Problems in making dc-dc buck converter..

Discussion in 'The Projects Forum' started by CB_nitd, Jun 4, 2014.

  1. CB_nitd

    Thread Starter New Member

    Jun 4, 2014
    9
    0
    I am working on a project which requires a buck converter. But the converter seems not working above 15volts.

    I am using a 555 timer as astable multivibrator to generate rectangular waves of frequency 10khz at 62% duty ratio.

    The output of the 555 timer is fed into the gate of the mosfet(IRF840) and the drain of the mosfet is fed with input dc supply. The mosfet is getting heated which i have solved using a heat sink.when the drain voltage is 15volt the mean voltage of the buck converter which i am getting at the output resistive load is 7v(which should be 9.3V). But when i am increasing the gate voltage to 30v i am still getting output voltage of 7v. It seems that the output voltage is acting independent of the input. I am using inductor of 0.1mH and capacitor of 1.5nF and a resistive load of 196ohm.
     
  2. Experimentonomen

    Member

    Feb 16, 2011
    331
    46
    This is because the gate voltage has to be atleast 10V above the source pin voltage.

    You need a high side mosfet gate driver with a floating 12-15V power supply.

    The heating is the mosfet only partially turning on and operating within its linear region.
     
  3. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,771
    1,103
    The rated maximum gate-source voltage is 20V !!
     
  4. crutschow

    Expert

    Mar 14, 2008
    12,981
    3,221
    This is a nit but there often is confusion in the operating point definitions between a BJT and a MOSFET since some of them are opposite. When the MOSFET is not fully turned on it is said to be operating in the saturated region. When it is fully turned on it is operating in the linear or ohmic region.

    To the op: Please post your circuit.
     
  5. Experimentonomen

    Member

    Feb 16, 2011
    331
    46
    You got that backwards, when a mosfet(and igbt) is fully turned hard on its saturated while partually turned on is in the linear ohmic region.
     
  6. CB_nitd

    Thread Starter New Member

    Jun 4, 2014
    9
    0
    my circuit it like as shown in the picture... while the matlab simulation is giving correct results but i m nt getting correct result practically...

    Experimentonomen:
    ''the gate voltage has to be atleast 10V above the source pin voltage.''
    I am unable to get it?? plz elaborate...
     
  7. Experimentonomen

    Member

    Feb 16, 2011
    331
    46
    Google high side mosfet gate driver.
     
  8. DickCappels

    Moderator

    Aug 21, 2008
    2,647
    632
    The schematic is very helpful. The use of an N-Channel MOSFET in this manner is very unconventional and not recommended. What a buck converter needs is a saturated switch and what you have drawn is a source follower that is limited by its drain voltage.

    If you use a P-Channels MOSFET with the source connected to VS+ and then ground the gate to turn the MOSFET on and connect the gate to VS+ to turn it off, you will get a good approximation of an ideal switch. To wit:

    [​IMG]

    That's the basic idea. If you use a high side driver as mentioned by Experientonomen instead of the bipolar transistor and resistors you will obtain better results.
     
  9. CB_nitd

    Thread Starter New Member

    Jun 4, 2014
    9
    0
    but i am simply using the mosfet for switching operation. It is getting gate voltage around 11v which is well above it threshold voltage. then why do i need a driver?? basically i dont want my converter output to get saturated above 15v drain voltage.
     
  10. crutschow

    Expert

    Mar 14, 2008
    12,981
    3,221
    Nope, you're the one confused. Only a BJT is fully turned on in the saturated region. Read this.
     
  11. crutschow

    Expert

    Mar 14, 2008
    12,981
    3,221
    That is your matlab circuit. What does the circuit look like with the 555 timer?
     
  12. CB_nitd

    Thread Starter New Member

    Jun 4, 2014
    9
    0
    yes..now i m clear..u mean that my mosfet goes to saturated region instead of linear region and thats why my output current gets saturated..right?? can u suggest me any ways to eliminiate dis? does igbt works better in buck..because it is not possible for me to increase the gate voltage more than 15v by 555 timer but i want my converter to work at atleast 100v.
     
  13. CB_nitd

    Thread Starter New Member

    Jun 4, 2014
    9
    0
    my timer looks like this... i am connecting the 3rd pin to the gate of the mosfet...
     
  14. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,281
    1,232
    Yes, a cruel joke they played on us when they did FETs. :eek:

    CB, with the NFET when the gate voltage rises the FET will start to turn on so the source voltage will rise. There will come a point where the gate is no longer be above the threshold voltage unless you use a gate voltage 8 or 10 volts above the drain voltage. That is why Dick C. suggested a PFET. Then it is easy for the gate to be 15 volts below the source and it will turn all the way on instead of just 7 volts or so.
     
  15. Experimentonomen

    Member

    Feb 16, 2011
    331
    46
    Prollem with the P fet is that input voltage cannot exceed the max gate voltage if the control circuitry is ground referenced.

    So i prefer N channel and drive circuitry referenced to the mosfet source pin.
     
  16. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,281
    1,232
    Hmm, didn't see that 100 volts - thought it was 15. So the FET is the least of the problems.

    Vin = 100v
    Vout = 62v
    Iout= .32a
    L = 100uh
    C= 1.5 nf

    Wont work.

    Use this to pick the inductor, capacitor, frequency combination.

    http://www.daycounter.com/Calculators/Switching-Converter-Calculator.phtml

    Maybe, 40Khz., 1mh, 12Uf

    So yes it will take a high side driver or a 3 watt zener and 10 watt resistor.
     
  17. CB_nitd

    Thread Starter New Member

    Jun 4, 2014
    9
    0
    is it practically possible for 555 timer to achieve 40khz frequency. any other method without using driver? because we dnt have driver ic here n it will get some time to get one from outside as i need it urgently.
     
  18. DickCappels

    Moderator

    Aug 21, 2008
    2,647
    632
    If you insist on using a N-channel MOSFET, drive the gate through a transformer or other high-side drive circuit so you can get the gate voltage higher than the input voltage.
     
  19. CB_nitd

    Thread Starter New Member

    Jun 4, 2014
    9
    0
    finally i got IR2110 driver ic but i dnt knw how to implement it in my circuit. can anyone please take the pain of explaining or post any suitable link for me. I have googled much but couldnt reacha fruitfull result.
     
  20. BobTPH

    Active Member

    Jun 5, 2013
    782
    114
    What do you not understand about the diagram on page 1 of the datasheet?

    Bob
     
Loading...