Patent Application
20080308294
None.
Embodiments of the invention relate to electronic wiring and cabling employed to conduct signals from point to point. Such embodiments fall under the category of wired interconnect components.
Twisted wire pairs and cables employing multiple twisted pairs are ubiquitous in the electronics industry. While twisted wire pairs are excellent from the standpoint of reduced EMI and reasonable consistency of impedance along the length of the wire pair, they are prone to other issues such as crosstalk and inter-pair skew. Prior art has attempted to minimize inter-pair skew, the difference in delay between signals transmitted on two adjacent signal pathways, by ensuring that both (or all) twisted wire pairs are constructed identically, with exactly the same twist rate. Whereas this ensures that the total physical length and the corresponding effective electrical length of the twisted wire pairs are the same or nearly the same, a side-effect is a dramatic increase in crosstalk between adjacent twisted wire pairs because of the very uniformity and consistency of twist that lends low skew. Again, prior art has addressed the increased crosstalk by adding shielding jackets around twisted wire pairs. While shielding helps to minimize crosstalk, it substantially impacts the impedance of the twisted wire pair, and leads to the necessity for thicker insulation for the wires of the twisted wire pair. The overall effect is a very substantial increase in the volume and mass of the wire pair per unit length, leading to bulky, physically inflexible and expensive cables.
Prior art twisted wire pair as well as standardized cables such as Cat-5e, Cat-6 (different categories specified by the Electronics/Telecommunications Industry Associations) addresses such concerns of electromagnetic coupling or crosstalk without shielding as well. A wire pair consists of two individual wires coupled strongly and placed close to each other providing a means for ‘differential signaling’, a technique whereby a signal and its complement are transmitted simultaneously and the corresponding symbol recognized as the difference between the two electrical quantities received. Any distant-source noise that couples electro-magnetically into this wire pair couples in very much the same manner into both wires, thereby retaining the difference signal the same, and causing no significant degradation in signal integrity as long as the receiver differential amplifier is capable of rejecting this ‘common-mode’ noise. But a wire pair lying adjacent to another wire pair may not see such a benefit, such as in a flat-tape cable where signal wires as arranged in a bonded fashion adjacent to each other. This problem is effectively addressed by twisting the wires of the wire pair around each other. Over a sufficient length, because of the twist, the coupled noise from any adjacent signal wire sums out to be the same on both individual wires of a twisted wire pair, thus again rendering such noise ‘common-mode’. But this effect does not help when two twisted wire pairs, adjacent to each other in a cable, are twisted identically, leading to physical proximity of wires between the wire pair akin to that which exists when wires are not twisted. To address this issue, prior art standard cables such as Cat-5e also offset the twists of wire pairs with respect to each other, starting with a low twist rate for one wire pair and tightening the twist rate for other included wire pairs in the cable assembly. Because the twist rates are different, the probability of physical proximity between two wire pairs akin to that which may exist in untwisted wire pairs is greatly minimized. The unfortunate consequence of such a design is that there is significant and unavoidable skew between the wire pairs of the cable.
Additionally, twisted wire pairs are also prone to impedance discontinuities that arise due to the physical separation of the wires of the wire pair that may arise due to assembly errors. As the frequency of data transmission through wire pairs increases, these impedance discontinuities become more significant and impact signal integrity. Attempts to correct such problems include very tight twisting as is done in improved cabling solutions in the industry [Ref. 5]. Such designs further increase effective electrical lengths of the twisted wire pairs, potentially increasing inter-pair (between wire pairs) skew and thereby increasing synchronization challenges between signals flowing in wire pairs within a cable assembly. Inter-pair skew is a problem usually addressed by realignment circuits in receiver systems. Values of inter-pair skew in Cat-5e cables resulting from twist offset are typically more than 1 nS per 10 meters of length.
A need therefore exists for improving cable architecture and the simultaneous minimization of inter-pair skew and crosstalk in a cost-effective manner.
As the definition and quality of 2-D and potentially, 3-D images and audio in multimedia transmission increases, there is a need for significantly higher data rates and correspondingly high frequencies of operation of such links as defined in the High Definition Multimedia Interface (HDMI) specification [1]. In order to provide such high-throughput data flow within homes and small establishments, there also exists a need to develop cables that are flexible and therefore easily installed and used without the concerns associated with thick, inflexible cables. Thus wire-pair shielding that leads to a need for thicker wires is not a preferred technology direction for high-performance cables.
The invention implements a method for simultaneous crosstalk and skew minimization through the variation of twist rate along lengths of twisted wire pairs. Wire pairs are twisted with twist rates chosen from a finite set of twist rates, feasible for the wire size, and for finite lengths, chosen from a set of lengths, such that the twist rate or twist rate and length corresponding to the twist rate are changed in a random fashion as the wire pair is twisted. In one embodiment, the algorithm that changes the twist rate and twist length randomly does so in order to ensure that any segment of the twisted wire pair of length to which the algorithm corresponds demonstrates approximately the same signal delay (or effective physical length of each wire of the pair) as any other segment. Additionally, the randomness in the choice of twist rate and twist length is designed to ensure that any segment of such twisted wire pair laid adjacent to any other segment does not share the same twist rate over a significant fraction of the twisted wire pair segment length. Such ‘swizzling’ or mixing of the twist rate and length leads to the crosstalk minimization benefit associated with varied twist rates for adjacent wire pairs while ensuring that both wire pairs have approximately the same physical length or electrical delay. Swizzled twisted wire pairs may also be shielded for additional cross-talk minimization. A cable consisting of multiple swizzled twisted wire pairs may also be shielded in its external jacket that maintains cable structure. Through these enhancements, the invention cable architecture minimizes inter-pair skew while simultaneously minimizing crosstalk without a necessity for shielding.
Prior art unshielded twisted wire pair (TWP) cables are illustrated in
Prior art wire pair twist and cable design introduces a significant disadvantage in the variation of the effective lengths between wire pairs. Twist rates for TWP's are made different to improve crosstalk between TWP's, and this leads to substantial variation in effective lengths of the TWP's. A significant disparity in the effective length of one wire pair with respect to others leads to what is called ‘inter-pair-skew’ that leads to limitations in maximum cable run lengths as well as the need for sophisticated re-alignment integrated circuitry. Prior art attempts at eliminating inter-pair skew include assembling cables out of well-shielded twisted wire pairs that have exactly the same twist rate, or shielded twin-axial or co-axial cable assemblies. Such cable designs are bulky and expensive, and therefore undesirable in conjunction with consumer electronics components such as multi-media devices that see steady erosion in sales prices.
To address the problem of crosstalk and the need for low delay skew simultaneously, the invention proposes varied twist rate or twist rates along the length of TWP's. This goes directly against prior art design and manufacturing practices for TWP's and introduces some challenges in the manufacturing process, which, along with possible solutions, are described in further paragraphs. By varying the twist rate along the length of a TWP, it is possible to both equalize the lengths of all TWP's in a cable while retaining the crosstalk benefit of incongruent twist rates along the length of the cable. As an example embodiment, consider a prior art unshielded twisted pair (UTP) category 5e (Cat-5e) cable length of 2.5 meters. Because of the varied twist rate between the four TWP's in this cable, there exists a deterministic (defined by design/manufacturing variations) delay skew or inter-pair skew between the TWP's. Say the delay numbers for the TWP's are A-tightest twist rate, B-next lower twist rate, C-second-from-the-last twist rate and D-loosest twist rate TWP. Now if an identical Cat-5e cable of 2.5 m were to be spliced to the first, such that the tightest twist rate TWP of the first cable is joined to the loosest twist rate TWP of the second cable, the second tightest or next lower twist rate TWP of the first cable is joined to the second-from-the-last twist rate TWP of the second cable and so on, we obtain a 5 meter cable with delays in the TWP's being (A+D), (B+C), (C+B) and (D+A). It is obvious to one skilled in the art that if the twist rates are designed such that (A+D)=(B+C), all the TWP's of the jointed 5 meter Cat-5e cable will display exactly the same signal delay values, or negligible inter-pair skew. Simultaneously, the crosstalk benefits of varied twist rate along the length of the cable are retained as the same per unit length as in the original 2.5 m Cat-5e cable. In this embodiment of the invention, each TWP sees a single variation in the twist rate at the 2.5 m mark along the length of the 5 m cable, or at the midpoint of any length of this ‘swizzled’ twisted wire pair (SZTP) cable.
Illustrations of the embodiment described above are shown in
With reference to
The junction at 16 in
With reference to
The invention embodiment described above deals with jointed twisted wire pairs of different twist rates. While this derives the benefits of the invention method, it may be undesirable as a manufacturing process due to the presence of a joint in the cable, that while being an additional manufacturing step, may also introduce impedance discontinuities that may impact very high speed data throughput.
An alternate embodiment of the invention employs random variations of the twist rate along the TWP length. Such a design is truly ‘swizzled’, or is an embodiment where the twist rate is ‘agitated’ along the length of the wire pair. A SET of a number of TWIST RATES are chosen, such that the twist rates are sufficient to maintain the two wires in close proximity to each other despite bends in the TWP, and such that any small length of wire pair of one twist rate laid adjacent to a length of wire pair of another twist rate effectively cancels out or substantially minimizes crosstalk. When this alternate embodiment of the swizzled twisted wire pair (SZTP) is made, the twist rate is varied, chosen at random or pseudo-random from the SET of TWIST RATES, such that the twist rate changes within wire pair length that is a small fraction of the desired length of cables assembled from this wire pair roll. For example, if it is desired that cables of length 1 meter be made using multiple segments of SZTP, the twist rate is varied at least every 10 centimeters. The probability of selection of any particular twist rate out of the SET is made the same as the probability for any other twist rate. Because of the random variation of the twist rates along the length of the SZTP, and equal probability for all twist rates, any reasonable cable-length of wire cut out from the SZTP roll will contain approximately the same content of different twist rates, thereby ensuring that the effective physical length of wire of the cable-length will be the same as any other such cable-length cut out from the SZTP roll. Additionally, since the twist rates are chosen at random, the correlation in twist rate between any two cable-lengths of SZTP wire, or the fraction of the cable-length for which the twist rate is identical when these cable-lengths are placed adjacent to each other can be made small. In other words, two cable-lengths of SZTP will not only exhibit almost the same physical length of wire and electrical signal delay, but also behave as if the twist rate is different along most of the cable-length, thus ensuring very low crosstalk in unshielded cable architecture.
One skilled in the art can also appreciate that to ensure a degree of certainty in terms of effective physical length, the sub-length for which a certain twist rate is maintained may be correlated to the twist rate itself. In other words, for a tighter twist, one can control the sub-length to be small, and for a looser twist rate, the sub-length may be made proportionately larger. This will ensure that any choice of twist rate, made at random, will result in a constant increase to the effective physical length of the SZTP. In other words, a SET of SUB-LENGTHS may be mapped one-to-one to the SET of TWIST RATES during manufacture. Alternately, the set of sub-lengths may only contain one value used for all the twist rates.
Through the use of a SET of SUB-LENGTHS, one skilled in the art can also appreciate that another dimension of randomness may be introduced into swizzling. If SUB-LENGTHS are also chosen at random, instead of being chosen in accordance with the chosen twist rate from the SET of TWIST RATES, one may derive a further benefit in the form of a reduction of correlation between any two cable-lengths of wire from the SZTP roll. Whereas this may provide further crosstalk benefit, it may also widen the DELAY SKEW or inter-pair skew distribution. A correlated (mapped) swizzling algorithm that matches a twist rate with a twist sub-length will, on the other hand, display an extremely narrow delay skew distribution, while displaying most of the crosstalk benefits of the invention.
While the architecture and design of machines that implement a swizzled twisting algorithm are beyond the scope of this disclosure and specification, a discussion on the modifications necessary in order to accomplish swizzling is appropriate. As indicated previously in this specification, changing the twist rate dynamically as wires are twisted to form a wire pair can be difficult due to the mechanical momentum and inertia of the machines employed for this purpose. In its simplest form, a machine that twists wires in a double helix may use a spinning disk with tensed wires fed through holes situated at matched radial distances from the axis of rotation, while the twisted wire is rolled up by another spinning mechanism. Such machines are best calibrated to run at a constant rate specifically because of the inertia of the spinning systems. The rate at which the twisting disk and all associated mass rotate may be controlled by a motor that in of itself may be incapable of changing this rate rapidly in order to transition to another twist rate. This change may be facilitated by any number of mechanisms, such as mass that can be repositioned radially (at different distances from the spin axis) through electronically controlled movement away from, or towards the spin axis, or changes in spin speed limiting fluid viscosity that may be controlled electronically, or rapid changes in mass by the expulsion or injection of fluid into the spinning system. There will be a delay in transitioning from one spin rate to another, thus making the change in twist rate continuous. Recalibration and machine modifications may be necessary.
Alternately, swizzling may be accomplished by changing the rate of ‘pull’ on the wire pair as it is rolled after being twisted, which may be controlled primarily by an electronic motor and gear mechanism, and may pose less of a challenge as compared with changing the spin-speed of the twisting disk. Again, machine modifications and recalibration may be necessary. The inventor believes this to be a preferred modification for best implementation of the invention.
Although specific embodiments are illustrated and described herein, any method, process or component arrangement configured to achieve the same purposes and advantages may be substituted in place of the specific embodiments disclosed. This disclosure is intended to cover any and all adaptations or variations of the embodiments of the invention provided herein. All the descriptions provided in the specification have been made in an illustrative sense and should in no manner be interpreted in any restrictive sense. The scope, of various embodiments of the invention whether described or not, includes any other applications in which the structures, concepts and methods of the invention may be applied. The scope of the various embodiments of the invention should therefore be determined with reference to the appended claims, along with the full range of equivalents to which such claims are entitled. Similarly, the abstract of this disclosure, provided in compliance with 37 CFR §1.72(b), is submitted with the understanding that it will not be interpreted to be limiting the scope or meaning of the claims made herein. While various concepts and methods of the invention are grouped together into a single ‘best-mode’ implementation in the detailed description, it should be appreciated that inventive subject matter lies in less than all features of any disclosed embodiment, and as the claims incorporated herein indicate, each claim is to viewed as standing on its own as a preferred embodiment of the invention.
