2. misc-audio
  3. 1146359124

Isolation xfmr correct polarity orientation test

I pulled out an old single phase 2KVA xfmr I have. I thought I would do some bench testing for the correct primary to secondary polarity orientation of the windings. I was especially curious to see if the primary was connected to 240 volt if the secondary would read equal AC voltage from each ungrounded lead to ground as I thought.

The Xfmr, transformer.
2KVA
Dual primary windings.
Dual secondary windings.
Data plate on xfmr:
Primary 240V/480V
Secondary 120V/240V

I first connected the primary to 120V. (Parallelled the primary windings). (Hot H1&H3), (Neutral H2&H4) Line voltage 123V.
I configured the secondary windings for a balanced output. X2&X3 tied together but left floating from earth ground for the test. I energized the xfmr and measured the secondary output voltage across X1 and X4, 124.7 VAC. I need to stop here and mention the meter I first used for the test. My Fluke 23 series Multimeter was at work so I used a cheapo Sears Digital Multimter #982015 I had at home. So for the polarity test. I first touched one meter probe from earth ground to X4. This should be the correct polarity as H2&H4 that is connected to the neutral from the source 120V. This reading should be the lower of the readings if correct. The reading was, 0. I them measured from X1 to earth ground, again 0. What did I do wrong. Check earth ground being used, was ok. Double check again, Still 0. This can't be I have done this test several times in the past. Darned if I did not try again, still 0 from each secondary lead when measured to earth ground. Note here you do know when I refer to earth ground it is the equipment ground. But really the measurement is being made back to the source grounded conductor the neutral. Sorry, back to the test. What was going on. I was puzzled to say the least. What was different from any other time I had checked for the proper polarity orientation of commercial power xfmrs. The cheap multimeter.

To make a long sory short I went and got my Fluke meter. Energized the xfmr measured the voltage from X1 to X4, 125.3 Vac. I then measured from earth ground to X4, 50.5V. I then measured from X1 to earth ground, 163.2V. I should mention here as all of you may know this is not a usable AC voltage. Will it bit? Well I for one did not try it....

The 240V primary test.
For this test I connected the two primary windings in series, H2&H3 connected together. 240Vac source connected to H1 and H4 repectively. My power runs about 246VAC as measured at the time of the test.

Secondary wired the same as the first test.
Output secondary voltage, 125.2Vac.
X4 to earth ground reading, 36.6V.
X1 to earth ground reading, 149.8V.

I was surprised these two reading were not closer. I do not have an explanation why.

In either of the two tests when I bonded the center tap of the secondary to ground, X2&X3 the voltage from X1 and X4 repectively to ground was 60V nominal.

Final test.
Primary wired for 240V. windings parallelled.
Secondary 120Vac with secondary windings parallelled. Full 2KVA rating of the xfmr.
(X1&X3), (X2&X4) X1 will be the hot ungrounded conductor. X4 the grounded conductor,neutral.
X2&X4 is left floating for the test.
Primary energized 247Vac.
X4 to earth ground measured 21Vac
X1 to earth ground measured 135Vac
X4 has the lowest AC voltage reading to ground. This will be the secondary lead that will bond to earth ground.
Jim

jea48
zargon's avatar
zargon
453 posts
 
Jea48,
Thanks for the test procedure for polarity of an ISO. Yes, a lot of us are interested in using ISOs on our systems, and this should be useful. I would have assumed that a manufacturer would mark the polarity somehow on the case. I guess not?
It might be helpful if you explained the labling conventions (X and H) to the general audience, especially if these are commonly used across manufacturers.
Also, I am interested in whether you are leaving the secondary floating (no connection between common and ground on the output side) when connected to your CDP? I believe that is the recommended approach in order to maximize noise isolation.
Finally, can you report to us in more detail the differences in sound now that you can listen both with and without the ISO?
jea48 OP
2,944 posts
 
>>"I would have assumed that a manufacturer would mark the polarity somehow on the case. I guess not?"<<
[Zargon]
>>>>>>>>

It is important that the user follow the wiring configuration diagram when hooking up the xfmr. Usually the winding leads are marked in some fashion. Most commercially sold transformers sold for power applications are marked with letters and numbers.
Example,
H = high voltage, primary side of the transformer.
X = low voltage, secondary side of the transformer.
Numbers on leads,
if the xfrm only has one primary winding the the leads may be identified by H1 and H2. If the primary has two winding the leads will be identified by H1 and H2 first winding, and H3 and H4 the second winding.

For the secondary of an isolation transformer most power application transformers have two secondary windings. Winding one leads are identified X1 and X2. The second winding X3 and X4.

If the manufactured labeled the windings correctly, and the big boys do, then the correct polarity of the primary to secondary winding should be,
H2 > X2
H4 > X4
==========

>>"Also, I am interested in whether you are leaving the secondary floating (no connection between common and ground on the output side) when connected to your CDP? I believe that is the recommended approach in order to maximize noise isolation."<<
>>>>>>>>>>>>

No, per NEC the center point must be bonded to earth ground. When using an isolation transformer for power application you are creating "A separately derived AC system." In simple terms you are creating a new electrical service, basically like the one feeding your home. NEC requires all AC power systems shall be AC grounded systems.
So if the secondary is 60/120V or as many call it
60V-0V-60V the center tap of the two windings shall be bonded, connected, to earth ground. What this means is the center tap shall be connected to the grounding electrode system of the main electrical panel system of your home. You can run a separate ground wire all the way back to your main incoming metallic domestic water line and install an aproved water pipe ground clamp there if you want.
Why does NEC require the secondary to be grounded? Safety, in the event of a leak or ground fault from one of the two hot conductors the fault current will travel back to the source, the isolation transformer. If the fault current is large enough it will cause the secondary over current device to open. If there is not overcurrent protection on the secondary the transformer will overload and cause the overcurrent device of the circuit feeding the transformer to open, a breaker in your main panel.
=======
Phase relationship
http://www.ibiblio.org/obp/electricCircuits/AC/AC_9.html

>>"Finally, can you report to us in more detail the differences in sound now that you can listen both with and without the ISO?"<<

I will try to report back this week

herman
1,974 posts
 
I agree that the post is confusing but something like this begs for drawings to help us visualize the set up. Since Audiogon won't let us post pictures then it is probably as good as it gets.

If the secondary is floating i.e. nothing on the secondary is conected to earth ground, and you measure from ground to a secondary terminal, then this reading will not make any sense.

Your meter measures the difference in potential between it's leads. You can't get a meaningful measurement between 2 points that don't share a common reference point. Since the floating secondary is isolated from the ground they do not have a common reference and any measurement is invalid.

You proved this by measuring with the secondary center tap grounded and you got what you expected, but got nonsense when it was floating. The difference in the meters can probably be explained by their impedance as suggested above.
jea48 OP
2,944 posts
 
>>"Your meter measures the difference in potential between it's leads. You can't get a meaningful measurement between 2 points that don't share a common reference point. Since the floating secondary is isolated from the ground they do not have a common reference and any measurement is invalid.<<"
>>>>>>
Thats funny Herman, I and many other electricians have used this test to find the correct phasing orientation of secondary windings of transformers for years. I have used the procedure where winding lead identication are missing. The test is reliable. Don't knock it till you try it.
========

>>"Since the floating secondary is isolated from the ground they do not have a common reference and any measurement is invalid.<<"
>>>>>>>>

AHhh, but how about the grounded AC system that is feeding the primary of the isolated transformer? And that difference of potential that exists and can be measured when the floating secondary leads, is due to capacitive coupling between the primary and secondary windings. The fact is the test will match the actual polarity, phase, identifications that are known for the transformer every single time.
==
Herman have you ever checked for the proper AC polarity orientation of a preamp or power amp power supply transformer? You are preforming the same test. The only difference is the secondary of the transformer is fixed so if it is reversed you change the primary. Correct?
herman
1,974 posts
 
I guess I got lost in your post. I thought you wanted to know why you didn't measure 60 - 0 - 60 referenced to ground when it was floating.
More to discover