[Boatanchors] Boatanchors Digest, Vol 141, Issue 29

[Donald Chester]

>> I bet if I plug in my isolation xformer and touch one end of the secondary
>> with one hand and touch a water pipe with the other, I'll feel something.
>> Don't know for sure but I'll try it.
>>
>> Rob
>> K5UJ

> or loss on isolation between primary and secondary...

>> You will feel nothing unless your isolation transformer is defective with an
>> internal short.
>>
>> Tom k3tvc

It depends on how the transformer is constructed.  There doesn't have to be a short. Capacitance between windings may be enough to pass a  small amount of current, even if the insulation is perfectly sound.  My house was wired in the 1930s, and none of the original outlets or light switches have a separate ground wire, just hot and neutral.  With the light switch upstairs, I can stand bare-footed on the 150 years-old wooden floor, and when I brush my finger across the brass switch plate, I don't feel any tingle, but the friction between my finger and the plate feels modulated by the 60~ a.c.

Sometimes a piece of equipment that uses a perfectly good power transformer will exhibit a "hot" enough chassis that you can feel a tingle, due to the capacitive coupling between windings of the power transformer and the core, since the primary winding is usually the one wound first, closest to the laminations.  If the "start" lead (to the bottom layer of wire wound directly over the core) happens to be connected to the a.c. mains "hot", a small a.c. potential may appear on the chassis.

Some transformers are wound symmetrically, so that each winding is split into two sections, each a mirror image to the other.  This is more common in audio transformers, but some power transformers are wound the same way.  That should balance out any capacitive coupling.

Ideally, the isolation transformer should have a symmetrically wound balanced secondary, with the mid-tap grounded both to earth and the a.c. safety ground. This would result in two 60-volt "hots" and no neutral, treating the  a.c. power mains just like a balanced audio line, greatly reducing hum and ground loop problems. In addition, this would be safer, since the greatest jolt one could get by coming in contact with one of the wires and ground at the same time would be 60 volts instead of the full 120. Although I have never seen any documentation to that effect, I have been told that U.S. Navy ships are wired that way for that exact reason — personnel safety.  Actually, the 240-volt outlets in our houses are already wired in similar fashion, with the third wire, the neutral,  solidly grounded at the meter and entrance panel. Most 240-volt appliances have a safety ground, but no need for a neutral wire. 


> When I got home from work I did an experiment. I plugged in an
> isolation transformer to a wall outlet (my line v. is up to 128 v.
> now!) and got out my Fluke 115 and measured 128 v. from one end of the
> secondary to the other. I put the fluke on the "hot" side of the
> secondary and the outlet box metal ground and measured about 82 v. I
> measured my body resistance from one hand to the other as about 1.5
> megohms. Then I held the fluke probe connected to the isolation
> transformer secondary "hot" and put my other hand on the outlet box
> and saw the v. on the meter drop from about 82 v. down to 17. But I
> felt no current through my body at all. So evidently I am mistaken
> but I don't know why I didn't feel any current when the v. dropped so
> much. Well, I guess 65 v. / 1500000 ohms is a very small amount of
> current, too small to feel. 

The current isn't zero; it's just too small to feel but your meter shows it's still there.  I'd bet if the outlet box had a smooth bare surface like the brass light switch plate does, you could feel 60 hz modulation of the friction as you lightly brush the surface with your finger.


> The 1:1 isolation transformer must
> provide only the current the load uses and no more. That is the key
> thing I was missing I think.

It provides a little more, since no transformer is 100% efficient.  It takes a little bit of extra energy to magnetically excite the core 120 times a second even when there is no load on the transformer, plus core losses due to hysteresis and leakage inductance, core-to-winding and winding-to-winding capacitive coupling, etc.

Don k4kyv