each corresponding to a cell of  switching, recesses being provided in  said first conductive track (111) and in  said insulating support to allow, for each  switching cell, an anode connection  and the cathode of each diode (116a, 116b,  116c) respectively with the second track  conductive (115) and the conductive strip of  link (114a, 114b, 114c) associated with said  cell, and a connection of a terminal of the  capacitor (117a, 117bu 117c) associated with said  cell with said second conductive track  (115), each controlled switch (113a, 113b,  113c) being connected between said first track  conductive (111) substantially rectangular and  the connecting conductive strip (114a, 114b,  114 (c) associated with it.  7. Method according to claim 6, characterized in that it further comprises, for each cell, the establishment of a second capacitor (136) between the first conductive track (138) and the second conductive track (133) at near the first capacitor (137), an additional recess being arranged in the first conductive track (138) and the insulating support so that this second capacitor (136) is electrically connected to said first and second conductive tracks (138, 133) at the head spade relative to the first capacitor (137).  8. Method according to one of claims 6 or 7, characterized in that it further comprises the establishment of two control lines (156, 157) parallel to the supply lines (152, 153) in a third plane substantially parallel to the planes of the conductive tracks, each controlled switch (154a, 154b, 154c) having its control electrode connected to one of the two control lines (157) and its source electrode connected, via the first conductive element (159a, 159b , 159c) to the second command line (156).  9. Method according to claim 1, characterized in that it comprises the following production steps:  - on a first face of a material support  insulator, a first conductive comb (162)  including a first supply line  (172) and several first bands (174a, 174b,  174c) each associated with a  switching, and a second conductive comb  (161) comprising a second supply line  (173) and several second bands (175a, 175b,  175c) substantially aligned with said  first bands (174a, 174b, 175c), the first  and second supply lines (162, 161) being  parallel and placed side by side  - on the other face of said support, several strips  connecting conductors (165a, 165b, 165c)  electrically independent of each other,  each associated with a switching cell,  and placed opposite a first (174a,  174b, 174c) and one second (175a, 175b, 175c)  tapes, and in that it further comprises, for  each switching cell, the locations  following  * the controlled switch (167a, 167b, 167c) is  connected between the first strip (174a, 174b,  174c) and the connecting strip (165a, 165b, 165c)  * the diode (163a, 163b, 163c) is connected between  the second strip (175a, 175b, 175c) and the strip  binding (165a, 165b, 165c); and  * the capacitor (164a, 164b, 164c) is connected  between the first strip (174a, 174b, 174c) and  the second strip (175a, 175b, 175c).  10. The method of claim 19, characterized in that it further comprises, for each switching cell (140), the establishment of a second capacitor (C2) connected between the first (141) and second (141) supply lines near the first capacitor (C1), connections (143, 144) between the terminals of said second capacitor (C2) and the supply lines (141, 142) being arranged so that the second capacitor (C2 ) is electrically connected upside down with respect to the first capacitor (C1).  11. Method according to one of claims 9 or 10, characterized in that it further comprises the installation of two control lines (169, 168) parallel to the two supply lines (172, 173) and contained in a plane substantially parallel to the plane of the two conductive tracks (162, 161), each controlled switch (167a, 167b, 167c) having its control electrode connected to a first control line (168) via a control resistor (171a, 171b , 171c) and its source electrode connected to the other control line (169) via a first strip (174a, 174b, 174c) and a source resistor (170a, 170b, 170c).  12. Application of the method according to any one of claims 1 to 8 for the production of a switching power supply (40) including an asymmetrical bridge comprising a set of switching cells placed in parallel (41a-43a, 41b-43b ), characterized in that the switching cells belonging respectively to each switching arm are arranged symmetrically on either side of the power source (25).  13. Application of the method according to any one of claims 1 to 8 for the production of a switching power supply (60) including an asymmetrical bridge comprising a set of switching cells placed in parallel (61a-63a; 61b-63b) , characterized in that the power source (25) is carried outside of said converter and in that the switching cells are located substantially in line on the same support and constitute an in-line structure comprising two power lines (66, 67) to which the power source (25) is connected.  14. Application according to claim 13, for the production of a switching power supply, characterized in that the output stage comprises a transformer (64) comprising as many primary windings (61, 62, 63) as pairs of switching cells, each pair of switching cells (61a, 61b; 62a, 62b; 63a, 63b) consisting of a switching cell (61a, 62a, 63a) belonging to said first switching arm and a switching cell switching (61b, 62b, 63b) belonging to said second switching arm which are respectively connected to the two terminals of that of said primary windings (61, 62, 63) which is associated with said pair of cells, the connections of said primary windings being alternated.  15. Application according to claim 14, characterized in that the transformer further comprises several secondary windings, discharging in particular on diode cells in parallel, and in that the connections of said secondary windings are alternated.  16. Application of the method according to any one of claims 9 to 11 for the production of a chopper including at least one switching arm comprising a set of switching cells placed in parallel.