Connect with us

Minimal SMPS regulator

Discussion in 'Electronic Design' started by [email protected], Jan 8, 2006.

Scroll to continue with content
  1. Guest

    Hi


    In response to a query yesterday, I tried to sketch from memory my old
    minimal sm reg circuit. This is a minimal parts sm regulator.

    Its asynchronous, as that saves lotsa parts. In plain english, the
    timings are all over the place.

    This version is a minimum parts switched mode regulator. Its optimised
    for build cost rather than best performance.

    I'm running from distant memory here... cross fingers:




    .. c e
    .. IN _________________ _ ___________O_O_O____________________ OUT
    .. | (_) | | |
    .. | | | L1 | |
    .. R1 > | b | ___ |
    .. > | | ZD /^\ |
    .. | | | | |
    .. | | ___ R2 > ___ +
    .. | | /^\ > ---
    .. |_________|_____ | ____________ | |
    .. | c |_ | |
    .. | (_)-+ |
    .. | e | b |
    .. | | |
    .. 0v __________________________________________________________ 0v



    Tr P_diss is not so easy to calculate, as it depends on switching slew
    rate, frequency, load, how hard it switches on and off, etc. Guess,
    allow margin, then check its not too hot.

    Output reservoir cap should be as small as is consistent with
    satisfactory operation. Increasing the cap will increase the tr's max
    current, and make any startup overshoot last longer. Smaller cap also
    means less inductance needed. Realise the max tr i is only controlled
    by L1, C, i_load and the in and out Vs, and will peak during startup
    for regs without soft start.

    Worst case V overshoot (during startup with zero load current) is apx
    V_out + 2x (Vin - Vout). IRL it'll be much less due to resistances and
    load.

    Maybe that was it, the 2 tr version that worked well. But I really cant
    remember. Just 2 Trs, 2 Rs, L & C.

    Feedback welcome...


    NT


    PS I dont remember the L details at all, so heres a diy one for the
    @home hobbeasts: 4" of 1mm insulated gardening soft iron wire can be
    folded and apx 50 or 100 of turns of enamelled copper wrapped round it
    for an inductor.
     
  2. Ken Smith

    Ken Smith Guest

    Some may remember that I posted a worse regulator's schematic a while back
    but it wasn't that much worse.


    Using a PNP for the switching device would save some losses. The way you
    have it, R1 has to be a very low value.

    You need to add some better positive feedback at the control section so
    that the thing oscillates at some reasonable frequency.

    In low volume, LM741, LM339s and etc cost the same as a single transistor.
     
  3. Ken Smith wrote...
    The same can be said for the very useful TL431 reference/regulator/
    amplifier IC, e.g., $0.08 qty 250, for On Semi parts from Allied.

    .. s d L1 _____
    .. IN -----+------+----+-, ,------+---------------+--UUUUU--+----+----- OUT
    .. | | _|_^_|_ | | | |
    .. R4 | --- | | | |
    .. | |/v | | _|_ | |
    .. +----| |_______ | __________ /_\ | |
    .. | |\ | | | | | |
    .. | |______| _|_ | | | _|_ +
    .. | | /_\ R3 | R1 ---
    .. |____________ | ______ | _____ | | | |
    .. | | _|_ | | | |
    .. | | /_\---+-- | --------+ |
    .. R5 | | | | |
    .. | | | | R2 |
    .. 0v __________________|________|______|________|_________|____|____ 0v
     
  4. Ken Smith

    Ken Smith Guest

    For better preformance add a capacitor in series with R3 as shown:

    Modified:
    Without it, the load regulation is a bit funny. The voltage rises with
    increased load.
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-