0
\$\begingroup\$

My appplication require mosfets to switch between two power sources from adapter to batter source and battery to adapter source. There is a confusion in my mind relating to turn on / turn off time of mosfets. Turn on time of mosfets is significantly smaller then turn off time due to its physics. My questions is if a lower mosfer is not switched off and upper mosfet gets switched on then current might flow from first power source (adapter) to second source (battery) through lower mosfet.In this case circuit can malfunction, am i right. :)

Regards;

\$\endgroup\$
5
  • \$\begingroup\$ Why Mosfets. They are not fully off with no input control so current can flow with dangerous results in your application. Yes they can operate like low voltage drop diodes but your control needs to be very precise. \$\endgroup\$
    – kingchris
    Commented Apr 1, 2016 at 6:23
  • \$\begingroup\$ Why not use something designed for the task (you will need to add external MOSFETS): linear.com/products/powerpath_controllers_*_ideal_diodes \$\endgroup\$ Commented Apr 1, 2016 at 6:38
  • \$\begingroup\$ Yes, application required low drop out. Yes, control cct is there but even if turn off of mosfet is larger then on time then current from upper mosfet of adapter might still flow to lower mosfet of battery. \$\endgroup\$ Commented Apr 1, 2016 at 6:56
  • \$\begingroup\$ @peter smitch, yes it looks that i have to go with that \$\endgroup\$ Commented Apr 1, 2016 at 6:57
  • \$\begingroup\$ one thing, how frequently u switch b/w rails. Battery generally used in case power off. they have lots of time in between (Turn OFF to ON or vice versa). Now u should make sure about the current paths in all cases. Use Back-Back MOSFETs in each leg. Linear technology is master in this. U can use LT4416,LT4412,LT4417 etc. TPS2110,2111 etc. Check this link powerelectronics.com/discrete-power-semis/… \$\endgroup\$
    – user19579
    Commented Apr 1, 2016 at 9:39

1 Answer 1

1
\$\begingroup\$

Normally the on/off switching times of a MOSFET are largely determined by how fast you can charge /discharge the gate. So you need a strong driver circuit. Then the "physical reasons" why switching off takes longer become insignificant. I mean, this is in the order of nano seconds and only relevant in certain applications !

Depending on your circuit indeed an NMOS and PMOS in series (like in an inverter) do "short" the supply very briefly. I write short but actually it is not a short, just a low resistance for a very short time. It happens in all digital ICs all the time. That is why supply decoupling is needed to provide that short burst of current.

This effect gets more troublesome at very high switching frequencies. Switching some supply lines is usually not done at high frequencies but at extremely low frequencies, so low that you can almost neglect it.

To prevent the effect altogether you need a "break before make" circuit, the second transistor can only switch on after the first has had time to switch off. Some switching MOSFETs in switched mode power supply ICs use this principle.

\$\endgroup\$
0

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.