0
\$\begingroup\$

I'm trying to replicate a circuit that switches couple of inductive solenoid switches at ~65 V but having some occasional destructions on my MOSFETs. I also added an overcurrent protection using a hall-effect sensor, a comparator and a latch so that when an overcurrent occurs, it has to be manually reset by the MCU.

The switches consume about 1.5 A when triggered and it doesn't have to be triggered very frequently. I have 1000 V diodes across switch MOSFETs (U1, U2) but not my protection MOSFET (U3). T0 and T1 has 10 kΩ pulldowns before gate resistors.

The original circuit uses TIP122G transistors. Should I choose a MOSFET with a similar avalanche rating?

Will the voltage spike reach the protection MOSFET in this case?

Also Is it better if the protection MOSFET is located in the positive side rather than the ground side?

Do I need a snubber in addition to the diodes?

Can the overcurrent protection circuit be the culprit of the destruction since it reacts fast and instantly closes the solenoid before it settles?

AP5N10BI Datasheet

Schematic

\$\endgroup\$
3
  • \$\begingroup\$ The symbols appear to be erroneous. What is the +65V, a power supply? How long are the wires? What is the sensor? Why not a sense resistor? What is the nature of the overcurrent: how does it happen, how fast can the system respond in the worst case, what is your SOA calculation with respect to fault current and response time? Is the fault condition even due to overcurrent? How does it fail? \$\endgroup\$ Jan 25 at 15:06
  • \$\begingroup\$ From @RohatKılıç comments and your own on his answer below, the schematic in your question is wrong as it shows an incorrect FET symbol. Please can you correct that ASAP, along with any other errors you find in the schematic. Otherwise, work put into answers by site members may well be wasted, as his appears to be. Thanks. Downvoting for that reason until corrected. \$\endgroup\$
    – TonyM
    Jan 25 at 15:26
  • 1
    \$\begingroup\$ The APM datasheet also uses the wrong P-FET symbol for the AP5N10BI. \$\endgroup\$
    – Jens
    Jan 25 at 16:57

1 Answer 1

0
\$\begingroup\$

You are using P-Channel MOSFETs.

In the configuration shown the body diodes are forward-biased which basically means that the load current will flow through the MOSFET body diodes regardless of whether the MOSFETs are driven or not - although I can't see how you can drive the MOSFETs.

In the following image I tried to show the current flow of CH2:

enter image description here

The current flows through the load and completes through U2's and U3's body diodes, even the MOSFETs are off.

Anyway... Replace them with NMOS.

\$\endgroup\$
4
  • \$\begingroup\$ It's an N Channel Mosfet as you can see from the datasheet. I believe the symbol is incorrect. \$\endgroup\$
    – Berkays
    Jan 25 at 14:39
  • 1
    \$\begingroup\$ @Berkays and your diagrams contain the "incorrect" symbol which leads to my observations above. \$\endgroup\$ Jan 25 at 14:41
  • \$\begingroup\$ The circuit works without any issue under a resistive load and no current is drawn when mosfets are off. This problem only happens infrequently when using the solenoid switch and only after triggering. \$\endgroup\$
    – Berkays
    Jan 25 at 14:58
  • 3
    \$\begingroup\$ @Berkays please edit your question with the correct schematic. The first thing that one's eyes will catch is the MOSFETs and their body diode arrangement. \$\endgroup\$ Jan 25 at 15:04

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.