Isolation Transformer

[A-day]

 Hi all !
  i picked up 2 Tripp lite IS-250 Transformers a few months ago hoping to use them on my bench to protect my gear. but most IS transformers that i see in use have more features and i figure may be purpose built to be used on a repair bench so i'm wondering if the tripplite 250 i picked up will be any good to me on my bench. thanks for any help in advanced....

[Ian.M]

The datasheet says:

* Neutral to ground bonding at the secondary eliminates common mode noise and provides an isolated ground reference for sensitive equipment

so they should be suitable for testbench use.  However its always a good idea to check that there is no continuity between output and input ground terminals before initial use.

[tautech]

For them to be truly useful on the bench you need to ensure the Earth connection terminates/ends at the transformer and does not connect to the Earth pin on the transformer output socket.
Use them for the DUT not your equipment.

As Ian has found it the datasheet. 

[ptricks]

 Two power transformers with the secondaries connected to each other also works.

[Cliff Matthews]

I had a similar question a while back. I tried some 80va units back-to-back (as previous post mentioned) and with 122v in, I lost 12 volts at approx. 80% load. So it appears traversing 2 magnetic couplings gives poor regulation.
https://www.eevblog.com/forum/beginners/diy-isolation-transformer-question/msg806281/#msg806281

[mmagin]

In general you'll find two kinds of isolation transformers out there:  Those which are intended to isolate the secondary so that it is not earth ground referenced and you will not be shocked by touching the line (only).  Those which tie ground and neutral of the secondary to the earth ground of the input.  This is more intended to provide an isolated supply for sensitive equipment.  Thus there will be insignificant potential difference between neutral and ground on the output (as opposed to what you might find on a heavily loaded building circuit), and there is usually a grounded electrostatic shield between the windings. 

This second type is what you'll most commonly found among medical grade isolation transformers cheaply available surplus.  Without some modification, it is not useful for preventing shock or allowing probing a circuit with a grounded scope probe, but it is nice for isolating more sensitive devices, and it can also help to isolate devices which cause nuisance GFCI/RCD trips.

[Ian.M]

Its useful to be able to break the secondary side N-E link if you are working on a SMPSU, as you will probably be grounding the negative side of the reservoir cap with your scope probe ground (assuming you don't have an isolated probe), and if the D.U.T. has a metal chassis, you really don't want the resulting half-wave rectified mains on N-E connected to anything you might lean on!   

As the O.P. is USA based, a three pin to two pin grounding converter adapter, with its ground lead clipped off short, would be suitable to disconnect the N-E link.   *NEVER* use such a 'ground lifter' without the isolating transformer.

[johansen]

So the cool thing about all these new plastic electronic OScopes is the housing is plastic, the SMPS provides fairly low capacitive coupling between L1 and L2 and the oscope "ground".  and all you have to do is not shock yourself on the plastic buttons.

so

I had a similar question a while back. I tried some 80va units back-to-back (as previous post mentioned) and with 122v in, I lost 12 volts at approx. 80% load. So it appears traversing 2 magnetic couplings gives poor regulation.

80VA! Replace them with 800VA and your 10% voltage drop might be 2% instead.

[tautech]

So the cool thing about all these new plastic electronic OScopes is the housing is plastic, the SMPS provides fairly low capacitive coupling between L1 and L2 and the oscope "ground".  and all you have to do is not shock yourself on the plastic buttons.

If you are implying that it is OK to use an isolation transformer to provide scope isolation from mains ground....it is not OK, in fact quite dangerous. All exposed metal parts (BNC's) of the scope are then at whatever potential the user clips the reference lead to. 
For the safety of the user the use of any means of removing the mains earth connection from a scope must be frowned upon. 

[Cliff Matthews]

So the cool thing about all these new plastic electronic OScopes is the housing is plastic, the SMPS provides fairly low capacitive coupling between L1 and L2 and the oscope "ground".  and all you have to do is not shock yourself on the plastic buttons.

so

I had a similar question a while back. I tried some 80va units back-to-back (as previous post mentioned) and with 122v in, I lost 12 volts at approx. 80% load. So it appears traversing 2 magnetic couplings gives poor regulation.

80VA! Replace them with 800VA and your 10% voltage drop might be 2% instead.

That post was 2015. I have since started using back-to-back 200va units for when the isolation need arises, but still actively troll around for something north of 1000VA.

** Sage advice dept **
Tim (from Seven Transistor Labs) had a great anecdote worthy of note for this application: "bigger is better"
Why? Fuses need to trip quickly: https://www.eevblog.com/forum/beginners/diy-isolation-transformer-question/msg806377/#msg806377

[Vtile]

Two power transformers with the secondaries connected to each other also works.

They work what comes to electricity, but they are not safety isolated. Their typical isolation (isolating sheets or some other mechanical barrier) between the primary and secondary is not enough to be sufficiend SAFETYisolator.

The difference comes in situation of the transformer meltdown or excessive HV transients that go trough a thin isolation foil. Also the capacitive coupling should (I have not measured or read on it) be smaller, since the extra distance between secondary and primary.

[Cliff Matthews]

Two power transformers with the secondaries connected to each other also works.

They work what comes to electricity, but they are not safety isolated. Their typical isolation (isolating sheets or some other mechanical barrier) between the primary and secondary is not enough to be sufficiend SAFETYisolator.

The difference comes in situation of the transformer meltdown or excessive HV transients that go trough a thin isolation foil. Also the capacitive coupling should (I have not measured or read on it) be smaller, since the extra distance between secondary and primary.

So maybe an answer (for some techs) is to make a DIY hipot tester? Like this one:
http://www.lb3hc.net/hipot

[ebastler]

Two power transformers with the secondaries connected to each other also works.

They work what comes to electricity, but they are not safety isolated. Their typical isolation (isolating sheets or some other mechanical barrier) between the primary and secondary is not enough to be sufficiend SAFETYisolator.

The difference comes in situation of the transformer meltdown or excessive HV transients that go trough a thin isolation foil. Also the capacitive coupling should (I have not measured or read on it) be smaller, since the extra distance between secondary and primary.

So maybe an answer (for some techs) is to make a DIY hipot tester? Like this one:
http://www.lb3hc.net/hipot

No, I don't think a hipot tester can address the safety problem highlighted by Vtile. Even if you were to test your homebrew isolation transformer every time before you use it (which you won't do in practice!), it would still be unsafe, because the transformer can fail while you are using it.

[Ian.M]

Its possible to work around it - add a ground connection between the two transformers, with current sensing to trip a pair of two pole latching contactors, (one on the feed and the other on the output) if the leakage to ground becomes excessive.  That will detect any insulation failure and disconnect it very quickly.
However if you are using thermally fused split bobbin transformers that have passed a Hi-Pot test within the last year and haven't been abused, the risk is vanishingly small anyway.

[ebastler]

However if you are using thermally fused split bobbin transformers that have passed a Hi-Pot test within the last year and haven't been abused, the risk is vanishingly small anyway.

Agree, if we postulate the presence of thermal fuses as an additional feature.
Or you might just buy a proper isolation transformer. 

[Cliff Matthews]

However if you are using thermally fused split bobbin transformers that have passed a Hi-Pot test within the last year and haven't been abused, the risk is vanishingly small anyway.

Agree, if we postulate the presence of thermal fuses as an additional feature.
Or you might just buy a proper isolation transformer. 

That's the problem.. big iso transfo's cost big $$. Some say hard to get too (unless you've got no currency problems), so I'd like to know if what Ian.M mentioned can be done maybe with a GFI or something a budget tech can afford.

[Cliff Matthews]

Its possible to work around it - add a ground connection between the two transformers, with current sensing to trip a pair of two pole latching contactors, (one on the feed and the other on the output) if the leakage to ground becomes excessive.  That will detect any insulation failure and disconnect it very quickly.
However if you are using thermally fused split bobbin transformers that have passed a Hi-Pot test within the last year and haven't been abused, the risk is vanishingly small anyway.

If it's not too much, could you sketch something? Perhaps a GFI could sense the leakage? (It would be cost effective).

[Ian.M]

A pair of GFCIs, before and after the paired transformers, with the connection between the transformers grounded (ideally via a center tap on one of the secondaries) should work.  If the line side transformer breaks down, the input GFCI will trip, and if the load side transformer breaks down, and anything grounded contacts the load circuit, the output GFCI should trip (subject to its sensitivity rating and the resulting fault current).

A better method would be to hack a three phase GFCI (normal current balance sensing type) with an appropriate trip current rating to act as an 'old-skool' Earth fault breaker, tripped by actual current in the earth conductor.   If it was modified to separate the current balance sensing from the contactor, one could use two poles to isolate both sides of the output and the remaining pole to disconnect the Live of the line in.  The ground wire from between the transformers would go through one phase of the current balance sensing coil, with the other phases left unused (or wired in series to double or triple the sensitivity).   You'd also need to be certain how its control circuit was powered didn't introduce a sneak path between the input and the output poles.

[BrianHG]

Ignoring the potential issues, if you must have the isolation & cheap.  I would use a 'Split Bobbin' 120/240v-120/240v transformer.  The split bobbin will keep the primary and secondary side coils well isolated compared to 1 winding on top of the next with noting but isolation tape between the windings.  In the past, I used to use such transformers for my audio pre-amps and dac units as they offered superior isolation without the cost of special grade transformers.

Another trick if you cant find a 120v-120v transformer, I at times would use a 120v-48v and then a second using it backwards to get back to 120v, or even a 120-240v, then another one to step back.  But once again, having that all important Split Bobbin at the time I was working and had to scope a telephone line powered device I was designing where even a lab grade isolater without the split bobbin transformer leaked 60hz hum and current to the phone line.

[Gregg]

Years ago when I was in the US Navy on board an aircraft carrier we had isolation transformers as standard equipment in damage control lockers.  These were portable units epoxy potted inside a fairly thick steel case so they were essentially water proof and well shielded.  We would get the avionics people to hi-pot test them for us.  They were specifically for isolating the output from the ship’s hull which is about as good of a ground as possible. They were for use in case electric tools or lighting was needed in an emergency situation.  I took apart one that didn’t pass hi-pot because I just couldn’t help myself and what else does one do for entertainment in the middle of the ocean. If I remember correctly it was about 600VA in size and the two windings were on separate spools with at least one probably the primary having a copper grounded shield around it.  There may be some of these in military surplus stores.

[Cliff Matthews]

Years ago when I was in the US Navy on board an aircraft carrier we had isolation transformers as standard equipment in damage control lockers.  These were portable units epoxy potted inside a fairly thick steel case so they were essentially water proof and well shielded.  We would get the avionics people to hi-pot test them for us.  They were specifically for isolating the output from the ship’s hull which is about as good of a ground as possible. They were for use in case electric tools or lighting was needed in an emergency situation.  I took apart one that didn’t pass hi-pot because I just couldn’t help myself and what else does one do for entertainment in the middle of the ocean. If I remember correctly it was about 600VA in size and the two windings were on separate spools with at least one probably the primary having a copper grounded shield around it.  There may be some of these in military surplus stores.

Wish we had abundant military surplus in Canada, but our good neighbour's south of the 49th have us living in safety. I suppose if I wanted something safe and beefy (as per T3sl4co1l's recommendation), I could split-wind a laminate toroid I have in my shop, but I'm not sure it's worth all the work or what power I could get out of it (OD 135mm x 65mm x 35mm, approx. 4.5kg)

Edit ** The photo is the condition I found it in... maybe it's good for 600VA? (I counted about 95 turns of iron sheet)

[sequoia]

Any pointers to model/make of isolation transformer that has option to isolate ground as well?


I've seen lot of references people simply not connecting ground to isolation transformer (one that has input/output grounds bonded) to avoid ground path "problem" when probing DUT with (non-isolated) oscilloscope probe, which doesn't seem very safe...

Would better solution be to modify such isolation transformer to have switch to disconnect grounds, so that the transformer (and its case) will always be properly grounded?
I've head also references to use "bleeder resistor" to keep the isolated supply from "floating" too high? (How does that exactly work?  Is isolaeted output common/ground bonded together and then resistor between those and input ground?)

[Vtile]

If bleeder is what I suppose, then it is also sometimes called as a signal grounding (as a free translation) for those equipment that needs to have somesort of reference to the ground to work properly. It is basicly a many megaohm resistor to allow some tiny amount of current to flow and therefore reference the DUT to mains ground. It might also be for some odd case where ie. some vande graaf generator or something is generating high amounts of electric fields, which lifts the DUTs potential over the isolationtransformer specs as a whole, but this goes over my (current) knowledge.

Don't ask more details. I haven't read any standards or regulations (they are so awfully expensive  ) which goes to specific to safety equipments and their design requirements in laboratories and repair shops to even ballpark a reasonable values, besides US propably have own regulations, even though electric safety standards are pretty universal.

In general safety isolation gives only that much safety. If you touch two potentials in DUT which is floating you still get exactly same shock that non floating DUT. If you have that bleeder the amount of shock and where and how you get it is more complex.

Edit. Also worth to mention is that when you put that non isolated scope ground to galvanicallyisolated (aka floating) DUT, then that point is ground and if that point you did ground were ie. the DUTs nominal 120Vac line then between your scopes case and DUTs nominal neutral  lead do have 120Vac potential difference. That means if you have as idiotic electron tweaking workbench like I do where grounded heat radiator is in reach of your legs, it will effectively zap you if you are foolish and touch the DUTs case if it happen to be in contact with the nominal neutral. Since it does have 120Vac potential difference between your scopes ground lead and that is in same potential as the grounded radiator under your workbench.

Why my brains always try to put US mains as 120Vdc and not AC. :/
Sorry, million edits I should have thought first and only then post.