GFCI AND ISOLATION TRANSFORMER

[smur]

Hello
I am designing a 120 v circuit that has an isolation (1:1)transformer in it. On the secondary 120V circuit, I would like to install a GFCI breaker for 5mA. But all the breakers I looked at needs a grounded system while my transformer secondary needs to be ungrounded and floating. How can I provide a GFCI 5mA breaker ?
thanks in advance
murli

 

[itsmoked]

You can't! How can the breaker reference a ground if you don't have one? And why would a current go to ground if there is no ground connection point?

I hope this isn't in some structure!

Wanna explain?

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[waross]

What Keith said.
However;
You can install a GFCI breaker. In the event that there is a fault to ground in the transformer secondary and someone made inadvertent contact between ground and one of the transformer leads, the breaker may or may not offer protection. I wouldn't bother.
But;
Life safety is sometimes assured with isolating transformers by the use of a monitoring circuit.
Twin voltage detector circuits may be connected from either transformer leg to ground and the voltage measured.
Normally, insulation resistance and capacitance to ground of the transformer windings and the circuit wiring will roughly divide the circuit voltage into equal voltages to ground.
A shift in voltages to ground indicates that there may be a problem.
The last time I selected and installed such a circuit was about 15 years ago. My information is 15 years out of date, and my old literature is not readily available.
At the time, I was guided by publications of the Canadian Standards Association.
The standards were the requirements and recommendations for the installation of small isolated circuits for use in operating theatres.
The allowable leakage current for such systems is less than 5 ma.
Although 5 ma. and below is often assumed to be a safe level, the CSA standards indicated that an electrical current that flows in the organs that are exposed during surgery may be lethal at levels much less than 5 ma.
Consider the case of a surgeon holding an organ with one hand to gain better access to the site of the surgery. He reaches up with the other hand to adjust the surgical light (fed from an isolated circuit).
If there is a small amount of voltage on the surface of the light fixture from insulation leakage, the surgeons body may conduct enough current to the patients organs to be fatal, without the surgeon feeling a shock.
I was surprised to read that, but the CSA standards are compiled by engineers with a lot more experience in their fields than I have.
The standards for floors for use with explosive anethsetics were a lot of fun also.
hope this helps
respectfully

 

[ScottyUK]

A residual current device, which I believe is the European equivalent to the GFCI, uses two coils wound on a toroid carrying the load current, plus a third sensing coil which drives the tripping circuit. Under normal conditions where line current in the 'Live' conductor is exactly balanced by the return current in the 'Neutral' conductor the net flux in the toroid is zero and there are no flux linkages to the sensing coil. If current finds another path which bypasses the toroid to return to the source, the sensing coil will 'see' flux linkages due to imperfect cancellation of the line and neutral currents. If the current induced in the sensing coil is large enough the RCD will trip.

In your installation, where would the potential return path bypassing the toroid occur? If the transformer secondary is not earth referenced the current can not pass through ground.


----------------------------------

Sometimes I only open my mouth to swap feet...

 

[Skogsgurra]

If your intention is to detect ground faults, you should use an insulation supervising device. The simplest one is a voltage divider to create an artificial zero half-way between the two phases. As mentioned before, the line capacitances are usually equal, or close to equal, so the voltage between your artificial zero and ground will be low. Add a voltage sensing relay between artificial zero and ground and adjust it to trip when there is a 5 mA leakage to ground. The latter may be difficult if you have a small system - the total system capacitance may not be large enough to creat 5 mA. I think that you may need to revise your system requirements. If you have lots of money for the project, there are Bender and similar devices for insulation supervision.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[smur]

http://bg.ecmweb.com/ar/electric_replacing_wire_ungrounded

Pl see this above. Is this an acceptacle solution to this problem ? thanks

 

[waross]

We seem to be switching horses in mid current (to mangle a metaphor, but all in good pun).
The original post dealt with GFCI protection of an isolated, ungrounded circuit derived from an isolating transformer. This would be an ungrounded system. The link given by smur relates to grounded systems where the grounding conductor is not present at the receptacle.
You have been given the same good advice, in different words, by all the responders.
Please explain your application so that we know what we are dealing with.
The more information that you provide to us, the more useful our responces will be.
respectfully

 

[ScottyUK]

No. The supply in your house is referenced to ground somewhere even if the ground conductor isn't brought through to the wall socket. Usually the neutral is connected to ground at either the service entrance to the property or at the origin of the supply, or in some instances via a system called PME. Rarely an isolated source is provided for special applications. Your transformer is in this category. It does not have a 'live' or 'neutral'. It has two terminals which are entirely interchangeable in function - you could ground either one and designate it 'neutral', thus allowing you to use a RCD (GFCI) or leave both ungrounded and not use an RCD. You can not do both.

Without wishing to cause offence, are you sure you should be designing a safety transformer circuit? Your understanding seems a little too sketchy for someone to have to put their life on the line based on your limited knowledge plus some information you have gathered on a website, however knowledgable the contributors to that site are. You should also consider your own personal liability if you design something which ends up being dangerous.

----------------------------------

Sometimes I only open my mouth to swap feet...

 

[itsmoked]

Is this an acceptacle solution to this problem ?

We say again... [highlight]What problem?[/highlight]

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[smur]

Apologies. I did not read the link fully. The requirement is for a ungrounded transformer. no grounds anywhere. Is there any means where we can detect a ground in an isolaed system?

 

[itsmoked]

Why can you not tell us briefly why you need this system, so our guidance isn't a speculative shot-in-the-dark?

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[smur]

This is to supply 120v to some instruments which will not tolerate a ground. And since this goes to some other cabinet, I would like to add a GFCI. Any ideas?

 

[Skogsgurra]

Like using an isolation transformer to "float" an oscilloscope?

If that is so, and the problem is that you have a distributed isolated 120 V. I think that you could as well use local isolated 120 V - one transformer for each instrument. That will not only put an end to your concern about ground faults, it will also handle the much more difficult problem of having several isolated instruments that "will not tolerate ground" being connected through a common "ground", albeit isolated. Transformers are cheap, special solutions to your problem are expensive and hard to implement in a way that satisfies code and good practice.

I have been working with lots of different measuring equipment for decades and never met a situation where it was necessary to do what you are trying to do. Differential front ends and isolation amplifiers are there for situations like yours. An instrument that cannot be used without floating ground is either badly designed or badly used.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...