Doable buck converter design ?

Discussion in 'General Electronics Chat' started by hobby16, Oct 13, 2010.

  1. hobby16

    Thread Starter Active Member

    Aug 30, 2010
    30
    4
    Hi all,
    I've got an idea of a simplified buck converter I want to design for the charging of a high amp lead battery (first 1A for testings then 10A and probably 30A), V1 in the schematics below.

    Instead of a usual high side switch, I want to use a low side Mosfet (M1), the stepped down voltage is generated with a fixed frequency PWM (uC out) and monitored with 2 ADC inputs of a uC. Since I don't need to have the battery grounded, I want to avoid the unecessary hassle of commuting a high side mosfet.
    [​IMG]

    I would like to know if the topology principle is OK and/or if such design already exists. Any suggestion and critique is welcome.
    Thank you in advance.
     
  2. schmitt trigger

    Active Member

    Jul 12, 2010
    36
    7
    I've seen low-side-switch buck converters in the past. It can be done.

    As long as you can accurately solve sensing the battery voltage from a positive common, it is doable. I assume that you are doing precisely that with the two resistor dividers and then substracting (in the microcontroler) In2 from In1. Your resistor dividers would have to be identical, though, meaning the use of 0.1% tolerance resistors.

    Another option would be to use an instrumentation amplifier.
     
  3. hobby16

    Thread Starter Active Member

    Aug 30, 2010
    30
    4
    Thank you for confirming it exists. It was so simple I wondered if it works and why it's not seen more often :D
    As to the voltage feedback to adjust the uC's PWM ratio, I've done it before. Since the load is nearly constant (a batterie), it's quite easy to do the software feedback loop.

    BTW, I don't see the point of 0,1% tolerance resistors. The resistor dividers are only needed because the input range of the ADC is 2.56V (uC's internal voltage reference). The ouput voltage accuracy is the resistors' accuracy and 1%, even 5% resistors are enough for me.
     
  4. schmitt trigger

    Active Member

    Jul 12, 2010
    36
    7
    You require high accuracy resistors NOT to set your output voltage, but to reject the common-mode voltage.
     
  5. timrobbins

    Active Member

    Aug 29, 2009
    318
    16
    You may also want to consider some reverse current protection from the battery - typically using a series diode - not needed if you have fusing or the battery circuit and a nice protected and controlled 'situation'.

    You will also need to start deploying standard switchmode power conversion techniques - such as local capacitor filtering of the raw input supply, and judicious selection and layout of parts and drive circuit.
     
Loading...