Isolation transformer and oscilloscope

To my understanding you should not have a grounded secondary side.

You are looking for total isolation for safety.

I found a 4-page very interesting document on archive.com : search for “Understanding Ungrounded Oscilloscope Measurements” by Raymond E. Herzog.

@paulvo Nice article.  Thank you.

@larry_n7luf I realized I got it wrong with the layout of the secondary side. The earth wire is connected in the middle of the secondary winding. 

New schematic

Does this make a difference when measuring with a scope?

I have two Tektronics 

453a and

503 which is a differential scope. 

@paulvo Thanks for all the help. Nice article. I see my self as a beginner with scopes and don’t want to mess upp. Learned a lot from that article. Especially interesting with differential measurements.

@mika The key think I know is not to have a pathway for current to flow to the earth ground.
If there is a path to ground, it is not isolated.

Remove all routes to ground form the secondary side.

Again this is my understand and was told many years ago to keep one hand behind you all the time when the unit was under power and like Mr. Carlson always say, you do it at your own risk.

The transformer winding is the isolation to the unite under the test.

Hope this helps.

The most important point for having a Ground inside the transformer’s Box is protection against electrocution. Say the transformer’s primary for any reason could get some turns shorted but still keeps working for a while (getting hotter and hotter by the second) and as the winding gets hot it will inevitably melt the varnish (the winding insulation layer shorting more and more turns as it melts resulting to an increased current consumption) and because the transformer’s core is made of metal, it would at some moment come in contact with the LIVE (by now) bear copper winding. So, if the the ground is present and I mean the transformers box/enclosure is grounded then,  at that very moment the live current would automatically find a return path to ground therefore triggering the resettable fuse in your house fuse box. So as a result of this very quick event (that actually happens almost in an instant) would prevent you from getting electrocuted and very possibly injured! if you were touching the metal box at that moment. Now, the fact that the ground wire (according to the presented drawing) extends further to one side of the secondary winding that makes the transformer  “NOT BE COMPLETLY ISOLATED FROM MAINS !” therefore defeating the whole reason of having an isolated power supply in the first place. Like Larry and others have said, there shouldn’t be any wires that lead back to mains from the secondary side. So again, in order to have a 100% isolation from mains, just simply disconnect the ground wire from the secondary winding but still keep it connected to the chassis (the metal enclosure) so the protection against electrocution (in the event like the one described above happens) still remains!.

@ovi4 Thanks. I plan to do that. Still my question about why in the first place the secondary was connected is somewhat interesting. This is an old commercial IT so there must have been some purpose in doing so.

Is it about having AC with no interference from other equipment on the same power line? 

@mika Mr. Carlson I believe has videos dealing with this.

There are other videos talking about having the probe grounding lead get vaporized and having your scope input damaged.

Sometime the chassis can be about ground like an old radio and you do not want to feed the short thru your scope and really be very upsetting.
Like the old saying is not to let the smoke out of your parts.

Like @ovi4 said ground the Primary and box of your Isolation transformer.

Always remember the part you are working on can be above ground and be hot.

That is why Mr. Carlson always say, if you do not know what you are doing, and you do it at your own risk.

Funny story, as a teen I was working on our neighbors TV, I got zipped and stood up and put my head through top of his TV.

I have to say, I love sparks and became a welder, and yes I have let smoke out of things.

Couldn’t you determine if the primary of an isolation transformer is “isolated” from the secondary using a continuity checker?  If you compared plug (input) leads with the socket (output) leads including the case/chassis and got no response, can you conclude the transformer is solely linked by a magnetic field when in use and has no direct wire connection.

I understand the ground prong may be connected to the center tap or chassis or to the neutral at the circuit panel.

So I’m wondering if this is a stupid question, else I would have seen it before, or if it falls into the “unanswerable” category.

Regards,

Mike

Apart from the fact that the signal ground of an electronic circuit, due to a transformer issue, may get connected to the mains (as pointed out by Ovi4), there is also the earth connection prong, which is either or not connected to the chassis of a measuring instrument. This is of course not applicable to equipment with just a 2 prong plug and no earth connection (*).

Some signal generators and power supplies have an earth connection connected to the chassis, but isolated from their signal ground. Sometimes they even have a metal strip to allow the earth ground (also called protective ground) to be connected to the signal ground. On the other hand, most oscilloscopes have the signal ground (the outer ring of the BNC-connector) connected to earth ground (=protective earth PE).

With a multimeter you can easily verify this: measure if there is continuity between the earth prong of the wall plug and the outer ring of the BNC connector or the signal ground plug on the instrument.

And that’s where equipment with a ‘hot’ signal ground and an earth connection prong, will damage a measuring instrument with its earth connection connected to the signal ground when the mass (or signal ground) of its probe is touching the ‘hot’ signal ground of the equipment under test: due to a short circuit created between the earth ground and the hot signal ground, through the chassis of the measuring instrument.

In this case, the mains circuit is not concerned: it is a different mechanism to the (also very important!) electrocution risk, caused by the contact of signal ground with the mains (leakage current).

(*) A must-read article about the differences between Signal Ground, Chassis Ground and Earth Ground is this. Highly recommended 🙂