Connect with us

Need suggestions for SMPS ICs for a DIY Lab power Suipply

Discussion in 'Electronic Design' started by hrh1818, Jun 6, 2013.

Scroll to continue with content
  1. hrh1818

    hrh1818 Guest

    I am looking for suggestions for SMPS ICs for a DIY Lab Power Supply.
    Required outputs are:

    Output 1
    Adjustable over the range of 2 to 15 VDC
    0 to 2 amps load current
    Adjustable current limiting
    18 to 24 VDC input
    Output 2
    Adjustable over the range of -2 to -15 VDC
    0 to 2 amps load current
    Adjustable current Limiting
    18 to 24 VDC input
    Output 3
    5 VDC
    0 to 5 amps load current
    Adjustable current limiting
    18 to 24 VDC input
    Output 4
    3.3 VDC
    0 to 5 amps load current
    Adjustable current limiting
    18 to 24 VDC input

    Load Regulation plus and minus 0.5%
    Line Regulation plus and minus 0.5%
    Through hole mounted ICs preferred

    I ask this question because there is an over whelming number of SMPS ICc and maybe another member of the sci.electronics.deign group has some favorite SMPS ICs for this type of application.

  2. Robert Macy

    Robert Macy Guest

    wow, 2A at 15Vdc

    If you want great performance and not have to concentrate on the
    design, I'd use SMPS followed by the variable SMPS's, followed by
    'tracking' LDO followed by the most humongous RFI/EMI filtering you've
    ever seen. That way you have efficiency and quiet. Note the tracking
    LDO means transitory inefficiency. Once you set the Vdc, the tracking
    takes the LDO up near the value and you keep the efficiency but gain
    over 10:1 on the output noise.

    If you're going to use this as a lab supply, I'd not free run the
    SMPS, but use sync mode and then vary the frequency while watching the
    effects on the breadboard as a check to make usre the PS is NOT the
    problem. It gives you a 'handle' to work with.

    I may be preaching to the choir, but WATCH OUT FOR MAGNETIC PICKUP!
    Use small loops, copper tape shielding etc. Noise goes around all your
    filtering by either capacitive pickup [which pepole understand] or by
    magnetic pickup [which most forget about]

    I'd concentrate on the vendor:
    First, I lean towards Linear as a source of IC's that have support.
    LTC1871 use LTspice to see what's going on. Completely simulate the
    supply and store with your documentation. LTC1871 I think they have
    AppNotes that take these up to 10A.

    2nd: National has LM3668, which only goes to 1A, they're support I
    found anemic, with their online simulation. To me, I lose control of
    the design using tools provided only online.

    If you're really brave look at the TI line, I found high quality, and
    now they're going after high performance like some really knock your
    socks off LDO's.

    I hope you come back and share the project with all of us. May be post
    it on the DIY website too.
  3. hrh1818

    hrh1818 Guest

    Hello Robert,

    Thank you for sharing your experience and knowledge. However, it is not obvious to me why you recommend using a SMPS with a fixed output as a pre-regulator and connecting the inputs of the variable output SMPS to the output of the fixed SMPS. What advantages are gained by adding a pre-regulator to this DIY Lab supply.

  4. Robert Macy

    Robert Macy Guest

    vary the output of the second one.

    with the first one fixed, the second one's range of output is
    input Vac +/15%, including 'brownout'?
    input ratio 1.35
    output 2:15Vdc?
    output ratio 7.5:1

    no fixed 'input' ratio is then over 10:1
    Yes, not sure why tandem occurred to me, but, this is a lab supply,
    one of a kind, and the overall efficiency suffers with two in series,
    but blocks with simpler requirements in a system are easier to design
    seems better to have a lower ratio, you can count on what's going on.
    Especially when you get into current limiting.

    Plus, the filtering at the output won't need to get rid of AC mains
    modulation, well as much.

    When you do design your filtering, watch out for the Zseries to spike
    at 'cross-over' The damage that can do is to make you 'think' you have
    an excellent filter because the noise you look at on the scope is so
    low. But, when you attach an active load with some clock harmonics
    right on top of that Z spike; you no longer have a low Zimpedance
    looking back into your power supply, you have something like over 10
    ohms! and now you can't understand why the BBD is acting wonky. Easy
    to use LTspice to see the effect. I've often had to NOT follow the
    parts list given in vendors' AppNotes, because that effect was not
    looked at. Make your own filtering, and watch out for PCB layout
    getting in there. A little bit of magnetic field coupling can undo
    your best filtering.
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