confirming that supply is isolated from main ground

Discussion in 'General Electronics Chat' started by bob800, Dec 31, 2013.

  1. bob800

    Thread Starter Member

    Dec 10, 2011
    I purchased a 240/480 to 120/240 distribution transformer, hoping it would serve as a cheap yet robust isolation transformer when connected in reverse (of course raising the voltage to 240V). Several transformer manufacturers claim that these may be safely reverse connected, with only minor consequences (larger inrush currents and slight voltage drops).

    However, when I measure the voltage with my digital multimeter from mains ground to the output, I measure voltages of 120V and 94V depending on which output wire I use. Does this indicate that the supply is actually unisolated? My transformer has a shielded ground plane in the middle of the core--this shouldn't make a difference in terms of isolation, correct? I get the same result when I leave the center plane floating.

    As a comparison I measured the voltage between GND and a rewound microwave transformer that I knew was isolated--this measured around 20V.

    I would really appreciate if anyone could explain where these voltages come from and if my distribution transformer setup is indeed providing isolated power.
  2. MikeML

    AAC Fanatic!

    Oct 2, 2009
    What does Ohmming between the primary and secondary show on the unpowered transformer?
    bob800 likes this.
  3. studiot

    AAC Fanatic!

    Nov 9, 2007
    This is too high a voltage (and probably current capacity) to get wrong.

    Post a diagram of all your connections, when making these measurements.
    bob800 likes this.
  4. crutschow


    Mar 14, 2008
    As Mike noted, you just need to measure the resistance between the output and input of the transformer (when unconnected of course). If the resistance is very high or unmeasurable then the transformer is isolated. The voltage you measure is likely due to capacitive coupling between windings.
    #12 and bob800 like this.
  5. MaxHeadRoom


    Jul 18, 2013
    IMO you should never be using ground as a reference or do you mean neutral?
    But measuring an apparent isolated secondary to earth ground any reading is usually due to the sensitivity of modern VOM's which is reading stray induced voltages.
    bob800 likes this.
  6. russ_hensel

    Well-Known Member

    Jan 11, 2009
    Connect first one side of the output then the other to a ground through a high value resistor. Measure the voltage drop across the resistor.

    Be careful, a slip on the probes could be dangerous, might want to use clips and keep your hands off.
    bob800 likes this.
  7. inwo

    Well-Known Member

    Nov 7, 2013
    Nothing to do with your question, but there may be other connection possibilities more suited to your purpose.

    120 X 120

    Or even 120 X 120 X 240 X 480

    How will this be used?

    Is it X1-X2-X3-X4
    and H1-H2-H3-H4

    Control transformer or distribution?
    VA/KVA rating?
    bob800 likes this.
  8. bob800

    Thread Starter Member

    Dec 10, 2011
    I attached a diagram (please excuse the MS Paint work...)

    My meter gives no reading when measuring the resistance between primary and secondary (indicating open circuit), but I also tried russ_hensel's suggestion.

    With a 220K resistor between ground and the output, I measured voltage drops of 9.4 and 7.4 VAC. However, I also measured the current through the resistor (which might eliminate the problem of overly sensitive voltage readings) which was 0.03 mA. By ohm's law 0.00003 * 220000 = 6.6V, so I assume this just due to capacitive coupling as crutschow said.

    Would it be safe to assume at this point that probing circuits powered by this transformer with an oscilloscope would not cause issues? I have the transformer connected to a variac so I can do initial testing at safer voltages.
    • xfmr.png
      File size:
      274.1 KB
  9. #12


    Nov 30, 2010
    This kind of transformer will protect the oscilloscope from an earth ground on the secondary side. Just be sure you don't install an earth ground on the secondary side out of habit.

    That would be the habit of connecting "bond" (the green wire) to the chassis, and then connecting the common of the circuit you're testing to the chassis or neutral.
  10. inwo

    Well-Known Member

    Nov 7, 2013
    I would suggest the opposite, although further testing is moot.
    Transformer seems fine.

    Connect a low value resistor or fuse to frame and/or 120 volt neutral.
    (For a test only)

    If transformer winding is isolated there will be no problem grounding a phase.
    If transformer has a ground fault you will lose a 10 ohm 1/4 watt resistor or fuse in a puff of smoke.

    Suggestion in my previous post would be to use 120 volt windings only for 120 to 120 volt isolation transformer.
    Plus 240 and 480 if you felt comfortable at those dangerous levels.
    Alternately use the 240 to 240 connections @ 120 volts for a current limited isolation transformer.

    Should have enough isolation for testing at any voltage you are likely to encounter.
  11. studiot

    AAC Fanatic!

    Nov 9, 2007
    That's a really good safe way to proceed.

    Thank you for posting the pic, you can see from the replies that it was really helpful to the discussion.

    MS Paint is just fine.

    I assume from the red lines that you have earthed the electrostatic screen so I guess that the previous comments about induced pickup are correct.

    Incidentally the current policy at AAC is against mains+ voltages so this thread is an object lesson in safety and how to discuss it.