This invention relates to a device as in claim 1 , to a yarn feeder as in claims 35 and 36, and to a method as in claim 37, and is destined to be used to control yarn tension when a running yarn is subjected to stresses along its path that vary extremely rapidly.

Controlling yarn tension is very important in order to achieve maximum effectiveness in many weaving operations, such as insertion of the weft yarn in a high- performance loom, like the modern shuttleless looms, or feeding of a group of yarns to a warping machine, or still other operations. In all these activities, the way in which the yarn is drawn off is such that yarn tension may vary considerably. A particular case is insertion of weft yarn in a shuttleless loom, where the yarn is fed in at extremely high and extremely variable speed, causing tension peaks in the yarn that may even result in the yarn breaking.

Mechanical devices are known which control yarn braking in relation to yarn tension, so as to maintain a substantially constant tension in the yarn, far from very high values and exempt from sharp variations. These are equipped with brakes that react mechanically to the tension existing at any moment in the yarn, so that the braking force exerted by the brake on the yarn decreases as the tension increases and, conversely. Braking is mechanically modulated in relation to the yarn tension value. Mechanical devices generally have a slow response time in detecting variations of tension occurring at high frequency. These drawbacks derive from the fact that the yarn tension is detected to modulate braking by using mechanical parts engaging with the yarn and that are prone to move, during the yarn tension variations. Though of very small mass, these mechanical parts nevertheless have an inertia and an

elasticity that are such that, they interact dynamically with the yarn, exchanging kinetic and elastic energy with it, especially in cases of a sudden variation in yarn tension. These dynamic variations slow down the response and diminish the efficaciousness of these devices to control the brake and, as a result, the yarn tension. Moreover, the dynamic interaction of mechanical parts and yarn has an unsettling effect on existing yarn status, and may thus considerably modify yarn tension, especially when the tension varies extremely rapidly and the dynamic interactions between these mechanical parts and the yarn occur continuously. These known devices are, in addition to their slow speed of reaction to tension variations, never completely passive with respect to the yarn in detecting its tension and always alter its value to a certain extent, thereby reducing efficaciousness of braking modulation. Furthermore, these known devices do not measure the absolute or numeric value of yarn tension but are limited to detecting relative variations of tension in order to control yarn braking.

As known devices do not possess a feature whereby the effective tension value is visualised this prevents them from being set accurately and permits only an empirical regulation (generally performed 'by feel' by an operator).

US 5,316,051 shows a truncated cone-shaped cap fitted on the front end of the drum of a yarn feeder to press directly on the yarn coming from the reserve and passing between cap and drum. This brake exerts a braking force on the yarn that is responsive to the tension in the yarn when exiting from the brake itself. In fact, this exiting yarn tension gives rise to a tension component oriented axially and apt to act retroactively on the cap for being subtracted from the braking force exerted by the cap on the yarn. The axial tension component causes a reduction of the braking force. Therefore, if yarn tension tends to rise, the brake reduces its braking force proportionally. This brake reduces yarn tension peaks and smoothes the tension pattern, but cannot completely cease from braking on the yarn, i.e. the yarn can never be disengaged from the brake completely, to reduce yarn tension as much as

possible. This brake cannot nullify entirely the tension at the brake outlet. The yarn always remains engaged with and thus braked by the cap.

Devices are known to control yarn tension that are used on a yarn feeder, and comprise a yarn tension gauge, attached to the eyelet coaxial with the feeder drum. Tension measurement is obtained by measuring the variation of parameters due to axial displacement of the eyelet, because of the force dependent on yarn tension. The eyelet is thus left free to move in the axial direction under the action of this force. These devices have the tendency to exchange kinetic and/or elastic energy with the yarn, thus leading to the risk of the eyelet starting to vibrate. These inconveniences mean that tension measurement is generally affected by error and is thoroughly unreliable when yarn is drawn off from the drum at extremely high and rapidly varying speed.

One task of this invention is to provide a yarn tension control device having a very high response speed and high precision in measuring yarn tension. This is achieved by the device recited in claim 1.

A further task of the present invention is that of providing a yarn tension control device that visualises the effective tension values, so that the latter may be of use to an operator for convenient setting of the desired tension threshold values and also make it possible to subsequently control deviation in time of the tension from the desired values, so that action may be taken if the deviations are found to be anomalous. This is achieved by a yarn tension control device stated in claim 11.

Another task of this invention is to provide a yarn tension control device which, as well as measuring yarn tension, also has a very high response speed and high precision in varying braking force on the yarn, so as to intervene in effective and timely manner to correct yarn tension and maintain it substantially within a programmed pattem, especially when the yarn is subjected to stresses varying extremely rapidly while it proceeds along its path and which is also able, as required,

to disengage from the yarn and thus completely annul braking on the latter. This is achieved by a device defined in claim 12.

According to another aspect of the invention, the movable guide element engaging the yarn is associated with elastic yarn retracting means, which can be regulated manually and which are arranged between the movable guide element and the yarn tension sensor means in order to retract the yarn and keep it constantly under tension, preventing it from becoming slack.

According to another aspect of the invention, the motor of the yarn tension sensor means may be used for positive driving of the guide element so as to retract and tension the yarn in cases where it tends to become slack as, for example, at the end of a cycle of weft yarn insertion in a shuttleless loom.

The device of the invention is applied to particular advantage on shuttleless looms thanks to its capacity to control weft yarn braking during the insertion stage in function of the tension effectively acting in the yarn and not, as is the case in known systems, in a way dependent on or synchronised with the loom cycle. The inventive concept permits to control yarn tension at the outlet of a yarn feeder, feeding yarn at high speed into a weaving machine. The yarn passes through an eyelet at high speed and exerts a force on the eyelet that is dependent on the tension of the yarn.

In combination with a yarn feeder, a further embodiment is disclosed in which the device of the invention to control yarn tension, is used, e.g. as an embodiment as in claim 21.

The device and method offer considerable advantages, such as that of not altering the path normally followed by the yarn in the yam feeder as it unwinds and that of being able to obtain a yarn tension measurement from a tension component apt to assume a significant value, i.e. similar to that of actual tension of the yarn. In fact, the device of the invention utilises to advantage the deviation that the yarn

undergoes in passing through the eyelet coaxial with the drum on its way to the weaving machine. This arrangement of the eyelet with respect to the drum is such as to produce a very high angle of deviation of the yarn, in correspondence with the eyelet, as the yarn reaches the eyelet along a path that is radial with respect to the drum, coming from its outer cylindrical surface, whereon the yarn reserve is located. The effect of said high deviation is that the force the yarn exerts on the eyelet and which is proportional to both the tension to be measured and to the deviation undergone by the yarn, assumes a very significant value, similar to that of the tension and apt to permit a precise and reliable measure of the said tension.

In the drawings, are:

Fig. 1 a perspective view of a device for controlling yarn tension;

Fig. 2 an electrical block diagram of the device of Fig. 1 ;

Fig. 3 a perspective view of a variant of the device of Fig. 1 ;

Fig. 4 a top view, partially sectioned, of the variant of Fig. 3;

Fig. 5 a cross-section taken along the line V-V of Fig. 1 or in the variant of Fig. 3;

Fig. 6 an enlarged scale view of a detail of Fig. 1 ;

Fig. 7 a variant of the brake of Fig. 5;

Fig. 8 a graph showing the time course of operating parameters of the device;

Fig. 9 a perspective view of the device of Fig. 1 equipped with mechanical coupling means;

Fig. 10 a variant of some details of Fig. 9;

Fig. 11 a diagrammatic representation of a first form of a damper useful on the device of Fig. 1 and Fig. 3;

Fig. 12 another form of construction of the damper of Fig. 11 ;

Fig. 13 a variant of the block diagram of Fig. 2;

Fig- 14 a lateral and partial cross-section of a variant of the device of Fig. 1 ,(yam retracting means); Fig. 15 a front view of the variant of Fig. 14;

Fig. 16 a flow diagram of one mode of operation of the device of Fig. 1 , to retract yarn;

Fig. 17 a longitudinal section of the device associated with a yarn feeder;

Fig. 18 a first variant of the device of Fig. 17;

Fig. 19 a second variant of the device of Fig. 17;

Fig. 20 a third variant of the device of Fig. 17;

Fig. 21 a front view in enlarged scale of some details of the variant of Fig. 20;

Fig. 22 an enlarged scale view of a different embodiment of the variant of Fig. 20;

Fig. 23 a variant of the device of Fig. 17 with the addition of a brake;

Fig. 24 a different construction of the variant with brake of Fig. 23;

Fig. 25 an electrical block diagram of the devices and variants of Figs. 17-20 and Fig. 23-24;

Fig. 26 a further variant of the device of Fig. 17, with a cap brake;

Fig. 27 a time pattern of parameters applied to control yarn tension during insertion in a loom;

Fig. 28 an electric diagram representing some parts of the electrical block diagrams of Fig. 2, Fig, 13 and Fig. 25 in greater detail; and

Fig. 29 a variant of the electric diagram of Fig. 28.

In Fig. 1 , a device 30, acts on a yarn 31 , which moves along a path in a direction indicated by arrow 32. The yam is guided by eyelets 33 mounted on a fixed structure 34 of the device 30. Yarn tension sensor means 36 suitable for measuring the yarn tension, also called "tension gauge" hereinafter, are provided to act, by means of a guide element engaging the yarn 31 , on a portion of the yarn. The tension sensor means 36 comprise electronic control means having a movable control element driven by an electromagnetic force.

The electromagnetic control means and the relative movable control element can be constituted by a direct current motor 37 attached to the structure 34 and its rotor 40. The latter is apt to rotate about a casing of the motor 37 and has a shaft 38 protruding from the casing. The movable guide element consists, for example, of a

rod 41 , attached to the shaft 38 and apt to deviate the yarn portion, upon which the yarn tension sensor means 36 act, from a rectilinear trajectory between the two eyelets 33.

A generic deviation of an angle α (Figs. 3 and 5) produced by the rod 41 on the yarn with respect to the rectilinear trajectory, automatically results in application of a force F by the yarn 31 on the rod 41. The value of F, assuming negligible friction between yarn 31 and rod 41 , is equal to 2Tsin(α), where T is the tension in the yarn. The tension gauge 36 also comprises a disk 42 integral with the rotor 40 of the motor 37, and which is equipped with a slot 43 (Fig. 6). In the application, the latter is aligned with a position sensor 44 and, under these conditions of alignment, establishes a determined measuring position (hereinafter also called reference position) of rotor 40 and rod 41. In this measuring position of rod 41 , there is a given angular deviation α of the yarn 31. To signal alignment of the slot 43 with the sensor 44, the latter is equipped with optical elements 46 apt to detect each displacement, of the slot 43 from the alignment condition, and hence of the rod 41 from the measuring position.

Other types of sensors, for example, Hall effect or inductive sensors may be used to signal displacements of the rod 41 from the measuring position. This position may correspond to a very low angle of deviation α of the yarn, of only a few degrees, so that the additional tension induced in the yarn 31 by its sliding on rod 41 is practically negligible.

In Fig. 2 blocks of the diagram refer to elements already described in the foregoing and are therefore indicated with the same numbers. A control circuit 51 is apt to receive from the sensor 44 a signal S indicative of displacements of the disk 42 with respect to the position where it is aligned with sensor 44 and hence also of the displacements of rod 41 , with respect to its measuring position. On the basis of this signal S, the circuit 51 powers the motor 37 with a current I of intensity and direction such as to maintain disk 42 substantially motionless in the alignment position,

instantaneously balancing the resistive torque that the force F applies on shaft 38 through rod 41 , whatever the value of force F and hence of the tension T acting on the yam 31 may be (F = 2Tsin(α) ).

Oscillations ought to be avoided from being generated in the current I whenever it varies rapidly in value. The oscillations can normally be attributed to secondary oscillation phenomena and could, if highly marked, give rise to instantaneous yarn tension measuring errors. On this subject, a particularly advantageous form of construction of the control circuit 51 , and to which reference will be made by describing the variant of Fig. 13, is that of using the digital control technique.

As rod 41 , rotor 40 of motor 37 and disk 42 remain practically motionless during variations of tension, this prevents their inertia and elasticity from having a negative, unsettling impact on detection of tension T. Rod 41 , rotor 40 and disk 42 behave as if they were virtually still and free of inertia and elasticity as regards detecting the tension T of the yarn 41. The control circuit 51 can drive motor 37 in a PWD-mode (Power With Modulation) to supply power to the motor 37 in pulsing mode and cause I to vary analogically. When the yarn 31 is not subjected to any tension, rod 41 and rotor 40 are not under stress either, and hence the current I assumes a null or extremely low value that corresponds to a minimum current threshold for controlling position of rotor 40, when there is no torque acting thereon. When there is a tension T in the yarn, the rod 41 is subjected to a force F proportional to the tension and applies a resistive torque on rotor 40 of motor 37. The resistive torque is equal to the product of F by its arm with respect to the axis of motor 37, and is balanced by an active torque generated by the current I. As the active torque generated by a direct current motor is proportional to the current with which the motor is supplied, the force F and hence the tension T in the yarn 31 are proportional to the current I.

Current I and tension T are proportional for all values of the angle α of deviation of the yarn 31. Assuming T remains constant, current I tends to drop as decreases, since F also decreases correspondingly. The device is apt to supply a current I

indicative of the tension T, even for very low values of angle α, under 10°, so that sliding of the yarn 31 on the rod 41 is practically incapable of influencing tension T of the yarn 31.

A circuit 52 (Fig. 2) for detecting the current I is connected to the line supplying the motor 37 with current I and has high impedance so as not to influence the value of I. In turn, circuit 52 is apt to generate a signal C, indicative of the current I. Preferably, signal C is processed by a filter circuit 53, to filter signal C of the oscillations and noise typically inherent in operation of the control circuit 51. The filter 51 is apt to generate at the output a filtered signal indicative of the current I, which is sent to a tension measuring circuit 54 of yarn tension. The measuring circuit 54, in turn, is arranged to obtain from this filtered signal, the effective value of yarn tension and to command visualisation of this value on a display 58. The circuit 54 is apt to update the tension value on the visualizer 58 with a frequency much lower than the measuring frequency in order to allow the tension value to have a certain stability on the display 58 and thus to be read with ease.

For calibrating the measuring circuit 54, a potentiometer 55 is provided for being regulated. The tension value visualised by the indicator 58 ought to coincide with a sample value for tension of the yarn 31. Provided by an external measuring appliance, a dynamometer for example, applied on the yarn 31 to pull it in the direction of the arrow 32 while potentiometer 55 is regulated.

A setting circuit 56, is apt to receive a yarn tension measuring signal from circuit 54 and is connected to a push-button 59, or to a functionally equivalent element. The push-button 59 may be actuated by the operator to memorise a data TREF, corresponding to a set tension value, in the setting circuit 56. In practice, the operator varies yarn tension manually and, on observing a tension value on indicator 58 equal to the value that he wishes to set, actuates push-button 59, thereby memorising the said TREF value in circuit 56.

Fig. 3 represents a device 60 similar to Fig. 1 , in which the tension gauge 36 uses an electromagnet 61 in place of the direct current motor 37. This electromagnet 61 is attached to the fixed structure 34 and comprises a core 62, a coil 63 wound around the core 62 and an armature 64 disposed between two extensions 66 and 67 (Fig. 4) of the core 62. The armature 64 has an end portion defining an air gap 68 with the end surfaces of the extensions 66 and 67. The core 63 is supplied with a current P (Fig. 4) by the control circuit 51 (Fig. 2) and generates a magnetic field apt to be conveyed by the core 62 to the air gap 68, to cross it and to produce in correspondence therewith a magnetic force FE suitable to attract the armature 64 towards a position corresponding to a state of minimum reluctance of the said air gap 68. Armature 64 is attached to a shaft 69, rotatably coupled to the fixed structure 34 such that the armature 64 is apt to move longitudinally with respect to the end surfaces of extensions 66 and 67.

An example of armature 64 and of expansions 66 and 67 and their disposition in defining the geometry of the air gap 68, is illustrated in Fig. 4. Magnetic force FE tends to move the armature 64 towards the inside of the space between extensions 66 and 67, increasing the area of the armature sides which faces the end surfaces of the extensions 66 and 67. Furthermore, this geometry of air gap 68 is such as to determine a substantially constant course of force FE for a given current I' supplied to coil 63, along a fairly wide longitudinal displacement of armature 64 with respect to extensions 66 and 67. This property of air gap 68 depends chiefly on the fact that it does not vary substantially its geometry, while armature 64 moves, since the sides of the latter stay at a constant distance from extensions 66 and 67. The shaft 69 is arranged substantially parallel to the portion of yarn 31 on which the tension gauge 36 acts and supports at one end the rod 41 engaging the yarn 31 to deviate it by the angle α. A position sensor 71 , of the optical type for example and composed by an emitter part 72 and a receiver part 73, is aligned with one corner of the armature 64 to signal each, displacement of the latter from a fixed reference position with respect to extensions 66 and 67. Said reference position is selected within the field of

displacement of armature 64 with respect to extensions 66 and 67, within which field the pattern of magnetic force FE assumes a practically constant value.

In electromagnet 61 , the magnetic force FE acting on armature 64 is apt to assume a single direction, which is that of reducing reluctance of the air gap 68. Unlike the motor 37 which is apt to exert on its rotor 40 active torques having opposite directions to move the rotor in both directions, the magnetic circuit of electromagnet 61 is usually capable of activating magnetic forces suitable to act on armature 64 in a single direction, irrespective of the direction of the current I' in coil 63.

In order to cooperate with the magnetic force FE in keeping armature 64 substantially motionless in its reference position, armature 64 is associated with elastic positioning means, for example, two springs 74 disposed. The springs 74 are fastened to posts 76 extending from the structure 34, whilst they are attached to rod 41. One or both of these posts is apt to be adjusted in the direction of arrow 77 to regulate rod 41 in the exact measuring position corresponding to the desired angle of yarn deviation α from the rectilinear feed trajectory of yarn 31 between the two eyelets 33.

The electrical diagram relative to the device 60 is substantially equivalent to the one shown in Fig. 2. Current I, motor 37 and sensor 44 in the diagram of Fig. 2 needs to be replaced with current I', electromagnet 61 and sensor 71 ; in additional, the control circuit 51 feeds the current I' to the electromagnet 61 of device 60 differently with respect to how it feeds current I to the motor 37 of device 30. In device 60, control circuit 51 does not need to feed current I' to electromagnet 61 in two opposing directions. This is unlike the case of device 30, in which the current I may be fed to motor 37 in two different directions in order to generate deflecting torques apt to act in opposition on rotor 40 to hold it in the measuring position.

Control circuit 51 , in response to signals received from sensor 71 indicating displacements of the armature 64 from the measuring position due to variations of

the tension T of the yarn 31 , feeds the coil 63 with the current I ^* so that the magnetic force FE activated by this current I' is apt to generate a torque with respect to the shaft 69 sufficient to contrast the torque produced, again with respect to shaft 69, by the force F that yarn 31 exerts on the tip of the rod 41.

Regarding activation of the magnetic force FE by the current I' to contrast the force F, force FE, which follows a substantially constant course for a wide field of displacement of armature 64 with respect to its measuring position, may easily be controlled by the current I' to rapidly assume a value apt to contrast the force F.

The signals used by circuit 51 to command the current I' are solely those indicating displacements of armature 64 in the direction opposite that of the force FE, namely the displacements that can be referred directly to variations of the tension T of the yarn 31. Any signals of displacements of armature 64 in the same direction as that of the force FE are used by the control circuit 51 to cancel the current I', so that it is only the springs 74 that bring armature 64 back to the measuring position. Accordingly, the action of springs 74 suffices to make up for the impossibility of the force FE produced by the electromagnet 61 to act from opposing directions on armature 64, as has already been stated on several occasions.

The magnetic force FE is apt to act in a single direction. The elastic force of the springs act symmetrically from two opposing directions on rod 41 cooperate and integrate their actions on rod 41 co-operate in order to contrast all its displacements away from the measuring position during variations of the tension T of yarn 31. The equilibrium with respect to the shaft 69 of the torques produced by the forces FE and F and by the elastic force of springs 74 is constantly restored by varying the current I' in function of displacements of armature 64 from its reference position and hence rod 41 , shaft 69 and armature 64 remain substantially motionless during variations of the tension T of the yarn 31. Thus the current I' is indicative of the tension of the yarn 31 and is measured by the measuring circuit 54 (Fig. 2) in a way substantially equivalent to that already described for the device 30.

A brake 81 (Fig. 1 and Fig. 3) may be coupled with the yarn tension sensor means 36 to take action on a yam portion before the latter reaches rod 41 and consequently its tension is detected by the yarn tension sensor means 36.

In Figs. 1 and 5, brake 81 comprises two pressure elements 82 and 83, facing and pressing against each other with the yarn 31 interposed between them. In one example, the elements 82 and 83 can be plate-shape. Element 82 is fixed and integral with the structure 34. Element 82 is free to move to vary its distance from element 83 in function of cross size of the yarn 31. Element 82 bears a block 84 equipped with a slot 86 so as to be apt to slide along a pin 87 of an arm 88 connected to the structure 34.

Pressure elements 82 and 83 are pressed together by way of a magnetic circuit comprising the element 82, a core 91 and a coil 89 wound on the core 91 , wherein the latter are both integral with the structure 34 and are disposed beside element 83. Pressure element 82 is of a ferromagnetic material, unlike element 83 which is of a non-magnetic material, stainless steel for example, and thus element 82 permits closing on core 91 of a loop of magnetic flux 92 generated by the coil 89 when the latter is supplied with a current If. With core 91 , the element 82 defines two air gaps 93 corresponding to the areas in which the distance between the pressure elements 82 and 83 is smallest and in which a force of attraction is generated magnetically between the latter elements, constituting the braking force FR effectively exerted on the yarn 31. Attraction members 94 in ferromagnetic material are mounted on plate 82 in correspondence with the air gaps 93 and allow force FR to be increased.

The push-button 59 (Fig. 2) is used to condition operation of the brake 81. Push¬ button 59 can be activated in two modes or positions, in the first of which in addition to memorising the TREF value, it also arranges for a braking circuit 96 of the brake 81 to operate under a modulated braking condition. After activating push-button 59 in this first position, the TREF value remains constant and does not undergo any

alteration until the push-button 59 is activated again in this first position In the second position, push-button 59, by way of setting circuit 56, switches operation of the braking circuit 96 from the modulated braking mode to a constant braking mode, in which brake 81 is powered with the constant current If and, consequently, generates a constant braking force on yarn 31

A comparing circuit 57 is apt to receive both the TREF value set and a value TIST supplied by the measuring circuit 54, which corresponds to the instantaneous yarn tension value, and to provide as output for the braking circuit 96 a signal indicating which of the two values, TREF or TIST, is the higher If circuit 57 signals that the TIST value is higher than the TREF value and the braking circuit 96 is set for modulated braking operation, said braking circuit 96 automatically reduces either entirely or in part the current If supplied to brake 81 , thereby cancelling or at least reducing braking on the yarn 31 Conversely, if comparing circuit 57 signals that TIST is below the value of TREF, the braking circuit 96 starts powering the brake 81 again with the current If, restoring it to its nominal value either instantaneously or progressively and thereby increasing braking on the yarn 31

When the braking circuit 96 is adjusted by push-button 59 for constant braking operation, the result of the comparison made by circuit 57 does not condition circuit 96 and the latter continues to exert a constant braking on the yarn Circuit 96 can be regulated manually by a potentiometer 97 to set the braking in constant braking mode, or to set the nominal value of the current If and thus of braking which, in modulated braking mode, corresponds to the threshold value TREF In the latter case, the potentiometer 97 and the push-button 59 may be used together, by first operating on potentiometer 97 to vary yarn braking manually and change the tension value visualised on the indicator 58 until the value visualised is that of the TREF value desired and then, at this point, by activating push-button 59 to set this value for TREF

Fig. 7 represents device 101 (variant of brake 81 ) in which the magnetic circuit is disposed completely on one side with respect to a fixed pressure element or plate

102 in non-magnetic material. The magnetic circuit comprises a pressure element 103, also plate-shaped, of ferromagnetic material which acts as a movable armature, a core 104 and one or more coils 106 wound on the core 104 and apt to be supplied with a current to generate a magnetic flux 107. The core 104 and the plate 103 define between them air gaps 108, which are traversed by the flux 107 to generate magnetic forces FM. Unlike the brake 81 of Fig. 5, the air gaps 108 do not close the magnetic flux through the surfaces of plates 103 and 102 in contact with the yarn.

Plate 103 is pushed by a spring 109 against plate 102 that is fixed to the structure 104 to exert a pressure on the yarn 31 passing between plates 103 and 102. The charge C of the spring 109 on plate 102 can be adjusted manually, by varying length of the spring 109 by means of a knob 11 equipped with a leg screwing onto a support 112. The latter is apt to rotate on the structure 34 in order to move plates

103 and 102 away from each other. The movable plate 103 is supported to move perpendicularly with respect to the fixed plate 102 and also to contrast the tangential force to which it is subjected by the yarn 31 because of the sliding friction thereof on plate 103. For example, plate 103 can be equipped with a pin which is apt to slide along a corresponding hole (not shown) of the core 104.

In device 101 , the magnetic forces FM generated by the flux 107 are apt to remove plate 103 from plate 102 against the action of the charge C of spring 109 in order to reduce yarn braking. In this case, the magnetic forces FM have a negative action on tension T of the yarn 31 in that they increase, the more braking and therefore tension T of the yarn 31 decreases. Brake 101 can also operate both in constant braking mode, in which braking is fixed and is set by varying the load C through the knob 111 , and in modulated braking mode in which the coil 106 is supplied with current to generate the forces FM in response to the comparison of a reference value TREF with an instantaneous tension value TIST as described for brake 81.

Compared to brake 81 and in the constant braking mode, the brake in device 101 has the added advantage of not absorbing energy as no current flows through coil 106. Moreover, should the magnetic forces FM prevail over the force C of spring 109, the plates 103 and 102 would open and the yarn 31 would be freed completely from their grasp.

This characteristic of being able to open and close the plates with respect to the yarn 31 may also be achieved on the brake 81 described earlier simply by providing the latter with a spring (not shown in the drawings) apt to exert on plate 82 a force with direction opposite that of the forces FR in such a way that when the latter become null, plate 82 moves away from plate 83. For example, this spring may be housed inside the slot 86 and act between the block 84 supporting plate 82 and the pin 87 to make block 84 slide on pin 87 away from the brake 81.

Plates 82, 83 and 103, 102 of brake 81 and brake 101 respectively can be rigidly associated to vary braking force when the brake 81 and the brake 101 work in modulated braking mode. In both brakes the air gaps 93 and 108 are placed adjacent to the surfaces of the plates which are in contact with the yarn. The magnetic force generated in the air gaps acts directly on these surfaces and in fact avoids the plates from behaving in an elastic fashion, which could disturb control of the braking force during its variations resulting from the corresponding magnetic force variations. An elastic behaviour is unfavourable and depends essentially on the fact that the plates may bend or become deformed, thereby acquiring elastic energy for increasing their reciprocal pressure. This elastic energy is then released by the plates when they reduce their pressure. In modulated braking mode, this elastic energy reduces the rapidity with which braking force is controlled, and this renders braking modulation less efficacious. According to the above-mentioned characteristic, no parts of the plates are subject to deformation during braking force variations nor can they store elastic energy, which is what would happen had the plates elastic parts arranged between the air gaps in which the magnetic force

determining braking force is generated and the surfaces of the plates that effectively brake the yarn and are in direct contact therewith.

Fig. 8 includes diagrams indicating different yarn tension pattern when the yarn is braked by a brake which, in the modulated braking mode described earlier, works in rigid mode, with respect to the case where the yarn is braked by a brake working in elastic mode.

The unbroken line diagram 113 refers to the brake operating in rigid fashion and shows that when the tension T exceeds a threshold value Ts corresponding to the set tension TREF, and consequently the braking current If is cancelled or at least reduced, the yarn tension continues to increase only for a very short time Δt before returning rapidly to below the threshold value Ts.

The dashed line diagram 114, regarding the brake that operates in elastic fashion, clearly shows that, after Ts is exceeded, the tension tends to continue increasing for a time Δt1 considerably greater than Δt, and has a lengthy excursion above the threshold value Ts before returning below it. Therefore, in the case of the rigid brake, the current If will be made vary with a high frequency (diagram 116 with the unbroken line), braking force will respond immediately to the current If and as a result a course of tension T will be obtained with minimum deviations from the threshold value Ts. On the contrary, in the case of the elastic brake, the current If will vary at a much lower frequency (diagram 117 with the dashed line), braking force will follow current If with a delay time and the deviation of tension T from the set value Ts will be greater.

For example, in place of a stepped variation of current If, like that shown in Fig. 8, where the current If goes instantaneously from a set, nominal value Is to a null value when the threshold tension Ts is exceeded, and vice versa, goes instantaneously from zero to the value Is when the tension drops below Ts, a ramp type variation

may be arranged where the current If is made vary gradually, both when the threshold tension Ts is exceeded and when tension drops below the threshold.

Moreover, in brakes 81 and 101 the parts arranged between the surfaces making contact with the yarn 31 and the gap in which the magnetic forces are generated, can be very light. Their lightness, together with their aptitude to operate in a rigid way without deformations, makes the brakes 81 and 101 apt to control quickly and exactly the braking of the yarn 31 , in response to the commands of the yarn tension sensor means 36. In this way the device of the invention is suitable to modify instantly, that is in "real time", the normal course of yarn tension and to compel it to follow a desired course, even when the normal course is characterised by fast, frequent and high tension peaks induced by corresponding quick variations of the yarn advancing speed, as is the case of the weft yarn inserted in a modern shuttleless loom, of the gripper, or projectile or air jet type.

Fig. 9 shows device 120 which differs from the version 30 of Fig. 1 in the addition of mechanical coupling means between the rod 41 and the motor 37, with the purpose of making shaft 38 rotate relatively to rotor 40 with an angular rotation amplified with respect to that of rod 41. Device 120 accordingly acquires greater speed and precision of response with respect to device 30, while at the same time motors 37 of smaller size may be used. The greater speed and precision depend essentially on the fact that the angular displacements of disk 42 are amplified with respect to those of rod 41 and therefore the sensor 44 can signal with greater rapidity each angular displacement of rod 41 from its reference position as a result of variation of tension T of the yarn 31. In addition, smaller size motors 37 may be used given that the torque acting on them is lesser in inverse proportion to the amplification of the angular displacements of rod 41. The coupling means comprise two meshing gears 121 and 122, of which gear 121 has a greater pitch diameter than that of gear 122 and is attached on a shaft 123 rotatable on the structure and the gear 122 is attached to the shaft 38 of motor 37.

Fig. 10 illustrates mechanical coupling means, constituted by an arm 124 divided into two parts 126 and 127, which are attached by one end respectively to shaft 123 and to shaft 38 of the motor 37 and have their other two ends connected together inside arm 124, by means of a deformable element 128. The part 128 is longer than part 127 and thus arm 124 transfers, after amplifying it, the angular displacement of rod 41 to shaft 38, simultaneously deforming element 128.

To damp possible mechanical vibrations, induced in the rod 41 and the rotor 40 by variations of tension, damping means 129 in Fig. 11 are provided; apt to cooperate with rod 41 and rotor 40 to dampen displacements thereof. The damping means may comprise a rubber element 131 placed between the rod 41 and the shaft 38 of the motor 37.

According to Fig. 12, the damping means 129 comprise a case 132 connected to the shaft 38 and bearing the rod 41 , hermetically sealed and filled with a viscous fluid 133, and a mass 134 which is free to move inside the case 132, though the displacements thereof are braked by the viscous fluid 133, which occupies the space between case and the movable mass 134. The displacements of mass 134 take place later than those of the case 132 and the rod 41 because they are damped by the fluid 133 and, jointly by way of reaction, the rotating displacements of rod 41 and case 132 also suffer damping.

Fig. 13 illustrates one variant of the electrical block diagram of Fig. 2. In this variant, some circuits which in Fig. 2 were separate are now integrated in a single circuit or microcontroller 141. The latter is equipped with a computing section (CPU) 142, a memory 143 and an analogic/digital conversion section (A/D) 144, and is apt to receive, through the A/D section 144, the values sampled by a sample & hold type circuit 146 of the signal C, which is in turn provided to circuit 146 by the current detecting circuit 52. On the basis of the sampled values, the microcontroller 141 commands the braking circuit 96 and the display 58, visualising on the latter the tension value measured for the yarn 31. Furthermore the microcontroller 141 is

connected by way of a line 147, whereupon digital signals transit in both directions, to a position control circuit 148, which is destined to digitally control motor 37 The microcontroller 141 and the position control circuit 148 are parts of a digital type position control which, has undoubted advantages over analogic type controls, such as having greater precision and speed and being practically insensitive to noise In the memory 143 of microcontroller 141 , a program is stored that manages the different operations of the device 30 of the invention This program may carry out a plurality of functions, such as updating and visualising on display 58 the maximum value of tension T measured over a given period of time, the mean tension value calculated and other statistical data concerning behaviour of the tension over a given period of time, and still other functions The values cited above can be either only calculated and stored in the memory 143, in anticipation of a request from the operator to visualise them, or they can be visualised automatically at programmed intervals The microcontroller can be conditioned by a RESET push-button 149 to reset the device 30, causing the above-mentioned program to start again from the beginning, and by one or more setting push-buttons 151 , which are provided to set ,ιn a way similar to the push-button 59, the threshold value Ts for the tension of the yarn 31 , in the case of operation of the device in modulated braking mode, or to set the device for constant braking mode operation

The described electrical diagram is also applicable for device 60 of Fig 3, in which the electromagnet 61 is used in place of the motor 37 Fig 13 illustrates, close to the motor 37 and by way of alternative to the latter, electromagnet 61 with the sensor 71 The way in which current is provided by the position control circuit 148 differs from motor 37 to electromagnet 61 , bidirectionally in the case of the motor 37 and unidirectionally in the case of the electromagnet 61

Rod 41 as the guide element for yarn 31 in detecting tension T can also act as a retractor for the yarn, when it tends to slacken Fig 14 shows that elastic yarn retracting means comprising a spring 161 are arranged between the rod 41 and the shaft 38 and are apt to exert on rod 41 a torque that tends to cause it to rotate in the

direction of the arrow 162 (Fig. 15). Therefore, this torque acts on rod 41 in a direction opposite that of the force F generated by the yarn 31. The spring 161 consists of a series of windings disposed around the shaft 38 and has one, adjustable end attached to an adjustment knob 163 and the other end connected to a sleeve 164, coaxial with and rotatable on the shaft 38 and whereon rod 41 is attached. The knob 163 is used to regulate the charge of the spring 161 on sleeve 164, by rotating the adjustable end of spring 161 with respect to the shaft 38 so as to provide its windings with a varying amount of torsion. In order to lock the knob 163 in numerous adjustment positions, locking means 166 have been provided, comprising for example a small ball housed in the knob 163 and pushed by a spring to couple with recesses located in the outer surface of the shaft 38 along its circumference.

Whereas the shaft 38 is held motionless in the measuring position defined by the alignment between disk 42 and position sensor 44, rod 41 being connected by elastic means to shaft 38 through spring 161 can oscillate during variations of the tension of the yarn 31. To limit these oscillations and to transmit to shaft 38 a torque that is as indicative as possible of tension of the yarn 31 , the sleeve 164 is slidingly mounted on shaft 38 by means of a unidirectional coupling 167 and a viscous joint 168. The unidirectional coupling 167 is disposed between the sleeve 164 and the viscous joint 168 and is apt to render the rod 41 and the sleeve 164 integral with the viscous joint 168 only when the rod rotates in the direction opposite that of the arrow 162, under the action of the force F dependent on the yarn tension.

The viscous joint 168 connects the unidirectional coupling 167 with the shaft 38 and comprises a rotating flange 169 closed inside a case 171 , and coupled with internal surfaces thereof through a viscous fluid. The viscous fluid fills the spaces between the rotating flange 169 and the case 171 , and gaskets 172 prevent the viscous fluid from seeping out of the case 171. The latter is fixed to shaft 38, which is therefore subjected to the same torque that the flange 169, while rotating in viscous manner with respect to the case 171 , exerts on it. This torque is, in turn, determined by the force F of the yarn 31 on the rod 41 and therefore also by the yarn tension T. By

damping the rotations of rod 41 , the viscous joint 168 ensures that the latter rotates at a substantially constant speed during variations of the tension of the yarn 31 and thus permits a yarn tension measurement to be obtained that is little suffering from disturbance due to the inertia of rod 41 Furthermore, the unidirectional coupling 167 leaves the sleeve 164 free to rotate with respect to the viscous joint 168 according to the direction of the arrow 162, namely when the charge exerted by the spring 161 prevails over that F of the yarn tension on rod 41 , and rod 41 thus rotates to retract and put in tension the slackened yarn Accordingly, when rod 41 has to act as a yarn retractor, it turns extremely rapidly and in a way similar to that of known retractors, having rods that act solely to retract the yarn

As rod 41 can rotate with respect to shaft 38, the force F acting thereon will vary not only in function of the tension T of the yarn 31 , but also of the yarn deviation angle α or, by the same token, of the angle θ (Fig 15) which defines angular position of the rod 41 For obtaining an exact measurement of the tension that takes into consideration the angular position of rod 41 , a sensor of angular positions 173 is provided, apt to transmit to both the measuring circuit 54 and the microcontroller 141 a signal POS indicative of the various positions of rod 41 , in order to condition measurement of the tension of the yarn 31 The position sensor 173 may comprise an optical group 174, attached to the fixed structure 34, and a blade integral with sleeve 164, in which a plurality of notches 177 have been made corresponding to different positions of the rod and are apt to create optical interference on group 174 to generate the POS signal

As an alternative to using elastic means 161 to make rod 41 act additionally as a yarn retractor, it is possible to use the measured value of tension T of the yam 31 to drive the motor 37, so that it acts positively on the yarn 31 , by means of rod 41 , and exerts a force of preestablished value on it and apt to retract and put in tension the yarn 31

Fig. 16 is illustrating operation of the motor 37 as a yarn retractor. A program apt to produce this form of operation may be stored in the memory 143 of microcontroller 141. Block 181 indicates the nominal state of motor 37, in which the rotor 40 thereof is held motionless in the measuring position in order to measure tension of the yarn 31. The decision-making block 182 measures if the yarn tension has been kept null or at least very low for a fixed time TPREF, and if so activates the block 183 in which motor 37 acts as a yarn retractor; if not, on the other hand, it continues to maintain motor 37 in the measuring condition 181. In block 183, motor 37 generates a torque to apply a preestablished force on yarn 31 through rod 41 , which consequently rotates and leaves the measuring position. In block 184, the signal S supplied by the position sensor 44 is used to detect whether or not rod 41 passes through the measuring position again. If so, a return is made to block 181 where the motor locks onto and holds the rod motionless in the measuring position so that tension of the yarn 31 can be measured. If not, the status corresponding to block 183 is maintained, whereby motor 37 acts for retracting and tensioning yarn 31.

A typical situation in which tension of the yarn may tend to cancel itself out for a certain time may occur at the end of a cycle of insertion of weft yarn. During such time, therefore, motor 37 is made operate as a yarn retractor, before returning to act as a tension gauge immediately at the start of the subsequent insertion cycle, when rod 41 is abruptly made rotate by the yarn drawing and consequently the rod 41 passes through the measuring position, where it is locked.

The device to control yarn tension can be associated with a yarn feeder 190 (Figs. 17-20), for example, a weft yarn feeder for a shuttleless loom. The feeder 190 is provided with a drum 191 having an axis 192 and on which a plurality of windings 196 of the yarn 31 are wound to form a reserve of yarn. The windings are apt to be unwound to feed the yarn 31 to a weaving machine not shown in the drawings, in the direction indicated by the arrow 193. The device to control yarn tension is arranged frontally to the drum 191 , where it is borne, at least partially, by a support 194 integral with the fixed structure of the feeder 190. The device also comprises yarn

tension measuring means, ("tension gauge") which include electromagnetic control means provided with a movable control element.

In the yarn feeder 190 the yarn 31 , in order to unwind regularly from the reserve of windings 196 without creating entanglements or slipping of the windings or disorderly movements in general, needs to a certain extent to be held adherent to the drum 191 while it is drawn off from the reserve and moves radially with respect to the drum 191 towards the axis 192. This can be generally achieved by lightly braking the yarn 31 against the drum 191 along the trajectory of the yarn immediately following the reserve of windings 196. A flexible cap 197 (or similar element) is arranged against a rounded end of the drum 191 in order to brake the yarn 31 , which pass between drum 191 and cap 197, while it moves circumferentially along the rounded end as it is drawn off. The cap 197 is adjustable along the axis 192 so as to vary its pressure on drum 191 and hence also braking of the yarn 31.

The device to control tension, may concern either only the yarn tension gauge, or a combination of the tension gauge with a brake, wherein the gauge and brake cooperate reciprocally to control yarn tension.

In the former case, as the yarn tension measuring means are practically unable to influence said tension but are confined to detecting its value, the cap 197 is the only effective braking element and is therefore used to regulate the operating tension with which the yarn 31 is fed by the feeder to the weaving machine. This means that, in the embodiments of the device without brake operating together with the yarn tension measuring means, the cap 197 is normally adjusted along axis 192 to exert a more consistent braking on the yarn 31 with respect to the strict minimum amount required to guarantee regular unwinding of the yam 31 from its reserve.

On the contrary, when the device is equipped with the brake associated with the tension measuring means, the brake is responsible for defining the tension at which

the yarn 31 is fed to the weaving machine and, in this case, flexible cap 197 is adjusted to exert on yarn 31 only the minimum braking already stated.

In the device 200 in Fig. 17, the yarn tension gauge 201 is disposed along the axis 192 and symmetπcally around it in order to convey the yarn 31 to the weaving machine through an outlet channel 202. In the yarn tension gauge 201 , the electromagnetic control means and the movable control element are constituted respectively by a magnetic circuit having a structure similar to that of a permanent- magnet linear motor of the type commonly known as "voice coil" and by a coil 203 coaxial with the axis 192 and partially immersed in the magnetic flux generated by the magnetic circuit. The yarn tension gauge 201 also comprises a sleeve 204 which supports the coil 203 by means of a flange 206 and which is arranged coaxially to the axis 192, and an eyelet 207 attached to one end of the sleeve 204 and apt to deflect the yarn 31 towards the axis 192. The eyelet is apt to receive the yarn 31 radially while the latter rotates around the eyelet, coming from the cap 197 (Figs. 17-20).

The magnetic circuit is apt to generate a magnetic field by means of a magnet 208 and defines a ring-shaped air gap 209, which is traversed by the magnetic field. A central part of the coil 203 extends along the gap 209. The sleeve 204 may be linked to the fixed part 194, on condition however that it maintains the capacity to move along the axis 192, in such a way that both the eyelet 207 and the coil 203 are apt to move along the said axis. For example, the sleeve 204 may be housed slidingly in a support 211 consisting in a ball bearing for axial sliding with low friction. In turn, eyelet 207 is subjected to an axial force FA, value of which depends on both the tension T in the yarn 31 and also on the angle of deviation of the yam 31 in correspondence of the eyelet 207.

Taking β (Fig. 17) to indicate the supplementary angle of the yarn deviation angle (the exact angle of deviation is in fact equal to (180°-β) ), axial force FA may be calculated using the formula: FA = 2 T (cos(β/2))2. Altematively to fitting slidingly on

support 211 , sleeve 204 may be linked elastically to the fixed part 194 by diaphragms 212 (Fig. 23), elastic in the direction of the axis 192 and rigid perpendicularly to it, or equivalent elastic elements, that guide the tube 204 and permit it to move along the axis 192 only. Position sensor 213 is apt to emit a signal S1 to signal to a position control circuit 214 (Fig. 25) axial displacements of coil 203 with respect to a given measuring position along the axis 192 and corresponding to identical axial displacements of the eyelet 207 caused by the axial force FA. In turn the control circuit 214 is apt to supply coil 203 with a current 11 , depending on the signal S1 received from sensor 213. Position sensor 213 may comprise an optical group constituted by an emitter and a receiver cooperating with a blade 216 integral with eyelet 207 and coil 203, wherein the blade 216 is apt to intercept radiation emitted by the emitter in the direction of the receiver in order to generate position signal S1. The magnetic circuit comprises two cores 217 and 218 of ferromagnetic material having an axially symmetrical shape with respect to axis 192 and arranged to the sides of the permanent magnet 208, in order to convey the magnetic field produced by the latter to the air gap 209, wherein the magnetic field is apt to generate a magnetic driving force on coil 203 in the direction of the axis 192 when said coil 203 is supplied with the current 11.

In operation of device 200, any axial displacement of coil 203 with respect to its measuring position or, similarly, of eyelet 207 with respect to a rest position thereof corresponding to the said measuring position, is signalled by sensor 213 to control circuit 214. In turn, the latter is apt to operate on the basis of the signals received from sensor 213 to prevent coil 203 from moving significantly far away from its measuring position. For this purpose, control circuit 214 sends into coil 203 the current 11 with a value such as to generate a magnetic driving force, acting on said coil 203 and apt to contrast and balance, in the area of eyelet 207, the axial force FA caused by the yarn.

Accordingly, coil 203 is held virtually immobile in the measuring position, rendering practically insignificant all exchanges of kinetic and elastic energy between yarn 31

and eyelet 207 during variations of the force FA and the corresponding variations of the yarn 31 tension. Current 11 thus adapts instantaneously to the value of the force FA and consequently is indicative of the tension T in the yarn 31.

Moreover, the angle β associated with the yarn deviation at eyelet 207, being close to 90 degrees, defines a force FA (according to the formula provided earlier FA=2T(cos(β/2))2) which is of the same order of magnitude as the tension T and which, in any case, for equal values of the tension T, is much higher than the corresponding force F of Fig. 1 (in that case, in fact, F=2Tsin(α) with α fairly low). Consequently, current 11 will also assume a more significant and higher value in order to balance force FA, with respect to the value of the current I associated with the force F, thus greatly advantaging the precision and reliability of the tension T measurements.

Current 11 , supplied to coil 203, is detected by a circuit 219 (Fig. 25) to generate a signal C1 , which is sampled by a sampling circuit 221 and then sent to a microcontroller 222, which provides for taking the effective value of tension T from the sampled C1 signals and for visualising this value on a visualizer 223, in the same way as already described in relation to the schemes of Fig. 2 and Fig. 13.

The microcontroller 222 comprises a computing unit (CPU) 224, a memory 225 apt to memorise a program planned to control the different operations of device 200, and an analogic-digital conversion section 226 (A/D) apt to receive the sampled values of signal C1. Finally the microcontroller 222 is connected to a RESET push-button 227, for resetting of the device 200.

In Fig. 18, device 230 includes electromagnetic control means of the yam tension measuring means 201 constituted by an electromagnet. Comprising a case 231 of metallic material, a winding 232 inside case 231 and apt to be supplied with a current 12, a sliding bush 233, and a cylindrical armature 234 coaxial with the axis 192 and supported slidingly along said axis by bush 233. Armature 234 constitutes

the movable control element. Eyelet 207 is connected to the cylindrical armature 234 by means of a sleeve 236 which transmits to the latter each axial displacement of eyelet 207 caused by variations in tension of the yarn 31. Also provided are elastic means, which are functionally similar to the springs 74 of variant 60 (Fig. 3) described earlier, constituted by two springs 237. Said springs are arranged between the case 231 and the group comprising the eyelet 207, the sleeve 236 and the armature 234, and act on said group to hold it elastically in a position of equilibrium along the axis 192, in which position the springs balance each other reciprocally (measuring position). The tension gauge 201 comprises a Hall effect sensor associated with the armature to detect axial displacements thereof, with respect to its measuring position due to variations of the tension T of the yarn 31.

Operation of device 230 with electromagnet differs from that of the device 200 in that, in order to keep the armature 234 substantially motionless in the measuring position during yarn 31 tension variations, it is not possible to exert opposing magnetic forces on said armature 234. In fact, similarly to device of Fig. 3, the magnetic force generated by the electromagnet and acting on armature 234 is apt to assume a single direction only, irrespective of the direction of the current 12 in winding 232. For this reason, the current 12 is supplied to winding 232 in one direction only, namely that indicated on the right in Fig. 25 with reference to device 230.

Armature 234 is subjected, in a single direction along axis 192, to a magnetic force apt to oppose force FA and activated by the current 12 in function of axial displacements of armature 234 which are directed in accordance with the force FA and, in both directions along said axis 192, to the elastic forces of springs 237. Accordingly, both the magnetic force activated by the current 12 and the elastic forces supplied by springs 237 cooperate to maintain armature 234 in its measuring position and permit current 12 to assume a value indicative of the tension T of yarn 31.

The yarn tension control device 240 in Fig. 19 includes yarn tension measuring means 201 , electromagnetic control means and the corresponding movable control element constituted by a motor 241 and by the rotor thereof. The motor is powered by a current 13 supplied by control circuit 214 (Fig. 25) and is attached to structure 194 of the yarn feeder 190 by means of a support 242, in a region disposed laterally with respect to the axis of drum 191. The rotor of the motor 241 is equipped with a shaft 243 to which one end of a rod 244 is attached. The rod in turn supports on its other end the eyelet 207. Shaft 243 is thus subject to a torque, value of which is equal to the product of the force FA acting in correspondence with the eyelet 207 and depending on the tension T of the yarn 31 by the arm of the force FA with respect to the shaft 243. This arm is determined essentially by the length of rod 244. Also provided is a position sensor 246 apt to collaborate with a blade 245 integral with shaft 243 and rod 244 to signal each rotation no matter how small, of the rotor of motor 241 with respect to a given measuring position.

In operation device 240 has the advantage over previous devices 200 and 230 that it does not create any encumbrance in the region around eyelet 207, and consequently permits comfort of operation in this region, in manually threading the yarn 31 into the eyelet 207, for example.

Fig. 20 represents device 250, in which the electromagnetic control means and the associated movable control element are constituted respectively by a flat-shaped motor 251 and by a winding thereof 252, also flat in shape. Motor 251 can be classified as a rotary voice coil actuator. Winding 252, which corresponds to the rotary coil, is apt to be supplied with a current 14 and is supported by a frame 253 which, in turn, is pivotally mounted on a fixed pin 254 to allow said winding 252 rotate between two permanent magnets 256 (Fig. 21 ), keeping at a constant distance from the two. The permanent magnets 256 are supported by a casing 257, which is attached to the fixed structure 194 by means of a support 258 and is constructed in metal sheet bent into a 'U' shape. The permanent magnets 256 are apt to generate a magnetic field suitable to interact with winding 252 supplied with

the current 14 so as to generate a magnetic force acting on the latter to rotate it around pin 254. The magnetic force may cause the flat winding 252 to rotate either clockwise or anti-clockwise, depending on the direction of the current 14 in winding 252. Frame 253 is connected to one end of a rod 259, which in turn supports the eyelet 207 at its other end. Finally an optical position sensor 261 is placed between the fixed structure 194 and the frame 253 to signal any displacements of winding 252 from a given measuring position, due to corresponding axial displacements of eyelet 207 caused by variations of the force FA depending, as repeatedly said, on the tension T of the yarn 31.

In Figs. 20 and 21 , the winding 252, the frame 253, the rod 259 and the eyelet 207 form a compact group, comparable to a balancer lever pivotally mounted on pin 254. Consequently the relationship between displacements of the eyelet 207 and the corresponding displacements of winding 252 is equal to the relationship between the respective radial distances from pin 254. This configuration gives the device optimum size, for achieving the maximum speed of response in measuring yarn tension.

In device 250, an important advantage consists in that its parts are apt to have angular displacements, even if very slight, around the measuring position, unlike other embodiments, such as 200, in which the corresponding parts are apt to have rectilinear displacements. The said parts of device 250, being pivoting on pin 254, are subjected to a very low rotary friction generically dependent on a yarn tension component acting on the eyelet 207. In device 200, the sleeve 204 is constrained to move solely along the axis 192, causing a low sliding friction to act on sleeve 204. In device 250, the position of the fulcrum along the lever, sections of the arms, moment of inertia with respect to the fulcrum etc of the balancer lever, lead to optimum performance. It does not possess parts arranged concentric with and adjacent to eyelet 207, so that the area around eyelet 207 is free. The operation of device 250 is essentially the same as described above. The electrical block diagram of Fig. 25 indicates schematically at the bottom right how the yarn tension control

devices 200, 230, 240 and 250 are interfaced electrically with the rest of the diagram. Devices 200, 240 and 250 operate in like manner whereas the device 230 with electromagnet has provision for supply of current 12 in a single direction.

Fig. 22 illustrates mounting of frame 253 on support 258, wherein frame 253 instead of being pivoting on the support 258, is connected thereto by means of an elastic plate 262. Said plate cooperates elastically with the magnetic forces acting on winding 252 to maintain the latter substantially motionless in its measuring position.

Device 270 in Fig. 23 comprises, in addition to tension gauge 201 , an electromagnetic type brake 271. Brake 271 is apt to cooperate with gauge 201 to brake the yarn 31 in function of the yarn tension measured by said gauge 201. The brake 271 is arranged along the feed path of yarn 31 , immediately before the tension gauge 201 , which thus receives the yarn 31 directly from the brake 271. Brake 271 comprises two disks 272 and 273 facing each other, coaxial with axis 192 and apt to press yarn 31 between them, the first of which is arranged near the drum 191 and the second near the yarn outlet channel 202. Disk 273 is also equipped with a central hole 274, through which yarn 31 enters the eyelet 207. The latter, in turn, is disposed inside the hole 274, in such a way as to guide yarn 31 during the majority of the deviation of its trajectory towards axis 192. Disk 272 is fixed, whereas disk 273 is movable axially along a guide 281 , formed in the support 194, so as to be able to adapt to the cross size of yarn 31. A magnetic circuit 276, comprising a coil 277, a core 278, and the disk 273 (constructed in ferromagnetic material) is apt to press the latter-named on disk 272 (which is in non-magnetic material) to vary instant by instant braking force on the yarn 31 , while it passes between disks 272 and 273.

Brake 271 is controlled by a braking circuit 279 (Fig. 25) and can operate either in a constant braking mode or in a modulated braking mode. Provided for this purpose (Fig. 25) are a regulator 282 of the current If sent to brake 271 , for manual regulation of the braking force on yarn 31 , and a setting push-button 283 for setting a reference

value for yarn tension T, at which modulation of brake 271 must be activated, in a way substantially equivalent to that described in relation to Figs 2 and 13.

Fig. 24 illustrates device 290 (similar to device 270), in which brake 271 is replaced by a brake 291 , apt to generate a magnetic force which is subtracted from the mechanical force exerted by a spring 292 to vary braking on the yarn 31 when working in modulated braking mode, similarly to the way described in relation to the device 81 and 101. Brake 291 comprises two disks 293 and 294 pushed against each other by spring 292, and whereon operates, detaching them, a magnetic circuit co-operating with disk 294 by means of an air gap 296, in turn traversed by a magnetic field generated by a coil 297 and conveyed towards the air gap 296 by a core 298. The pressure exerted by spring 292 on disks 293 and 294 is manually adjustable by means of a threaded ring-nut 299, which is apt to move along the axis 192, screwing on to a fixed part 302 of the device 290. Disk 294 is apt to slide axially on a guide 301 formed in core 298 in order to adapt to the cross size of yarn 31 and be able to move away from the fixed disk 293. Disk 294 has a central hole 303, wherein is housed eyelet 207 of the yarn tension gauge 201 (shown in Figs. 17 and 23).

Likewise brakes 81 and 101 , brakes 271 and 291 have parts arranged between the gap in which the magnetic forces are generated and the surfaces in contact with the yarn, which are characterised by a high mechanical transitivity. These parts are advantageously realised to be very light and to operate rigidly without elastic deformations.

The brakes installed on feeder 190 and responding to the yarn tension measuring means 201 (hereinafter also called yarn tension sensor means 201 ) are not to be limited to the brakes 271 and 291 described earlier. In another embodiment , the brake associated with the tension measuring means 201 comprises an annular cap brake 300 (Fig. 26) of similar shape to the cap 197 described earlier and, arranged

on one rounded edge of drum 191. The annular cap brake 300 may make use of other braking devices superfluous.

Modulation of the braking produced by brake 300 acts on the yarn as soon as it comes out of the reserve of windings 196 on drum 191 , and not in the area of eyelet 207, as in the case of brakes 271 and 291. To modulate braking, the annular cap brake 300 is associated with electromechanical actuating means which drive it and alter its pressure on the rounded edge of drum 191 , in response to the signals of the yarn tension sensor means 201. Operation and purpose of these electromechanical actuating means for driving the annular cap brake 300 are substantially similar to those of the electromechanical means associated with brakes 271 and 291 , i.e. to rapidly vary the braking of yarn 31 , even though the electromechanical actuating means for driving the brake 300 are structurally different from the electromechanical means associated with brakes 271 and 291 , on account of the different location and the greater braking surface of the annular cap of brake 300.

Device 310 in Fig. 26 is a combination of the yarn tension measuring means 201 with the annular cap brake 300 resting against a rounded edge 304 of one end of drum 191. Cap brake 300 comprises a thin, non-extendible, continuous body 305, that has the shape of a truncated cone and is symmetrical with respect to axis 192. The truncated cone body 305 is intrinsically flexible when free, but can take on a practically non-deformable configuration when pressed against the drum 191 by the thrust of resilient means 306 which place body 305 under traction. The body 305 is designed to remain substantially non-deformable during variation of the load applied on it by resilient means 306 and/or by the electromechanical actuation means associated with the body 305 itself.

The fact that it fails to bend but locally deforms where the yarn is passing during load variations characterises both the structure and operation of the body 305 with respect to the cap 197 employed in the embodiments described earlier. For a better understanding of the differences between the body 305 and the cap 197, it will be

remembered that cap 197 is designed for operation under conditions entirely different from those of body 305 of brake 300. The latter in fact has to be operated in order to modulate braking, whereas cap 197 is arranged so as to exert a constant braking force all during operation. At best, cap 197 may be regulated selectively, prior to operation, in order to vary and set its braking action. Cap 197 is normally constructed in such a way as to bend during regulation along axis 192, so as to vary pressure on drum 191. Note that this flexibility of cap 197 does not have a negative effect on its operation, since cap 197 maintains a substantially stable configuration during operation. Cap 197 may be comprised by a plurality of metallic strips designed to flex and vary their pressure on the drum during manual regulation or, in place of the metallic strips, by brushes that are also pliable.

By virtue of its substantially unbending configuration in the operation, the truncated cone shape body 305 can respond extremely rapidly, varying its pressure on drum 191 , to the commands of the electromechanical actuating means, which consist for example of a ring-shaped magnet 307, fastened to the fixed structure 194 and comprising a coil 308 and a ring-shaped core 309. Provided for performing this high responsiveness is a flange 311 , fastened to one end of body 305 and apt to be attracted by magnet 307 to reduce pressure of truncated cone body 305 on drum 191 , that is on yarn 31 , contrasting resilient means 306. Resilient means 306 may be adjusted with respect to frame 194 of yarn feeder 190 by means of a movable support 312, thus varying the pressure they exert on drum 191. It is also possible to regulate an air gap 315 between magnet 307 and flange 311 in order to set the attraction force of magnet 307 on flange 311.

A control unit 317 receives a position signal S from yarn tension measuring means 201 , supplies means 201 with a current IBAL to balance the forces acting on eyelet 207 caused by yarn tension T, in order to keep the same eyelet 207 substantially motionless in a predetermined position and, in function of the value of current IBAL, supplies magnet 307 with a current IM to modulate braking on yam 31.

With the device 310 of Fig. 26, braking force controlled by measuring means 201 is subtracted from a braking force imposed manually by way of resilient means 306. Other schemes may be envisaged, such as the one in which the body 305 is positively pressed against yarn 31 by electromagnetic means controlled by yarn tension measuring means 201 and, for example, housed inside drum 191 in front of body 305.

In other possible combinations of yarn tension measuring means 201 with braking devices, the latter may consist of brakes that produce their braking action essentially by deflecting the trajectory of the yarn. These brakes, also known as multiplying effect brakes, are specially equipped with guides that engage the yarn and cause it to follow a path deflected laterally from the rectilinear path. For yarn braking modulation, these brakes arranged to downstream of means 201 , along the direction of feed of yarn 31 , have movable guides apt to move with respect to fixed guides, in order to alter lateral deflection of the yarn path and thus the angle at which the yarn is wound on the guides, both movable and fixed.

All embodiments described control the tension T of weft yarn during insertion in the shed of a loom. Fig. 27 illustrates the pattern of weft tension (curve a) during insertion in a gripper loom, in conventional type operation without using the device of the invention.

With bold curve b, Fig. 27 shows the pattern obtained when the device of this invention is used. In conventional type operation, the cycle of yarn tension T is normally characterised by two distinct parts, which are a true reflection of the speed pattern (curve f) of the two grippers of the loom, the delivering one and the receiving one. The amount of the friction encountered by the weft yarn during insertion in the loom shed, and which is responsible to a large extent for its tension, is in fact in a first approximation proportional to the speed at which the yarn is picked by the two grippers. For this reason, in each part of the cycle there is a maximum tension peak roughly corresponding to the gripper maximum speed condition and, for the same

reason, the central part of the cycle normally has a point at which tension tends to be null, corresponding to the changeover of yarn between the two grippers, which takes place at a yarn speed which is practically null. At the start of insertion, there is generally a peak c caused by sharp acceleration of the yarn. Fig. 27 shows that the device serves as a means for avoiding the yarn tension T having a pattern closely linked to speed Vp of the grippers with resultant peaks. The device maintains tension T very close to a predetermined tension pattern, such as defined for example by a constant threshold level TMAX, independently from speed cycle of the grippers. According to curve d of the current absorbed by the brake and corresponding to the braking force FB exerted by the latter on the yarn, the brake intervenes repeatedly in each part of the loom cycle to modulate braking force FB and thus keep tension substantially close to the defined value TMAX, in spite of variations of the yarn speed. The brake intervenes instantly to vary braking force FB each time the threshold TMAX is crossed through by tension T, whether in the ascending or descending direction, respectively to decrease or increase braking force FB.

At the start of the cycle when the weft yarn is accelerated, the brake is deactivated almost immediately (part h of curve d), avoiding an initial tension peak. In the centre, when yarn is exchanged between the two grippers, the brake tends to remain active for longer (portion e of curve d), thus allowing yarn 31 to be under tension during the changeover. Further, in the time between one insertion cycle C1 and the next C2, the brake ceases modulation and generates a constant braking force (portion g of curve d). During this time, the yarn is motionless and thus subjected to a static tension of an essentially null or reasonably low value, given that despite the pressing action of the brake on the yarn, the latter does not slide with respect to the brake.

Similar considerations may also be made with reference to other types of loom, such as air jet or projectile looms.

With device 270, 290 and 310, yarn tension during insertion in the loom shed is controlled without imposing any lateral deviations on the yarn trajectory. The brakes have two faces or guides that press on the yarn from two opposite sides like a clamp in order to guide the yarn and control its tension during insertion, and that also remain substantially motionless during operation, with the result that these guides do not perform transverse movements with respect to the yarn path (fixed guide or clamp brakes).

Brakes referred to earlier altering yam trajectory laterally by means of movable guides in order to control yarn tension, may be used as an alternative to the fixed guide brakes (movable guide or yam lateral deflection brakes). These are different from fixed guide brakes. These differences might result in inferior performance of the movable guide brakes with respect to the fixed guide brakes in controlling yarn tension during insertion in the loom.

If the yarn lateral deflection, or movable guide type brakes are used, for example, with all other conditions being equal, a lower yarn tension control speed is implicitly obtained during insertion, on account of the far from negligible times needed for the movable guides to move from one position to the other in controlling yarn tension. That similar movable guides may consist of retractile guides, each arranged between two fixed guides and controllable in a positive sense to deflect a portion of the yarn sideways with respect to the fixed guides. In this way, it is possible to alter the angle at which the yarn is wound on the guides, whether fixed or retractile, for varying, as a result, the braking exerted by the friction of the guides on said portion of yarn and consequently the yarn tension. Similar movable guides may also consist of guides having a passive operation, such as for example guides that yield under the yarn tension action. In both cases, the rapidity of the device's response in controlling yarn tension is certainly conditioned by the time it takes the movable guides to complete their movements. No such conditioning is imposed by the fixed guide brakes.

The action of movable guide brakes, operating to vary yarn braking by altering the angle at which the yarn is wound on movable and/or fixed guides, is achieved substantially through a multiplying effect on the yarn tension. These movable guide brakes operate substantially as amplifiers of the tension in the yarn at the brake input, so that yarn tension at the brake output depends on the input tension value. If the latter is null, for example, variation of the winding angle described earlier does not produce any yarn tension variation effect. Performance of these brakes may be limited by the fact that a non-null yarn input tension, better still if controlled, is required.

In fixed guide brakes, the fixed guides are suitable for co-operation with the yarn in a relationship of pressure. As being in contact with the yarn and associated with electromagnetic means to vary their pressure on the yarn, these fixed guides can in fact put in tension the yarn without depending on the existence of any yarn input tension, as well as on its value. The fixed guide brakes also have the advantage of not causing any lateral movements of the yarn which could, if rapid, cause corresponding tensions in the yarn because of its lateral inertia. Operation of the clamp brakes that do not laterally alter trajectory of the yarn, is substantially exempt of any dynamic phenomena that could have a negative effect on speed and precision of the yarn tension control. With high and rapidly varying insertion speeds, fixed guide brakes intervene to vary the braking force FB on the yarn during all parts of the insertion cycle, with a promptness and frequency that ensure a practically continuous yarn tension control in time and rapid restoration of the yarn tension programmed pattern after even the slightest deviation of the yarn tension from it.

A surprising effect obtained from the device (particularly, though not exclusively, when equipped with fixed guide brakes) is that yarn tension may be controlled during insertion in a way that is completely independent of the loom cycle, i.e. without synchronising with this cycle and without receiving any commands from the loom cycle control unit, in virtue of the very high speed and precision with which the

device intervenes to control tension. Obviously this does not exclude combinations in which the device of the invention is operable in such a way as to interact with the loom cycle control unit for intervening, for example, at predetermined times during the cycle.

In Fig. 28, balancing current CUR1 supplied to the electromagnetic control means 320 in order to keep the relative movable control element 321 (corresponding to rotor 40, armature 64, coil 203, winding 252 of the movable coil rotating actuator 251 , etc) motionless, is produced by a driving circuit 322 in function of an analogic offset signal S1. The latter signal can be obtained from comparator circuit 324 by detecting the difference between the analogic signal S issued by position sensor 323 and a constant value reference signal RIF, corresponding to the signal generated by position sensor 323 when movable element 321 is exactly in its predetermined position. Current CUR1 is controlled by circuit 322 in order to prevent any substantial displacement of movable element 321 from its predetermined position and its value is measured by sampling the voltage on the terminals of a resistor R across which current CUR1 flows, using a high impedance measuring circuit 326 so as not to significantly alter the value of CUR1.

Fig. 29 illustrates a comparator circuit 327 for comparing the signal S coming from position sensor 323 with a sawtooth signal generated by the circuit 328 and outputting a logic signal S2 that is dependent on outcome of the comparison. This logic signal S2 is suitable for controlling a gate 329 placed between a switching circuit 331 and a clock circuit or time base 332 generating a constant frequency signal CK. Accordingly, gate 329 enables or disables passage of signal CK through it and controls activation of switching circuit 331 by signal CK to permit or inhibit a current generator 333 to supply electromagnetic means 320 with a balancing current CUR2. Finally a count circuit 334 determines the percentage of time for which the gate is active and allows passage of signal CK, so that size can be calculated of the balancing current CUR2 supplying electromagnetic means 320 at any time.