The invention relates to a method for actuating a power switch according to the preamble of Claim 1, in particular a semiconductor power switch arranged between two energy storage devices in a wiring system of the vehicle equipped with an integrated starter generator. It also relates to a device for implementing said method according to Claim 2.
In a wiring system of the vehicle with ISG, switching processes are necessary between the energy storage devices—accumulators of different nominal voltages and capacitors (intermediate circuit capacitors, double layer capacitors)—via static frequency changers or switching regulators by means of power switchs which are carried out on the commands of a control unit. A requirement in this case is that before a switch is opened, the switch current flowing through it is brought to 0 A and that before a switch is closed, the switch voltage between its switching contacts is brought to 0V so that the switch can be actuated in a zero-power state. A switch current 0 A can be implemented, for example, by disabling an AC/DC static frequency changer or a DC/DC switching regulator and causes no problem in practice. Regulation to the 0V switch voltage, i.e. no potential difference between the poles of the (opened=non-conductive) switch, usually takes place by deliberately reversing the charge of one of the energy storage devices, for example, an intermediate circuit capacitor, since this is usually the smaller of the energy storage devices. In principle, this regulation can also be carried out by means of a static frequency changer or a switching regulator positioned between said static frequency changer and the wiring system of the vehicle.
The intermediate circuit capacitor for example has a capacity of several 10.000 μF, the double layer capacitor for example a capacity of 200 F and the accumulators a capacity of several Ah. The switch voltage to be equalized can be up to a voltage of 60V.
However, determined by the unfavorable ratio of the power of the static frequency changer (e.g. 6 kW) or the switching regulator (e.g. 1 kW) to the energy required for charge equalizing (up to 40 joules), stringent limits have also been set in practice for voltage equalizing.
If now for example, for reasons of reliability and space requirements, semiconductor switches are used as switches, the accuracy of voltage equalizing which can be achieved in this way is not sufficient.
Currents and powers occurring during normal operation require the use of components (capacitors, switches) with very low resistances. In the case of existing voltage differences, the equalizing currents are accordingly high across the switch to be closed. In extreme cases, this leads to a destruction of the semiconductor.
A limitation of the equalizing current flowing through the switch to a safe value requires a current measurement which necessitates a cost-intensive current sensor at the peak of the occurring currents. In addition, the equalizing process cannot be carried out time-optimized because in the case of an excessive switch voltage, the power loss in the switch is high which represents a further possible limitation.
The object of the invention is to create a method and a corresponding device for actuating a semiconductor power switch which functions without a cost-intensive current sensor and in the case of which the transient effect and the closed circuit condition are controlled in such a way that, even in the case of a high voltage difference at the switch, damage to the semiconductor is excluded.
This object is achieved according to the invention by means of a method in accordance with the features of Claim 1 and a device in accordance with the features of Claim 2. A method and a device for switching a semiconductor power switch are proposed in which the outstanding feature is that the resistance of the switching path of the semiconductor power switch is controlled so that the chip temperature of the semiconductor power switch does not exceed a predetermined nominal temperature, in this case, when the predetermined nominal temperature is reached, the resistance of the switching path of the semiconductor power switch is increased, which on the one hand causes the power dissipation to fall and on the other hand lowers the chip temperature as a result of the reduced power dissipation. Advantageous further developments of the invention can be taken from the subclaims.
The inventive method includes the technical theory to control the resistance of the switching path of the semiconductor power switch by a control voltage so that the temperature in the power switch (chip temperature) does not exceed a predetermined value or is held at a constant value, with the control variable acting as a control signal for generating the control voltage.
With an arrangement for executing this method, provision is made for embodying the switch as a transfer gate with special semiconductor transistors into which diodes for recording the chip temperature are integrated, and for controlling it in such a way by means of a charge pump, that the chip temperature of the transistors is regulated and can be limited to a predetermined nominal value.
Advantageously the predetermined nominal temperature lies in the operating range of the power semiconductor. On the one hand this prevents the power switch being operated above its permitted temperature and on the other hand, because it is operated in this way, prevents the lifetime of the power semiconductor being reduced.
Advantageous further developments of the invention can be taken from the subclaims.
An embodiment of the invention is explained below in greater detail on the basis of a schematic drawing. The drawings show:
The transfer gate TG consists of two MOSFET transistors Q1 and Q2 connected in series whose source connections s and gate connections g are interconnected in each case. The drain connections d serve as input E or output A of the switch. Because in the wiring system of a vehicle, the voltage differences Vdiff and the current directions at the switch can have any leading sign or any direction, two transistors or groups of transistors connected in series must be used, of blocked state of the power switch. Such an arrangement is known as the transfer gate, which performs the actual switching function.
Such a switch embodied as a transfer gate is controlled by applying a control voltage between the source connection and the gate connection. In order to reduce the control voltage, a resistor not described in greater detail in this case is provided between the gate and the source connection.
There is provision in accordance with the invention, as indicated in
In
The charge pump LP known per se (capacitors C2 to C5 and diodes D3 to D5) sets up a control voltage between the source connection and the gate connection s, g of the transfer gate (switch 2). It is supplied by a gate oscillator (logical circuit elements U1 up to U4) having an enable function. In this way, both the oscillator and the charge pump LP can be enabled and disabled by a logical control signal En (enable). The generation of this control signal En is explained further below.
By enabling the charge pump LP by means of a signal En (enable), a positive control voltage is set up between the source connection and the gate connection as a result of which switch S2 (transfer gate) accordingly becomes conductive. After the disabling process, this voltage is again reduced as a result of which switch S2 again becomes non-conductive. The enabling and disabling takes place controlled in time, i.e. by means of explicitly enabling and disabling the charge pump, the transfer gate can be kept in an similar conductive state.
This temperature recording unit consists for each transistor Q1, Q2, of a series circuit at the poles of a voltage source (which can be an existing 5V supply), consisting of a resistor R7, R8 and the temperature-sensitive diode DT1, DT2 (which corresponds to the diode D1B, D2B in
The connection point between resistor R7 and diode DT1, or resistor R8 and diode DT2, is linked in each case to the non-inverting input of a comparator K1 or K2, at the inverting input of which lies a nominal voltage VTsoll assigned to a nominal temperature Tsoll. The outputs of the two comparators K1, K2 are connected to the inputs of a first logic element NAND, of which the output is connected to the input of a second logic element NOR, to the other input of which an ON/OFF signal is fed, which will be discussed in more detail below. The enable signal En, which is fed to the gate oscillator of the charge pump. appears at the output of the second logic element NOR. The diodes DT1, DT2 for recording the chip temperatures have a negative temperature coefficient, i.e. as the chip temperature increases the flux voltage reduces monotonously at around 1.6 mV/° C. The value of the flux voltage at 25° C. is 660 mV for example.
As a result of the structure of the transfer gate one transistor is operated with reverse polarity in each case (drain-source voltage), whereas the other carries the major part of the switch voltage. The chip temperatures also develop in a correspondingly different way during the switch-on process. It is thus necessary to record the temperatures of the transistors Q1, Q2 separately and to align the regulation to the higher temperature in each case.
The following information can be taken from the tables shown below (in which high=H and low=L; an underscored reference signal means that the signal at its output is meant):
Table A: Provided the diode voltage VT1ist, VT2ist generated by the relevant chip temperature T1ist, T2ist is greater than a predetermined nominal voltage value VTsoll assigned to an increased but permitted chip temperature Tsoll, the output of the assigned comparator K1, K2 is at a high signal.
Table B: As soon as the diode voltage VT1ist, VT2ist assigned to the relevant chip temperature T1ist, T2ist falls below the predetermined nominal voltage value VTsoll, the output of the assigned comparator K1, K2 goes to a low signal and the output signal of the first logic element NAND jumps to a high signal. Table C: If the output signal of the first logic element NAND goes to a high signal, the output signal of the second logic element NOR which follows it (enable signal En) jumps to a low signal, whereby the charge pump LP stops and the transfer gate becomes non-conducting.
An ON/OFF signal can be taken from
This ON/OFF signal is identical to the signal Dis in
It can also be seen from Table C that the charge pump LP can only make the switch conducting if on the one hand the control device for it gives its permission (ON=Dis=low) and if on the other hand the output signal of the first logic element NAND signals by its low state that no chip temperature has exceeded the nominal value. The output signal En of the second logic element NOR then goes to a high level and the subsequent gate oscillator (U1 to U4,
Overall this produces a two-state controller of which the oscillator frequency and amplitude depend on the delay times of the control elements.
Number | Date | Country | Kind |
---|---|---|---|
10247111.8 | Oct 2002 | DE | national |
Filing Document | Filing Date | Country | Kind | 371c Date |
---|---|---|---|---|
PCT/DE03/03090 | 9/17/2003 | WO | 10/13/2005 |