FR2786001A1 - Power controlling for power supply of electric illumination dim switch by allowing increase of temperature according to this conduction using smaller increase steps as long as measured temperature is below temperature threshold - Google Patents

Abstract

If a temperature exceeds a threshold value a commutation reduction of conduction is forced, to bring back the corresponding point below the threshold. It is followed by allowing increase of the temperature according to this conduction using smaller increase steps as long as the measured temperature is below the temperature threshold. An Independent claim is included for: (a) a device for regulating temperature of a switching cell

Description

"Automatic temperature drift correction device for

  The invention is in the field of devices for controlling the power of electrical apparatus, it more particularly relates to the temperature regulation of low-frequency switching cells of the semiconductor converter type such as a dimmer, etc. The cells of FIG. This type of switching may present a fault related to a temperature variation, which originates either in overloads which cause a heating of the switching cell, or in the rise

  abnormal ambient temperature of installation.

  The appearance of this defect may represent a danger to the user (fire

  for example), or can destroy the product.

  The present invention therefore aims to overcome the aforementioned drawbacks, by proposing a device and a method of automatic correction of

  temperature drift for switching cell.

  To this end, the invention proposes a method of temperature regulation of a switching cell of the semiconductor converter type, feeding a load, said cell defining a curve representing the temperature in the cell as a function of the conduction time, characterized in that it comprises steps: - measurement of the temperature in the switching cell, - comparison of this temperature to a predetermined threshold temperature Os, and, - if the measured temperature exceeds this threshold temperature Os, of decrease according to a first law of the conduction time of the switching cell, to reduce the point corresponding to the torque (conduction time, temperature) in an increasing part of the curve representing the temperature as a function of this conduction time at a point corresponding to a temperature below the threshold temperature Os, then re-increase at regular intervals according to a second lo i conduction time as long as the

  measured temperature is below the threshold temperature Os.

  According to a preferred embodiment, if the temperature curve as a function of the conduction time O = f (conduction time), reaches a maximum for a predetermined conduction time called the folding time TR, the first law of reduction of the time of T1 conduction is of the form: transition from T1 to T = 2.TR-T1, if TI> TR, transition from T1 to T = T1-P1, if T1 <TR, P1 being a time called no descent. and, after stabilization of the temperature in the cell, reduction of the T1 conduction time by P1 descent step as long as the temperature is

  always higher than the threshold temperature Os-

  and in that the second law of modification of the conduction time comprises an increase of the conduction time T1 by pitch P2

  until it reaches substantially the threshold temperature Os.

  According to one advantageous implementation, the folding time TR is worth substantially five milliseconds, the pitching step P1 is worth substantially one millisecond, and the pitch P2 is substantially equal to half the pitch of

downhill Pl.

  In this way, the point corresponding to the torque (conduction time, temperature) is brought back to the increasing part of the curve representing the

  temperature as a function of this conduction time.

  According to a preferred embodiment, the method also comprises steps of: - computation at certain times of the temperature gradient in the switching cell, - comparison of this gradient with a predetermined curve giving the

  maximum permissible temperature gradient according to the temperature in the cell.

  and if, for the temperature measured in the cell, the calculated gradient is greater than the ceiling gradient, stopping the operation of the cell of

  switchover so that the electrical device is no longer powered.

  According to a preferred embodiment, the method also comprises steps of: comparing the temperature measured in the switching cell to a predetermined cut-off temperature,, and, if the measured temperature exceeds this cut-off temperature,,

cut off the current in the load.

  This arrangement allows better protection of the device and the load. The invention also relates to a device for temperature regulation of a switching cell, comprising a switching transistor, a driver and the terminals of a supply network, characterized in that it also comprises a linear response sensor. CTN type arranged near the transistor, whose signal is transmitted to a

  microcontroller, mounted upstream of the pilot.

  According to a preferred embodiment of the invention, the microcontroller comprises a prerecorded function for regulating the switching time of the

  transistor according to the signal from the CTN sensor.

  The description and the drawings of a preferred embodiment of the invention,

  given below, will help to better understand the aims and benefits of

  the invention. It is clear that this description is given as an example, and does not

  limiting nature. In the drawings: - Figure 1 is a block diagram of the temperature control device; FIG. 2 is a representation of the temperature curve as a function of time; FIGS. 3 and 4 illustrate an example of operation of the cell of

  switching over time, when a temperature fault occurs.

  In the description that follows, reference is made to a switching cell

  low frequency semiconductor converter type, such as those used in the dimmers available to users to enable them to select any light level included in

  a predetermined range of values.

  In such a switching cell 1, a transistor 2 is arranged upstream of a load 3 (a lamp for example) to be fed on a network 4, 5. This transistor 2 has the function of interrupting the current flow in the charge 3 on a regular basis according to the mains voltage. During a half-grating period (1/100 second), the current curve in transistor 2 thus follows the mains voltage curve (sinusoid) during a portion of the voltage change from 0 to Umax when transistor 2 is closed, then switches to the open position with a current that returns to zero. The cycle is repeated at each half-period, which limits the current flowing in the load 3. The switching instant is controlled by

  a control device (driver) 6 slaved to a command of the user.

  The switching of the transistor 2 is not instantaneous, but on the contrary is slowed down, to avoid causing medium frequency disturbances that would appear in case of steep switching edge. The power dissipated in the transistor 2 in the form of heat is then equal to the sum of the power dissipated in conduction and the power dissipated in commutation, as a function of the switching time. The maximum of this power is reached at each vertex of the sinusoid (for example 5

  milliseconds after switching to zero voltage, according to semi-

driver used).

  The value reached by the temperature of the transistor 2 because of this heating, partially linked to the choice of a slow switching, can therefore be greater than a limit value. It is therefore desirable to protect the transistor 2 in temperature (for low overloads, of the order of 20 to 30%

maximum of the rated current).

  In the device according to the invention, as shown diagrammatically in FIG. 1, on which the switching transistor 2 is recognized, the load to be supplied 3, the driver 6 and the terminals 4, 5 of the power supply network. , a

  NTC type linear response sensor 7 is disposed near the transistor 1.

  The signal from this sensor CTN 7 is then processed by a microcontroller 8, of a type known to those skilled in the art, mounted upstream of the driver 6. The microcontroller 8 includes a prerecorded time regulation function.

  T1 conduction of transistor 2 as a function of the signal from sensor CTN 7.

  In the method according to the invention, the microcontroller 8 compares the temperature signal from the CTN sensor 7 (measured for example at regular intervals of seconds) with pre-recorded information in a conventional type of memory device, not detailed herein: On the other hand, a threshold temperature Os causing the regulation, on the other hand, a cut-off temperature Oc, causing the cut-off of the current in the load, finally, a curve of maximum temperature gradient.

  admissible according to the temperature in the cell.

  1 / The comparison with a predetermined threshold temperature Os allows to regulate small overloads or a rise in the installation ambient temperature. Regulation is effected by decreasing the conduction time of the cell of

switching 1.

  Knowing that the temperature curve as a function of the conduction time 0 = f (conduction time), represented in FIG. 2, reaches a maximum for a certain conduction time called the folding time TR, here equal to 5 ms, it is reduced this conduction time T1 according to the following law: Transition from T1 to T = 2.TR -T1, if TI> TR Passage from T1 to T = T1 - Pl, if TI1 <TR, P1 being a time called no

  descent and worth in this example a millisecond.

  In this way, the point corresponding to the torque (conduction time, temperature), is reduced in the increasing part of the curve representing the temperature as a function of this conduction time. For example, assuming a conduction time of 7 ms (point 9 in Figure 2), we reduce the time to

conduction at 3 ms (point 10).

  After stabilization of the temperature in the cell (for example after a new measurement of temperature in the cell carried out after 30 seconds), one decreases by step of descent Pl, for example of 1 ms, if the

  temperature is always higher than the threshold temperature Os.

  In the opposite case, the conduction time is increased by step P2, with a duration substantially equal to half of the descent step (in this case 0.5 ms), until reaching the threshold temperature Os (with a certain tolerance range). In the example of FIG. 2, the temperature measured at the instant corresponding to point 2 being greater than the threshold temperature Os, the microcontroller further reduces the conduction time to 2 ms (point 11). Then, the conduction time T1 is increased by no rise P2 (part 12 in the figure) until the temperature is in the tolerance range around the

threshold temperature Os.

  2 / the comparison with a predetermined cutoff temperature Oc, causing the interruption of the current in the load, makes it possible to take into account

  abnormal increases in ambient temperature, for example.

  For this purpose, the microcontroller 8 regularly compares the temperature measured in the switching cell with the stored cut-off temperature O 2, and, if the measured temperature is greater than this cut-off temperature Oc, the microcontroller 8 interrupts the supply of the load 3 by

  via the switching cell 1.

  3 / Comparison with a predetermined curve of maximum permissible temperature gradient as a function of the temperature in the cell allows temperature regulation with respect to high overloads (typically 50% above the nominal current), the rise in temperature being

fast in this case.

  For this purpose, the microcontroller 8 regularly calculates the temperature gradient, and, if the instantaneous variation in temperature is greater than a gradient characterized by said curve (depending on the temperature measured in the cell), the microcontroller 8 interrupts the supply of the charge 3 by

  via the switching cell 1.

  The maximum permissible gradient curve is stored in a conventional manner, and for example in the form of a parameterized function, or in the form of a series of pairs (temperature, maximum admissible gradient) between

  which the microcontroller 8 performs linear interpolations.

  This type of overload is here treated quickly enough to

  avoid any runaway or thermal destruction.

  It is clear that the temperature limit gradient at the initial point is chosen to be higher than the rise in temperature that appears at the start of operation.

load 3.

  FIGS. 3 and 4 illustrate an example of a temperature regulation sequence, for the case of a predetermined threshold temperature Os of 110 ° C.

  (line 13), and a cutoff temperature Oc of 130 C.

  The curves 14, 15 in FIG. 3 illustrate the powers dissipated by the transistor 2 in conduction and in commutation respectively, as a function of the conduction time, expressed in milliseconds. The sum of these two curves gives the total dissipated power curve 16, whose peak corresponds to a temperature in the switching cell of 100 C (point 17). These curves, strictly given by way of non-limiting example, correspond to a case of power in the load of 1000 W, with an ambient temperature of C. Another curve 18 corresponding to the case of power in the load of 1200 W, with an ambient temperature of 35 C is also illustrated. This curve 18 intersects the line 13 of threshold temperature Os for a conduction time

  slightly less than 4 ms (point 19).

  In these hypotheses, FIG. 4 illustrates the evolution of the temperature in the switching cell 1 over time. After switching on the load by the user (point 20) with an initial conduction time of 6 ms (predetermined by the microcontroller 8), the temperature rises according to the curve

  21, and passes above 110 C after a certain time (point 22).

  The microcontroller 8 then reduces the conduction time to 4 ms, according to the

  reduction law explained above.

  The temperature in the switching cell 1 then follows a curve of lower progressivity 23, and the temperature stabilizes at a point 24 of value assumed here greater than 110 C (corresponding to point 24b in FIG. 3).

  value chosen as threshold temperature Os.

  According to the law of modification of the conduction time, the microcontroller 8

  reduces the conduction time again to 3 ms.

  The temperature in the cell descends along the curve 25 to the stabilization point 26 (corresponding to point 26b in FIG. 3). According to the method of the invention, the conduction time is then gradually re-increased so much

  that the temperature remains below 11 0 0C (curve 27).

  When the temperature becomes greater than 1 10 C (point 28), the microcontroller 8 again reduces the conduction time by 1 ms, which causes the temperature drop (curve 29) and so on. The line 30 illustrates the maximum slope corresponding to the threshold temperature gradient not to be exceeded from the cell stop temperature (according to the second temperature regulation criterion). Curve 31 represents the maximum permissible temperature gradient curve as a function of the

measured temperature.

  This slope is chosen higher than the normal temperature rise rate of the switching cell 1, under temperature conditions.

normal environment.

  The originality of the principle is to compare the sensed temperature signal with prerecorded information: a) a threshold temperature Os to be regulated, b) a cut-off temperature Oc never to be exceeded, b) a maximum temperature gradient curve. of the present invention is not limited to the details of the above embodiments considered by way of example, but extends instead

  to modifications within the reach of those skilled in the art.

Claims (10)

  1. A temperature control method of a switching cell (1) of the semiconductor converter type (2), feeding a load (3), said cell defining a curve representing the temperature in the cell as a function of the time of the conduction, characterized in that it comprises steps: - measuring the temperature in the switching cell (1), - comparing this temperature to a predetermined threshold temperature Os, and, - if the measured temperature exceeds this temperature threshold (Os), of decreasing according to a first law of the conduction time of the switching cell, to bring back the point corresponding to the torque (conduction time, temperature) in an increasing part of the curve representing the temperature as a function of this time conduction at a point corresponding to a temperature below the threshold temperature (Os), then to a second increase in conduction time as long as the measured temperature   is below the threshold temperature (Os).   2. Method according to claim 1, characterized in that if the temperature curve as a function of the conduction time 0 = f (conduction time) reaches a maximum for a predetermined conduction time called the drawdown time (TR), the first law of reduction of the conduction time (T1) is of the form: transition from T1 to T = 2.TR -T1, if TI> TR, transition from T1 to T = T1 -P1, if TI <TR, P1 being a time called no descent. and, after stabilization of the temperature in the cell, decrease of the conduction time (T1) by step of descent (Pl) as long as the temperature is   always higher than the threshold temperature Os.   and in that the second law of modification of the conduction time comprises an increase of the conduction time (T1) by step of rise (P2) up to   reach substantially the threshold temperature (Os).   3. Method according to claim 2, characterized in that the time   the crimping factor (TR) is substantially five milliseconds.   4. Method according to any one of claims 2 to 3,   characterized in that the descent step (Pl) is substantially one millisecond.   5. Method according to any one of claims 2 to 4,   characterized in that the rise pitch (P2) is substantially half of the no descent (P1).   6. Process according to any one of claims 1 to 5,   characterized in that it also comprises steps of: - computation at certain times of the temperature gradient in the switching cell, - comparison of this gradient with a predetermined curve giving the maximum temperature gradient admissible according to the temperature in the cell.   and if, for the temperature measured in the cell, the calculated gradient is greater than the ceiling gradient, stopping the feed of the electrical appliance.   7. Method according to any one of claims 1 to 6,   characterized in that the method also comprises steps of: - comparing the temperature measured in the switching cell to a predetermined cut-off temperature ()), and, - if the measured temperature exceeds this cut-off temperature   (Oc), of breaking of the current in the product.   8. Device for regulating the temperature of a switching cell, comprising a switching transistor (2), a driver (6) and the terminals (4, 5) of a supply network, characterized in that also comprises a CTN type linear response sensor (7) arranged near the transistor (1), the signal of which is transmitted to a microcontroller (8), mounted upstream of the driver (6) and means for implementing the steps: - measuring the temperature in the switching cell, - comparing this temperature to a predetermined threshold temperature Os, and, - if the measured temperature exceeds this threshold temperature (os), decreasing according to a first law of time for conduction of the switching cell, to bring back the point corresponding to the torque (conduction time, temperature) in an increasing part of the curve representing the temperature as a function of this conduction time at a point corresponding to a temperature e lower than the threshold temperature (Os), then re-increase according to a second law of the conduction time both   that the measured temperature is below the threshold temperature (Os).   9. Device according to claim 8, characterized in that the microcontroller (8) comprises a prerecorded function for regulating the switching time of the transistor (2) according to the signal from the sensor CTN (7).   10. Device according to claim 9, characterized in that the microcontroller (8) compares the temperature signal from the NTC sensor (7) to pre-recorded information on the one hand, a threshold temperature (Os) causing the regulation, on the other hand, a cut-off temperature (oc), causing the current to be cut off in the load, finally, a maximum temperature gradient curve permissible as a function of the temperature in the cell.