1
\$\begingroup\$

I am attempting to make a PCB that high-side switches an inductive load (water pump) that draws 4.6A peak. The switching scheme will not include any PWM, just basic on/off functionality.

My concern is clamping the voltage spikes from the load during turn-off. I want to avoid going into avalanche mode on the MOSFET to ensure maximum reliability as the board will be operating in temperatures of 80-90 C ambient.

The MOSFET I am considering is here and I am looking at this gate driver.

My current circuit idea is shown below:

schematic

simulate this circuit – Schematic created using CircuitLab

D2 will conduct below the maximum Vds of the MOSFET, during switch off of the load where the source voltage will go negative.

I am less confident in the behavior of D1. My intention with D1 is to conduct during shut off of the load, so that Vgs does not exceed the maximum allowable. My uncertainty stems from the gate driver - how will it behave in this scenario?

\$\endgroup\$
13
  • \$\begingroup\$ D1 just acts as a voltage clamp for the gate-source junction to tame funny business from high dV/dT working its way through parasitics. It is not always required. \$\endgroup\$
    – DKNguyen
    Jul 7, 2019 at 3:03
  • \$\begingroup\$ Regarding the gate driver, what parameter will tell me the most negative source voltage that it can maintain Vgs of 0V for? \$\endgroup\$
    – user202237
    Jul 7, 2019 at 3:05
  • \$\begingroup\$ I don't understand your question. \$\endgroup\$
    – DKNguyen
    Jul 7, 2019 at 3:08
  • \$\begingroup\$ assets.nexperia.com/documents/application-note/AN90001.pdf page 11 of this doc states "Vs will go negative until current flow is maintained, however the source is tied firmly to the gate by the driver which is operating at 0 V." \$\endgroup\$
    – user202237
    Jul 7, 2019 at 3:11
  • \$\begingroup\$ Are you just asking what the gate driver's minimum Vs is under absolute maximum rating in the datasheet? Or Note 1 at the bottom of the page? \$\endgroup\$
    – DKNguyen
    Jul 7, 2019 at 3:14

3 Answers 3

Reset to default
2
\$\begingroup\$

L1 should have a freewheel diode across it .Any fast diode will work .It is not good to allow negative backswing on the fragile highside driver chip .The freewheel diode also ensures that M1 will not see much more than 12VDC .

\$\endgroup\$
1
\$\begingroup\$

This is all the protection you need

schematic

simulate this circuit – Schematic created using CircuitLab

D_Freewheel prevents the output of the inductor from going more than 1v below ground. It's got to be able to handle the full current that the motor takes. Even 'slow' diodes like 1N540x are quite sufficient, they're only slow on turnoff, they turn on quickly enough for this application.

You'll notice I've drawn it directly across the inductor terminals. This is intended to show that I'm minimising the area enclosed by the inductor-diode loop. If you grounded the diode at some distance from the motor ground, the excess inductance comprised of that loop may still give you trouble.

D2 was completely superfluous.

D1 is a reasonable precaution to have anyway, FET gates are fragile things.

\$\endgroup\$
1
  • \$\begingroup\$ I have long leads going from the pcb to the motor so this isn't possible for my case \$\endgroup\$
    – user202237
    Jul 7, 2019 at 10:49
0
\$\begingroup\$

I found a solution better suited to my application: https://www.nxp.com/docs/en/data-sheet/MC33981.pdf

\$\endgroup\$
1
  • \$\begingroup\$ That should work as it includes a built-in charge pump along with the bootstrap cap so you can have 100% duty cycle operation. \$\endgroup\$
    – John D
    Jul 7, 2019 at 18:53

Your Answer

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