Connect with us

12VDC to 18.5VDC

Discussion in 'Electronic Basics' started by [email protected], Aug 7, 2006.

Scroll to continue with content
  1. Guest

    I have an old laptop (366MHz) which takes in 18.5V @ 3A by external
    (switching?) power supply. (It didn't have a battery with it when I
    bought it - got it 2 years ago for about $200.)

    I recently built an external battery pack, powered by 16 Energizer 1.2V
    2500mAh NiMH AA cells, to power my laptop in lieu of the external power
    supply, and it does work. So far I have been able to run the laptop
    for about an hour and a half, and still going strong. (Decided to stop
    stress-testing at 1:15am, since I had work the next morning.) A far
    cry from paying $150 for a "new" laptop battery, when the laptop only
    cost me $200. Although, my 16 NiMH AA battery pack screams "GEEK!" at
    anyone who glances at it... ;-)

    For my next project, a DC-DC converter.

    So... I'd like to convert 12VDC from a jump-start car battery to
    18.5VDC, at 3A.

    After reading a recent thread about avoiding Maxim like the plague, I
    went over to TI's website, entered my parameters (Input V: 12VDC;
    output V: 18.5 VDC; current: 3A.)

    I found this from the search result:

    http://focus.ti.com/lit/ds/symlink/tps5430.pdf

    but I was a bit confused why this would have been suggested, since it
    seems to be a step-DOWN regulator.

    Any suggestions...?

    Thanks,

    Michael
     
  2. Guest


    Eh, no response from sci.electronics.basics; belatedly cross-posting to
    SED. Better than multiposting... ;-)

    Michael
     
  3. Rich Grise

    Rich Grise Guest

    Well, for one thing, don't power it from a "car" battery, unless the
    battery is on a continuous float charge. And _especially_ don't use a
    jump-starting battery. They're designed to provide hundreds of amps,
    for a minute or so. Use a golf-cart battery, wheelchair battery, boat
    battery, whatever - the operative term would be "deep-cycle" or "deep
    discharge".

    I can't adivse you on a boost regulator, but I do think it seems simpler
    to just use 24V of batteries, and regulate it down with a linear regulator
    on a mongo heat sink.

    Or did you want this thing to be portable? ;-)

    Good Luck!
    Rich
     
  4. Guest


    Well, portability was kind of the idea, since I already have the 110
    AC-to-DC switching power supply... ;-)

    I was planning on using the jump start battery, without using my
    square-wave inverter (Vector 75W I think is what it is) to power the
    switching power supply, and exchange this extra bulk with a homebuilt
    DC-DC converter.

    So a jump-start battery isn't a deep-cycle, eh? Thanks for the info...

    Michael
     
  5. Joerg

    Joerg Guest

    Hello Michael,


    Now don't show up at airport security with that concoction :)


    Look at the LM3478 from National. Nice chip, have used it myself but
    AFAIR it only comes in the teeny MSOP package. Besides some mundane
    parts around it you'd need a stiff N-channel FET, a beefy inductor and a
    fast diode that can handle the amps. Since it is usually operated at
    several hundred kHz your inductor won't have to be huge.
     
  6. Joerg

    Joerg Guest

    Hello Michael,
    At Walmart I saw a deep-cycle 12V last week for around $65, about the
    size and capacity of a car battery.
     
  7. Guest

    You could use an isolated 12 V to 6 V DC-DC converter and add the
    output to your 12 volts from the battery. You could probably run it
    with a full square wave without regulation to simplify feedback. Just a
    high frequency transformer driven by something like a UC3526 and a
    couple MOSFETs, and a simple rectifier and filter on the isolated
    output in series with the battery voltage to get what you need. You
    could use the PWM to make it regulated if you want.

    Paul
     
  8. Guest

    Oh, man! It's a pretty straightforward concoction - two 8-AA battery
    holders, each with a 9V clip on top. I could disassemble it, but I
    wouldn't want the clip on top to get shorted by any bare metal - could
    cause a fire. The two 9V clips are soldered to a "C"-style Adaptaplug
    from Radioshack.

    http://www.radioshack.com/product/i...daptaplug+c&kw=adaptaplug+c&parentPage=search

    I guess I could take apart the 16 AA's... I'd feel like a guerilla
    soldier, though.

    Then again, while I would be happy as a clam for maybe an hour or so on
    the plane, once the batteries died, recharging would be problematic...

    Hmm... would operating at that frequency cause bad things to happen to
    the pilot's collision avoidance system, etc.? Encase it in aluminum
    foil...?

    Thanks,

    Michael
     
  9. Guest


    Oh, that's elegant! Thanks!

    Michael
     
  10. Joerg

    Joerg Guest

    Hello Michael,
    .... and then carry them in a leather ammo belt with an NRA sticker :)
    You aren't supposed to let any more leak out than the FCC alloweth.

    But the same goes for use elsewhere. If you build something and then it
    interferes more than permitted it could lead to problems.
     
  11. Guest


    Say... how would I get the 6V DC isolated? I was thinking of using a
    transformer driven by a PWM circuit I built awhile back and just happen
    to have in a box, but where could I get a [email protected] (primary) to
    [email protected] (secondary) transformer? Jameco doesn't seem to carry such a
    beast...

    Michael
     
  12. Tam/WB2TT

    Tam/WB2TT Guest

    The thing that TI came up with is a buck regulator. You want a boost, or a
    buck/boost. The boost regulator looks almost the same, but the inductor is
    connected in a different way. The same IC may work, but you would be better
    off with one that shows the boost use on the data sheet. You might try the
    TI configurator again, or look at National or Linear. If you Google on buck
    boost regulator, you will get gobs of hits.

    Tam
     
  13. You will probably need to "roll your own", but it should really be nothing
    much more than a small toroid about 1" diameter, wound with a couple dozen
    turns of wire, if you use about 50 kHz PWM. #22 to #26 AWG wire should be
    OK for 3 amps or so. Otherwise, you could probably use parts of an old
    computer power supply. Of course, you can just buy a DC-DC converter, but
    18 watts will probably cost at least $50.

    Paul
     
  14. Guest

    Paul E. Schoen wrote:

    ....

    Oh, thanks! Good idea... roll my own... what kind of torus... powdered
    ferrite?

    Where could I get more detailed information on the gauge of wire,
    versus current... I'd like to build in a safety factor, so, if I want
    3A I'd like it to be able to handle 4A without getting too hot.

    Thanks again,

    Michael
     
  15. Guest

    Paul E. Schoen wrote:

    ....

    My PWM controller is basically this:

    Probably a silly question to ask, but just making sure: should I make
    C1 equal to 200 (or the more commonly available 220) pF to get 50 kHz?
    Any special type of cap, or would a simple ceramic do?

    Thanks,

    Michael
     
  16. Guest

    silly, forgot the link:
    http://www.cpemma.co.uk/555pwm.html

    (sleep deprived - woke up 4:30am this morning to fly out to LA for a
    meeting.)
     
  17. The 555 will handle 50 kHz, but 200 pF is marginal. It would probably be
    best to change the pot to 10k, and then use a 2 nF capacitor for C1.
    Ceramic or film types are OK.

    You cannot just put a transformer in place of the motor, because you will
    have DC which will saturate it. The easiest method is a push-pull design
    with a center tapped primary, and the best IC is a regulating PWM
    controller such as SG3526, UC3526, or TL494, which is designed to drive two
    MOSFETs or bipolar transistors. It has provisions for soft start,
    overcurrent shutdown, dead time, and voltage regulation. They are used in
    many switching power supplies.

    For wire size, you can use my spreadsheet at:
    http://www.smart.net/~pstech/WireSize.xls

    It shows that #22 AWG should handle 3.8 amps. #20 allows 5.3 and #18 will
    handle 7.6 amps. These values are for a conductor in free air with
    something like 30 C temperature rise. Use the heaviest wire that will
    easily fit the core, for best efficiency.

    Good luck,

    Paul
     
  18. joseph2k

    joseph2k Guest

  19. Guest

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

-