Buck Converter Output Stabilization problem!!!

Discussion in 'The Projects Forum' started by Mustafatarhan49, Aug 16, 2012.

  1. Mustafatarhan49

    Thread Starter Member

    Aug 16, 2012
    56
    0
    Hello Guys,

    This is the first time that I am posting to this forum. I hope I will enjoy the forum :)

    I have couple of question regarding to a project that I am working now.
    First of all, Let me describe what really I want to do.
    I am designing efficient and cost effective buck converter for the >>12V systems and uCs.

    I have these below parameters;
    Vin= 14V up to 48V
    Vout =5V ~ 3.3 Volt
    Iout = 30mA
    Fsw = 150kHz


    According to the above values I have find my output inductor value 150uF
    and for the output capacitor I have used 2200uF (I may change this capacitor, I am willing to hear your advise to which value I need to replace:) )

    I am controlling my output with the voltage controlled PID compensator and also using Comparator to change my duty cycle.
    But Unfortunately I am having a lot of problems with selection of mosfet!
    First of all, I used n-channel mosfet and It was working good (at least I thought that it works good :) ) When I change it to logic level P-Channel, I could not switch the Mosfet!
    Any suggestion to which mosfet brand or model should I Use ? Since I wanna have cheap solution, I don't want to use n channel due to its' drive circuit. Please suggest some p-channel :D

    Also for the controlling of output voltage, can you advise me some cheap controller name?
    I also would like to ask you if you can give me some application note or tutorial for the Buck converter?
     
    Last edited: Aug 16, 2012
  2. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    That last part won't work.

    Can you post a schematic of what you have built up?
     
  3. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    With only a 3.3V supply, that gives you maybe 2.5V of drive to switch the FET. It should switch but will not switch as well as it would with more gate drive voltage.

    Please post schematic.
     
  4. Mustafatarhan49

    Thread Starter Member

    Aug 16, 2012
    56
    0
    Please see the attached file.
    I actually, did not get why it may switch or not :)


    converter.png
     
  5. Mustafatarhan49

    Thread Starter Member

    Aug 16, 2012
    56
    0
    Why does this forum stay that much quite ? Any response guys? :mad:
     
  6. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    656
    To turn a PMOS device OFF, Vgs must be ≈0V. You must connect the source to +50V, so the gate voltage would have to swing from 50V when off to about 45V (with a logic level device) when ON.
     
  7. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    This looks like something from a simulator. If the fET is not switching correctly, the problem is in the gate drive. I have no idea why the gate drive's "source node" shown in the schematic seems to be referenced to the switching side of the FET (inductor side) which should be the DRAIN of a PFET. The statements Ron made:

    are correct but voltages with respect to ground not with respect to the inductor side of the FET.


    If you have a real schematic, hopefully from a circuit you built up and are trying to make work, I think we could help.
     
    Last edited: Aug 18, 2012
    Mustafatarhan49 likes this.
  8. Mustafatarhan49

    Thread Starter Member

    Aug 16, 2012
    56
    0
    Blinked! You are right. The reason of referencing source side of the mosfet was the give my pulse between Gate-Source. But since it is always 5 Volt, It will not switch my Mosfet. It is suppose to be 0 V between Gate and source to switch!


    I also would like to ask you guys, is it that important to select Vds in case of p channel Mosfet, I mean how does that effect to my design. Since I am using input voltage level up to 50V, Do I really need to have -50 Vds to switch my Mosfet?

    For the selection of the components, what Mosfet do you guys suggest me to use for this design ?
     
  9. Mustafatarhan49

    Thread Starter Member

    Aug 16, 2012
    56
    0
    Don't you think that I can find cheap controller :( ? Or do you mean something else ?
     
  10. bountyhunter

    Well-Known Member

    Sep 7, 2009
    2,498
    507
    FYI, if the FET in the drawing is a P-FET, it is installed backwards. I believe the source has to be to the input side and the drain goes to the switching (inductor) side of the circuit.
     
  11. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    656
    Have you Googled "high voltage buck converter"?
     
  12. Mustafatarhan49

    Thread Starter Member

    Aug 16, 2012
    56
    0
    Yes, I did search couple of time. But In my case I am drawing really small current (30mA). In most of the application, people use high currents.
     
  13. Mustafatarhan49

    Thread Starter Member

    Aug 16, 2012
    56
    0
    I actually, did what you said, replaced source with the drain but I am getting really small output voltage (2mv). Do you think I have some problem in the pulse?
    Please see the attached figure.


    buck forum.png
     
  14. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    656
    You are exceeding the common mode input range of the LT1721 by about 45V. In real life, it would be dead.
     
  15. Mustafatarhan49

    Thread Starter Member

    Aug 16, 2012
    56
    0
    Indeed, I will not use LT1721 regulator. I just put that regulator to reject output ripples. Thanks for your advise ;)

    By the way, I just want you to be clear about one thing for the Mosfet selection. Are the below parameters enough to select my mosfet ? I always use these below parameters to select my MOSFET. But I am not sure If I need something else.

    Temp = 50-125C
    Vgs = 50 (at least Max input voltage)
    Rds(on) = small as possible (have no clue if it has anyother effects rather than power dissipation)
    Ids = related to my output current ?
     
  16. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    656
    LT1721 is the comparator, not the regulator. The input common mode range is from ground to (vcc-1.2v). You are applying 45-50V to the -pin. The +pin is connected to VOUT. I can't find that anywhere else in the schematic.
    However, none of this matters, if you want to do this the easy way. See my next comment.
    I don't think you will find a MOSFET with Vgs(max)=50V. The highest I have seen is 30V. Most are 20V. Some are less.
    I am not an expert on buck converters. I have never built one. Most of the schematics I have seen use an N-channel MOSFET, with the drain connected to Vin and the source connected to Vout. The gate drive is typically 10 or 12V pulses, bootstrapped to the input voltage. See this note.
    See AN-978, page 22, for an example. Here is the IR2117 datasheet.
     
  17. Mustafatarhan49

    Thread Starter Member

    Aug 16, 2012
    56
    0
    Hello Again Ron,

    Thanks a lot for your ideas! It helps me too much!

    I found some results and at the end I can turn the MOSFET on and off :)
    I am getting desired voltage value after 14Vin but even though I apply more than 50V it looks still 5V (I don't know if this result is realistic? what do you think about it? )and also the output voltage varies between 4 volt and min 2.7Volt and my current value is in between 15mA up to 30mA. Please see the attached picture.
    To be sure I just posted the below picture In the one of these pictures, I tried to apply 70V to the input but it looks still fine, and I am thinking about the limits of this circuit. How will I define upper and lower input voltage limits ?

    And also Do you think the results look fine ? :)

    forum.GIF

    Now, What I want to ask apart from above discussion is, what I can use for sleep option for this circuitry (Because I don't want this concept to work all the time) and also how can I measure my efficiency value on the schematic? Is it simply output Power Divided by input power or do I need to use some other formulations ?

    Thanks a lot In advance
     
  18. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    656
    You can make circuits work in simulators that will never work in hardware. You can't build that circuit and make it work. I told you how to do this in post #16, and you totally ignored me.
    I'm done.
     
  19. Mustafatarhan49

    Thread Starter Member

    Aug 16, 2012
    56
    0
    .
    Noo! :( I did not ignore you...

    My supervisor wants me to use p-channel, otherwise I know it is better to use N-channel mosfet. I had no other option to use n-channel...
    But anyway, thanks a lot for your help...

    I will try to make this design with the method that you suggest in #16.
    I am still in the learning process, I also have not designed any buck converter, please forgive if I say something wrong...
     
  20. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    656
    You didn't say anything wrong. Why is your supervisor wanting you to use PMOS?
     
Loading...