Below is my response to a member on another forum. I attemped to to speak more in layman terms.
======================
Jon, I have to admit I did a lousy job on my presentation of this thread. I should of kept it simple and to the point. I should of left out all the voltage measurement values. They only confused people, and the issue.
The test is a simple way to check the polarity, phase relationship, of the secondary output of an isolation transformer to the AC grounded system the transformer is fed from. I have used this test procedure for years in commercial and industrial applications. It has never failed.
========
Example:
Test procedure.
Primary
I connected the primary of the single phase isolation xfmr to 123V from the electrical panel in my home, connected to (L1) and the neutral. The two primary windings are parallelled, (H1&H3) connected to the Hot. (H2&H4) connected to the neutral.
Secondary
I had to series the two secondary windings to achieve 120V. X2&X3 tied together insulated. X1 and X4 floating with a difference of potential between them of 124Vac.
I measured the AC voltage from each of the two floating secondary leads, X1 and X4, one at a time to the equipment ground or to the line neutral. The lowest AC voltage reading of the two will become the neutral and be bonded to earth ground. The lowest voltage reading was from X4 to equipment ground.
I then bonded X4 to the equipment ground. X4 is now the neutral, the grounded conductor. X1 the Hot conductor. X1 to ground measures 124Vac.
=======
Now if every thing is correct I should be able to measure from the new X4 neutral to the electrical panel neutral and the reading should be 0. The reading was 0.
Next measurement to take is from new Hot conductor, X1, to L1 of the electrical panel. If the secondary polarity of the isolation transformer is correct, in phase, with the electrical panel L1, the difference of potential should be 0 or very close to it. The difference of potential was 1.8Vac unloaded. Good.
Is this really of any importance?? Probably not if all of the audio connected ic equipment is fed from the output of the xfmr. What if the transformer is only to feed the preamp, and the Amp is plugged directly into the wall recept. Notice I used L1, line one, for my test. The xfmr is fed from L1 and I would assume the power Amp would be connected to L1 from the electrical panel.
======
What would the result if I had Bonded X1 to ground instead of X4 of the secondary of the transformer?
I did so,
X1 bonded to ground now the neutral, the grounded conductor.
X4 is now the Hot conductor. X4 measured to ground 124Vac, should be...
Measurement from X1, new neutral, to the line neutral, electrical panel, 0 Vac. Ok.
Measurement from X4, Hot, to electrical panel neutral 124Vac..ok
Measurement from X4, Hot, to L1 of the electrical panel, 248Vac....
Real power?? You bet! I connected a table lamp with the three way bulb set on 150 Watts across the Hot X4 and the panel neutral, it lite full brilliance. I wished I had a couple pigtail lamp sockets on hand, but no such luck. I took the plugs of two lamps and series two of the blades for the next test. With the two 150 watt bulbs in series and a couple of make shift pigtails I connected one end of the series two lamps to the hot X4 of the xfmr secondary, and the other to L1 of the electrical panel. They lit full brilliance.
=======
I tried to use the most direct way of explaining the test procedures.
I hope it came through that way. My attempt was to explain in layman terms best I could.
The actual test was preformed with two three wire cord and plug connected to the same electrical duplex outlet. The outlet is fed from L1 but the results would of been the same for L2, of the electrical panel. Primary of xfmr was connected from one cord and plug. Test measurements for the electrical panel reference taken from the three leads from the other cord and plug. Polarity, continuity, of both cords and plugs are the same.
=============
If we have learned from experience that audio equipment that is connected together by ICs should be connected to the same line. Then should not all isolatation transformers have their output polarity, phase relationships, in sync the same? Or at least as close as possible?
I would appreciate your comments Jon.
=====================================
>>"It can take just a bit of a load to read near correct voltages, and this is related to how much capacitive coupling there is between the primary and secondary, and how much capacitive imbalance there is between the various windings/taps. Very little capacitive coupling (which usually indicates a higher quality transformer, or one designed specificaly for a low amount of capacitive coupling) require very little loading to get nearly correct voltage readings."<<
>>>>>>>>>>>
Jon as I had stated earlier in a post I did connect the three way lamp set at 150W for a load on the xfmr. The only other heavier load I could think of was a 1200W cloths Iron. One hell of a transformer all I can say. Unloaded the output was 125V. With the Iron connected the voltage dropped to 100V With the secondary windings in series my 2kva xfmr was a 1kva. I quickly took my test readings from the two floating 120V secondary leads though the readings were a little lower when measured to the equipment ground one was high and the other was low. The readings matched the polarity orientation of the xfmr leads.
=============
Jim