Method of Broadly Distributing Termination for Buses Using Switched Terminators

1. Field of the Invention

The present invention relates to buses and more particularly to termination of buses for use in information processing systems.

2. Description of the Related Art

In computer and information processing systems, various integrated circuit chips communicate digitally with each other over a common bus. The signal frequency at which this communication occurs can limit the performance of the overall system. Thus, the higher the communication frequency, the better. The maximum frequency at which a system communicates is a function not only of the time that it takes for the electromagnetic wavefronts to propagate on the bus from one chip to another, but also of the time required for the signals to settle to levels than can be recognized reliably at the receiving bus nodes as being high or low, referred to as the settling time.

The length of the settling time is a function of the amount of reflection and ringing that occurs on the signal bus. The more effective the termination of a bus system, the smaller the effects of reflection and ringing in the system and the shorter the overall settling time of the signal.

It has been discovered that a bus may be provided with broadly distributed signal termination by using switched termination logic where the pull up resistance of the driver corresponds to the characteristic impedance of the line and the pull down resistance of the driver corresponds to the number of drivers coupled to the line. Accordingly, signals being transmitted over the bus suffer relatively few reflections thus advantageously producing a shortened signal settling time, thereby increasing the attainable signaling frequency.

In a preferred embodiment, the invention relates to a method for terminating a line having a number of drivers coupled thereto, the line having a characteristic impedance. The method includes the steps of pulling up a signal on the line with a pull up resistance substantially corresponding to the characteristic impedance of the line; and pulling down a signal on the line with a pull down resistance substantially corresponding to the number of drivers coupled to the line.

Additionally, in a preferred embodiment, the invention relates to a method for terminating a line having a plurality of ends, the line having a characteristic impedance. The method includes the steps of coupling a plurality of drivers to the plurality of ends, each driver including a pull up circuit, and a pull down circuit, the plurality of drivers being divided into a driving driver and a plurality of receiving drivers, pulling up a signal on the line with the pull up circuit of the driving driver, the pull up circuit of the driving driver having a pull up resistance corresponding to the impedance of the line, the receiving drivers having a resistance corresponding to the impedance of the line when the driving driver is pulling up the signal; pulling down a signal on the line with the pull down circuit of the driving driver, the pull down circuit having a pull down resistance corresponding to the number of drivers coupled to the line, the receiving drivers having a resistance corresponding to the impedance of the line when the driving driver is pulling down the signal.

The present invention may be better understood, and its numerous objects, features, and advantages made apparent to those skilled in the art by referencing the accompanying drawings.

FIG. 1 shows a block diagram of an information handling system having a bus as well as driver circuits in accordance with the present invention.

FIG. 2 shows a block diagram of broadly distributed termination using switched termination logic of the information handling system of FIG. 1.

FIGS. 3A, 3B, 3C and 3D show schematic block diagrams of a variety of bit elements of the driver circuits having switched termination logic in accordance with the present invention.

The use of the same reference symbols in different drawings indicates similar or identical items.

Referring to FIG. 1, information handling system 100 includes a plurality of components 102 such as processor 102a, memory controller 102b, and I/O controller 102c. It will be appreciated that these components 102 may be any type of component commonly found in an information handling system. Each of these components 102 is generally configured as an individual integrated circuit chip. However, it is known to combine various components into a single integrated circuit chip. Components 102 are coupled via bus 104. Bus 104 includes a plurality of parallel lines which are coupled to individual signal outputs of each of the components 102. Components 102 are also coupled to a common reference voltage (REF)

Each component 102 includes a plurality of input/output circuits 108 which are coupled to individual signal paths of bus 104. Each input/output circuit 108 includes a receiver circuit 109 and a driver circuit 110. Each receiver circuit 109 is also coupled to the common reference voltage. Component 102 also includes impedance control circuit 112 which is coupled to each driver circuit 110. Impedance control circuit 112 controls driver circuits 110 so that the impedance of each pull up driver circuit is matched to that of the bus 104 and the impedance of each pull down driver circuit is equal to the impedance of the bus 104 divided by the difference of the number of drivers coupled to the bus minus one.

In operation, driver circuits 110 include switchable termination logic for controlling whether each driver circuit 110 is driving or terminating and thus each driver circuit includes circuitry which controls the pull up and pull down termination resistance of the driver circuit.

Referring to FIG. 2, each driver circuit 110 of information handling system 100 includes a plurality of signal nodes coupled to bus 104 having a characteristic impedance of Z.sub.O. Attached to each signal node at the end of a transmission line trace or stub is a the push-pull driver 110. Driver 110 includes a pull up driver circuit 202 and a pull down driver circuit 204. Pull up driver circuit 202 includes a pull up resistance that is substantially equal to (i.e., within 5% of) the characteristic impedance Z.sub.O of the transmission line. Pull down driver circuit 204 includes a pull down resistance that is substantially equal to(i.e., within 5% of) ##EQU1## where n is the number of nodes in the information handling system 100. The receiver, if any, functions as a small capacitive load; however, no DC current flows between the receiver and the transmission line. Accordingly, the receiver does not contribute to the output impedance of the driver.

When a node is driving signals onto line 104, the driver 110 of the node either pulls the signal up or pulls the signal down, as appropriate to send the desired digital signal. When a node is receiving, the pull up portion of the driver remains active continuously, thereby functioning as a terminating resistance which is matched to the impedance of the transmission line.

However, because the pull down resistance at the driving node is equal to the net resistance of the parallel combination of the pull up portions of all of the receiving nodes, the signal swing on the bus 104 is from VDD to VDD/2. Accordingly, the receivers are capable of resolving a swing of VDD/2.

Because the impedance at each receiving node is matched to the impedance of the transmission line, signals arriving at the receiving node terminate without reflection. The impedance at the driving node is also matched to the line when driving high. However, the impedance at the driving node is mismatched when driving the signal low because the impedance is equal to Z.sub.O /(n-1) instead of Z.sub.O. Accordingly, the only stub on which reflections are created is the stub with the driver when the driver pulls the signal low. Because there is an impedance mismatch at the intersection of this stub with the rest of the bus, only a fraction of these reflections, which fraction is already a fraction of the initial signal, are transmitted to the receiving nodes. This condition is preferable because the settling time of the receiving node is the crucial parameter to bus performance. The net effect of the absence of reflections from the stub ends can be a reduced overall settling time when compared to systems without termination at each receiving node.

In this configuration, the pull down resistance is Z.sub.O /(n-1) because, by having the pull up resistors on chip, the pull up resistance at the driving node is switched off when the driver is pulling low. Accordingly, this system advantageously consumes less overall current and power than a system that has an off chip pull up resistor that is always drawing current.

Additionally, providing the termination within each component 102 improves signal integrity when compared to having a termination resistor that is not within component 102. When the termination resistor is not within component 102, the terminator is actually separated from the receiving node by some distance along a transmission line and thus parasitics are introduced in the connection to the termination resistor. Also, because of the separation such a system can have reflections from the intersection of the stubs of the transmission line. However, when the termination is within each component, the termination resistance is placed right at the receiver, thereby reducing reflections and ringing.

Referring to FIGS. 3A, 3B, 3C and 3D, the pull up and pull down elements of driver circuit 110 may be of a variety of configurations. For example, as shown in FIG. 3A, the driver element may be a PMOS transistor. Also for example, as shown in FIG. 3B, the driver element may be the parallel combination of a PMOS transistor and an NMOS transistor. With this parallel combination, it is the resistance of the parallel combination that would be equal to the desired bit element resistance. Also for example, as shown in FIG. 3C, the driver element may be an NMOS transistor. Also for example, as shown in FIG. 3D, the driver element may be the parallel combination of two NMOS transistors. In a preferred embodiment, the pull up driver element includes the parallel combination of the PMOS transistor and the NMOS transistor and the pull down element includes the parallel combination of two NMOS transistors. It will be appreciated that a driver circuit may have other circuitry that contributes to the overall pull up and pull down resistance of the driver.

Other Embodiments

Other embodiments are within the following claims.

For example, while a variety of configurations are disclosed for the pull up and pull down driver elements, it will be appreciated that other driver configurations may be used so long as the appropriate driver and termination resistances are maintained.

Also, in the present invention, a MOS transistor may be conceptualized as having a control terminal which controls the flow of current between a first current handling terminal and a second current handling terminal. Although MOS transistors are frequently discussed as having a drain, a gate, and a source, in most such devices the drain is interchangeable with the source. This is because the layout and semiconductor processing of the transistor is symmetrical (which is typically not the case for bipolar transistors). For an N-channel MOS transistor, the current handling terminal normally residing at the higher voltage is customarily called the drain. The current handling terminal normally residing at the lower voltage is customarily called the source. A sufficient voltage on the gate causes a current to therefore flow from the drain to the source. The gate to source voltage referred to in an N-channel MOS device equations merely refers to whichever diffusion (drain or source) has the lower voltage at any given time. For example, the "source" of an N-channel device of a bi-directional CMOS transfer gate depends on which side of the transfer gate is at a lower voltage. To reflect the symmetry of most N channel MOS transistors, the control terminal is the gate, the first current handling terminal may be termed the "drain/source", and the second current handling terminal may be termed the "source/drain". Such a description is equally valid for a P channel MOS transistor, since the polarity between drain and source voltages, and the direction of current flow between drain and source, is not implied by such terminology. Alternatively, one current handling terminal may be arbitrarily deemed the "drain" and the other deemed the "source", with an implicit understanding that the two are not distinct, but interchangeable.