Do DC/DC converters behave like batteries in series/parallel?

Discussion in 'General Electronics Chat' started by strantor, Dec 3, 2010.

  1. strantor

    Thread Starter AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
  2. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    No.

    Unfortunately not. That would be great if they did.

    If you had two 12v to 100V converters and tried to connect them series, you would blow the input circuit of the second converter to smithereens.

    The second would be designed to have 12v input NOT 100v.
     
    strantor likes this.
  3. strantor

    Thread Starter AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
    I assume that you are talking series there, but what about parallel? Could I double my wattage capacity by putting 2 in parallel? That's the one I'm really interested in.
     
  4. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    Should be possible. You would double your current while retaining your voltage.

    That is not the same as doubling your wattage.

    In series you double voltage keeping current.

    In parallel, you double current keeping voltage.

    At 12v, 3 watts is .25A

    If you paralleled that, you would have 12v @ .5A = 6 watts
     
  5. strantor

    Thread Starter AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
    looks like a double to me! lol.

    Thanks again
     
  6. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    See what happens when you type faster than you think.

    ..Or see what happens when I type faster than I can think? :eek:


    The 'not' shouldn't be there. ;)
     
  7. tom66

    Senior Member

    May 9, 2009
    2,613
    213
    It might be possible, provided the 1kV supplies are fully isolated converters (including feedback wise); if so, tying the positive to the common of the other supply would work. However, they are probably not isolated, so you will just see a bang.
     
  8. retched

    AAC Fanatic!

    Dec 5, 2009
    5,201
    312
    True that.

    Protection diodes, like those used for solar panels, would be a requirement.

    Use diodes rated for twice the expected voltage and amperage.
     
  9. tyblu

    Member

    Nov 29, 2010
    199
    16
    The supplies would have to be well matched in terms of feedback response and set point, otherwise the fastest reacting or lowest set point unit would be taking on all changes to the load. I am currently designing a 3x parallel SMPS with a single controller to counter this...
     
  10. thatoneguy

    AAC Fanatic!

    Feb 19, 2009
    6,357
    718
    Make sure the output of the supplies explicitly states "Floating outputs", or "Floating Ground", otherwise it won't work.

    What is the application of the high voltage? It may be easier/cheaper to start with AC and use a voltage "ladder" of capacitors and diodes.
     
  11. strantor

    Thread Starter AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
    Not sure, but I think you might be the only one talking parallel. What if you put a big capacitor in parallel with your 2 (or 3) HV converters? Then would the capacitor be the one reacting to load changes? I guess probably not, or you wouldn't be designing a controller to counter it.
     
  12. strantor

    Thread Starter AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
    I'm making a HV continuity/leakage tester to monitor cables as they go through a certain phase of production. I could not find any 1KV or better DC supplies, but I agree, if I could find one it would be better. Since it's for work, time is more important than money. I am trying to buy as much off the shelf stuff as possible and avoid too much tinkering with components, as it appears to my direct supervisor as "playing." He would rather see me with a wrench in my hand.

    All I need is a 1KV 3W or better supply. AC/DC or DC/DC doesn't matter
     
  13. tom66

    Senior Member

    May 9, 2009
    2,613
    213
    There is no way to say good things about your supervisor. Seriously, designing a circuit looks like playing? Either way, consider a microwave over transformer. They give out 2kV, but if you need 1kV consider rewinding the secondary. Or, you could buy/scavenge a 230Vac trafo, and run it on 115Vac mains.

    Be very careful. The transformer is often designed to run a 600W+ microwave, so you can pull significant current on the secondary side. Remember if you harvest the transformer from an old microwave to make sure the boost capacitor is discharged. If you want DC, scavenge the diode and the capacitor from the microwave to build a half-wave rectifier.

    How accurate does it need to be? 0.5%, 5%, 10%, 20%?
     
    Last edited: Dec 3, 2010
    DBlue likes this.
  14. timrobbins

    Active Member

    Aug 29, 2009
    318
    16
    The datasheet shows that the output is not isolated - they are three terminal devices. So no, their outputs cannot be connected in series, but it is likely the outputs can be paralleled but that is not specifically stated in the datasheet (eg. email the manufacturer).
     
  15. tyblu

    Member

    Nov 29, 2010
    199
    16
    A big cap (not BFC) would definitely help. It looks like you're looking to use commercial products in series/parallel; I wouldn't recommend it; not without looking at system stability and component tolerances, first. Some DC/DC up-converters can be stacked quite easily, such as in Marx generators. They don't do parallel very well, though!
    The one I'm doing is only 19V, made for small size, so no large caps allowed.
     
    strantor likes this.
  16. timrobbins

    Active Member

    Aug 29, 2009
    318
    16
    tyblu, those modules are semi-regulated and short-circuit protected. Paralleling them is highly likely to be similar to the little Newport style dc/dc's, where you should achieve 10-20% matched load sharing, and would use an inductor or diode in series with each output feeding a common cap to accomodate the non-synchronous output ripple, and they would be un-conditionally stable .

    They would be a very convenient simple solution to the application, as the 12VDC input is regulated.

    Ciao, Tim
     
    strantor likes this.
  17. tyblu

    Member

    Nov 29, 2010
    199
    16
    'Suppose I should have looked at the datasheet beforehand: they are indeed unregulated, so stability is guaranteed*. They won't be perfectly matched, so don't expect a full 2x power rating. The manufacturer may know better.

    Good points.

    *Why the asterisk? I feel this statement needs some kind of liability clause!
     
  18. strantor

    Thread Starter AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
    Timrobbins, Am I getting what you are saying?....
    [​IMG]

    If so, Any suggestions as to what value of inductors I should try? and Capacitor value? FYI I had already planned to have a large capacitor in parallel with my power supply(s) to give a better chance of arcing to detect indirect shorts. I have not looked into what value of cap to use.
     
  19. strantor

    Thread Starter AAC Fanatic!

    Oct 3, 2010
    4,302
    1,988
    Any reccommendations on a cap size? I already planned on having a big mambajamba in there make sure It will jump across air if given the opportunity.
     
  20. timrobbins

    Active Member

    Aug 29, 2009
    318
    16
    Yes the schem is corrent. I suggest you email the manufacturer for recommendations or check their website (I haven't). It will depend on switching frequency (the inductor will need a SRF significantly higher) and output voltage, and somewhat on the module output capacitance as it will form a pi filter with the LC you are adding. As you are attenuating high frequencies - the layout of the parts and part choice is important (eg. a surface mount inductor) and surface mount ceramics (possibly in series).
     
Loading...