Connect with us

Ground loop isolator recommendations?

Discussion in 'Electronic Basics' started by davidd31415, Jul 20, 2005.

Scroll to continue with content
  1. davidd31415

    davidd31415 Guest

    Hi,

    I'm trying to remove a 60Hz hum from the outputs of a 3phase amplifier
    used to control an electro-dynamic shaker. I am guessing that the hum
    is being caused by a ground loop between the amplifier controller and
    the amplifier itself (coax connection).

    Are there many differences between the brand/styles of ground loop
    isolators? How do these devices work?

    I had a Radio Shack isolator which recently seemed to be introducing
    more noise than it removed from the system. I'm hoping the isolator is
    broke and not something else in the system.

    I would like to find something of industrial durability to use; if
    anyone has recommendations please post them!

    I've been told that the noise could be introduced by other devices
    leaking current to ground, or noisy current leaked between AC mains
    transformers through capacitance. I don't quite have an understanding
    of the last reason there, but if the problem is with the building
    wiring what methods could I use to check/fix this?

    Thanks,

    David.


    Thanks,

    David.
     
  2. Jamie

    Jamie Guest

    isolation xformer?
     
  3. Chris

    Chris Guest

    Hi, David. Try to be concise, and provide complete information when
    posting with problems. It makes it easier when you've taken a minute
    to reread your post and made sure all of the information you think
    someone might need to answer your problem is there. It would have been
    very helpful to know the equipment you're working with.

    If it's any consolation, though, this is not an uncommon problem.
    Shaker tables seem to want to be bad as far as hum. Here's a list of
    checkables off the top of my head and limited experience:

    * It _did_ work right before -- something changed. It's basically
    working right now -- there's just been a change of some kind. Look at
    the changes first, particularly those which have been introduced from
    outside the system.

    * Check with the house electrician/maintenance person for any recent
    house wiring changes. That can cause this kind of problem.

    * I suppose the signal generator operates off 120VAC, and the shaker
    amplifier off 240 or 3-phase -- they're usually high power beasties.
    Look for yourself that they have a common GND, and that no GND
    connections are resistive. Look to see that L2 has not been swapped
    with L1 on the 120VAC line, and that it's not floating. If you don't
    know, or don't know how to check, ask the house electrician/maintenance
    person.

    * Try to determine what changes the problem. That's frequently a good
    indicator of where to look. Try giving your amplifier a 0V signal
    input by replacing the signal generator with a signal terminator, and
    see what happens (using proper terminating resistors, of course). This
    will tell you something very important.

    * Check _all_ GND straps, connectors, &c., on everything. Reseat all
    connectors in their sockets, check all nut/lockwasher connections by
    giving them a twist. If you see anything loose, fix it, then retest.
    A shaker table can be a high vibration environment for electronics.

    * The coax between your signal generator and the amp could be bad.
    Flexing the cable and checking with an ohmmeter isn't the best check,
    but might tell you something. Try swapping in a known good cable, and
    see if it helps. Make sure you've got the right coax.

    * If the chassis of the signal generator is not bolted to a common rack
    or otherwise tied into the system, try physically and electrically
    isolating the signal generator chassis from the rest of the system.
    (This is where your power isolation transformer comes in.) Test. Does
    it eliminate the problem, do nothing, make it better, make it worse?

    * Remove all unnecessary test equipment. Test. If something is only
    required for setup, use it and then disconnect it. Does that make a
    difference? (Note: I once worked on a shaker table where an old
    DuMont scope permanently installed in the amplifier system and used
    solely for monitoring the input signal and accelerometer feedback had
    gone flaky. The scope itself was injecting hum into the system! I'd
    be embarassed to say how many other things I checked first! ;-) )

    * While you're on the subject, check any coax going to an accelerometer
    or other system stuff, too.

    A couple of other points. The RS audio isolator (270-054) probably
    won't help you. A signal isolator might, though. If you gave more
    specific information about what you've got, you might get a more
    specific answer. A signal isolator provides transformer isolation of
    the signal. It's very dependent on the input and output impedances
    (which you didn't mention). They are also made to work over a given
    frequency range (which you also didn't mention).

    Power transformers have many turns, both on the primary and the
    secondary. These windings are in close proximity, and separated by an
    insulator. The definition of a capacitor is two conductors separated
    by an insulator. The value of the capacitance is dependent on the
    details. Look up transformer capacitance to find out more. The thing
    is, it's possible that this capacitance is injecting a 60Hz AC
    component into the power supply of the signal generator or the
    amplifier.

    Oh, yes. When all else fails, RTFM (Read The Fine Manual). It's a
    free education, and the answer you need just might be in there.

    Happy hunting
    Chris
     
  4. John Larkin

    John Larkin Guest


    Here's the good stuff:

    http://www.allenavionics.com/VHumElim/VHumElim.htm


    The classic loop killer is just a lot of turns of coax wound on a
    high-mu transformer core, with response down to DC and no actual
    electrical isolation. This adds a lot of common-mode impedance but
    doesn't affect the normal-mode signal path.

    John
     
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

-