The following circuit has been designed to drive a contactor. In addition to the primary relay the contactor has a secondary auxiliary (AUX) relay that can be used to latch the contactor on.

Circuit Diagram for Contactor Drive

The circuit is operated by microcontroller that drives the gate of the MOSFET for 500 ms. This powers the contactor coil which closes both relays of the contactor. Once the AUX relay closes the MOSFET is bypassed and the contactor is latched on. A normally closed button connected in series with the AUX relay can be used to de-latch the contactor.

A failure regularly occurs where the MOSFET fails short preventing the contactor from ever being disengaged. The MOSFET has been replaced several times, so this is not an isolated issue.

What could be causing this?

In the next schematic I have added TVS diodes to the gate, drain and source of the MOSFET, but this has not prevented the issue. TVS Diodes Added to Contactor Drive Circuit

  • 1
    \$\begingroup\$ Link the data sheets for the TVS diodes you tried and, link the data sheet of the MOSFET that fails please. \$\endgroup\$
    – Andy aka
    Nov 21, 2023 at 11:30
  • \$\begingroup\$ With the time constant of a 10k resistor and 2u2 capacitor, I’d suggest you are exceeding the SOA of the mosfet. Use pwm if you want to activate the contactor gently. \$\endgroup\$
    – Kartman
    Nov 21, 2023 at 12:18

1 Answer 1


It's probably exceeding the "Safe Operation Area (SOA)" of the Mosfet. Have a look at the last graph on page 4 of the datasheet. Assuming the "Contactor En" port signal change is fast enough, the R2 and C1 make around 20 ms delay in turning on the Mosfet. That is really slow. If the current of the load is more than 200 to 300 mA (not easy to get the exact number out of that graph), it exceeds the SOA.

  • \$\begingroup\$ There's no trouble with using 5 V on the gate of this part. (Here's the datasheet.) It's not the threshold voltage alone that matters; better for this kind of application to look at the Rdson specification (page 2, near the bottom of the "static" section), which shows a maximum on-resistance of 42.5 mΩ with 4.5 V on the gate. That's certainly low enough for me to consider the part fully on. \$\endgroup\$
    – Hearth
    Nov 21, 2023 at 13:43
  • \$\begingroup\$ Right. The 5v is always bigger than the maximum of the threshold. I removed that from the answer. Thanks. \$\endgroup\$
    – Saadat
    Nov 21, 2023 at 14:14
  • 1
    \$\begingroup\$ "bigger than the maximum threshold" does not guarantee that it will be fully on at 5 V. You need to look at other specifications to determine that; threshold voltage is nigh useless for determining what voltage you need to drive the FET fully on. \$\endgroup\$
    – Hearth
    Nov 21, 2023 at 14:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.