Power Supplies

Discussion in 'General Electronics Chat' started by richbrune, Nov 19, 2005.

  1. richbrune

    Thread Starter Senior Member

    Oct 28, 2005
    106
    0
    I'd like to know if anyone would help me find a design for a power supply that will change a 28v RMS sine wave into 12v dc. Does anyone know if there's a voltage regulator that does a lot of this in a "package"? I'm trying to avoid using too much space and creating too much heat, caused by large capacitors and resistors. I'm currently using a 7812 voltage regulator, which requires that there be no more than 30v applied to it. I can't seem to hook that up directly to the bridge rectifier without the 7812 overheating. Is the 30v max requirement of the 7812 is actually a peak value, not an RMS value? The total current draw of the circuit need be no more than 250mA, so it seems that someone would have made something off the shelf for this type arraingement but I don't know where to look,

    Thanks, Rich
     
  2. kubeek

    AAC Fanatic!

    Sep 20, 2005
    4,584
    770
    Best solution seems to be switched supply like TOPxxx, but they cost quite much.

    Or use two 78xx regulators and connect the first one´s ground through zener (about 15V or so), so on the first one you decrese the voltage for the second regulator to something like 25V.

    Kubeek
     
  3. n9352527

    AAC Fanatic!

    Oct 14, 2005
    1,198
    3
    The heat would always be the same for linear based regulator (e.g. 7812, etc.). The power dissipated is P=VI, which is the voltage drop multiplied by the current. In your case this would be close to 10W. You either need to cool your regulator sufficiently or distribute the power to 2 or more drops (e.g. 1 drop resistor + 1 regulator).

    If you want lower power dissipation you could use switch mode power supply (smps) instead. If you dont want to design the smps yourself, you could get smps in hybrid or small packages for various input voltage ranges and output voltages. Have a look at Newport components dc-dc converters, CandD technologies acquired Newport awhile ago so Newport's site might already be merged into CandD now.

    Edited: CandD has 7812 footprint compatible dc-dc converter with 15-36V input, 12V 0.4A output for around 10 bucks. But you might need to limit the input voltage because your circuit input voltage is above 36V.
     
  4. italo

    New Member

    Nov 20, 2005
    205
    1
    no matter what you use a good designed power supply can achieve a max 80-85% efficiency PWM is the way to go and forget 7812 it will make it reduntant. just design with a stable feedback loop. magnetic amplifiers are more efficient but not may people use them and i don't know why. i think is the driving output current.
     
  5. n9352527

    AAC Fanatic!

    Oct 14, 2005
    1,198
    3
    What is a magnetic amplifier?
     
  6. italo

    New Member

    Nov 20, 2005
    205
    1
    it is an old technology basically an amp witha coil and feedback. i used it once for on an ultrasonic device is actually good for power distribution where you have multiple diferent power requirement like 12v 1amp 7.5 .5a things like that.
     
  7. richbrune

    Thread Starter Senior Member

    Oct 28, 2005
    106
    0
    I'd really like to see a schematic or diagram or something, this sounds very interesting. Got a link? If you were to ask why people don't use them more, my first impression is that anything with a coil is going to take up space, make EMF noise, and use copper, which might be expensive. I'm very interested in the physics of something that evades these rules.
    Thanks for your input!

    Rich
     
  8. Brandon

    Senior Member

    Dec 14, 2004
    306
    0
    How much juice u need from the 12Vdc source anyways?

    Mmm... Maybe a 2:1 xfmr to get it to 14RMS, then the Vdrop for a regulator is much less so ur disapated heat will decrease.

    The 30v max should be listed in the specs if its RMS or signal. Regardless, 28RMS if the parts max was 30RMS is SO CLOSE to the max value, ur begging for a burn out. If its 30 peak, 28RMS is over 30 peak.

    Hmm... If you can get 2 higher V diodes (15v or so) and connect them in anti-parallel to clamp the sin of 28RMS to a lower value. It will intriduce higher frequency components, but sinces it going to be regulated, should be ok.
     
  9. mozikluv

    AAC Fanatic!

    Jan 22, 2004
    1,437
    1
    hi,

    are you amenable in using another voltage source? i agree to the suggestion of "brandon" there are lots of 12v stepdown transfo at 300ma-500ma. this way space and heat will come in, in a compromise manner. ;) the board will not be bigger than 7.5 sq. in. likewise the 7812 will not necessarily need a heatsink assuming your device uses only 250ma.

    moz
     
  10. rjenkins

    AAC Fanatic!

    Nov 6, 2005
    1,015
    69
    Have a look at the National Semiconductor 'Simple Switchers' range.

    The LM2575HVT-12 should do the job, it's a 5 pin TO-220 package and all you need with it (over a linear regulator) is a small inductor & schottky diode.
    It will give up to 1A at 12V output.

    If you use a normal bridge rec & small smoothing cap on the 28V, you will get around 45V DC at it's input which is well within the 60V limit for that device.
    Try 470 or 1000uF @ 63V.

    The input current is going to be much lower than it would with a linear regulator, so the smoothing cap can be 1/3 or 1/4 the usual size.

    You also need a cap on the output to smooth it, 100uf or larger 16V should be OK.

    As it's a high efficiency switched mode device, the regulator heat output will be quite low; around 2W @ max load at a guess, compared to 30W+ for a 7812 under the same conditions.
     
  11. richbrune

    Thread Starter Senior Member

    Oct 28, 2005
    106
    0
    Thanks for reading my inital post thoroughly! That sounds just like what I'm looking for and I'm on my way to check it out!

    Thanks again,
    Rich
     
Loading...