Winding my own transformer

Discussion in 'General Electronics Chat' started by microelectronix, Feb 3, 2010.

  1. microelectronix

    Thread Starter Member

    Jan 21, 2008
    20
    0
    Hi all,

    So I'm making an audio amplifier, and I've certainly seen how different transformers produce different levels of noise. One I had, which would be perfect (36V output with center tap, at least 300VA) produced so much hum I could hear it in the next room. I switched it back to a smaller 25.6V / 2A one I had and it went away.

    Anyways, I have a toroidal core I want to use to wind my own transformer with for use with 120V mains.

    1) Why are toroidal cores better for audio? I know this is the rule but I don't understand why.

    2) How many turns/what gauge would this require? I want ~30V output w/ center tap, 3-4 amps would be nice.

    Thanks!
     
  2. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Toroidal cores are better because the EMF is concentrated around the core. Very little energy is lost, if the core choice is proper.

    But you haven't given nearly enough specifications to even venture a guess as to what might be appropriate.

    No specs on the toroid you have. What is it's AL value, or at least manufacturer and model number?

    What is the input impedance desired? (this is the output impedance of the amplifier)

    What is the impedance of your speaker network?
     
  3. 3ldon

    Active Member

    Jan 9, 2010
    82
    3
    Toroidal cores are better for audio (though I haven't seen very many, and they do have other issues) because there is significantly less iron loss, and therefore less distortion. Reason is, almost all electrical steel has a preferred direction for magnetic exciton, and a toroid can take advantage of that.
    Also, they have less leakage reactance, therefore better higher end response.

    Are you sure that your 36volt 300 va transformer is an audio transformer? was it saturating at lower frequencies causing the hum, or was the "hum" some multipule of 60Hz?

    As far as turns per volt for line operation... this is an unknown, as you do not know what type of iron the core is made from.
    I give it 50/50 odds that it will either saturate at 1.7T, and you can run it at 1.6T at 125Vac; or it will saturate at 1.2T, and you will need to run it at 1.1T at 125 Vac.
    10% more field strength than the "knee" in the saturation curve and you will overheat the transformer/double your losses.
    the amount of power you can get out of said transformer is related to the square of the flux density.
    If you have an ac amp meter, volt meter and a varac you can easily wind up 100 turns and plot the voltage vs current, (a watt meter really helps) and find the saturation point in the curve, back off 10 or 20 % depending on what you consider the maximum no load temperature to be, and calculate the turns accordingly.

    if you want to play it safe and just wind the transformer, then go with 1.1T and you will need about this many turns
    [​IMG]
    B is Teslas, A is square meters of core area, N is turns and f is 60Hz.
     
    Last edited by a moderator: Feb 7, 2010
  4. rjenkins

    AAC Fanatic!

    Nov 6, 2005
    1,015
    69
    I'm guessing the 'hum' is the mechanical noise from the transformer itself.
    This is commonly due to the laminations vibrating with the changing magnetic field, if the transformer was not properly resin impregnated.

    It can also come from nearby metalwork due to the magnetic leakage, however another problem may be inductive pickup on low level signal wiring in the area.


    Toroids are usually completely silent and have virtually no magnetic leakage, so less problems all round.
     
    Last edited: Feb 3, 2010
  5. t06afre

    AAC Fanatic!

    May 11, 2009
    5,939
    1,222
    Have you tried to move the noisy transformer around in respect to your circuit board. Twist it to see if noise disappear if it has some angle to your circuit board also. You could also try to box it in, with meatal.
     
  6. iulian28ti

    Member

    Dec 4, 2009
    42
    0
    Do you know what is the typical efficiency of toroidal transformers ?
     
  7. 3ldon

    Active Member

    Jan 9, 2010
    82
    3
    Small toroids in cheap sound systems run fairly cool, and only consume one-two watts at no load. full load losses are significanly higher than E core transformers, because it is difficult and not cost effective to fill the core with copper as full as E cores run at.

    Transformers scale to the 4 thirds power, so a small 50 watt toroid may hit 90%, but it is easy to hit 96 with a larger 500VA unit.

    however it is pointless to cite standard values, there are so many exceptions and cheap knockoffs.


    1) Why are toroidal cores better for audio? I know this is the rule but I don't understand why.

    Do you mean for impedance matching as in a tube output transformer or as in the power supply?

    Toroids are common in power supplies for audio amps because there is less leakage reactance, and therefore less "hum" induced into every single wire in the audio amp, and most of the time they do not need an extra kilogram or so of sheetmetal to trap said magnetic field.
     
  8. microelectronix

    Thread Starter Member

    Jan 21, 2008
    20
    0
    The transformer is for the power supply, not output.

    I salvaged it from a variac with bad windings. Its a donut of rolled steel sheet, about 2.75" OD / 1.25" ID / 1.75" tall. Back in the day I tried using it as a flyback transformer... more the less to try it, not for any use. Got a few kilovolts out of it before the transistors fried!

    Another note is that the amp is currently on a breadboard with lots of wires a few inches long, which I'm sure contributes to the hum. And its the output of the speakers, not anything mechanical. However the big transformer did also contribute the mechanical hum... that thing, even if it hadn't been noisy, was pretty big overkill for the power I'm aiming for (about 20W/channel).

    Thanks for the formula's. Unfortunately (as far as I'm aware) there aren't any classes for transformer design, aside from what professors could tell me.
     
  9. microelectronix

    Thread Starter Member

    Jan 21, 2008
    20
    0
    BTW is the "a" for core area the cross section of the toroid?
     
  10. 3ldon

    Active Member

    Jan 9, 2010
    82
    3
    yes, a is square meters.
     
  11. t06afre

    AAC Fanatic!

    May 11, 2009
    5,939
    1,222
  12. microelectronix

    Thread Starter Member

    Jan 21, 2008
    20
    0
    Yes, but WHAT area... the product of what dimensions?
     
  13. t06afre

    AAC Fanatic!

    May 11, 2009
    5,939
    1,222
    Then you get older and more experienced you will know...... I find the paper quite conclusive then it comes to specifying transformers connected to a rectifier with filter. Like your amp
     
    Last edited: Feb 4, 2010
  14. lmartinez

    Active Member

    Mar 8, 2009
    224
    6

    Utilize the following formula:
    Area of a circle is equal to: [​IMG]
     
  15. Bernard

    AAC Fanatic!

    Aug 7, 2008
    4,173
    397
    Do you know the wire size of origional Variac, or how many turns, or wattage? Turns/120= turns/V, then you can calculate the secondary turns.
     
Loading...