Wave transmission lines and networks – Plural channel systems – Having branched circuits
Reexamination Certificate
1998-07-31
2002-06-11
Metjahic, Safet (Department: 2858)
Wave transmission lines and networks
Plural channel systems
Having branched circuits
C333S02200F, C324S646000
Reexamination Certificate
active
06404299
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates generally to termination of reflected waves or otherwise propagated waves that are unwanted in a network, and in particular to use of partial termination by resistance located at the network hub for a star network.
BACKGROUND
The simplest network consists of only two sites or nodes. In the case of a two node network, each end of a single transmission line is terminated with a characteristic impedance, which results in no reflections of signals sent in either direction.
This ideal situation is compromised as soon as a third site is added to the network. In the situation of three or more nodes, the intermediate site must not load the transmission line or reflections of transmissions may occur. Thus, for n sites, each of the n sites must be prevented from loading the line, and use of proper termination is necessary to eliminate reflections and other unwanted wave propagation within the network.
Preventing loading and eliminating reflections can be accomplished by using a network configuration in which the transmission line starts at one node and loops through all intermediate sites, terminating at another end, such that the network constitutes essentially a chain of nodes. Such a network is often referred to as a “ring” network or a “token ring” network. The ring network, however, is not suitable for many applications, such as wiring in homes and houses and offices, which must connect to a central location.
In contrast to the ring network, which is ideal with regard to loading and reflections, typical networks include wire connections to all sites via a central point, referred to as a “hub”. This type of network that has a common hub is referred to as a “star” network. The star configuration of nodes simplifies both the wiring of the network and the process of adding or removing a site, but at a great loss in high frequency performance. If the transmission lines are wired in parallel to the hub, from the point of view of any of the transmission lines, the hub appears as a discontinuity of Z
0
/(n
1
) ohms (where Z
0
is a positive resistance value and n is the number of nodes), which for large networks can result in almost all the energy being reflected from the hub back to the sending station.
Commonly used integrated circuits for signaling, such as those falling under the broad classifications RS-422 and RS-485, for use with networks, are useful with twisted wire connections for a large number of sites. These integrated circuits typically require that no terminating resistance be placed at each site. As a result of this approach, the energy reflected from the hub back to the sending site is almost totally reflected, yet again, from the sending site back to the hub. For a large network having random lengths of connecting wire from the hub to each site, the ensuing cacophony of reflections increases the statistical possibility that the reflections will combine to create a brief reversal of signal polarity and hence a false transition at some site.
Further, in a typical network, transmission lines typically include two wires close together, such as two wires side-by-side, two wires arranged in a twisted pair, or concentric wires (e.g., a coaxial cable for a television). This wiring arrangement therefore constitutes conductors at a fixed distance apart for some considerable length of the conductor, producing a fixed capacitance and inductance per unit length, and thus a corresponding impedance. A characteristic of transmission lines so arranged is that, unlike single wires in a circuit, the wire pairs cannot carry an independent amount of voltage and current: in a single wire circuit, an applied voltage on the wire produces a given current running through the wire, and the voltage and the current typically are totally unrelated (i.e., do not affect each other); however, in a transmission line, the voltage and current are fixed relative to each other in a specific ratio.
Thus, all transmission lines have an impedance associated with them, which is the ratio of the voltage applied to the current that flows in the wire. For example, a transmission line having one volt applied may carry 20 milliamps. The impedance of the cable is the voltage divided by the current. In this example, 1 volt divided by 0.02 amps equals 50 ohms, which is the impedance of the cable.
Fifty ohm cables are commonly used with radios; televisions typically use 75 ohm cables; for twisted pairs such as those used for a 10 Base T network for a computer, the impedance is generally about 100 ohms, but typical impedance is not so precisely defined in this use as for radios and televisions. Category 5 wiring is generally used in businesses for computer networks. This type of wiring includes pairs of wires in which impedance is carefully controlled and repeatable so that each pair of wires has the same impedance.
With a simple network having two nodes connected by a transmission line, in order to minimize impedance and reflection, the transmission line may be terminated such that the resistance equals the impedance of the line. When a wave transmitted from one node reaches the far end of the wire (the receiving end), all of the energy of the wave is received and dissipated in that resistor and, importantly, none of it is reflected. If the resistor at the far end of the line does not match the characteristic impedance of the line, a reflection is produced. The reflection is positive if the resistor value exceeds the impedance of the line and negative if the impedance of the line exceeds the resistance of the resistor.
Thus, matching the resistor to the impedance is simple for a network consisting of two nodes. For example, for a 50 ohm impedance transmission line coaxial cable, a 50 ohm resistor is used at each end of the network; or the transmission line is a twisted pair, 100 ohms is used at each end. Whichever type of transmission line is used, signals travel in both direction between the nodes, reach the respective resistors and are fully dampened.
The appropriate amount of resistance to apply when more than two nodes is involved, however, is much more complicated that the situation with two nodes.
In the “star” network, the central point is referred to as the “hub” of the network or the “network hub.” A problem with the star network is that, once a third site is added, determining the appropriate resistance to place at each node in order to prevent reflection becomes complicated. For example, with a star network having three nodes “a,” “b,” and “c,” and a central hub, the three sites radiate out from a center point, along three lines. If a signal is sent from “a” to “b” and “c”, the signal first proceeds down a spoke to the central hub. At the point of reaching the central hub, the signal must proceed out two spokes to nodes “b” and “c.” Thus, there is now not a single transmission line of 100 ohms, but two transmission lines. As a result, at the central point, the wave encounters half the impedance of the first spoke: in the example above, the spokes to “b” and “c” in parallel produce an impedance for the wave coming down leg “a,” of 50 ohms, rather than 100 ohms. The twisted pairs of wires are around 100 ohms impedance each, but two of them wired together in parallel have an impedance of 50 ohms each (R=1/(1/R
1
+1/R2)=R
1
R
2
/(R
1
+R
2
)=R
1
/2, since R
1
=R
2
100 ohms). Similarly, for four spokes, three spokes to which a wave is transmitted produces each have 33 ohms of impedance; with five spokes, each has 25 ohms impedance. This situation becomes even more complicated to the extent that any differences in impedance exist with respect to the different lines.
As a result of the arrangement of the star network, equivalent impedance of the transmission lines tends toward zero as the number of nodes on the network becomes very large. A problem thus exists in the prior art with overcoming the situation in which a signal is sent from “a” to all the other sites, while preventing, once the signal reaches the central hub, due t
Kerveros J
Metjahic Safet
Russound/FMP, Inc.
Weingarten Schurgin, Gagnebin & Lebovici LLP
LandOfFree
Devices, methods, and systems for network hub partial... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Devices, methods, and systems for network hub partial..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Devices, methods, and systems for network hub partial... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2929941