6
\$\begingroup\$

enter image description here
(A typical circuit from FAN7390 datasheet.)

I have high-side and low-side MOSFETs to drive. I read dozens of MOSFET driver datasheets, all of them have two separate inputs for controlling high and low side MOSFETS independently.
However, my control signal is a single PWM signal. I want the high side MOSFET to turn on when the PWM signal is logic-1, and the low side MOSFET to turn on when the PWM signal is logic-0.

enter image description here

As you see in the circuit above, the HIN pin can be directly connected to the PWM, and LIN signal can be obtained by inverting the PWM by a BJT. But I can make sure that this won't cause a shoot through or any other problems.

What is the proper/best way of obtaining LIN and HIN signals from the PWM signal?

\$\endgroup\$

3 Answers 3

6
\$\begingroup\$

The reason why the drivers have two signals for the high-side and low-side MOSFETS (or other power switches) is because usually you want to first switch both sides OFF for a short period of time and then to switch the needed MOSFET ON.

Usually, the switching OFF is slower than switching ON. This way if you switch the input signals in the same time, there will be short period when both transistors will be open and the power source will be shorted.

In result, the power switches will dissipate big amount of power/heat and the efficiency of the circuit will be very low (well, or both switches will be burnt).

For low power schematics, where the switching goes faster and the resistance of the switches is higher, such "shortcuts" are not dangerous, but for high power (above several watts) such behavior is not acceptable.

So, you need to form two pulses, where the positive edge of every pulse will be little delayer in order to provide a gap, where both signals are 0.

\$\endgroup\$
0
5
\$\begingroup\$

Consider using a FAN73932 instead. It is comparable to the FAN7390 except that it takes a single-ended PWM signal and adds dead time between the high-end and low-end output signals.

\$\endgroup\$
2
  • \$\begingroup\$ FAND73932 is indeed a good suggestion. However, I'm going to purchase all of my circuit elements from the same hardware vendor (Farnel), because shipping between countries costs too much. Is there any other alternative which I can find there? \$\endgroup\$ Nov 7, 2013 at 11:07
  • \$\begingroup\$ @hkBattousai: Not sure. Check Fairchild's website. \$\endgroup\$ Nov 7, 2013 at 11:11
4
\$\begingroup\$

You could try using one of these: -

enter image description here

The trouble is with this type of driver for the FETs is that you cannot turn both FETs off; one or the other is always on. The saving grace of the 4449 is that when the PWM signal is inactive it will turn both FETs off.

The other way is this (using Schmitt trigger gates): -

enter image description here

The RC network (220R and 1nF for example at 100kHz switching frequency) will delay the rising edge from the AND gate and extend the falling edge from the OR gate. To keep polarity correct when feeding a conventional high-side N channel FET, one of the lines has to be inverted - use an exor on both lines (to preserve delays) and have one set as an inverter.

\$\endgroup\$

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.