CA2011477A1 - Process for producing electrical connections on a universal substrate and apparatus for the same - Google Patents

Abstract

ABSTRACT

The process for wiring substrates with a universal pattern of conductors and a large number of connecting possibilities is characterized in that from the quantity of all the connecting possibilities is removed the quantity necessary for the particular application and this quantity is tested for further wiring possibilities realizing the circuit, said additionally found possibilities quantity being made available as a correction quantity during wiring.

The advantage of such a correction potential is that layout or contacting errors can be corrected in a second passage, so that wastage is reduced.

Description

PROCESS FOR PRODUCING ELECTRICAL CONNECTIONS ON A UNIVERSAL
SUBSTRATE AND APPARATUS FOR THE SAME

The invention is in the field of electrical circuit technology and relates to a process for producing electrical connections on a universal substrate having a plurality of connecting possibilities, as well as to an apparatus for performing the process.

Universal substrates are used in preferred manner in elec-trical circuit technology for prototypes and small-scale production. These are "circuit boards" with a plurality of applied connection and contact possibilities, which if necessary can be subsequently interconnected and contacted.
In this way it is possible without any specific conductor pattern layout to have a p~urality of different electronic circuits on the same board type. However, these "circuit boa~s" are not printed circuit boards in the conventional sense, but substrates very densely occupied by lines and onto which in hybrid technology the IC's and other compon-ents are directly bonded. Generally such a substrate has a ceramic base with a conductor pattern. In the case of universal substrates at least two conductor patterns are provided in different planes in such a way that there is a maximum number of circuits thereon. The desired conductor structure is constructed by subsequent contacting and/or separating conductor crossings and por~ions belonging to two different planes. A slbstrate of this type is e.g.
described in the Applicant's European Patent 167 732.

A user is confronted with completely new problems as a result of the multiplicity of circuit possibilities of all the conductors and their contacting points in the case of such unive~sal substratesO Thus, from the large number of possibilities, it is necessary to find the applic-ation, i.e. the desired connection, wiring or cabling by contacting and/or separatingO However, this is not all, because it is necessary to assume that corrections in the form of alternative or auxiliary lines will be necessary in the case of wiring errors in the layout phase on the substrate. In order to convert this into facts an optimization process is unavoidable. In addition, during the physical performance of the wiring, i.e. the contacting and/or separating operations, such a process must be perfor~ed as rapidly as possible through an adequate machine control. This must all be carried out in a micro-medium, in which the human operator cannot work without optical aids. Thus, the activity of "wiring" a substrate must be completely performed by a machine, which also forms an object of the invention.

However, the main problem is the optimum dealing with the large number of wiring possibilities, together with a simultaneous operation. It is therefore the problem of the invention to provide a process enabling an optimum utilization of a universal substrate both from the circuit and performance standpoints, which are closely interlinked.

Firstly a few introductory words on the process procedure in general. Universal circuit substrates e.g. used to be produced by wire-wrap technology. The latter makes use of the number of circuit planes required in order to produced the desired connection. The aim was always to produce electrical connections between two or more points without short-circuits with the remaining connections.
Fundamentally no~hing has changed in this connection, the aim still being to produce short-circuit-free electrical connections between two or more points, i.e. to carry out an elec~rical wiring, except that in the present case the "wires" already exist and must be interconnected in a special way. However, the systematic application of predetermined contacting points prevents the use of the hitherto known wiring methods either directly or in modified form. It is instead necessary to use 2 new wiring technology, which forms the subject matter of the following discussion.

2 ~

Considered topologically a universal substrate has a certain variation potential through its variable switching points with regards to the existing and realizable connections.
Before a single contact can be made, the switching point variation potential, or varia~ion potential for short is at a maximum. The magnitude of this variation potential is dependent on the construction plan of the substrate.
If the latter is fixed, then the upper switching point limiting potential (maximum) is predetermined. With each irreversible switching point change, there is a decrease of the variation potential to the lower switching point limiting potential (minimum) of zero. The quantity o~
all the possibilities of obtaining a given electrical circuit plan (an application) on the substrate, is the application or useful potential, which is naturally between the upper and lower switching point limiting potential.
In the most favourable case, the application potential can extend close to the limiting potential and in the least favourable case can be very small. The unusable potential of the "unusable" switching points in the specific case is the switching point loss potential. The switching point limiting potential as the number of circuits which can be made (quantity of maximum usable switching points) contains the switching point application potential as the quantity of usable circuits for the given application (quantity of usable switching points). The difference, i.e. the quantity of the unusable circuits or switching points gives the switching point loss potential. The circuits or switching points usable, but not required for a selected realization give the switching point correction potential. The latter is a very important quantity, because it permits the subsequent correction of manufacturing and/or design wiring errors. The correction potential is a subquantity of the application potential. However, if there are substrate defects or design errors, there is a modified Yariation and/or application potential and the correction potential can extend into the range of ` ` 2 ~ 7 ~

the previous loss potential. With reference to the following drawings information is given on how it is possible to cope with this. In the drawings show:

Fig.lA and lB a pictoral representation of the switching point potential consideration in terms of set theory.

Fig.2 a diagrammatic representation of the essential parts in the process sequence during contacting of a universal substrate and simultaneously repres-enting the essential apparatus parts of a micro-contacting device.

Figs.3A,3B,3C a flow chart as an example of one of several possible procedures.

Fig.4 an apparatus (derived from fig.2) for microcontacting and micromeasurement of universal substrates.

Fig~5 the principle of a mechanical multi-microcontactor or a multi-microtest head.

Fig.6A,6B,6C a first contacting process.

Fig.7A,7B,7C a second contacting process.

Fig.8A,8B,8C a third contacting process Fig.9 a line pattern of a universal substrate, as used as exemplified embodiment of a con~acting process according to the invention.

Fig.10 the line pattern according to Fig.9 in modular form to show that the ~rocess can be used on line patters of random size within the substrate dimensions.

:,' , . .

As stated hereinbefore, prior to using a universal substrate and the electric circuit to be produced thereon, it is necessary to have a switching position potential consider-ation. The next stage is the association of the switching points t Fig.lA shows in the form of a rectangle as a quantity the variation potential of a specific substrate, here its limiting potential GP. This can merely relate to a par~ or a detail of the substrate, if e.g. for a specific circuit of several on the same substrate the application potential is to be exposed. Or alternatively it relates to the complete substrate if only one circuit is to be produced on it. In this quantity representing the GP are incorporated the potentials in the form of quantities of three different applications APl, AP2, AP3.
The residual quantity represents the loss potential VP.
It is pointed out that these potentials in each case rep-resent the combinatorial number of possibilities of producin~ -circuits from switching points and do not represent the actual number of switching points physically present on the substrate.

It can be seen that Application 1 has more possibilities (APl) for realizing the circuit on the substrate than the two other applications~ Application 2 is restricted, although the same number of switching points exist on the substrate as for Application 1 and this is even more marked in the case of Application 3. The reasons are e.g. more connections, less favourable wiring configurationS
(crossing), circuit nesting, etc., which are all disadvant-ageous in connection with wiring on a universal substrate.
It can also be gathered from the overlap that although application 1 can be real~zed with each of the two other applications on the same substrate, it cannot be realized with both of them simultaneously and infact applicationS
2 and 3 can be realized together. For a simultaneous realization of applications 1 and 3 on the same substrate, the application potential is AP13.

.

~, ~3 ~

The selection of the circuit to be obtained for the plurality of given possibilities takes place on the basis of an optimization criterion with weighted factors such as e.g.
surface requirement, number of contacts, overlap of differ~t applications (reciprocal influencing), line lengths, etc.
An important aspect is the remaining conection potential.
Each realization reduces in its way the variation potential and the application potential, because a contact produced at a specific point renders impossible a number of other solutions. The correction potential can also be incorporated into the optimization. However, there is only something to be corrected if a realization has proved to be faulty or erroneous. It is then to be assumed that a connection or separation not intended at a provid~ point or which has undesirably been brought about cannot be reliably used or remedied and therefore must not be used. Thus, a fault will always further reduce the correction potential and in some cases drastically if the fault point is a key circuit point.

Fig.lB pictorially shows this process by means of the quantity diagram. On the basis of a limiting potential GP with an application with the useful potential AP, the realization R brings ahout a reduction of the application potential AP to the correction potential KP, which is in turn further reduced by additional/ error-caused loss potential FVP. During processing or wiring of the substrate, the faulty substrate can be returned, so as to undergo recontacting. However, this is only possible if there is a sufficiently large correction potential.

Fig.~ shows diagrammatically as in Fig.l a process flow of a uni~ersal substrate introduced as~a blan~ into the process on the left hand side, whilst on the right hand side it lea~es with the application made as a customer-specific product. The completely unwired substrates are prewired in a contacting machine or de~ice (final wiring - ~J~ 7~

taking place by bonding on the circuit elements), i.e.
they are physically treated and then passed on to a testing means. In the latter the substrate is measured and checked with respect to the preset in~ormation, i.e. the wiring plan from the computer. If the preset requirements are fulfilled, the substrate is passed on as ~good~'. If the preset requirements are not fulfilled, the correction potential is determined and related to the necessary correc-tions, i.e. to the electrical replacement paths. If these criteria are positively ful~illed, i.e. it still has a possiklity for electrical replacement paths, the substrate is returned and sent back into the con~acting device for reprocessing. ~owever, if these criteria are negatively fulfilled, then the substrate is disposed of as waste, although it still need not be completely unusable.

The real correction potential is generally only obtained when the fault is present. If the fault is detected by the testing device, e.g. an inadequately formed contact point, then the computer must establish whether among all the remaining circuit possibilities there is still a "path", which can functionally compensate the error point. In certain circumstances, there can be more than one path for overcoming the difficulty. If this is the case, a sufficiently high correction potential is indicated and the substrate, which is a very expensive component, is returned to the process. If there are several possib-ilities or paths and the correcting contacting fails again, then the substrate is introduced a third time into the contacting or wiring machine.

The contacting machine, which will be described in gxeater detail hereinafter, e.g. operates with a laser for separating or welding, or with a microcontact welding device or combined with both. Apart from these two contacting procedures, this process also permits contacting by means of electrically conducti~e materials, e.g. adhesive between two conductors - ?, ~ r~

and for this purpose the apparatus must be correspondingly designed.

The contacting machine can perform contacting simultaneously-or sequentially, in parallel or series, or in mixed forms.
Contacting takes place according to the circuit diagram.
A distinction is made between the circuit diagram of the electronic connections SPE, the circuit diagram for the substrate (SPS) (the actual wiring) and the "circuit diagram"
for the electrode control SPK in the contacting machine.
From the SPE is obtained the SPS and from the latter the SPK. The inYentiVe procass connects between the SPE topo-graphy and the SPK electrode control and in an additional step for the transport means for the material processed in the machine.

If for contacting and/or separating laser welding is required for contact formation or conductor interruption, then the apparatus will probably operate in a serial manner.
However, acceptable process speeds can be obtained, because the laser control operates very fast 1~ a microcontact welding is required, e.g. by means of ultrasonic or cold welding with pressure, which is a function of the substrate type, at least a partly parallel procedure is desired.
A separating laser producing the separating patterns is then used for separating the conductors.

Uni~ersal substrates have a regular, coordinate-related arrangement of contact points, so that it is possible to construct and use multiwelding heads with arrays of e.g. 10 x 10 welding points or 20 x 40 welding points.
Apart from the positioning routine bringing the multiwelding head~ into the desired position, there must also be a n from m selection, n being the number of contact points to be instantaneously welded and m the number of ~elding points of the multiwelding head used. Thus, the process speed can be considerably increased in the case of a once r;~

and for all construction expendi~ure for the multiwelding head.

As stated, the process switches between SPE and SPK and can covers in an additional process step an operator 0 , which switches between the contacting machin~ and the testing station. It carries out the following functions.
It links the test results with the wiring details and frees the real correction potential or at least assists in this. It establishes the reoptimization and determines any residual correction potential. Thus, this operator 0 is competent for the reoptimization of a wiring of the already processed universal substr~te. It operates in the processing operation in such a way that in the case of a single su~rate passage there is a v;rtual re-optimization, i.e. the correction potential is not used, or in the case of a multi-passage there is a real reoptim-ization and therefore the correction potential is used.
It always operates and flanks the process with respect to the malfunctions. However, it is not necessary that for each individual substrate, whether faulty or not, the correction potential is always determined. The calcul-ation provided for this purpose is only initiated if the coincidence of the preset SPS and the tested (realized) SPS does not exist.

Finally, if the correction potential is exhausbed, the substrate is rejected as waste, or is sup~lied recursively or iteratively to ano~her application. Thus, on such a substrate it is possible ~o house a circuit of a similar type because, as stated, considerable differences can exist between the applica~$on potentials. This is a further advantage of the present process, namely that naturally offline applications can be sought, which can s~ill be carried out in such a substrate, which merely has a reduced limiting potential. In certain circumstances, this is even possible with the desired correction potential.

7 ~

Thus, if the substrate has already been discarded as waste, then the determined and existing data of the wiring test-piece it is possible to seek offline further applications, which can still be carried out on the ~residual substratel'.
This add;tionally reduces losses, which is important in view of the high cost of substrates.

The diagrammatic representation of Fig.2 shows separate devices for optimization, contacting (wiring and testing).
These are on the one hand the essential parts in the process sequence on contacting a universal substrate and on the other hand the essential apparatus parts of a microcontacting device. Cbviously these functions can be partly or entirely combined. Thus, for example, the three calculating or computing parts can be combined into a common computer, whose output or outputs inter alia supply the control signals for the control part of the contacting electrodes or the measuring electrodes the testing part. It is possible that the same head part (e.g. a slide) carries both the contacting and the measuring electrodes, i.e. the contacting and testing machine have a common coordinate control.
The contacting of a substrat~ or workpiece and the testing of another substrate (e.g. the preceding substrate), which is now a testpiece, can take place simultaneously jointly covering the SPK and SPP ("circuit diagram" for the test control). In the case of contacting by cold welding, i.e. with pressure on the con~acting electrodes, the latter can optionally also be used as test electEodes. Thus, the contacting machine and testing machine would be sub-stantially identical, but from the process sequence stand-point, the process shown in Fig.2 generally applies to all conceivable cases.

In summary form the procedure is essentially as follows:

- fix the optimized wiring determining the minimum number of contacting operatisns and the minimum conductor 2 ~

lengths, etc. (computer), - fix control of microcontacting tool (computer), - carry out the positioning of the electrodes (control), - carry out welding and~or separation (physically), - carry out testing process (measuring and calculating), - decision for good, waste, return tcomputer), - return for new wiring (physical), - fix reoptimization of wiring (computer), - fix electrode control for correction (computer), - carry out electrode positioning (control), - carry out welding and/or separation for correction wiring (physical), - carry out retesting process (measuring and calculating), - decision for good, rejection, return, etc.

Figs.3A,3B and 3C show the process sequence in greater detail. Fig.3A shows the sequence without substrate return in the case of faulty wiring. In a first step the optimum connection design 70 is calculated wi~h data 63 of the electric circuit diagram (3) and data 60 of the universal substrate. The procedure is as hereinbefore for the poss-ibility potentials.

The use of a CAD system is recommended for the connection design. In connection with such systems there are manual routers (disentanglement by ha~d) and autorouters (auto-matic disentanglement)O Routers are calculation programs for the design of printed circuit boards. These programs are based on the idea that the plane on which the conductors L are to be applied are completely free and conductors can pass in any direction. It is presupposed for an optimum connection design that there is an optimum planning of the position of the components taking account of the minimum surface, minimum connections, short connection paths and optimum thermal heat dissipation. In the case of optimum connection of the elements via crossed contact points K, in the case of the discussed universal substrates, 7~

there is a degree of freedom such that in connection with the placed elements it is not yet necessary to decide on which bonding sur~aces B of the substrate they have to be bonded (Figs.9 and 10). It is merely necessary to respect the condition "no crossing of bonding wires"
and "bonding wire shorter than 3mm". On utilizing this variation potential, the router requires a corresponding addition, because in the discussed example on right-angled direction changes are appropriate. For conductors not used, or better not realized in the basic pattern, a "travel interdiction" must be imposed on the autorouter. These are certain important, specific boundary conditions for the autorouter to be used (e.g. based on the Bartels Router, Lee Router, etc. with the corresponding addition).
Then the minimized number of separations must be determined for maximizing the correction potential.

The data 67 of the use-specific substrate plan, i.e. the wiring data, are supplied to a calculation process 74 for determining the optimum conta~ing process, which also requires data 64 of contacting means (4). The wiring data are also supplied to a further calculation process 75 for fixing the testing process also requiring the data 65 of the testing device (5). From the two calculation processes result the control data 61 for the contacting device (1) and the control data 62 for the testing device (2~. These are the apparatus and use-specific preparation treatments 89, which are only carried out once for each application.

This can be followed by the contacting 71 of the unwired substrate in the contacting device and the testing 72 of the wired substra~e in the testing device. In the latter the substrate is measured and tested for preset information, ie. the wiring plan, from the computer (80).
If the preset in~orma~ion is fulfilled (81)~ then the substrate is passed on (83) as usable. If the preset informations are not fulfilled ~82) and no return is planned, rl 7 then the substrate becomes waste (84). This phase (88) is repeated for each new substrate to be contacted for the same applicati on .

Fig.3B shows the process sequence for the case that there must be a return in the case of unfulfilled preset data.
Following an identical initialization 89 and contacting 71 to that of Fig.3A, at testing means 72 there is an evaluation 76 of the existing connections. Thus, it is established which connections have been correctly made and which are faulty and therefore are unusa~le. The correction potential of a substrate separated according to the criteria of the sequence in Fig.3A on the basis of the data 66 of the actual, real connections. Appropriately, but not necessarily, this takes place in two ~ages. Firstly in a simpler process 78, it is established w~ther a correcting connecting design is possible 85 and, if this is the case 86, the more complicated optimization 70' is carried out utilizing all the criteria. This optimization 70' must naturally be based on the data 66 of the real, actual substrate and not, as for the initialization 89, on the data of the ideal, non-contacted substrate.

Optimization 70' supplies data 67' of the corrected substrate plan or the correction contactings to be carried out.
From this, on the one hand the optimum correction contacting process 74' is calculated therefrom together with the data 64 of the contacting device (4) and on the other hand, together with the data 65 of the testing device, (5) the testin~ process 75' is calculated and returned with the corresponding data 61', 6~' for contacting and testing. This "internal cycle" 87 can , as stated herein-before, be repeated, if once again a contacting or wiring error has occurred and the correction potential is still not exhausted.

In another substrate returning process variant, which treats each substrate as an individual at the start without presupposing a specific connection potential, the calcul-ating costs would normally be too high, but in special cases, particularly when using a substrate eliminated from a prior application, it can be appropriate to procede in this way, as shown in Fig.3C.

In the discussion a distinction is made bettween two stages of the correction potential determination. In one variant all the surfaces covered by the application are considered as forbidden zones and a new connection between peripheral contacting points is calculated. In the other variant all the conductor portions (including the faulty portions) occupied by the application are considered to be occupied and the router seeks the new connection from the residual potential of connection possibilities. This also applies for separation points. If e~g. the defect is a short-circuit, i.e. a missing separation, then there must be a further separating point which can be opened along the "unseparated~ portion. If this does not exist, this faulty connection must be declared occupied and the router must seek a new connection enabling this faulty conductor part to be exlcuded. Bypassing of faulty parts can e.g. be obtained by a meander-like or parallel conductor path.
The optimum correction layout is obtained b~ an iterative router algorithm, which optimizes with dif~erent priorities according to dif~erent criteria. The limiting conditions is then to maintain all the connections except for the modified connections, whils~ seeking an alternative connec-tion, or recalcula~ion of all the connections taking account of the ne~ parameters. The first limiting condition must be sought with maximum economic priority.
Initialization 89'' is simplified, compared with the prev-iously explained variants. In place of the data 63 of the desired electrical circuit diagram, the connections of the ideal substrate are predetermined as data 63'' of the elec-trical circuit diagra~. Thus, the calculation 70 of the 2i~

optimum connection design becomes extremely simple and leads to data 67" of the universal substrate plan. If the contact-ing process was calculated on the basis of data 67l, then there would be a "zero contacting", i.e. no connection change would be required. This is directly inputted at 1 by the data 61". The testing process is established 75 as in the previously described variants and supplies the control data 62 for the testing device.

In the first passage through the internal cycle 87", contact-ing 71 is a "zero contacting", i.e. no contacting takes place. If the testpiece is an unused, defect-free part, there is a positive result 81 following testing 72 and evaluating 76 in test 80, where it is established whether the substrate has the expected connection. However, this does not mean that the part can be allocated to the usable substrates 83, unless it was merely wished to test the unused starting substrates to ensure they were free from faults. Thus, at a branch 77, provided in addition as compared with the variant o~ Fig.3B, it is checked whether it is the first passage. If this is so 78, working continues on the process branch known from Fig.3B and there is no longer any difference in the process compared with section B-B. However, this variant is only apparen~ly less compli-cated; so that initially individually the data 66 of the actual substrate is taken and in each case there is an individual calculation 70' of the optimum contacting process.

This obviously not only applies for the contacting process, but also for the possibly necessary conductor separation process, which also enters the optimization. As stated, for this purpose a separating laser or some other separating device is used.

The optimized wiring process permits the following connec-tions or separations of conductors, which are superimposed in t~o planes separated by an insulating material. If S~ S'l '~

the insulating material is gaseous (air), then the connection can take place by means of a laser pulse, which is dimen-sioned in such a way that one of the conductors sags by heating and contacts the other conductor. For this purpose -the thermally formed contact can be assisted by a mechanical holding-down device. In addition, a first laser pulse can pass through the superimposed conductors and they can then be brought together again with a second, larger diameter laser pulse and welded together. The effect of the laser pulse can be assisted by a focussed laser beam, in place of a mechanical holding-down device. This gaseous beam can be heated to assist the thermal action.

In another procedure, the two superimposed conductors are perforated by a laser beam and into the latter is introduced an electrically conductive medium. A preceding pulse can be used to act on a conductor in such a way that it contacts the other conductor. The electrically conductive medium then improves the contacting action.

In another procedure spot welding can be carried out by means of a capacitor discharge via a conductor crossover.
This dlscharge can be obtained by means of two electrodes or by means of one electrode and a conductor as an electrode.

In anoth~ procedure the conductors can be cold welded, in that one conductor is pressed with high pressure onto the other conductor. This typa of welding can be assisted by using heated pressure needles or pressure needles with a sonotrode.

In another procedure there can be a connection by means of a microdispenser with which the conductors are connected by means of an electrically conducti~e material. In the case of the substrate according to the aforementioned European Patent the polyamide windvw is filled e.g. with gold or sil~er-containing resin, e.g. epoxy resin. ~he microdispenser can be constructed in such the same way as an ink jet spray head, so that it acts in the manner of a multiwelding head.

In another known procedure the conductors can be connected galvanically, in that the conductors can be exposed in a galvano bath to an antipole current, whose crossover is to be connected.

The separation of the conductors e.g. takes place by a laser pulse, which must be such that it does not heat the surrounding area and instead merely locally vaporizes a conductor portion.

Examples of laser wiring or laser-assisted wiring are given here, ~ecause lasers permit a highly planned and rapid operation. Thus, on the discussed substrate type laser pulses of 10 to 30 micrometers diameter areeto be used, which is difficult with mechanical means alone.
However, wiring with mechanical means is still an alternative particularly if use is made of a multi-microcontacting head, similar to a matrix printing head, permitting the simultaneous very rapid contacting of complete fields of a substrate matrix with only one positioning movement.

Fig.4 diagrammatically shows an apparatus enabling the aforementioned prscess to be per~ormed. In a computer 1 containing the data for the controls and also the circuit diagram data, the wiring optimization of the universal substrate is calculated and is converted into control data for the control 2 of microcontactor ~ and the control 3 for the testing device with the microsensor 5 and on the basis of said data these ~wo devices can be positioned over the substrate and the microcontacting head operated or the test data collected. Two transporting devices ~,7 are controlled by con'crol 2 of the microcontacting device and control 3 of the testing device. Drive 8 of ; ,~$~

transport device 6, which supplies the blanXs to the contact-ing means and the program substrates to the testing means, is here controlled bv the microcontacting device control 2 and the drive 9 of transport device 7, which returns the waste substrates ~or correction contacting, is here controlled by the testing device control 3. The transfer of waste substrates to the return and rewiring is indicated by the two arrows 10 and 11.

The microcontacting device 4 is here in the form of a multi-microcontactor. This relates more to the mechanical device than to a laser microcontactor. However, all adequate microcontacting processes can be incorporated into the process or ~n be realised in the apparatus. There now ~ollows a discussion of a mechanical multi-microcontactor with needles for deforming the contact bridges and connectin~
two conductors of a crossover, with which a complete array of the substrate can be simultaneously contacted. If the substrate matrixes are e.g produced in modulo 100, which means that an edge length has a complete multiple of 100 contacts, then a multi-microcontactor has at least one row of 100 contacting needles. Thus, up to 100 contact crossovers can be programmed with one positioning. Generally several rows of such contacting needles are combined in a multi-microcontactor, it naturally not being necessary to apply these needles with the density of the crossing point~ The needle density can e.g. relate to every fifth crossing point, so that the needles are further apart.
Mechanical criteria decide on the dansity of the needle matrix. The individual needles are axially movable and individually controlled. If such a multi-microcontactor is positioned over the substrate in such a way that the needle matrix coincides with the crossing matrix, so that through the axial mo~ement of individual needles the corres-ponding crossover points can be contacted, then several contacts can be craated at once. Following a repositioning, the needles are activated, whose associated crossover point is to be contacted.

In this way it is also possible to construct a multi-microdispenser enabling electrically conductive materials, to be introduced into the air gap between the two conductors.
It is also possible to produce a multi-spot welder with integrated electrode, in which the electrodes whose assoc-iated crossover points is to be contacted are always activ-ated. The counterelectrode is formed by the conductor.

In order to function economically, such multi-microcontactors have arrays of 100 x 10~ contacting points, so that thousands of contacts can be formed in one positioning operation.
The position is correspondingly optimized in such a way that the minimum number of positioning operations have to carried out.

Such an apparatus is an alternative to a laser microcontacto~, which operates serially, but at high speed. The substrate type also makes a contribution to the decision of whether mechanical or non-mechanical (with light) contacting is used. The costs of a mechanical multi-microcontactor of the discussed size is approximately the same as for a corresponding laser installation, the programming COEtS
also being comparable.

The operator 0', which has switched between the contacting machine and the testing station, as stated hereinbefore, has to carry out the following functions. It links the testing results with the preset wiring information and frees or at least helps to free the real correction poten-tial. It fixes the reoptimization and determines any residual correction potential. Thus, said operator O
is competent for the reoptimization of a wiring of the already worked universal substrate. It works in the process-ing operation in such a way that in the case of a single passage of a su~strate there is a virtual reoptimization, i.e. the correction potential is not used, whereas in the case of a multiple passage there is a real oreoptimiz-ation and consequently the correction potential is used.

r~

It always accompanies the movement and flanks the process with respect to malfunctions. Howe~er, lt is not necessary to always determine the correction potential for each individual, faulty or non-faulty substrate. The calculation-for this purpose is only initiated if there is no coincidence between the preset SPS and the tested trealized) SPS.
The computer 1 is competent for this process part. These partial functions can also be transferred to in each case one computer part in the two contacting and testing units.

Fig.5 diagrammatically shows a multi-microcontactor with mechanical m~ans, which can be needles or microdispensPrs.
A guide bo~ carries a matrix-like arrangement of perfor-ations 16 through which the contacting elements 17 are passed. These contacting elements are axially movable and are driven in controlled manner by a drive means 18, e.g. a piezoelectric element. The perforations are arranged with a spacing DN, which is an integral multiple of the crossover spacing DK of the substrate. DK represents the unit of the positioning steps, in which the contacting head is positioned over the substrate. In a substrate, like that described in the aforementioned patent, the orthogonal spacing between two crossovers is, as a function of the density, 50 to 100 microns. These values are stand-ardized and are respected with high precision during manuf-acture. Thus, the step length is 1/10 mm and below, which also makes high dema~s on the mechanical guidance of the contactor, i~ it is borne in mind that a substrate can ha~e an edge length o~ several cm.

It is therefore recommended to provide a positioning sensor enabling corrections to take place in ~he positioning in such a way that positioning errors are not summated and are instead always eliminated by a regulating process.

The testing device 5 ~s shown as a test electrode pair.
This is intended to indicate that testing can fundamentally 2 ~

take place with two electrodes, a multi-test electrode head naturally leading to similar ad~antages to a multi-microcontactor. A multi-test electrode head can in principle have the same construction to the multi-microcontactor, so that measurement can talce place through positioning and placing so~e of the plurality of measuring electrodes at the predetermined measuring points. ance again it is recommended to use a positioning sensor, because the requirements are much the same with the test head as for the contactor, although the testing points are generally on the periphery of the substrate and in certain circum-stances can be constructed in large-area form in order to reduce testing costs.

In principle, by function reversal, a multi-microcontactor can be used as a multi-microtester, because the matrix always remains the same. Such devices are then used for programming universal substrates, which in each case have the same matrix module, which is always sought.

The contactingftesting heads are e.g. fixed to a cross-slide used for carrying out the positioning. The workpieces i.e. the substrate to be programmed, are supplied on the transport device and stopped for the programming/testing process. As a function of the test result, the programmed substrate is returned and undergoes recontacting as a workpiece. The transport device is diagrammatically shown as a belt, but it can also be a robo~ which provides the workpiece and it can also be a controlled substrate perform-ing the positioning steps, the contacting and/or test head being mounted with the leads. With the needles raised the positioning corresponding to a horizontal movement relative to the needl~ matrix is carried out and subsequently the contacting movement is performed or~hogonally thereto.
If this process is performed,"overheadl~, then as an additional safety element the weight of the needles can be used for decoupling from the substtrate. Robot handling i9 recommended . ~

2 ~

for this, in which the substrate is placed on a crosstable and the latter i9 rota~ed by 180 and pivoted downwards with the substrate over the contacting or test head and being brought into the first position with a short-stroke lowering movement.

As microcontacting can last a few minutes, the setting-up time increased by robot handling is acceptable. A
preparatory treatment can be introducecl by using alternately loaded crosstables and this makes it possible to halve the setting-up time.

Figs.6,7 and 8 provide examples of a contacting device, showing three examples of a contacting process at a contacting point. Each figure group shows threé states A,B,C, which means starting position A, contacting process B and subsequent (optional) welding C.

The parts 71,81,91 are hollow guides, so that energy can be passed through the opening ~or possible su~sequent welding. A laser source 72,82,92 is provided for this purpose. The laser energy is supplied either directly through the cavity or via a light guide 83,93 ~o the previously mechanically contacted conductors Ll and L2. For mechanical contacting purposes, a hollow guide has a contac~ing ring 74,94 designed for the limited mechanical loading.

The first stage A shows the starting position. The two crossed conductors are spaced from one another by a special shaping, i.e. they are isolated by an air bridge. This isolation is removed by pressing the upper conductor L2 onto the lower conductor Ll and a galvanic connection is obtained. This pressing is shown by the second stage B. In Figs.6 and 8 pressing takes place with guides 71,91 or contact rings 74,94 and in Fig.8 with the light guide 83, whlch is passed through the guide. As only limited forceæ are required for deforming the conductor bridges, r~

it is possible to use known light guide materials or light guides.

This galvanic connection fulfils the objective of creating a contact between two superimposed conduc~ors. Welding is possible as an optional measure in a further stage C. For this purpose a correspondingly dimensioned pulse from a laser source 72~82~92 is deflected onto the inter-section and the galYanic connection is reinforced by welding.
A light guide may, but need not be used.

Apparatus 71-74, 81-83, 91-94 shown in exemplified manner here can be part of a multi-contacting head. Then, in addition to the simultaneous mechanical contacting process, there will be a serial, subsequent welding process, because a random numbe. of laser sources cannot be located in a small space. Thus, the lasers subsequently move up to the intersection and carry out welding where it is required.
This need not involve all the contact points. A certain parallel arrangement o~ the process can be achieved by increasing the number of welding points via light guides and more than one laser source can be used. The principle of a multi-microcontacting head has already been discussed relative to Figs.4 and 5. This also applies with regards to the spacing of the contacting points and the way in which a substrate is moved along the head for contacting purposes. This can take place overhead, underhead or in any desired position.

Claims (26)

1. Process for producing electrical connections and/or for the separation of conductors on a substrate with a universal pattern of conductors and connecting points, characterized in that from the quantity of all connection possibilities of the substrate is calculated the subquantity necessary for wiring a predetermined electric circuit, that from the data of said subquantity are produced control data for guiding at least one contacting means, that on the basis of said data a contacting or separating process is carried out on the conductors and contact points of the substrate, that the wiring carried out is tested and the test data are used for determining the quantity of additional wiring possibilities wiring the predetermined circuit diagram and which are made available for possible reworking of the substrate.

2. Process according to claim 1, characterized in that the wired universal substrate is returned for further wiring, which utilizes the additional wiring possibilities.

3. Process according to claim 2, characterized in that following a return, the wiring carried out is tested and the test data used for determining the quantity of additional wiring possibilities wiring the predetermined circuit diagram and made available for further reworking of the substrate.

4. Process according to one of the claims 2 or 3, charac-terized in that the return and determination of additional wiring possibilities and reworking takes place recursively.

5. Process according to one of the claims 1 to 4, charac-terized in that a multi-microcontactor is controlled with the control data.

6. Process according to claim 5, characterized in that the contactor is controlled in such a way that stepwise a modular spacing is obtained, the contacting points being reciprocally spaced by a multiple of the modular spacing.

7. Process according to one of the claims 1 to 4, charac-terized in that an operator (0?) defined for a reoptimization of a wiring of the already processed universal substrate, acts in the operating process in such a way that during a single passage of a substrate there is a virtual reoptim-ization, i.e. the correction potential is not used, or in the case of a multiple passage a real reoptimization takes place and consequently the correction potential is used.

8. Process according to claim 7, characterized in that the data (67) of the use-specific substrate plane are on the one hand supplied to a calculating process (74) for determining the optimum contacting process, which also requires for this purpose the characteristic data (64) of the contacting device (4), that on the other hand they are supplied to a further calculating process (75) for fixing the testing process, which also requires the charac-teristic data (65) of the testing device (5) and that from the two calculation processes are formed the control data (61) for the contacting device (1) and the control data (62) for the testing device (2).

9. Process according to claim 8, characterized in that following contacting (71) of the substrate in the contactor, testing (72) of the wired substrate takes place in the testing device and in the latter the substrate is measured and tested for preset information, i.e. the wiring plan or wiring of the electric circuit diagram, from the computer (80).

10. Process according to claim 9, characterized in that if the preset information is fulfilled the substrate is passed on as usable (83) and if the preset information is not fulfilled (82) and no return is planned, then the substrate is rejected as waste (84).

11. Process according to claim 7, characterized in that for the case, where a return is to take place in the case of unfulfilled preset information, following an identical initialization (89) and contacting (71), an evaluation (76) of the existing connections takes place at testing means (72).

12. Process according to claims 7 or 11, characterized in that the determination of the additional wiring possib-ilities takes place in two steps, in which firstly using a simple process (78) it is established whether a correcting connection design is possible (85) and, if this is the case (86), the more complicated optimization (70') is carried out, whilst incorporating all the criteria.

13. Process according to claim 12, characterized in that from the optimization means (70'), the data (67') of the corrected substrate plan or the correction contacting operations to be carried out are taken and from same on the one hand, together with the characteristic data (64) of the contacting device (4) the optimum correcting contacting process (74') is calculated and on the other hand together with the characteristic data (65) of the testing device (5) the testing process (75') is calculated and with the corresponding data (61',62') a new contacting and testing is carried out.

14. Process according to claim 7, characterized in that in place of the data (63) of the desired electric circuit diagram, the connections of the ideal substrate are preset as data (63") of the circuit diagram, whilst in a first passage there is no contacting or separating operation and consequently the relevant control data (61") are deter-mined without a calculating process and made directly available to the contacting process (71)(1).

15. Process according to claim 14, characterized in that in an additional branch (77) testing takes place to establish whether it is the first passage and if this is so (78), the procedure of claims (11 to 13) is adopted.

16. Apparatus for producing electric connections and/or for separating conductors on a substrate with a universal pattern of conductors and connecting points, characterized by a microcontacting unit (2,4) and by a micromeasuring unit (3,5), a computer (1) for calculating the control data for the microcontacting unit (2,4) and micromeasuring unit (3,5), as well as a device (6,7) or (6-11) for supplying universal substrates to the microcontacting unit (2,4) and for the latter to the micromeasuring unit (3,5) and from the latter to the microcontactor (2,5).

17. Apparatus according to claim 16, characterized in that the microcontactor unit (2,4) has a control part (2) and a multi-contactor part (4).

18. Apparatus according to claim 16, characterized by an arrangement with a computer (1), which contains the characteristic data for the controls and the circuit diagram data, calculates the wiring optimization of the universal substrate and converts it into control data for the control (2) of the microcontactor (4) and the control (3) for the testing device with the microsensor (5), with which data said two devices are positioned over the substrate, as well as two transporting devices (6,7), which are controlled by the control (2) of the microcontactor and the control (3) of the testing device, the control (2) of the microcontactor controlling a drive (8) for the transporting device (7), which suppolies the blanks to the contacting means and the programmed substrates to the testing device and the testing device control (3) controls a drive (9) of the transporting device (7), which returns waste substrates for correction contacting.

19. Apparatus according to claim 17, characterized in by a guide body (15) having a matrix-like arrangement of perforations (16), through which are passed the contacting elements (17) and are axially movable and are driven in controlled manner by a drive means (18).

20. Apparatus according to claim 19, characterized in that the perforations are arranged with a spacing (DN), which is an integral multiple of the crossover or conductor spacing (DK) of the substrate, the (DK) of the crossover spacing representing the unit of the positioning steps in which the contacting head is positioned over the sub-strate.

21. Apparatus according to claim 16, characterized in that the unit (2,4) contains at least one contacting element (17) in the form of hollow guides (71,81,91), in which through the internal opening can be introduced energy for subsequent welding following the mechanical deformation of the conductors.

22. Apparatus according to claim 21, characterized in that for the mechanical contacting of conductors (L1,L2) it has a pressure-exertable light guide (83) movable out of the hollow guide.

23. Apparatus according to claim 21, characterized in that for mechanical contacting, a hollow guide has a contacting ring (74,94) designed for the mechanical stressing.

24. Apparatus according to one of the claims 21 to 23, characterized in that a laser source (72,82,92) is provided enabling the laser energy to be passed either directly through the cavity or via a light guide (83,93) to the previously mechanically contacted conductors (L1,L2).

25. Apparatus according to one of the claims 21 to 24, characterized in that the hollow guides, light guides and laser source (71-74, 81-83, 91-94) form part of a multi-contacting head of the multi-contacting unit (5).

26. Universal substrate, contacted according to the process of claim 1, characterized in that it has substrate-free, mechanically deformed conductor bridges pressed onto a crossing conductor and/or welded to the crossing conductor and/or separated conductors.