US 5,081,602 AGrant
Computer Simulator for Electrical Connectors
Issue Date:1992-01-14
•7 Claims
•6 Drawing Sheets
Abstract
A computer model for an electrical connector of the type having multiple signal carrying paths utilizes both preselected, connector defining parameters which define electrical characteristics of the signal carrying paths of the electrical connector, as well as user selectable, application specific coded information that defines a set of signals and allocates each of the signals to a respective one of the signal carrying paths. The computer model models the electrical connector using the connector defining parameters to define the electrical connector and the application specific coded information to define the user selectable signals on the signal carrying paths of the electrical connector. This model can be used to investigate various pin assignments in order to optimize connector performance for an application specific set of signals. The disclosed connector model can also be incorporated into a larger model of an electronic system that includes at least one connector in order to emulate the electrical characteristics of the connector in the systems.
Metadata
Assignee
- AMP Incorporated
Inventor
- Douglas W. Glover
Application Information
Application Number:US 4336348
Filing Date:1989-11-08
Priority Date:1989-11-07
Art Unit:234
Classifications
IPC:
G06F 1500
Field of Search:
364578;480;481;488;489;490;491;801;802;580;200;900;482;483
Patent Drawings (6 sheets)
Description
Background of the Invention
A copy of the program listings are contained on 4 pages, 314 frames of microfiche film.
This invention relates to a computer simulator for electrical connectors and to a method for using such a simulator to allow a circuit designer to optimize pin selection for signals passing through the connector.
Electrical connectors have long been used in a wide variety of electronic circuits. In the past, electrical connectors have often been considered simply as an array of conductors, which do not contribute significantly to the operational characteristics of the electronic circuit in which they are used. However, as electronic switching speeds have increased, the electrical characteristics of electrical connectors have become an important part of circuit design.
The electrical characteristics of a connector are complicated by the fact that they are in part governed by the basic physical structure of the connector, but they are also in part governed by the manner in which the connector is used, and in particular by the pin selection made by the circuit designer The selection of the geometrical arrangement of the various signals among the signal paths provided by a connector can be an important part of an overall application design for an electrical connector.
In the past, computer simulation has been widely used to simulate individual electronic components, integrated circuits, and conductors on a printed circuit board. Such computer simulation allows a designer to simulate the operation of an electronic circuit before it is actually built. This conventional type of computer simulation is not, however, well suited for simulation of electrical connectors. This is because part of the electrical characteristics of the connector are determined by the manufacturer of the connector, who determines the connector geometry, and part of the electrical characteristics of the connector are determined by the user, who determines which signals will be applied to which pins of the connector. Thus, a need presently exists for an improved computer simulator that can be used to simulate both these aspects of an electrical connector, that will allow the user to simulate an electronic circuit including an electrical connector, and that will allow the user to select pins on the connector for desired signals. Such an interactive computer simulator for an electrical connector would greatly assist a user in optimizing pin selection and in simulating a complete electronic system, including both the electronic circuits and the connectors of the system.
In the past, non-interactive calculations have been used to predict the electrical characteristics of an electrical connector. For example, equivalent circuits for individual signal paths within a connector have been determined, and matrices have been used to determine capacitance effects of one signal path on another. However, such non-interactive calculations do not allow ready examination of the effects of different pin selections on the electrical characteristics of the connector.
Summary of the Invention
According to this invention, a computer model is provided for an electrical connector of the type having a plurality of signal carrying paths. This computer model comprises means for defining a plurality of preselected, connector-defining parameters which define electrical characteristics of the signal carrying paths of the electrical connector. The model also includes means for receiving application specific coded information that defines a set of signals and allocates each of the signals to a respective one of the signal carrying paths. The electrical connector is modeled using the connector defining parameters to define the electrical connector and the application specific coded information to define the user selectable signals on the signal carrying paths of the electrical connector.
The computer model of this invention allows a circuit designer to simulate an entire system (including both electrical circuits and connectors) and to explore alternative pin selection patterns for an electrical connector in order to optimize electrical performance of the connector.
This invention is also directed to a method for determining expected electrical performance characteristics of a circuit comprising at least one electrical connector comprising the step of providing a computer model for an electrical connector having a plurality of signal paths. This model has a first set of preselected, connector defining parameters which define a fixed geometry characteristic of the electrical connector, and a second set of user selectable, application specific parameters which define a plurality of signals applied to the signal carrying paths. The application specific parameters are then selected to define an initial set of signals applied to the signal carrying paths and the computer model is run to determine expected effects of the electrical connector on at least some of the initial set of signals. Then a second set of application specific parameters is selected to define another set of signals applied to the signal carrying paths, and the computer model is run to determine the expected effects of the electrical connector on at least some of the other side of signals. In many applications, the two sets of signals will comprise exactly the same signals allocated to differing signal carrying paths. In this way, a circuit designer can explore the optimum pin selection pattern to minimize undesired characteristics, such as cross talk.
The invention itself, together with further objects and attendant advantages, will best be understood by reference to the following detailed description, taken in conjunction with the accompanying drawings.
Brief Description of the Drawings
FIGURE 1 is a schematic view of a connector used to illustrate the preferred embodiment of the computer simulation method and apparatus of this invention.
FIG. 2 is a schematic view of the connector of FIG. 1 showing pin layout.
FIG. 3 is a schematic representation of a connector signal path model included in the connector model of this embodiment.
FIG. 4 is a schematic view of a capacitive matrix used in the model of FIG. 3.
FIG. 5 is a schematic representation of an inductive matrix included in the model of FIG. 3.
FIGS. 6a, 6b and 6c are schematic diagrams showing connector signal path models included in an interconnection model to simulate the respective signal path carrying a reference signal (FIG. 6a), used as an inactive signal carrying line (FIG. 6b), and used as an active signal carrying line (FIG. 6c).
FIG. 7 is a schematic representation of the computer simulation model of this embodiment.
FIG. 8 is a flow chart demonstrating one method for using the computer model of FIG. 7.
FIGS. 9 and 10 are schematic representations of the connector 10, showing two alternative pin selection patterns that can be modeled using the model of FIG. 7.
Detailed Description of the Presently Preferred Embodiments
Turning now to the drawings, FIG. 1 shows a schematic representation of a connector 10 that will be used for illustration in this discussion. The connector 10 can be any suitable connector, such as a right angle high density connector for connecting a daughter board to a back plane in a computer system. The connector 10 includes two connector halves 12, 14 which can be mated and separated in the usual manner. The connector 10 defines a plurality of signal carrying paths or signal paths 16 extending between respective input lines 18 and output lines 20.
FIG. 2 shows a schematic representation of the signal paths 16 in the connector 10, and the notation of FIG. 2 will be used in the following discussion. Note that each signal path 16 is characterized by a row and a column. The details of construction of the connector 10 form no part of this invention, and will therefore not be described in any detail. The example provided by FIGS. 1 and 2 is a relatively simple connector. It will of course be understood that the present invention can be applied to much more complex connectors if desired.
In FIG. 3, the reference numeral 22 is used to refer generally to a computer model of the connector 10 which incorporates the presently preferred embodiment of this invention. This computer model 22 can be thought of as comprising a plurality of signal path models 24, one for each of the 24 signals paths 16 of the connector 10. The signal path model 24 simulates the electrical characteristics of the connector 10 as determined by the geometry of the various conductive and non-conductive elements of the connector 10.
Each of the signal path models is organized around three respective nodes 26, 28, 30. Each node is numbered with a code number having three components. The node 26 carries the code number 1RCC, where R is the row number and CC is the column number of the respective signal carrying path 16. Similarly, the nodes 28 and 30 are identified by the code numbers 2RCC and 3RCC. For example, the signal carrying path 16 at the upper left hand corner of FIG. 2 (Row 3, Column 1) would use the following code numbers to identify the nodes 26, 28, 30: 1301, 2301, 3301.
Each signal path model 24 includes an equivalent circuit 32 for the respective signal path 16 that includes two series constants, RC and LC. RC indicates the series resistance characteristic of signal path 16, and LC indicates the inductance of the respective signal path 16. As shown, series element RC is interposed between nodes 26 and 28, and series element LC is interposed between nodes 28 and 30.
Additionally, each of the signal path models includes elements that together make up a capacitive matrix 34 and an inductive matrix 38, which will be discussed below in connection with FIGS. 4 and 5.
FIG. 4 shows a schematic representation of the capacitive matrix 34, which is made up of an array of capacitive elements 36. Each set of capacitive elements 36 is associated with a respective pair of signal paths 16. The capacitive elements provide the capacitive coupling constant for the respective pair of signal paths.
It will be understood that the most significant capacitive coupling is between adjacent signal carrying paths. In this model, a capacitive element is provided for each pair of adjacent signal paths, as indicated schematically by horizontal, vertical and diagonal lines in FIG. 4. In this embodiment, the capacitive elements are labeled CV, CH, CR and CL. CV is indicative of the capacitive coupling constant between a signal carrying path and the signal carrying path in the same column and the next larger row, as indicated by the symbol V in FIG. 4. The capacitive element CH is indicative of the capacitive coupling constant between a signal carrying path and the adjacent path in the same row and the next larger column, as indicated by the symbol H in FIG. 4. Capacitive element CR is indicative of the capacitive coupling constant between a signal carrying path and the diagonally situated path at the next larger column and row, as indicated by the symbol R in FIG. 4. Finally, capacitive element CL is indicative of the capacitive coupling constant between a signal carrying path and the diagonally adjacent signal path having the next larger row and smaller column, as indicated by reference symbol L in FIG. 4. It will be recognized that in this example there are sixty-five pairs of adjacent conductive paths as defined above, and the computer model 22 includes sixty-five capacitive elements.
The inductive matrix 36 is schematically shown in FIG. 5, and it is organized in substantially the same way as the capacitive matrix 34 discussed above. The inductive matrix 36 includes sixty-five inductive elements 40 in this example, each indicative of the mutual inductive constant between an adjacent pair of signal paths. In this embodiment, the inductive elements 40 are denoted by the symbols KV, KH, KR and KL, using the same symbols V, H, R and L as discussed above.
Table 1 provides a listing of portions of the computer model 22, and Table 1 will be used to clarify the foregoing discussion relating to the signal path model 24 of the computer model 22. In Table 1, consecutive lines are labeled with numbers between 1 and 461, and these line numbers will be used in the following discussion.
The portion of connector model 22 that specifies the connector defining parameters described above is found in Table 1 at lines 188 through 461. Lines 193 through 271 define the series elements RC, LC for each of the signal paths 16. In Table 1, rows A, B and C correspond to rows 3, 2 and 1, respectively. Each line defining one of the series elements RC, LC includes four entries. Note, for example, line 196 where the first entry (RC1101) identifies the series element being defined. The next two entries identify the nodes between which the defined series element is connected, in this example between node 1101 (node 26 of the signal path for row 1, column 1) and node 2101 (node 28 for the signal path for row 1, column 1). The fourth entry indicates the resistance (4.times.10.sup.-3 Ohms) for the series element RC being defined.
The notation for the inductive series elements LC is identical, except that the units of measurement for the last entry is Henries. Thus, line 197 defines the series element LC (element LC2101) as extending between nodes 28 and 30 for the signal path at row 1, column 1, and as having an inductance of 10.times.10.sup.-9 Henries. Lines 193 through 271 define the forty-eight series elements RC, LC needed for the twenty-four separate signal paths 16 of the connector 10.
Lines 272 through 365 of Table 1 define the sixty-five capacitive elements 36 for the capacitive matrix 34 discussed above. Once again, each line contains 4 entries. The first entry defines the element being modeled, and the second and third entries define the nodes between which the element is connected in the model. The fourth element of each line defines the capacitance in Farads of the defined element. For example, in line 276 capacitive element CV2101 is defined as extending between nodes 2101 and 2201 and having a capacitance of 0.39.times.10.sup.-12 Farads. This capacitive element CV2101 represents the capacitive coupling constant between the signal path at Row 1, Column 1, and the signal path at Row 2, Column 1. Similarly, the coupling constant defined in line 277 (CH2101) defines the coupling constant between the signal path at Row 1, Column 1 and Row 1, Column 2 as 0.45.times.10.sup.-12 Farads. The capacitive element CR2101 defined in line 278 represents the capacitive coupling constant between signal paths at Row 1, Column 1 and Row 2, Column 2, and defines this constant as 0.20.times.10.sup.-12 Farads.
Lines 366 through 459 define the inductive elements 40 of the inductive matrix 38 discussed above. The notation is similar to that discussed above, and the fourth entry on each line defines the ratio of current in one of the signal paths to induced current in the other. Thus, line 370 defines inductive element KV2101 (which represents the mutual inductive constant between the signal paths at Row 1, Column 1 and Row 2, Column 1) as a ratio of 0.18.
At this point, it should be emphasized that the signal path models 24 make up a connector model, and that this connector model is defined by the entries on lines 189 through 460 of Table 1. The series constants, capacitive elements and inductive elements defined here are determined by the physical geometry of the connector 10 itself and thus cannot easily be modified in a predictable way by the user. These constants are preselected constants which define the connector 10 and are outside the control of the connector user.
Of course, when installed in a circuit the connector 10 is interconnected with individual signal paths 16 carrying different types of signals. In order to properly simulate the operation of the connector 10, these interconnections must be modeled as well.
In this embodiment, each of the signal path models 24 can be interconnected either as a reference line carrying a reference signal such as a ground voltage, an inactive connected line which carries no active signal during modeling, or as an active signal carrying line which carries a signal generated by a voltage source. FIGS. 6a through 6c show these three possibilities and define certain parameters used in the example of Table 1.
If a particular signal path is being used to carry a reference voltage such as ground, that signal path is modeled as shown in FIG. 6a. The respective signal path model is connected via reference resistors RR at the input side to a ground node 0000, and at the output line to a node 4000. This reference resistance is meant to indicate a typical resistance between the respective signal path and ground and is typically made up of conductor resistances, solder joint resistances, and the like.
If a signal path is being used as a signal carrying rather than a reference path, yet is not actively carrying a signal, the respective signal path model 24 is connected as shown in FIG. 6b. The input line of the respective signal path model 24 is connected via a parallel resistance RT and capacitance CP to the ground node 0000. The resistance RT is meant to define the termination resistance of the associated transmission line, and the capacitance CP the appropriate capacitance, which is to a large extent associated with physical features such as plated through holes in a printed circuit board. Similarly, a terminating resistance RT and a capacitance CP are connected in parallel between the output line of the respective inactive line signal path model and the node 4000.
In the event a signal path is actively carrying a signal during the simulation, the respective signal path model 24 is modeled as shown in FIG. 6c as connected at its input end via a capacitance CP to the ground node 0000 and via a resistance RS to a node nn. The resistance RS is meant to simulate the source impedance. The node nn is then connected to the node 0000 via a voltage source VS. This voltage source generates a signal that is applied via the node nn and the resistor RS to the input line of the signal path model 24. The output line of the signal path model is connected via parallel resistance and capacitance elements RT, CP to the node 4000.
Thus, by allocating resistances RR to reference lines, resistances and capacitances RT, CP to inactive signal carrying lines, and resistances RS, RT, capacitances CP, and voltage sources VS to active signal carrying lines, the various signal paths 16 of the connector 10 can be modeled to simulate a particular choice of pins for the respective signals.
In the example of Table 1, this is done at lines 21 through 63 and lines 80-188. In Table 1 the signal paths indicated with an X in FIG. 9 are assigned as reference lines connected to ground with reference resistances RR. This is done at lines 81 through 103 of Table 1, where RR is set equal to 1.times.10.sup.-6 Ohms in each case. In this example the signal paths indicated with the letter A in FIG. 9 are active signal carrying lines, and the signal paths which carry neither an X nor an A are inactive lines. In Table 1, lines 104 through 145 define the terminating resistances RT for the input and output sides and lines 146 through 188 define the capacitances CP for the input and output sides. Once again resistances are expressed in Ohms and capacitances in Farads.
Lines 21 through 63 define the voltage sources VS and the source impedances RS for the active lines shown in FIG. 9. Source impedances RS are defined as shown for example in line 32 as each having an identifying number (RSO2 in this example) which is interconnected between two nodes (node 02 and node 26 of Row 1, Column 6) and having a resistance of 60 Ohms. Voltage supples are defined in terms of the lower and upper voltages, delay, rise time, fall time, pulse width and period. Note for example line 30 of Table 1 where voltage source VS02 is defined between the ground node 00 and node 02 as a signal making a transition from 2.0 to 4.0 volts with zero delay and a 3.0 nanosecond rise time. In this example, the signal is not periodic and fall time, pulse width and period are therefore not defined.
From this explanation it should be apparent that the circuit designer can easily select the signal paths in the connector 10 to carry selected signals by properly choosing the parameters RR, RT, CP, RS and VS. FIG. 7 shows an overview of the manner in which the computer simulation model 22 described above functions. As pointed out above, the computer simulation model 22 includes a number of preselected connector defining parameters including the series elements RC, LC, the capacitive elements CV, CH, CR, CL and the inductive elements CV, KH, KR, KL. These preselected connector defining parameters are illustrated in Table 1 at lines 189 through 460, and they define the predetermined geometry of the connector, as chosen by the connector manufacturer.
The computer simulation model 22 includes a second set of user selectable, application specific parameters that define the signals applied to the various signal paths of the connector 10. These application specific parameters include referencing parameters RR, termination parameters CP, RT, and excitation parameters RS, VS. The circuit designer can readily choose which pins are to carry which signals by appropriate selection of these application specific parameters. The computer simulation model utilizes these two sets of parameters as shown at blocks 42 and 44 to simulate the electrical characteristics of the connector, as used in the particular application being modeled. The computer simulation model 22 generates an output indicative of simulated signals on selected signal paths, as indicated at block 46 of FIG. 7. For example, the computer simulation model can be used to track voltage as a function of time on any one of the signal carrying paths of the connector 10.
The details of operation of the computer simulation model 22 do not per se form part of this invention. In this preferred embodiment the SPICE computer simulation program has been used. Those skilled in the art will recognize that a wide variety of models could be adapted for use with this invention. However, the code of Table 1 has been created for use with the SPICE computer simulation program, and a copy of the listing for this program is attached in microfiche Appendix 1. The listing of Appendix 1 is in FORTRAN source code, and can be compiled for various computers. Each sheet of the microfiche is entitled, "Computer Instructions," and includes four sheets numbered consecutively, 001-004.
FIG. 8 provides a flow chart showing how the computer simulation model 22 can be used to aid a designer in optimizing connector performance in a specific application. As a first step in the process illustrated in FIG. 8 the computer simulation model 22 is provided having preselected connector defining parameters (block 42 of FIG. 7) that characterize the physical connector being used. The user then selects application specific parameters for a first proposed signal path allocation (block 44 of FIG. 7). For example, FIG. 9 shows one such first proposed signal path allocation. As pointed out above, in FIG. 9 signal paths used to carry a reference potential are indicated with an X, signal paths that are actively carrying signals are indicated with an A, and inactive signal carrying paths are left unmarked. A box marked with the letter M is used to designate the signal paths which will be monitored, i.e, on which simulated signals will be calculated by the computer simulation model 22. Note in FIG. 9 that the reference lines are arranged in a staggered pattern on Rows 1 and 3. The user then runs the model 22 to obtain a first output indicative of the anticipated signals on the signal paths at Row 2, Columns 2, 4 and 5 and Row 1, Column 2 induced by the signals carried by the active lines.
The user can then revise the application specific parameters (block 44 of FIG. 7) as appropriate for a second proposed signal path allocation. One possible alternative is illustrated in FIG. 10, where the positions of the reference lines in Row 3 have been shifted, without increasing their number. The computer simulation model 22 is then run again obtaining a second output, which in this case is the simulated signals on signal paths at Row 2, Columns 2, 4 and 5 and row 1, Column 2. The user can then compare the first and second outputs as an aid to optimizing signal path allocation.
As pointed out above, a designer can materially enhance circuit operation in modern high speed electronic circuits by properly allocating reference, active and inactive lines to appropriate signal paths in the connector. It should be apparent from the foregoing discussion that the computer model 22 can greatly assist a circuit designer in this optimization.
The connector simulation model of this invention can also be used as part of a larger computer simulation model for an entire electronic system. Such a larger model would model electronic circuits included in the system, and would use the connector model of this invention to model the electrical connectors of the system. In this way the entire system can be simulated, and the effect of electrical connectors on system operation can be predicted. When used in this way, the connector model of this invention emulates the electrical characteristics of the physical connector as incorporated in the electronic system. In such applications, the larger computer simulation model generates the application specific coded information that defines the signals applied to the connector.
Of course, those skilled in the art will recognize that this invention can be implemented using a wide variety of computer simulation programs, and it can be adapted to simulate a wide variety of connectors. Furthermore, it should be apparent that the particular values used for the connector defining parameters and for the application specific parameters will vary depending on the one hand on the geometry of the connector and on the other hand on the specific application. The specific values used and the manner for determining these values do not form part of this invention, and are therefore not described in any detail here. Those skilled in the art will recognize that suitable values can be obtained either by direct measurement or by computer simulation. ##SPC1##
Claims
I claim:
1. A method for allocating a plurality of signals to specific signal carrying paths of an electrical connector, comprising the following steps: a) providing a computer system for modelling an electrical connector, said system comprising a computer and a computer implemented model for an electrical connector having a plurality of signal carrying paths, said model comprising a first set of stored, preselected, connector defining parameters which define a fixed geometry characteristic of the electrical connector and a second set of user-selectable, application specific parameters which define a plurality of signals applied to the signal carrying paths, said computer model arranged to execute on the computer; b) selecting the second set of application specific parameters to define an initial set of signals applied to the signal carrying paths; c) executing the computer model to determine expected effects of the electrical connector on at least some of the initial set of signals; d) selecting the second set of application specific parameters to define another set of signals applied to the signal carrying paths; e) executing the computer model to determine expected effects of the electrical connector on at least some of the other set of signals defined in step (d); and f) assigning a physical allocation of one of the initial and the other sets of signals to respective signal carrying paths of the electrical connector after comparing the expected effects of the electrical connector on at least some of the associated signals in steps (c) and (e) as an aid to optimizing signal allocation to selected signal carrying paths of the connector.
2. The method of claim 1 wherein the initial and other sets of signals comprise the same signals, and wherein at least some of the signals of the initial set are allocation to difference signal carrying paths in the respective sets.
3. The method of claim 1 wherein the plurality of pre-selected connector defining parameters comprise a plurality of series constants each indicative of at least a portion of an equivalent circuit for each of the signal carrying paths.
4. The method of claim 3 wherein the plurality of pre-selected connector defining parameters additionally comprise a capacitive matrix comprising a plurality of capacitive elements, each indicative of a capacitive coupling constant for a respective pair of the signal carrying paths.
5. The method of claim 3 wherein the plurality of pre-selected connector defining parameters additionally comprise an inductive matrix comprising a plurality of inductive elements, each indicative of a mutual inductance constant for a respective pair of the signal carrying paths.
6. A method for allocating a plurality of signals to specific signal carrying paths of an electrical connector, said method comprising the following steps: a) providing a connector simulator for simulating an electrical connector having a plurality of signal carrying paths, said simulator comprising a first set of stored, preselected, connector defining parameters which define a fixed geometry characteristic of the electrical connector and a second set of user-selectable, application specific parameters which define a plurality of signals applied to the signal carrying paths, said simulator operative to simulate on a computer and to generate an output indicative of changes in at least some of the signals in the electrical connector, including changes in said at least some of the signals causing by electrical influences of the signals on one another; b) selecting the second set of application specific parameters to define an initial set of signals applied to a first pattern to the signal carrying paths, wherein the initial set of signals comprises reference potential signals and active signals and allocates the reference potential signals and the active signals in said first pattern to respective selected ones of the signal carrying paths; c) using the simulator to produce a first output indicative of expected effects of the electrical connector and the associated signals on at least some of the initial set of signals; d) revising the second set of application specific parameters to re-allocate at least some of the initial set of signals to different selected ones of the signal carrying paths in a second pattern; e) using the simulator to produce a second output indicative of expected effects of the electrical connector and the associated signals on at least some of the initial set of signals as re-allocated in step d); and f) assigning a physical allocation of the initial set of signals to respective signal carrying paths of the electrical connector in correspondence with one of the first and second patterns after comparing the first and second outputs obtained in steps c) and e) as an aid to optimizing allocation of the reference and active signals to selected signal carrying paths of the connector.
7. The method of claim 6 wherein the application specific parameters of steps (b) and (d) define the same signals, and wherein the application specific parameters of step (d) reallocate at least some of the signals to different signal carrying paths than do the application specific parameters of step (b).