I'm designing a discrete 24volt high side switch circuit with Safe Operating Area (SOA) protection and fast current limit protection. The circuit uses a low RDS-on P-channel mosfet in a current limit configuration. That is, a sense resistor is in series with the source of the mosfet that will develop a voltage proportional to the load current coming out of the drain. This voltage is used to regulate the maximum current that will be delivered by the mosfet. I have set this current at 1amp. Less than one amp, the mosfet is saturated fully on, and drops very little V from source to drain so therefore dissipates very little wattage. If the load were to draw more than 1 amp then the mosfet begins to turn off and the voltage across at the drain will regulate to whatever value dependent on the resistance of the load so that only 1 amp is delivered. Worse case is a total load short, 0 ohms. Under this condition, the V across the mosfet is 24v and the current through it is 1amp - 24watts wow! It will heat up...
I will couple the thermal pad of the mosfet to thermal pad of the LM26L temperature sensor. Should the temperature reach the set point the LM26L will turn off the mosfet. If the short remains, the mosfet it will cylce between the set point and the hysteresis (turn back off) of the LM26L. This cycling is OK because this average temperature is below the SOA limit.
My question: Is it ok to couple the thermal pad of the LM26L via a copper pour to the mosfet drain thermal pad (soldered)? The mosfet drain will be wonder around 24volts - 0.65 volts max under typical load. Hopefully, the dielectric insulation between the LM26L thermal pad and its' other pins is up for the task.