0
\$\begingroup\$

i have a problem with driving a MOSFET in a converter, let me explain. I need to build a Buck Converter with the following requirements:

  • Input 180 V (From a 120 VRMS rectifier circuit)
  • Output 24 V
  • No Feedback, is a specific load for demostration purposes.

By designing formulas i came with the following circuit, 10 kHz switching frequency, the only change is the use of a RURG3060 as D1, proteus didn't have this model. The load is a set of 3 light incandescent bulbs in paralel, making 24 V and 120 W output. So a 5 A current is expected in the output. I simulated theorically with PSIM the circuit and the output and ripple requirements are met. Buck Converter circuit

Because of the high input voltage, i use an optocoupler 6N135 to protect the microcontroller from the circuit, and connected like this: enter image description here Here comes the problem: I notice that the NMos source pin is floating, i tried to connect the negative of the 12V from the optocoupler to the source pin of the NMos, thus, making a floating ground but that don't work, then, i connect the same negative pole to the circuit ground but i don't get any output even. Researching, i tried to connect a MOSFET Driver IC like the IR2112 to solve this problem but the MOSFET don't conmute. All of this is obtained by simulation, because of the use of High Voltage i want to make sure that the simulation works before making any PCB.

How i can drive the MOSFET then? Collegues suggested me to change to a PMos but how i can drive then? I'm not limited to use a MOSFET as switch, only to use a 6N135 as isolation between uC and the circuit. the output and load is fixed.

Any suggestion, advice, will be appreciated

Greetings

Edit1: Just to be sure, i delete the opto from the schematic and try it to float the negative PWM to source directly (5-0V PWM like a real uC) enter image description here

As well i make floating the load, both of them work. enter image description here

So, i think the problem is in the output side of the optocoupler, because i connected as i show both ground and floating referenced mosfet and don't work. I tested the output of the opto without the converter and i have a slightly slow PWM but Works, until i connected to the converter. How i can connect the output side of the opto then? I use this model because of availability and switching frequency used.

EDIT2: I finally found a way to make it work, i just connect the 6N135 as an example in his datasheet to the Q1's gate, using 5 instead of 12 V and with the floating ground. I get a working simulation and real implementation. The final circuit is shown: enter image description here

Thanks for the help!

\$\endgroup\$
  • 1
    \$\begingroup\$ in your schematic , the ground of input side is the same as the 12V ground. Is that really the case? the 12 used for optocoupler should be floating/isolated with its negative tied to source of the FET. \$\endgroup\$ – Navaro May 10 at 6:46
  • \$\begingroup\$ The N-channel MOSFET needs a gate voltage higher than the source. There are "bootstrap" ICs that will do this for you. Omit the lower MOSFET since you only need a high-side driver. Note that a load is necessary to charge the bootstrap cap. infineon.com/dgdl/… \$\endgroup\$ – Mattman944 May 10 at 7:00
  • 1
    \$\begingroup\$ Since the load is not required to be grounded, you can reference the switch to ground which would make driving it easier . As both MCU and switch have the same reference. see schematic here electronics.stackexchange.com/questions/436698/… \$\endgroup\$ – Navaro May 10 at 7:07
  • \$\begingroup\$ @Mattman944 I try simulating that IC with the circuit from the datasheet but i can't get it to work, i think my connections are bad, i connected the input PWM to HIN, with SD and LIN grounded. For the outputs, i only using the high side output, HO and VS pins, i guess the use of the low side output is needed too ¿Is that right? \$\endgroup\$ – Rafa1296 May 10 at 18:50
  • \$\begingroup\$ @Navaro, no, in fact, my first attempt was to connect the 12V negative to source as you mentioned, but that can't make the transistor switch. I will try the reference ground for the switch. \$\endgroup\$ – Rafa1296 May 10 at 18:53

Your Answer

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

Browse other questions tagged or ask your own question.