I know MOSFETs need to have a bootstrap circuit for N-channel on the high side of, say, an H-bridge; but if you're running AOK42S60L with this (as a general on-off switch, not as a fast switched component) PVI1050N x 2 (in series) @10V

Do you still need a large initial current supply to "charge" the gate capacitance fast enough for the first turn-on phase (so it doesn't spend any time in the linear region)?

  • \$\begingroup\$ You ALWAYS want a large current supply to turn on an N Channel MOSFET. The larger the better, within reason. \$\endgroup\$ – KyranF Apr 24 '15 at 23:08
  • \$\begingroup\$ Why do I need it? \$\endgroup\$ – ARMATAV Apr 25 '15 at 0:07
  • \$\begingroup\$ I'm aware you need high current for rapid switching to drive the gate capacitance as well as the bootstrap capacitor. It doesn't make sense that you would need that for a general switch. So yeah, why? \$\endgroup\$ – ARMATAV Apr 25 '15 at 0:10
  • \$\begingroup\$ For something that doesn't switch more than a few Hz or even just once every few minutes, you do not really need high current. But those optos doing only 10uA is a bit lame. I don't think that's really good enough, the linear region might be active for a few hundred milliseconds, which is enough for a fire to start \$\endgroup\$ – KyranF Apr 25 '15 at 0:27
  • \$\begingroup\$ Any recommendations? \$\endgroup\$ – ARMATAV Apr 25 '15 at 0:33

Figure out the switching time from the gate charge, then look at Figure 11 on the MOSFET data sheet (Safe Operating Area) and add some safety margin.

It can supply maybe 5uA, and the gate charge under high voltage conditions is a whopping 40nC. 5uA is 5uC/s so 40nC will take 8ms to switch. Let's call that 10-20ms to be safe

If it takes 10ms+ to switch, you probably need to keep the drain current well under an ampere, assuming a few hundred volts Vds, based on Figure 11.

enter image description here

These are calculated (not measured) single pulse curves, based on a Tj of 150°C and with the case held at an (generally impossible) 25°C. In other words.. very much the best possible case.

| improve this answer | |
  • \$\begingroup\$ So would it just be simpler to grab a DC wall wart at 12V and operate it with that? It's going to be passing AC in a solid state relay configuration. i.stack.imgur.com/vEOL7.jpg \$\endgroup\$ – ARMATAV Apr 25 '15 at 0:40
  • \$\begingroup\$ Alternatively, could I "stop" the h-bridge that is supplying the power by shutting down the driving circuit, then switch this on, then wait >10ms; then turn the bridge back on? \$\endgroup\$ – ARMATAV Apr 25 '15 at 0:42
  • \$\begingroup\$ It has to snap on and off very quickly to get the most out of the MOSFET. This could be accomplished with a wall wart, however it would require some circuitry to drive the gate quickly under all conditions. Something like a power supply supervisory chip controlling a MOSFET gate driver would do. \$\endgroup\$ – Spehro Pefhany Apr 25 '15 at 1:13
  • \$\begingroup\$ Alternatively would it be better if I just used a higher current, like a 10mA PV driver? Then it's basically fine when it comes to switching it on fast. \$\endgroup\$ – ARMATAV Apr 25 '15 at 1:34

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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