Buck converter (18.4V to 12V)

Discussion in 'The Projects Forum' started by viper550i, Nov 10, 2009.

  1. viper550i

    Thread Starter New Member

    Oct 30, 2009
    Hi everyone
    I need help in my simulation of a buck converter in Proteus ISIS
    I did the same schematic in PCIM software and its works fine.... but i need it in proteus so i can control the Mosfet using Microcontroller

    see the attached image
    its oscillating between 1.2V and 1.4 V instead of giving me 12V :confused::confused:
    • Buck.jpg
      File size:
      94.3 KB
  2. SgtWookie


    Jul 17, 2007
    You're using a TC4427 1.5A MOSFET driver, but this is not a high-side driver.
    The MOSFET you're using is an N-channel. In order to turn it fully ON, Vgs (voltage at the gate in respect to the source terminal) needs to be 10v. In order to turn it OFF, Vgs needs to drop to 0v.

    As the driver starts to turn on the MOSFET, the source terminal voltage (in regards to ground) rises. Since the driver can't go above Vcc/Vdd, Vgs will fall as the source terminal voltage rises.

    Take a look at something like this instead:
    TC4432; high-side MOSFET driver

    If this is merely an educational experiment, have at it. ;)
    If you're actually going to use this in an application, you will wind up using most of your uC's processing power attempting to keep the output regulated.
    In the latter case, I suggest that you look at using something like the MC34063A:

    It's a very easy to use IC that can perform as a DC-DC boost, buck, or inverter, and requires very few external components.
    Last edited: Nov 10, 2009
  3. viper550i

    Thread Starter New Member

    Oct 30, 2009
    Thank you "SgtWookie" for the advice about using MC34063A IC, but i have just to simulate the buck converter by the Proteus ISIS .

    so in my case, i should remove the MOSFET driver and replace the mosfet with other that can be fully on in 5v?

    sorry.. im still beginner in electronics things :confused:
  4. SgtWookie


    Jul 17, 2007
    See if Proteus has an IRL540, or IRLZ24.
    The "L" in the "IRL" portion usually means "logic level" for International Rectifier's components.

    Take a look at this application note:
    Refer to the schematic on the bottom of page 5; "Figure 4: Desktop CPU Power Supply" for ideas.
  5. viper550i

    Thread Starter New Member

    Oct 30, 2009
    Here what i did!
    1- I changed the MOSFET to IRL540
    2- I changed the diode since that zener diode give me a constant voltage drop , so i replaced it with other one
    3- I changed the RL resistance to 5 ohm since im going to charge a 12V battery which has low resistance

    see the attched image for my output. The graph shows the change in the current "ILOAD" and output voltage "VOUT" with repect to the duty cycly "VPWM"

    whats the problem? since i need to have a constant output voltage 12V?
  6. SgtWookie


    Jul 17, 2007
    OK, you have a few things going on.
    First, you seem to be feeding the gate of the MOSFET a 1MHz 50% duty cycle signal, of what amplitude I don't know - but it needs to go up to +18 relative to ground, or +5v relative to your maximum output voltage.
    1MHz is too fast. Bring it down to around 100kHz.

    Your load is 5 Ohms. In order to get 12v across your load, how much current do you think you will need out of the MOSFET?

    Don't forget that the MOSFET has a parameter called Rds(on); this is the resistance between the drain and source terminals when it is turned fully on.

    Since your input voltage isi 18.4v, the MOSFET Rds(ON) is unknown as of yet (you need to find a datasheet for an IRL540 and find out) and Rload is 5, what percentage of duty cycle (ON vs OFF time) will you need on the MOSFET gate to keep the average voltage across the load at 12v?

    Also, as you can see already, the load current will look like a series of triangle waves. You need to filter the output better.

    Look up "pi filter" to get an idea of what you need to do to smooth those triangle waves out.