WO2001038630A1 - Double twist twisting machine - Google Patents

Abstract

A double twist twisting machine imparts a true pretwist to one conductor of a twisted pair and backtwist modulation to another conductor of the twisted pair. The true pretwist is imparted by a uniquely threaded inside-out twister, while the backtwist modulation is imparted by modulating the position of a last pulley over which the other conductor passes before proceeding onto the formation point for the pair.

Description

DOUBLE TWIST TWISTING MACHINE

Cross Reference to Related Applications

This application claims priority to U.S. Provisional Patent Application Serial No. 60/167,547, filed November 24, 1999.

Background

1. Field of the Invention The invention relates to equipment and methods for producing cable with a pretwist applied to a conductor thereof, and to cable produced thereby.

2. Related Art

High speed data cables are required to be manufactured with an extremely stable impedance. Newer drafts of cable standards, for example Draft 11A of EIA/TIA-568, dated May 21, 1999 defined more stringent Return Loss requirements for enhanced CAT-5 cables than previously used. Older standard CAT-5 manufacturing processes are inadequate to construct the enhanced CAT-5 cables.

When standard manufacturing processes are used, Return Loss (RL), Structural Return Loss (SRL) and input impedance stability are influenced by the quality of the insulation, the quality of the twisting process and the quality of the stranding and jacketing processes. The cross section of insulated wire should be round, with a round conductor. Moreover, the conductor must be centered inside of the insulation. Any significant deviation affects Return Loss. The cross section from one point to another longitudinally should also be uniform. Wire having a high standard deviation cross section degrades Return Loss. In the twisting process, tensions, the relationship between capstan speed and flybar speed, the degree of crush between the wires, the position of the forming point relative to the twisting equipment, drag in flybar, and pulley size all can affect Return Loss. It is known that problems with the roundness and centering of the conductor within the insulation, as described above can be corrected by a technique called Pretwisting. Pretwisting consists of separately twisting the individual wires composing a pair prior to twisting them together to form a twisted pair. The result of the Pretwisting operation is that the center to center distance of the copper wires in the pair varies over a length period which affects electrical wavelengths well outside of the desired performance range for the cable. Without a Pretwisting process, the center to center distance of the copper varies somewhat randomly since there is nothing in the insulating and twisting process which can ensure that the insulated wires always enter the formation point of a pair with a constant phase, e.g., with the stripe of a striped insulated conductor always pointing in the same direction.

There are presently two techniques for Pretwisting used in the industry: Tandem Pretwisting and Separate Pretwisting. Tandem Pretwisting occurs in tandem with the pair twisting, for example using SETIC TA + DVD machines. In this technique two twisters pretwist the insulated wires before they are fed to the pair twister. The pretwisters are of the so-called "inside-out" type and the pair twisters are of the so-called "outside-in" type.

These types refer to the direction in which the product is twisted. Since Pretwisting can be done on the same twister as the assembly of the pair, a regular twister can be used to first pretwist each individual wires and then assemble a pair using two previously Pretwisted conductors. Pretwisting can be done by simply loading one conductor into a twisting machine.

A number of twisters currently in use are described in U.S. Patents No. 4,523,423, 4,525,993, 4,550,558 and 4,590,754.

Summary of the Invention

What is desired is therefore an improved technique for obtaining a twisted pair cable construction with improved electrical characteristics. The improved technique may include pretwisting. According to aspects and embodiments of the invention, a double twist twisting apparatus comprises a first wire twisting mechanism imparting a backtwist to a first wire; and a second wire twisting mechanism including a backtwist modulator, imparting a modulated backtwist to a second wire.

Brief Description of the Drawings In the drawings, in which like reference designations indicate like elements:

Fig. 1 is a partial cut-away schematic drawing of a double twister incorporating aspects of the invention;

Fig. 2 is a side view of the central section of the double twister of Fig. 1;

Fig. 3 is a schematic side view detailing the backtwist modulator; Fig. 4 is a schematic side view detailing operation of the backtwist modulator of Fig.

3;

Fig. 5 is a further detail of one backtwist modulator;

Fig. 6 is a detail showing the modulator in an extreme position; Fig. 7 is a detail of a one-pulley, space saving backtwist modulator;

Fig. 8 is a backtwist modulator installed on a tandem twister type of inside out twister machine; and

Fig. 9 is a modified backtwist modulator of another embodiment of the invention. Detailed Description

Aspects and embodiments of the present invention are now described in connection with a number of embodiments and examples thereof.

A twisting machine embodying aspects of the invention is shown in Fig. 1. The illustrative embodiments of the invention are constructed by modifying conventional double-twist twisting machines. The wire feed through a second flybar is pretwisted, in reverse direction producing a true backtwist, at a double twist rate.

Horizontal twisters can be perfectly symmetrical at both ends. That is, the yokes containing the pulleys and the flybar attachments at each end could be similar.

In a vertical twister the upper and lower yokes are quite different. The upper yoke contains the planetary gear system which keeps the cradle from turning. Thus, one reel is outside on its own give-up. The insulated wire is fed from one give-up into the twister. It will go through the opposite end of the twister from the end at which the twisted pair exits. It will go through the inside of the shaft used for the rotation of the lower yoke and the flybar. This shaft may be solid in the conventional machine, but it may be readily modified by drilling a V" hole through the axis thereof (item 1 in Figure 1). Since conventional vertical twisters are often raised on rubber pads for vibration cancellation, there is sufficient clearance to place a pulley underneath, in order to guide the wire into the bottom shaft. The first twist will happen between this added pulley and pulley no. 2.

The insulated wire is fed through the "balancing flybar" by using a different set of lower yoke pulleys (items 2 and 3) on the opposite side of the lower yoke pulley used for the formation of the pair. It is then fed into a flybar which contains guides (item 5). A guide mechanism guides the wire from pulley no. 3 to this flybar.

All twisters have two flybars. One is there only for balancing reasons. However, by using each flybar for guiding wire through the twister, the mechanics of the system are more efficiently and effectively used.

For the last century of twisting, twisters seem to have been used at only half of their potential.

According to the illustrative embodiment, he wire goes into the upper yoke, over a different pulley (item 6) than that used for the second twist of the pair. It then goes through another shaft (item 7 which is easily drilled at ^lΛ"). This is the axis of rotation between the upper yoke and the cradle (inside frame which does not turn). To get the wire out of shaft 7, a hole has to be made into the frame and a cut must be made in the frame in order for pulley 8 to direct the wire out of the center of the frame. The second twist happens between pulleys 6 and 8. Pulley 9 guides the wire downward. Should tension control not be critical, the now fully pretwisted wire could go directly to the formation point (13) of the pair.

Should tension control be important, we propose items 10 to 12.

Our embodiment of this idea uses a motorized capstan (item 10) and dancer (12) in order to control the tension on this pretwisted wire and make it independent of the tension added through the pretwist flybar which is product, linear speed and flybar speed dependent. This is very easy to do since there is a lot of space inside the twister because a reel has been removed to the outside of the bow envelope that had been previously located inside the bow envelope. The other conductor will come from inside the twister, in the conventional manner.

A newly developed motorized tension control system could alternatively be used. This tension control will be set so that the tension on both conductors will be perfectly matched at twisting.

According to this embodiment, the wire is effectively Backtwisted, as opposed to pretwisted, so it will become neutral inside the twister, i.e. having no twist. The actual backtwist ratio will be approximately 0.98 of the helix factor of the pair since the speed of the wire is slightly higher than that of the pair.

The modification of Figure 1 only improves one conductor. The other one remains without pretwist. A pretwist modulator improves the other conductor. This add-on to a current twister will "mix up" the eccentricities without using conventional twisting devices. It simply modulates the "backtwist effect" by using the fact that the rotational movement at the formation point backs up in the insulated wires. It is very innovative, small and compact and low priced. See Figures 3 to 9. The purpose of a backtwist modulator is to prepare individual conductors of a twisted pair in such a manner that its eccentricities will not have an impact on the impedance characteristics of the pair. This is done during the twisting process immediately before the formation of the pair. Adding the modulator to a pretwister such as described above yields a good quality input impedance, SRL and RL. Pretwisting techniques are also well known to improve the quality of the impedance/SRL/RL of twisted pairs and therefore the cable which contains them, as described in U.S. Provisional Patent Application Serial No. 60/110,739 and U.S. Patent No. 5,767,441.

In embodiments of the invention, each individual modulated wire is given differential twist frequencies without the use of supplemental twisting machines. This results in savings in terms of equipment purchase value, equipment maintenance, floorspace use, overall product output rate and therefore lowered product manufacturing costs. The twisting process hardens the copper slightly. That is, it lowers the elongation.

The twisting action that happens to each individual wire as it is twisted into the pair tend to spread backwards from the pair towards the give-up of each insulated wire. This phenomenon has been observed since long before. In the case of a twister, this was called a "backtwist" and in the case of a strander, called a buncher in the U.S., this was called a "backstrand".

On a strander this phenomenon has the consequence that before the formation point of the strand the measured lays on the pairs differ from those that can be measured after twisting. On a twister, if the insulation has a stripe we can actually see that at one point between the insulation reel in the give-up and the formation point that the wire is twisted with a lay which is close to LAY = V/N (V is the speed of the line in inches per minutes and N is the bow RPM).

We have found that the backtwist phenomenon goes backward until the wire goes over a pulley with a contact angle superior to 90 degrees. Backwards from this pulley the propagation of the backtwist (the twisting motion coming from the twister) is halted. Should the wire come in contact with a pulley for less than 90 degrees the backtwist motion only partly abates.

There is usually at least one pulley in a give-up system with a contact angle of more than 90 degrees. To simplify the current explanation, assume that this pulley is the last pulley before the formation point. This is the point to which the backtwist phenomenon propagates. This pulley is called the modulating pulley. This pulley acts as a "Virtual Twisting process." On one side, the wire approaches the pulley and is not twisted; on the other side, the wire leaves the pulley and is twisted.

There are two mechanical phenomenon happening to the wire explaining the phenomenon of backtwist. First is the elastic nature of metals which allows the twisting motion of the twister to be spread out over the "unblocked" length of the insulation. Second and more complex is the fact hardened twisted copper acts as a "drive shaft" to the twisting motion happening to the individual wires at the formation point of the twister. This last phenomenon tends to encourage the less twisted part of the insulation, usually the last feed after the modulating pulley, to twist first, like the weakest link of a chain. As pointed out by P. Pepin, this could all be explained by the natural tendency for forces and moments to spread themselves out over the modulated length.

Two important parameters of twisting processes are next described. Device rotation speed , meaning rotation of the twister itself, including the twister flybars, and line speed. As described above, the Backtwist lay in the wire approaching the formation point has basically the same lay as the first twist of the twister and follows the rule Lay = V/N.

Backtwist lay can be varied by varying the position of the modulating pulley in time. As will be explained below, this has the effect of modulating the backtwist lay. The backtwist lay on the insulated wire being different of the twist lay of the pair, we are now in a situation of differential twist lays. This results therefore in a situation where the eccentricities of the individual wires do not turn at the same rate as the pair, thus yielding a similar result to that given by traditional pretwisting techniques.

As mentioned, the backtwist lay follows the V/N law. By changing the position of the modulating pulley, we are changing the apparent "V" parameter of the process (line speed) at the exit of the modulating pulley thus a change in the backtwist lay. The average twist lay of the insulated wire, when measured over great distances, will remain the same as that of the twist lay of the twisted pair (this is not the case for traditionally pretwisted wires). Fig. 3 shows that the modulating pulley can be moved in such a way that the geometry of the formation process is never changed. The individual twisted wires keep on arriving at this formation point at the same speed as if there was no modulation going on and the twister keeps on twisting at the same rate (N).

There could be a very slight variation in lay caused by a slight non-constant elastic deformation of the copper. It is well known that formation point stability, which is related to twist lay stability, is not just affected by geometry and tensions but also by the homogeneity of the product being twisted. Because of the non-linear characteristics of the deformation of copper, the remaining elastic energy could be different from when it is more twisted or less twisted. Nevertheless, our first trials showed to us that when both modulators were operated in synchronization, i.e., in the same phase, both at the same ends at the same time, that no twist lay variations where observed. This is a unique feature of the backtwist modulator which could not be achieved if the single wires were oscillated in a previous part of the process.

The mathematical theory behind the above-described embodiments is now explored.

BacktwistJLay = V/N where N is rotational speed of flybar and V is the speed difference = (Wire speed) MINUS (Modulating pulley speed in the direction of the wire).

The only virtual twisting is assumed to be happening at the exit of the modulating pulley. Reality is much more complex. The "Drive shaft" explanation is not exact because this "drive shaft" is driven by an elastic and plastic metal phenomena and will twist more at its weakest point which is the part of the wire which is less twisted. In the case where the modulator starts approaching the formation point after going away from it, the tightening of the lay might very well start happening at the formation point instead of at the modulating pulley. More results can be seen from a simulator which was built for this purpose.

Let VL be the line speed, N be the flybar speed, Lay_tp be the twisted pair lay at the first twist (Lay_tp = VL/N) and is also the lay of the backtwist on the wires. Let L be the length of the stroke of the modulator. Let VI be the speed at which the modulator approaches the formation point and V2 be the speed at which the modulator leaves the modulator (V2 must always be negative, VI positive). For what follows, we will consider that the speed profile is constant when going in one direction. Based on the hypothesis that all the twisting is done at the exit point of the backtwist modulator we find that we have two different backtwist lays which will be obtained depending on the direction and these will give us two different backtwist ratios depending on the direction of the modulator.

VL - V Lay BT, = '-F_BT_ = \llay _BT,i = 1,2 (eq. 1)

where:

VI is the speed of the modulator when it is going towards the Formation Point (F.P.); V2 is the speed of the modulator when it is going away from the Formation Point

(F.P.); VI and V2 are referenced positively when going towards formation point so normally the numerical value of V2 should be negative;

VL is the line speed in inches per minute;

N is the flybar speed in inches per minute; N is the flybar speed in RPMs;

Lay_TP is the lay of the twisted pair being produced;

Lay_BTj is the lay that can be measured on the single insulated wire at its exit from the modulator (this is the real backtwist lay);

F_BTj is the multiplicative inverse of the above, the backtwist frequency; BTR is the effective equivalent "Backtwist Ratio". In the case of modulation, a t.p. lay of 0.5" which has a BTR of 10% on one of its conductors would yield a lay of approximately 4.5" or 5.5" which is the same as if on a separate pretwisting stage we would put a lay of 5" and then twist the pair with a lay of 0.5";

Vη Relative speed of the wire as seen from the modulator; Tj Time for the modulator to travel in direction I; and

Lj is the total length single wire which exits the modulating pulley as it travels in direction I.

Note that Lay BT; is not like a traditional Pretwist lay, it is the actual lay on the single insulated wire entering at the formation point of the twister.

F_BT, - F_tp _ V,

BTR = ^■i = l,2 (eq. 2) F _tp ^~ VL - V.

BT_ratio is defined here as being positive when the modulator goes towards the formation point. This is an overtwist situation on the wire and corresponds to "Forward twist" in the case of traditional pretwisting. We conclude that when the sign of BT ratio is negative that it corresponds to true "Backtwist" in the traditional pretwisting sense.

Equation 2 states that for this modulating process the obtained equivalent Pretwisting ratio (to compare this to regular traditional pretwisting) is simply the ratio of the speed of the modulator to that of the line speed as seen from the modulator. This interesting development shows that to obtain Pretwist ratios of the order of 20-30% (what we are accustomed to), the modulator needs only to move at a relative speed of 15%-45% of that of the pair. For most applications this would give a linear speed below 1 ft/sec. Unfortunately, eq 2 is non-linear for parameter Vj stating that in order to obtain a constant BTR the backward speed has to be bigger than the forward speed (towards forming point).

In the two following case studies, the theory of the backtwist modulator will be shown through two case studies.

CASE STUDY 1: Infinite L (infinite stroke of modulator)

The first example is the case where L is very long, say infinite. In this case, the modulator will always move in the same direction and the resulting backtwist lay on the insulation before the formation point will be Lay_BT = (VL-V1)/N. Numerical example N- 1800RPM, Lay_tp(first twist) = 1.0", VL=1800 inch/min VI = 6"/sec*60sec = 360"/min

Lay_BT = (1800-360)/1800 = 0.8" Pretwist ratio = (l-l/0.8)/l 25%.

The case of an infinite Length of modulator stroke described above will give the same effect as using a conventional Pretwist Process with pretwisters set at 25% BTR.

The purpose of this first example is to show that at the limit, Backtwist modulation becomes traditional Pretwist. It will be simple to understand that a twisted pair constructed with modulated singles becomes a transmission line that is a series of 2 types of sections of pairs. The two types of sections differ in only one aspect: the effective pretwist ratio is different (corresponding to the forward motion (VI) and backwards motion (V2) of the modulator). Both of these sections could be produced by current Pretwist techniques that resemble that of two different pretwist processes.

Lack of longitudinal homogeneity can become a problem on a transmission line. If this lack of homogeneity is cyclical it can affect RL, the parameter that we are trying to correct! It can be argued that there should not be a very big Impedance between a section of pair with a BTR of -20% and +20%. There is no "electrical explanation" which can be found to explain this. Mechanically speaking there are slight stress differences between both sections, but again, these are small.

Should there be any problem on RL created by the cyclical nature of modulation the two following techniques could improve the situation: making the length in one direction different than the length in the other direction. Slowly modulating the speed (VI and V2) of the modulator to ensure that every different subsection is of slightly different length.

CASE STUDY 2: Symmetrical Sections

As described above, this is not necessarily recommended to avoid peaks on RL but completes the theoretical framework of the theoretical part. To have the same absolute value of pretwist ratio in both directions, a first criteria must be that the desired BTR must be below 100% and realistically below 50%. We can use the following formulas:

BTR I *VL v_ =- 1+ I BTR I

(eq. 3a,b and c)

I BTR I *VL

F, = - 1- 1 BTR I

therefore the relationship in this symmetrical case of BTR would be:

_{v =} ^V^ ^VL

2V -VL

The travel times are different depending on the direction of the modulator.

1(1+ I BTR I)

Ά = - BTR I *VL

(eq. 4a,b) L(\- 1 BTR I)

7 =

² ' BTR I *VL

where L is the length of the modulator, Tl is the travel time in the direction of the formation point...Nevertheless, the length of each different section of direction of modulation once in the pair is the same. It has to be logically since the modulator does not add any twists and the overall average of twists on the conductor must match that of the pair so that in the case of +-BT symmetrical Pretwist ratios, the lengths of both sections must be matched. The calculation of the length of the modulated wire must be done by calculating how much wire goes through the modulator in each mode. This is because according to our simple model the twisting is being done on the insulated wire immediately at exit of the modulator. We must therefore calculate the speed at which the wire is leaving the modulator: VL

Vr. = VL - V_λ =

1+ I BTR I

(eq. 4a,b)

VL

Vr₂ = VL - V₂ = 1- I BTR

The total length of wire and therefore of twisted pair corresponding to each section is simply

^L- -^Vr→ *^τ- ⁼ n I →o^IK I ta- ⁵⁾

Eq. 5 has a very nice elegant result. It shows that each length of pair corresponding to each section (different direction of modulation) is equal to the Length of the modulator divided by the BT ratio of that section. This shows that since the typical ratios that we use are low that the length of each section will be long. The length of a cycle in the above case will be 2*L/BT. This formula encourages the use of low values of BT. For our TTs the typical value of L that we will use will be about 20" and by using a BT value of 0.10 (10%) we can expect a total cycle length of 400" . As far as RL is concerned 5 meters is a VERY long length. When we design cables and the processes, we never worry about cycles longer than 10 meters. Nevertheless, by slightly modulating the process through time, we will never have a couple of consecutive cycles of the same length so there already should be no issue for RL.

The present invention has now been described in connection with specific embodiments thereof. However, numerous modifications which are contemplated as falling within the scope of the present invention should now be apparent to those skilled in the art. Therefore, it is intended that the scope of the present invention be limited only by the properly construed scope of the claims appended hereto. What is claimed is:

Claims

Claims

1. A double twist twisting apparatus comprising: a first wire twisting mechanism imparting a backtwist to a first wire; and a second wire twisting mechanism including a backtwist modulator, imparting a modulated backtwist to a second wire.