DIGITAL CROSSPOINT SWITCH
FIELD OF THE INVENTION
This invention relates to a mirror pin-out arrangement for a digital data crosspoint switch.
BACKGROUND OF THE INVENTION
Crosspoint switches are electronic devices that are used to route inputs to outputs using a programmed configuration. The input/ output mapping can be altered at any time utilizing programming inputs to the device, so that a configurable switching matrix can be designed.
One of the challenges in digital data switching matrix design is to connect multiple crosspoint devices together to create larger switching matrices. In the past this has required the use of complex multi-layer circuit boards and high speed switches /multiplexers to multiplex inputs to multiple crosspoint devices.
Improvements in such crosspoint devices are described in the copending application of Eric Fankhauser and assigned to the assignee of the present invention, and filed concurrently herewith. (A copy of the disclosure and drawings of that application is attached as Appendix A hereto and is incorporated by reference herein.) However there remains a need for additional simplification.
BRIEF SUMMARY OF THE INVENTION
Accordingly, it is an object of the invention to provide a pair of crosspoint devices each having the same electrical characteristics, but in which the pin-outs have been mirrored across the vertical axis of one device as compared with the other. This allows simplified connection of the devices in matrices as will be explained.
In one aspect the invention provides a set of first and second crosspoint devices, each containing a crosspoint switch and each having
first and second opposed sides and third and fourth opposed sides, and wherein in said first device, said first side contains input pins, said second side contains control pins, said third side contains expansion data input pins, and said fourth side contains output pins, and in said second device, said first side contains control pins, said second side contains input pins, said third side contains expansion data input pins arranged in the reverse order from that of the expansion data input pins of said first side of said first device, and said fourth side contains output pins arranged in the reverse order from that of the output pins of said first device. Further objects and advantages of the invention will appear from the following description, taken together with the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
In the drawings: Fig. 1 is a plan view of a set of devices according to the invention and showing the mirror pin-out configuration;
Fig. 2 is a block diagram showing a 16 x 16 matrix using two sets of two devices each according to the invention;
Fig. 3 is a side view of a printed circuit board containing the devices of Fig. 2;
Fig. 4 is a block diagram showing two sets of two devices each according to the invention arranged in a 32 x 8 matrix; and
Fig. 5 is a side view of a circuit board containing the devices of Fig. 4.
DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS
Reference is first made to Fig. 1, which shows a device 20 having left and right sides 22, 24, and top and bottom sides 26, 28. Input pins 30 are connected to side 22, including eight sets of input data pins marked from INO, INO-B to IN7, IN7-B. The right hand side 24 of the device 20 contains control pins
32. The bottom side of the device 20 contains output pins 34, while the top side of the device contains expansion data input pins 36, for additional data inputs (as described in connection with said copending application).
Also shown in Fig. 1 is device 20', which is the mirror image of device 20 about the vertical axis of device 20. In device 20', pin-outs and reference numerals have been marked with the prime symbol ' to distinguish them from those of device 20.
In device 20', it will be seen that the input pins 30' are now on the right hand side 24 'of the device as drawn, while the control input pins 32' are on the left hand side 22' of the device. The expansion data input pins 36' remain at the top side 26 of the device but have been reversed in order as compared with device 20. Similarly the output pins 34' remain at the bottom side 28' of the device 20' but have again been reversed in order.
The arrangement shown in Fig. 1 allows switching matrix designers to deploy the two versions shown (the standard pin-out version
20 and the mirror pin-out version 20') on each side of a circuit board 40, as shown in Figs. 2, 3. Fig. 2 displays sides A and B of the circuit board 40, so that when the two halves of Fig. 2 are folded about vertical line 42, the complete circuit board 40 will be obtained (as shown in Fig. 3). Side A contains two devices 20 while side B contains two devices 20'.
As shown for side A of Fig. 2, the data inputs 46, 48 are received from the left hand side of the devices 20 and enter the devices 20 directly. In addition the traces which provide the inputs 46, 48 extend through vias in the circuit board 40 to the mirror devices 20' and serve as inputs to them as well. The remaining data pins of the devices 20, 20' are not connected together and therefore no additional vias are needed for them in the circuit board 40.
It will be seen that the arrangement in Figs. 2 and 3 permits an extremely compact 16 x 16 matrix using four devices (two standard devices and two mirror devices) connected together.
If additional vias in the circuit board are permitted to feed the output signal back and forth between the two sides of the circuit board,
then the arrangement shown in Figs. 4 and 5 may be used. Again Fig. 4 shows sides A and B of a circuit board 50, it being assumed that the two sides will be folded about vertical line 52 to produce the final circuit board, which is shown in Fig. 5. In Fig. 4, the inputs for standard devices 20-1, 20-2 come from the left hand side of the devices. These are inputs 0 to 7 and 16 to 23. On side B, the inputs 8 to 15 and 24 to 31 are received from the right hand side of the mirror devices 20'-l, 20'-2. This produces a 32 x 8 design, using four 8 x 8 devices. As shown in Fig. 5, the output from device 20-1 travels through traces 60 beneath device 20-1 (on the circuit board 50) and then through vias 62 of circuit board 50 to the expansion data inputs 36 -1 of mirror device 20'-l. The outputs of mirror device 20'-l travel through traces 64 beneath device 20 -1, through vias 66 on the circuit board to the expansion data inputs 36-2 of standard device 20-2.
The outputs 34-2 of standard device 20-2 again travel on traces 68 beneath device 20-2, through vias 70 in the circuit board 50 to the expansion data inputs 36 -2 of mirror device 20 -2. The outputs 34 -2 (outputs 0 to 7) of mirror device 20'-2 constitute the eight outputs of the matrix.
An advantage of the arrangement shown is that it permits an extremely compact design while minimizing the need for vias, complex printed circuit boards, and while reducing the resultant crosstalk which can occur in the circuit board. While preferred embodiments of the invention have been described, it will be appreciated that various changes may be made within the spirit of the invention and all such changes are intended to be included in the scope of the accompanying claims.
APPENDIX A
Title: CROSSPOINT SWITCH WITH IMPROVED PIN-OUTS AND
EXPANSION INPUTS
FIELD OF THE INVENTION
This invention relates to a crosspoint switch. In particular it relates to a unique pin-out arrangement for a crosspoint switch, and to expansion input port architecture for a crosspoint switch.
BACKGROUND OF THE INVENTION Crosspoint switches are electronic devices that are used to route data inputs to data outputs via a programmed configuration. The input/ output mapping can be altered at any time using programming logic inputs to the device, so that a configurable switching matrix can be designed. One of the significant challenges in digital data switching matrix design is the connection of multiple crosspoint switches or devices together to create larger switching matrices. In the past this has been accomplished by using complex multilayer circuit boards, and high speed switches or multiplexers which are connected to multiple crosspoint devices.
The use of external multiplexers and complex printed circuit boards has led to many difficulties, including difficulties in assembling the matrix into a compact configuration, crosstalk problems, and difficulty in terminating traces on the circuit board.
BRIEF SUMMARY OF THE INVENTION
Accordingly, it is an object of the present invention in one of its aspects to provide a pin-out arrangement for a digital data crosspoint switch which facilitates the design of a switching matrix using a number of individual crosspoint devices, and which at the same time reduces signal
APPENDIX A
crosstalk and circuit board routing complexity. In this aspect the invention provides a crosspoint device containing a crosspoint switch and having: (a) first and second opposed sides,
(b) a plurality of input pins on said first side,
(c) a plurality of second pins on said second side, a second pin being located opposite at least some of said input pins, (d) some of said input pins being data input pins coupled to said crosspoint switch, (e) the second pins opposite said data input pins being "no connect" pins which are not electronically connected to said crosspoint switch. In another aspect the invention provides expansion input ports for digital crosspoint devices, so that the devices can again be assembled in a matrix in a simple manner while reducing the need for complex printed circuit boards and also reducing input/output crosstalk. In this aspect the invention provides a crosspoint device containing a crosspoint switch and having:
(a) first and second opposed sides,
(b) a plurality of input pins on said first side,
(c) a plurality of second pins on said second side,
(d) some of said input pins being data input pins coupled to said crosspoint switch,
(e) said device also having a third and a fourth side each extending between said first and second sides, said third side containing output pins and said fourth side containing expansion data input pins coupled to said output pins,
(f) said expansion data input pins being self terminated. Further objects and aspects of the invention will appear from
APPENDIX A
the following description, taken together with the accompanying drawings.
BRIEF DESCRIPTION OF THE DRAWINGS
In the drawings:
Fig. 1A is a block diagram of a prior art multiplexed input array of crosspoint devices;
Fig. IB is a block diagram of another prior art input array of crosspoint devices;
Fig. 2 is a block diagram of a multiplexed output array for crosspoint devices;
Fig. 3 is a block diagram showing prior art bussed outputs for a crosspoint device array; Fig. 4A is a diagrammatic view of a pair of crosspoint devices according to the invention;
Fig. 4B is a diagrammatic view of a portion of a printed circuit board showing traces used with the device of Fig. 4A;
Fig. 5 is a block diagram of a crosspoint device having expansion input port architecture;
Fig. 6 shows in more detail a portion of the Fig. 5 arrangement;
Fig. 7 is a block diagram showing a matrix of devices connected together and using the expansion input port architecture of Figs. 5 and 6; and
Fig. 8 shows optional detail for the Figs. 5 and 6 device.
DETAILED DESCRIPTION OF PREFERRED EMBODIMENTS
Reference is first made to Fig. 1, which shows a conventional prior art crosspoint switch array having two n x n crosspoint devices 20, 22
APPENPIX A
having inputs 24, 26. In the past it was common to connect multiple crosspoint devices together to create larger switching matrices by the use of complex multi-layer circuit boards and high speed switches or multiplexers, such as multiplexer 28. Multiplexer 28 is connected to the inputs 24, 26 to create an n x m switching matrix, where n<m. Similarly, the outputs of the crosspoint devices were required to be multiplexed together to create a switching matrix, as shown in Fig. 2. In Fig. 2 the outputs 30, 32 of crosspoint devices 20, 22 are connected to another multiplexer 34 having outputs 36. (Alternatively, the inputs of crosspoint devices 22, 24 can simply be connected together as shown in Fig. IB, but this is not suitable for their outputs, as will be discussed.)
It is possible to provide devices which have outputs which are turned "on" or "off" utilizing a configuration interface. These switching matrix designs would not require external multiplexers in order to be connected together, and are illustrated in Fig. 3, where two switching crosspoint devices 40, 42 are illustrated as being connected together at connections 44 on a circuit board 46. Unfortunately, the design shown in Fig. 3 may suffer from crosstalk problems, particularly in the connection area 44 where the output traces 48 of one crosspoint device 40 cross the output traces 50 of the other crosspoint device 42. In addition, it will suffer from signal degradation from the interconnection of "off" outputs with "on" outputs. Further, since vias in the circuit board are required to connect the crosspoint devices' outputs to the output bus, additional crosstalk problems can occur. The output traces on the circuit board are also difficult to control and to terminate properly.
Reference is next made to Fig. 4A, which shows a crosspoint device 54 which facilitates connection of the device to a printed circuit board, and which also has other advantages which will be described.
The crosspoint device 54 of Fig. 4A is rectangular in plan, having left and right sides 56, 58 and top and bottom sides 60, 62. The
APPENDIX A
device 54 has input pins 64 at its left hand side, output pins 66 at its bottom side, control pins 68 at its right hand side, and self terminated expansion input pins 70 (to be described) at its top side. All of these pins, whether used or not, are soldered to the circuit board 72 (Fig. 4B) to which the device 54 is connected, for a secure mechanical connection.
It will be seen that the input pins 64 include a number of data input pins marked from INO, INO-B to IN7, IN7-B. According to the invention, the pins on the side of the device 54 opposite to these input data pins, namely pins 74, are "no connect" pins and are not used for any other purpose. As best shown in Fig. 4B, the circuit board traces 76 which carry the data inputs to the data input pins INO, INO-B to IN7, IN7-B are bussed across the circuit board and are connected to the "no connect" pins 74 and continue across the circuit board to be connected to the next crosspoint device 54A, as shown in Fig. 4A.
Because the only pins on the right hand side 58 of the device 54 which are connected to the circuit board traces 76 are the "no connect" pins 74, and since pins 74 are not connected to any internal circuitry in device 54, it is possible to route the circuit board traces 76 along a single layer of the circuit board 72. There is no need to have these traces routed down to a lower layer of the circuit board to go under the pins 74 and then to route them back up to the surface of the board for connections.
The arrangement shown in Figs. 4A and 4B therefore has a number of advantages. Firstly, it lowers the complexity of the circuit board 72 needed to hold a matrix of crosspoint devices, since fewer layers are needed for the circuit board. Secondly, it reduces the signal degradation effects which would normally be caused by vias and corners in the printed circuit board traces for high frequency digital data such as serial digital video data. Thirdly, it facilitates the simple design of large matrices made from individual crosspoint devices, and it allows higher crosspoint device density on a switching matrix circuit board. Finally, it allows all inputs to
APPENDIX A
have a single line termination, since any number of devices can be located on a single trace 76, and yet only one termination is needed, at the end of that trace.
Reference is next made to Fig. 5, which shows expansion input port architecture for a crosspoint device 54. In the past, crosspoint devices have normally included data inputs 80 (connected to the data input pins of Fig. 4 A), directed to an input buffer 82, which in turn is connected to the matrix (shown for example as an 8 x 8 matrix) 84 of crosspoints or switches. The matrix 84 is in turn connected to the matrix an output buffer 86 connected to data outputs 88 (the output pins OUT1, OUT1-B to OUT7, OUT7-B of Fig. 4A). The control inputs 90 are in turn connected to control input latches 92 which are used to control the matrix 84.
A difficulty with the prior art arrangement as described is that it has been very difficult to connect multiple crosspoint devices together in a matrix. Since external multiplexers were used, the architecture became cumbersome.
As shown in Fig. 5, the invention in one aspect provides self terminated expansion data inputs 96 on the top side 60 of the crosspoint device 54. These are connected to pins EXPO, EXP0-B to EXP7, EXP7-B of Fig. 4A. The reference to "self terminated" means that there is a transmission line impedance 97 (Fig. 6) located internal to the expansion port input. An impedance 97 is thus connected across each pair of pins EXPO, EXP0-B to EXP7, EXP7-B. This impedance can take the form of an active device (Bipolar, MOS, FET transistor(s)) or circuit or a passive device (resistor, capacitor, inductor) or circuit. Resistors 97 are shown as an exemplary implementation. Resistors 97 are integrated as part of the device 54 and are not external. The self terminated expansion data inputs 96 are connected through buffer 98 to a 2 x 1 multiplexer 100, which is in turn connected to the crosspoint matrix 84 and to the data output buffer 86. In Fig. 5 the control inputs 90 control the crosspoint matrix 84 as usual and
APPENDIX A
also control the multiplexer 100 so that the desired outputs (either from the data inputs 80 or directly from the self terminated expansion data inputs 96, as selected) are fed to the combined output stage and buffer 86 and hence to the data outputs 88.
It will be seen that in operation, the multiplexer 100, under control of the control latches 92, will direct either the outputs from the crosspoint array, or the expansion data inputs, to output buffer 86, and hence to the data outputs 88. Thus, data received on the self terminated expansion data inputs is routed directly through the device 54 to the data outputs, rather than being routed through the circuit board. This allows additional devices to be added easily to widen the matrix, as explained in connection with Fig. 7. Fig. 7 shows an example in which four crosspoint devices 54-
1, 54-2, 54-3, 54-4 are connected together in a matrix. As shown, the data inputs 80-1 for device 54-1 are connected to the input pins 64-1 of device 54-1 and are also bussed across device 54-1 and are connected to the input pins 64-2 of device 54-2 (using the arrangement of Figs. 4 A, 4B). A second set of data inputs 80-2 is connected to the data input pins 64-3 of device 54-3 and is also bussed across device 54-3 and connected to the data input pins 64-4 of device 54-4, again using the arrangement of Figs. 4A, 4B.
The data outputs 88-1 of device 54-1 are connected to the self terminated expansion data inputs 96-3 of device 54-3. Similarly, the data outputs 88-2 of device 54-2 are connected to the self terminated expansion data inputs 96-4 of device 54-4. The data outputs of devices 54-3, 54-4 are shown at 80-3, 80-4 respectively.
The arrangement shown in Fig. 7 avoids the need to run traces from the outputs of the crosspoint device to a common output bus, and therefore crosstalk between output channels, and signal degradation due to signal reflections on the transmission line, can be greatly reduced. In addition, in the Fig. 7 arrangement fewer circuit board layers are needed
APPENDIX A
since the data outputs of each device simply line up. Further, because the data outputs are routed from the top of the switching matrix to the bottom through the devices 54 themselves and not beneath the devices, inputs can be simply bussed across the board without concern about input/output crosstalk (since the data input traces 76 do not cross output traces in the circuit board). The use of self terminating expansion ports eliminates the need for external line terminating resistors, allowing a much denser interconnect between the data outputs of one crosspoint device and the expansion port inputs of another crosspoint device.
Reference is next made to Fig. 8, which shows a power saving arrangement. Normally each crosspoint device 54 has an output voltage of approximately 800 mv. However the devices 54 themselves can be made to operate with data input voltages of only 400 mv. Therefore each device 54 has the ability to provide data outputs at 400 mv when it is driving the expansion input part of another device 54, or to provide data outputs at 800 mv when it is driving an external device such as an ECL logic gate or other electronic component which requires ECL or pseudo ECL type input logic levels in the range of an 800 mv signal swing.
In addition to reducing the output voltage by one half, the associated internal current required to bias the output stage can also be reduced by one half for the same load impedance (i.e. 100 ohms), or by more than half if the load impedance is increased (e.g. to 200 ohms). For this purpose, and as shown in Fig. 8, each output stage and buffer 86 is controlled by bandgap reference control circuit 100, which in turn is controlled by an external resistor 102. Depending on the value of resistor 102, the control circuit 100 sets the data outputs 88 at 400 mv or 800 mv .
APPENDIX A
Typically there are three modes of operation, as follows:
Output
Value of Output Stage Differential
Mode Resistor 102 ECL Voltage Load Drive Power
1 2 KΩ 800 mv 100 Ω normal
2 4 KΩ 400 mv 100 Ω lower
3 6 KΩ 400 mv 200 Ω lowest
The load drive resistors of 100 ohms (not shown) will simply be external resistors connected across the differential data outputs. The 200 ohm resistors are the integrated resistors 97.
Using the Fig. 8 arrangement, the devices 54-1, 54-2 in Fig. 7 can for example be operated at lowest power, since they are simply driving further devices 54-3, 54-4. The devices 54-3, 54-4, whose outputs are connected to external circuits, will be operated at normal power. This arrangement saves power and reduces heat dissipation.
While preferred embodiments of the invention have been described, it will be appreciated that various changes can be made within the spirit of the invention, and all are intended to be within the scope of the appended claims.