OPA541: Dammage of OPA541AP with inductive load when the output goes quickly from -Vs to +Vs

Hello Gilbert,

Sorry you are having problems with your OPA541 application. Driving an inductive load can be much more thermally challenging compared to a when a purely resistive load is being used. The OPA541 datasheet in the 8.1.3 Safe Operating Area section states, "Reactive or EMF-generating loads such as DC motors can present difficult SOA requirements. With a purely reactive load, output voltage and load current are 90° out of phase. Thus, peak output current occurs when the output voltage is zero and the voltage across the conducting transistor is equal to the full power supply voltage." It is difficult to know at this point if the failure is due to a thermal condition, or something else.

You have established that the OPA541 input sense is being destroyed. Does that mean that the input circuits of the OPA541 are being damaged, or is this a broader term meaning the OPA541 used in the sense function is being damaged?

There are two common causes for the failure of power Op amps; thermal induced failure of the output transistors, and electrical overstress (EOS) of one of the pin functions. We would need to know more about the load being used in the application, the input signal characteristics and more details about the output waveform as it slews from -Vs to +Vs. Does the damage occur to the OPA541 if the output slew is from +Vs to -Vs?

You should be able to mount several OPA541 Op amps on the a common heatsink providing they are all operated in like conditions. I assume your OPA541 devices are in the 11-pin TO-220 KV package, in which case the package thermal tab is at an internal potential of -Vs. Therefore, if the TO-220 packages are not electrically insulated from the heatsink, the heatsink must be connected to the -Vs potential, not ground. Please check how you have the tabs connected to the heatsink.

Regards, Thomas

Precision Amplifiers Applications Engineering 

Hello Thomas,

In fact if I send the signal to the OPA541, it has no reaction. the output gives always 0,009 A, that's why we thaught it's the input sense that have been destroyed but it's not sure.

The damage occurs at a dirac response of the output of OPA541 from +Vs to -Vs (or -Vs to +Vs). If we make a pulse with a slope from +Vs to -Vs there is no damage.

The charge is inductive 194µH 0,9ohms( with small capacitors for cables).

Yes we are using 11-pin TO-220 KV package. The 6 thermal tabs are electrically connected to the heatsink but the heat sink is not connected to -Vs, it's connected to nothing.

Have you an idea of a solution to prevent the damage (protective component) ?

From my side : i will change the heatsink to decrease the temperature to 70° and use a slope instead of dirac pulse to make the commutation.

Gilbert