Power supply inrush current modeling and limiting

Discussion in 'The Projects Forum' started by hrs, Nov 27, 2015.

  1. hrs

    Thread Starter Member

    Jun 13, 2014
    82
    7
    Hi,

    I want to make a 230VAC to 25VDC power supply using a 160VA transformer to drive four LM1875 audio amplifiers. They are rated 20W at 25VDC. While I have read here and elsewhere that you shouldn't worry about inrush current below 500VA, just use a proper bridge, fuse, etc. I wanted to look at it anyway to get a more quantitative feel for it than "don't worry". Also, I want to add some voltage regulators to the supply (L7818 and L7918), can inrush current be a problem for those or will the current just rush past them?

    The transformer primary measures around 8 ohms, the secondaries around 0.3 ohms each. I have no idea what the actual inductances are at 50Hz so I just picked 163H and 1H to get from 230VAC to 25VDC in my spice model. This yields a peak current of 33A which is about 10 times the current for normal duty. Results and model(Draft1.asc) are attached. Is the model reasonable or is it completely of the mark?

    I found a DC-DC soft-starter here and fiddled with it to see if it might work with the pulsed current from the rectifier. Results and model are also attached. Do you think this would work or is it an over-complicated waste of time?

    People tend to use a power resistor and a relay on the primary side but I would like to avoid working on the primary side as much as possible for now. Some modeling of a resistor between secondary and capacitors suggests that this would work fine too. Is there any reason why this is/seems to be uncommon?
     
  2. eetech00

    Active Member

    Jun 8, 2013
    650
    112
    Hi

    In LTspice, transformer inductance is "the square of the turns ratio".
    So...

    230/25 = 9.2:1 (turns ratio)
    9.2^2 = 84.64:1 (inductance ratio)

    pri = 84.64, sec1=1, sec2=1
    For sim estimation purposes, the unit chosen should present a low impedence at 50hz.
    So, don't use uH for example.
     
  3. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,292
    1,255
    ="hrs, post: 927079, member: 241581"]Hi,

    The inrush is from charging the caps, so the regulators don't know anything about it.

    Seems reasonable

    Waste of time I think. The inrush is only one cycle - not much different than turning on a large light bulb.

    Putting it there increases the ripple voltage. If you add a load to your model you will be able to see the problem. So what ever you do needs to be switched out after the caps are charged.[/QUOTE]
     
  4. crutschow

    Expert

    Mar 14, 2008
    13,042
    3,243
    A negative resistance thermistor is often used as a current surge suppressor.
    It has a relatively high impedance to limit surge current when cold and a low impedance when it warms up due to normal load current to minimize voltage drop.
     
  5. hrs

    Thread Starter Member

    Jun 13, 2014
    82
    7
    In case of an NTC wouldn't you need to deal with the possibility of switching the supply off and on again in quick succession or does an NTC cool very quickly?

    Thanks everyone for your responses.
     
  6. crutschow

    Expert

    Mar 14, 2008
    13,042
    3,243
    It would take several seconds for the NTC to cool down so, if you turned it back on before that cool down, there would be a higher surge current.
    If that's a real problem, then you would need to go to some type of delayed switched series resistor approach that would reset immediately when the power is removed.
     
  7. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,564
    2,379
    160va seems rather minor to worry about inrush, if it was 1Kva then I would say Maybe.
    Max.
     
  8. hrs

    Thread Starter Member

    Jun 13, 2014
    82
    7
    Yes, I will just be using the recommended ratings for fuses etc and deal with inrush current in the unlikely event that I need to. But I thought it was interesting to look at the numbers.

    The only problem might be when you have a temporary outage and all mains connected devices simultaneously switch back on. Then you would have a massive surge! But there's not much you can do about that hypothetical problem I guess.
     
  9. MaxHeadRoom

    Expert

    Jul 18, 2013
    10,564
    2,379
    Although 160va would be rather insignificant in the event of a 'Massive Surge'.;)
    Max.
     
Loading...