960,374. Photo-electric inspection devices; testing curl of paper. WIGGINS, TEAPE RESEARCH & DEVELOPMENT Ltd. Aug. 25, 1960 [Sept. 16, 1959], No. 31650/59. Headings G1A, G1M and G1S. [Also in Divisions D2 and G3] During the manufacture or processing of a continuous web of paper or other material, the amount of curl of a longitudinal edge of the web is measure or detected and an output dependent on this amount' is provided. This output may be visibly displayed or may be used to control an earlier stage in the process, e.g. the heating of upper and lower backs of drying cylinders, so that the web has a desired amount of curl, e.g. no curl. As shown in Fig. 1, a paper web 21 is travelling into or out of the plane of the drawing. Flash-tubes LP1 and LP2 are mounted above and below the web and are 180 degrees out of phase with one another. A photo-electric cell PC1 receives reflected light originating from one tube or the other according to whether the edge of the web curls upwards or downwards. The output from this cell is fed to an amplifier A1, and the amplified output to a gating circuit comprising alternative routes 23 and 24 which are made one conducting and the other non-conducting alternately in synchronizm with the flashing of tubes LP1 and LP2. The gated signal is fed to a differential voltmeter 26, the output of which is used to operate a differential relay 25 and a meter 27. In a modification (Fig. 4, not shown), a second photo-electric cell, PC2 placed beside the flashtube LP1 is connected in a negative feed-back loop to compensate for variations in the output of cell PC1 that would otherwise arise from variations in colour or surface texture of the web. In another modification (Fig. 7 not shown) two nozzles are mounted one above and one below the web, and direct air jets on to the edge of the web. Two collector tubes, connected with differential bellows receive the air jets reflected from the web. Pressure differences in the bellows cause, through a linkage, movement of a pointer. In other modifcations : the amount of curl is detected by mechanical means; opposite surfaces of the web are illuminated by light sources differing in colour or polarisation, and discriminated by these characteristics; and the light sources are filament or discharge lamps and intermittent illumination is obtained by shutters or mirrors. The electrical circuits are shown in Figs. 5 and 6. T1 is a centre tapped transformer which charges condensers C7 and C8 through rectifiers X1 and X2. These condensers discharge through the tubes LP1 and LP2 to produce the light flashes. The transformer also provides a high-tension supply between lines L1 and L2 through resistors R17 and R18 and rectifiers X3 and X4. The anodes of the two photo-cells PC1 and PC2 are connected through a semi-conducting diode D1 with the junction of a potential divider comprising resistors R1 and R2. Their cathodes are connected with line L1 through resistors R3 and R4. The cathode of cell PC1 is also connected through a small capacitor C1 with the grid of a pentode V1, which in turn is connected with the line L1 through a leak resistor R5. Each flash of light detected by cell PC1 is therefore fed as a positive pulse to the grid of pentode V1, and produces an amplified negative pulse at its anode. This pulse is fed through small capacitors C4, C5 to the junctions of the grids of triodes V2A and V2B and unidirectional circuits of low resistance connected with the line L1 through semi-conducting diodes D3 and D6 and resistors R11 and R13. Capacitors C7 and C8 are connected through capacitors C9 and C10 and resistors R14 and R15 with the junctions of D3 and R11 and D6 and R13, so that as each of capacitors C7 and C8 discharges a negative potential is established across resistor R11 or R13, so interrupting the low-resistance circuit theough that resistor, so that the pulse is fed to the grid of triode V2A or V2B respectively. Only positive signals of substantial duration affect the anode current of valves V2A and V2B. The negative pulses fed in from valve V1 produce a negligible effect, but when the pulse is passed to the line L1, the grid of V2A or V2B goes positive producing a positive pulse in the anode current. The outputs of these valves are taken from their cathodes through terminals 101 and 102 (Fig. 5) and 103 and 104 (Fig. 6). Two triodes V4A and V4B have their grids connected through a capacitor C110 and their cathodes through a selected one of the resistors R19, R20 and R21 which serve to smooth the incoming signals. Their anodes are connected through the coil 30 of a moving coil relay (25 in Fig. 1) and a meter M1 (27 in Fig. 1). Connected in parallel with the relay coil are a semilogarithmic potentiometer RV1 and a resistor R22, and through contacts RL2 and RL3 a shunt resistor R23. The moving coil relay has a contact RL1 movable between the stable position shown in Fig. 6 and positions in which line L3 is connected through R24 with the grid of a triode V5A or V6A. It takes up the central position when no current is flowing in coil 30, and one or other positions when a current is flowing, according to the direction of that current. The triodes V5A and V6A are connected in parallel with triodes V5B and V6B respectively, and the two pairs form a flip-flop circuit. When the contact RL1 is in the position shown, the network of resistors R25-32, R49 and R51 holds the grids of triodes V5A and V6A sufficiently positive for the vaves to be conducting. When the contact moves to the left or right, the grid of triode V5A or V6A goes positive, and triode V5B or V6B becomes conducting and a current flows through relay coil 31 or 32. These relays have contacts controlling solenoidoperated steam valves regulating the supply of steam to upper and lower cylinders in a drying apparatus. These coils also have normally closed contacts RL2 and RL3 by which the shunt resistor R23 is connected across the relay coil 30. When a current flows in either of relay coils 31 and 32 the shunt path is interrrupted, and the current in coil 30 and the contact pressure of the relay contact RL1 are increased. The cathode of photo-cell PC2 (Fig. 5) is connected through a capacitor C2, with the grid of triode V3A. As the upper flash tube produces flashes of light, positive pulses are produced at the grid of triode V3A, and corresponding amplified negative pulses at the anode. These are fed to the grid of triode V3B through a small capacitor C11, the grid side of which is also connected with line L1 through a semi-conducting diode D2. The negative pulses fed to the grid of triode V3B thus pass to line L1, and subsequent recovery of the triode makes its grid positive. It then floats back to the potential of line L1 due to the leakage resistance of diode D2, The short negative pulses at the grid of triode V3B therefore produce longer positive pulses at the anode. The anode is pulled positive by line L2 through resistor R9, but prevented from rising above a predetermined potential by the diode D1 and potential divider R1/R2. The anode potential of triode V3B depends on the magnitude the signal from cell PC2 and is applied to the anodes of both photo-cells.