Connect with us

transformer buzz

Discussion in 'Electronic Design' started by eromlignod, Aug 19, 2003.

Scroll to continue with content
  1. eromlignod

    eromlignod Guest

    Guys:

    I am designing a device that will be used in a musical recording
    studio environment. It involves a transformer (1 kVA) that buzzes.
    The primary is 120 Vac and there are several low-voltage/high-current
    secondaries--about 2V @ 2A ea.

    Is there a standard, reliable, simple way to eliminate this buzz?

    Don
     
  2. Ian Stirling

    Ian Stirling Guest

    Vacuum impregnation with lacquer.
    Where is the buzz coming from?
     
  3. N. Thornton

    N. Thornton Guest


    Hi Don. Yes, put it outside.

    There are other approaches that can be tried with varying success:
    1. paint it to stick the core laminations together
    2. whack the core lam edges with hammer and chisen to jam them
    3. place TF on sound absorbing thing
    4. Place in labyrinthine carpeted box with a silent cooling fan

    With 1,3,4 take adequate precautions against fire risk.


    Regards, NT
     
  4. Yzordderex

    Yzordderex Guest

    You can try banging in a small piece (or pieces) of hardwood shim
    between the core and windings. Sometimes it's the windings moving
    around that creates the buzz.

    73
    Bob
    N9NEO
     
  5. Ben Bradley

    Ben Bradley Guest

    5. Use a toroid core transformer.
     
  6. Tweetldee

    Tweetldee Guest

    If the buzzing is audible, and is a result of the transformer laminations
    being loose, and if the transformer has holes in the corners of the
    laminations, it's probably an easy fix. Install machine screws in each of
    the holes (along with appropriate washers and nuts, of course), and tighten
    until the noise is under control.
    There's also a chance that the buzz is a result of the winding bobbin being
    loose. If this is the case, get a couple of thin wedges of wood and drive
    them between the core and the bobbin.
    Cheers!!
    --
    Tweetldee
    Tweetldee at att dot net (Just subsitute the appropriate characters in the
    address)

    Time is what keeps everything from happening all at once.
     
  7. Spajky

    Spajky Guest

    mount it inside a metal case on fine rubber washers & fill it with a
    very liquid sewing machine oil ... you wont hear it anymore ...

    -- Regards, SPAJKY
    - http://freeweb.siol.net/jerman55/HP/Spajky.htm
    Celly-III OC-ed,"Tualatin on BX-Slot1-MoBo!"
    E-mail AntiSpam: remove ##
     
  8. Yzordderex

    Yzordderex Guest

    Put it in a pail of beach sand
     
  9. Something doesnt add up here, unless you have 250 secondaries!

    You could have the transformer vacuum-impregnated with lacquer,
    or some more advanced thermally-conductive potting epoxies.
    Or choose a different transformer design.
     
  10. Roy McCammon

    Roy McCammon Guest

    If you are intending to operate at 60Hz,
    then try using a transformer rated for 50Hz,
    or use a primary voltage about 10% low (use
    a 140V tap on 120V for example).

    Power transformers are designed to run close
    to or maybe a little into core saturation. That
    saves metal and weight, but can cause audible
    hum. Using a lower frequency transformer or
    derating the primary voltage will give you a little
    margin. Might help.
     
  11. Tweetldee

    Tweetldee Guest

    Don, now that I understand what the objective is, let me recommend a
    different approach. Chuck the transformer and use a switching power
    supply.. They operate at frequencies that are out of the audible range, and
    will run cooler than a transformer. If you rectify the transformer outputs,
    and the end result is a DC source, then the switcher would serve your
    purpose well.
    Of course, if you need the 2V/2A at 60Hz, then you may be stuck with the
    transformer approach. Since it's obvoiusly a custom build, then you should
    be able to specify a maximum noise level from the transformer, and let the
    manufacturer worry about the design & construction details.

    Good luck on your project... sounds like it may actually work. Do you have
    a prototype yet?
    Cheers!!
    --
    Tweetldee
    Tweetldee at att dot net (Just subsitute the appropriate characters in the
    address)

    Time is what keeps everything from happening all at once.
     
  12. A toroidal transformer might be a better fit for the job.

    But a switching supply would mean that ripple on the supply would be
    much less likely to cause the strings to vibrate audibly, possibly at
    harmonics of the mains frequency.

    Best regards,
    Spehro Pefhany
     
  13. Spajky

    Spajky Guest

    it is expensive & the first one is also mineral one & sure does not
    burn before 200°C (IMHO before reaches that temp means that the house
    burned already down by some other reason...)

    -- Regards, SPAJKY
    - http://freeweb.siol.net/jerman55/HP/Spajky.htm
    Celly-III OC-ed,"Tualatin on BX-Slot1-MoBo!"
    E-mail AntiSpam: remove ##
     
  14. Active8

    Active8 Guest

    aside from fixing the buzz, this sounds like something where you'd wan't
    to control the current through the strings, but not by controlling the
    voltage that produces the current. figure as the string temp rises, so
    will the resistance and the strings aren't the same size. it might make
    the processor's job easier or at least the job of designing the
    processor.

    raise a controlled V, and I,R, and T raise. the higher R would lower the
    I and T. maybe that doesn't really mess things up. that's a physics prob
    i never considered. but if you apply a controlled constant current
    source, it stands to reason that the current controller would adjust for
    the increased R and the controller's job is easier.

    so if you haven't already thought of this...

    feel free to send money :)

    BTW, you mention a guitar pickup with a magnetic core. actually, it's a
    magnetized core with magnetic pole pieces. the magnetic vibrating string
    alters the magnetic field. but you knew that.

    good luck and brs,
    mike
     
  15. Active8

    Active8 Guest

    nyuk, nyuk, nyuk.
     
  16. You can pick up some kinds of used transformer oil really cheap. ;-)

    Best regards,
    Spehro Pefhany
     
  17. EROMLIGNOD

    EROMLIGNOD Guest

    Hi Mike:

    Actually I recently found that an ordinary ferrite-core inductor works great
    for pickups. You can permanently magnetize the core by passing it through a
    magnetizer/demagnetizer. I have found that the core only needs a very weak
    field to produce the signal I need.

    Actually, the crappier the signal the better. As the pickup gets cheaper and
    cheaper, the signal loses its overtones until you are left with an almost
    perfect fundamental sine wave, which is what I want. It would make a terrible
    guitar pickup, but then an actual guitar pickup would be terrible for what I
    want.

    I can get the inductors for $0.35 ea. and solder them to a PCB in the exact
    pattern to place them under the strings precisely. I solder my multiplexing
    chips to the same PCB.

    Thanks for the help!

    Don
     
  18. Active8

    Active8 Guest

    i just meant that magnetic and magnetized are 2 different things. iron
    is magnetic, but it isn't magnetized unless you align the magnetic
    domains in the same direction.
    makes sense. figure that the fundamental is the stronger signal and a
    pick up with a weak magnetic field won't be as sensitive.

    brs,
    mike
     
  19. Ben Bradley

    Ben Bradley Guest

    This is not only an excellent idea, but I wonder how the current
    (or power) into each string is controlled now. You need something like
    a linear device (power transistor or FET) per string, and it would
    dissipate up to half a watt, and you'd also lose substantial power
    through a bridge rectifier that would power each circuit.

    I might use an off-the-shelf off-line switching supply that gives
    10 to 25 volts (dependent on the max input voltage of the other
    regulators I use) at the needed 250 watts, then use a small
    2-amp-output switcher for each string. There seem to be lots of small
    switching regulators that will give the required output (0 to 2V at up
    to 2A) from Linear Technology, TI and such. I'd sync up the
    oscillators of all the regulators so there's no chance of audible
    difference frequencies being generated.

    To your original question, are you sure the buzz is from the
    transformer, and there's absolutely no buzz or hum from the wiring or
    the strings themselves? Running (filtered and regulated, not pulsed)
    DC through the strings would fix that.
    I also have to admit it's a pretty neat idea.
     
  20. EROMLIGNOD

    EROMLIGNOD Guest

    Hi Ben:

    Thanks for the ideas. The power is controlled with pulse-width modulation
    switching a TRIAC on and off on induvidual duty cycles. I'm kind of proud of
    the way I did it, so I'll tell you.

    I daisy-chained a row of Fairchild 74F675A sixteen-bit
    serial-in/serial-parallel-out shift registers. It's like having one giant
    224-bit shift register. It also has an output register so when the 224 bits
    have shifted into place, they can be clocked into the register and held while
    the next word is shifted in. I connect each of the bits to the control line of
    a TRIAC for each string in the piano that connects the string to the voltage
    source.

    By feeding successive words to the shift register in a cycle I can control the
    PWM duty cycles of all of the strings simultaneously using only a serial
    output, a serial clock and a store clock line. In other words, I feed 100
    numbers of 224 bits each to the shift in a repeating cycle. So if I wanted
    string #48 on the piano to be on 62% of the time I would feed a 224-bit number
    that had a "1" in the 48th place. Each number after this would also have a "1"
    in the 48th digit until I reach the 62nd number and then I would start feeding
    0's to the 48th place until the end of the cycle. In this way string #48 is
    held "on" for 62 words and "off" for the remaining 38 words, or a 62% duty
    cycle.

    I store the big words in 16-bit stacks in memory--one stack for each of the 100
    PWM words. I simply pop each stack and feed the bits to the serial line and
    then strobe the store clock to transfer the word to the output register and go
    on to the next word, and so on. What do you think?

    I don't have any trouble with the strings vibrating. They are simply
    conductors with no core, so there are no magnetostriction problems. If there
    were the wires in you house would hum!

    Don
     
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

-