Using isolation transformers.

[Hunts smoothing bomb]

But Al, why did the current want to flow to earth

If your transformer secondary is isolated from earth then it won't want to flow.

The only reason that mains live will flow to earth is because the neutral is tethered to earth at the transformer in the sub station. (and possibly at your service connection) I don't think your RCBO is operating in a true circuit loop unless you have connected one side of your secondary to common earth and then there is no point in having an isolating transformer surely because that is exactly how mains 230v is configured or have I missed something?

Cheers

[Argus25]

Originally Posted by Herald1360

You're no better off than with an RCD on the mains though and still have the same problem with grounded test gear as with RCD protected straight mains.

On balance, isolating the set completely is probably safer, and you can connect your earthed test gear to it. At that point, though, you CAN get belted by the HT to Earth so keep one hand in pocket and stand on the rubber mat, same as for an isolated set.

Herald1360, you have a very good understanding of this.

So the belt and braces approach is the isolating transformer, no mains earth coupled through and with an added RCD on its output to cover the implication of chassis earthing which occurs under the attachment of test equipment scenarios might help.

[PETERg0rsq]

Surely the whole point of an isolating transformer is that it isolates the output from earth! The output is "floating" and has no "potential" to earth.

If you touch any one point of the live secondary circuit, there will be no path to earth, so no current can flow (it is isolated from earth) and so you will not get a shock.

If you touch any two points on the secondary circuit, then you will risk electrical shock, as there is a potential difference between these two points, so current can flow.

An RCD will not help in this condition as all the current leaving the isolating transformer secondary will return back to the transformer secondary (non can be lost as there is no path to earth). If an RCD tripped on the secondary, then I would be very doubtful the transformer was truly an Isolation transformer!

Earthing one leg of the secondary totally negates the use of an isolation transformer, as there is now a path via earth, so current can flow to earth.

Peter

[Skywave]

Originally Posted by Hunts smoothing bomb

But Al, why did the current want to flow to earth

Good question! And I understand the logic of your reasoning. Short answer is I don't know - which, I admit, is a bit embarrassing - and, more significantly, is rather concerning. Capacity current in the earth wiring back to the consumer unit, etc.? Whatever, I'll have to have a look inside my isol. xfrmr. unit and see how its put together. Nevertheless, my experiment did trip the RCBO - as required - so I'll live with things as they are, pro tempore.

Al.

[Skywave]

Originally Posted by PETERg0rsq

If an RCD tripped on the secondary, then I would be very doubtful the transformer was truly an isolation transformer!

Quite - and that has started me thinking about the health of the xfmr. in my 'isolation xfmr' unit. Must examine it soon.

Al

[mhennessy]

The trouble is, there are many different way to use a double-wound transformer. The three main ones:

1. As a genuine "isolation" transformer.
2. As a way to reduce the magnitude of an electric shock.
3. As a way to isolate leakage currents, providing discrimination for cascaded RCDs.

I'd suggest that number 1 is what most people think of in a workshop environment. But, what are we trying to achieve here? Some folk think it's to reduce the likelihood of getting a shock in the first place, and there is perhaps some merit to that, but I would seek to deprecate that thinking. The primary reason is to allow the safe connection of earthed test equipment, such as an oscilloscope, to a DUT.

For #1, the rules are simple, and are covered by the link that Lawrence provided earlier, which is a HTML version of a BBC EGN that I've posted as a PDF on many forums over many years. Basically:

A. Only one DUT connected to the secondary.
B. Do not carry the earth through, for a NE swap results in a live case (if Class 1).

In general, an RCD is safer than a transformer, but as mentioned, the transformer allows the connection of earthed test gear. In other words, the isolation transformer is used only used when necessary. And actually, an active differential scope is safer than using a transformer, and would be the preferred method today.

As soon as you've hooked up something to the DUT, it is no longer floating, so "no touchy". But the point is, you've made that connection safely with no drama.


Moving to item #2 on my list, the familiar 110V yellow "site transformers" are a good example of that. The secondary is centre-tapped and connected to mains earth, so you have no more than 55V w.r.t. mains earth. Variants exist that produce 115-0-115, again with the centre-tap connected to mains earth, resulting in no more than 115V w.r.t. mains earth, so the severity of a shock will be reduced accordingly.

The output from such a device is clearly not "isolated", as in case number 1, and such a transformer won't allow you to connect earthed test equipment to something with a "live" chassis, such as the primary side of a switched-mode power supply. However, multiple devices can be connected to this, and an RCD on the output of this transformer should provide the expected safety enhancement.


Item 3 is something that I see at work a fair bit - the example I usually give is on OB vehicles, which are full of technical gear which cumulatively have a fair bit of leakage current that might well trip an RCD at a venue. One end of the secondary winding is connected to chassis (so you have L and N) so the local RCDs can work inside the vehicle, but the transformer has removed the leakage current from the POV of the upstream RCD. Neat, but you pay for it - the transformers are expensive and heavy.


My list stopped at #3, but there are several others. For example, complete earth-free environments are sometimes encountered in some places - these are complicated and likely to be outside the scope of this thread.

Anyway, in summary, I feel that any conversation about the use of isolation transformers should be clear about which type of usage scenario is under discussion. It's most likely #1, but if folk are thinking about other scenarios, confusion could result.

[Skywave]

Mark: interesting report: thanks.

One question though, please. For your 'case 2' scenario, with multiple items connected, doesn't that require that the mains on / off switches on all the equipment so connected, to be double-pole types rather than single-pole switches (which is probably more common), since with single-pole, one 'side' of the incoming 115-v. will always be at 115-v. w.r.t. earth / chassis, even if the item's mains switch is turned off? And if so, couldn't that lead to a possible false sense of safety?

Or have I over-looked something (again) ?

Al.

[Boater Sam]

The only thing that should be connected to earth is the box the isolation transformer is in.
There were many ex BT transformers sold not so long ago, I bought one.
What BT used them for is a mystery as the earth was incorrectly wired through to the outlet socket. THAT IS WRONG!
An isolation transformer should have a 3 core lead to the mains input only.
The output should not have any earth connection to anything, i.e. 2 core output even if to a 3 pin socket.
An RCD will not trip if in the output of a correctly connected isolation transformer, if it does, its not isolated - don't use it. Similarly if the rcd on the input trips when using the output, there must be a fault path on the input of the transformer, not the output, as there should be no earth unless the transformer is faulty or wired incorrectly with an earth path.

[Radio Wrangler]

One of the uses of an isolation transformer is to allow the chassis of something that wasn't normally isolated from line and neutral to be safely earthed so that normal earthed test equipment can be used on it. Of course, while doing this things like the HT line, the heater string and the mains input live remain dangerous, but the whole task becomes identical to working on a set with a proper mains transformer in it.

The isolating transformer becomes a temporary mains transformer and the set only differs in having string-connected heaters.

The biggest risks come from lack of thought and lack of care.

David

[ms660]

I wonder if Al's transformer is center tapped?

Lawrence.

[mhennessy]

Originally Posted by Skywave

Mark: interesting report: thanks.

One question though, please. For your 'case 2' scenario, with multiple items connected, doesn't that require that the mains on / off switches on all the equipment so connected, to be double-pole types rather than single-pole switches (which is probably more common), since with single-pole, one 'side' of the incoming 115-v. will always be at 115-v. w.r.t. earth / chassis, even if the item's mains switch is turned off? And if so, couldn't that lead to a possible false sense of safety?

Or have I over-looked something (again) ?

Al.

If you're asking about the switches on the outlets, there might well be regs and guidance about this, but intuitively, I'd say that sockets with double-pole switching is very much a Good Thing.

Today, both types are available but most are DP switched. And unswitched sockets are readily available.

In case you're thinking about mains switches inside equipment, we know,that either type is possible. And a lot of modern gear has sections which remains powered even when the device is ostensibly "off" - the humble PC being an obvious example.

IMHO it's safest to simply assume that everything is live irrespective of the specifics. Unless I can see a very real "air gap" (i.e., the thing is unplugged), then that's the only safe assumption.

After all, Neutral is often slightly above ground anyway, and although the voltage difference might be small, significant currents could flow through a short if you were to introduce one (tripping the RCD, hopefully).

[Skywave]

Thank you, Mark. In my workshop, many sockets are DP switches; I was thinking about switches that are integrated in equipment. But I take your point about modern equipment whereby although the 'mains switch' is off, some parts within the item could still have mains voltage available. So yes: if it is necessary to get one's fingers inside, simply unplug it.

Al.

[Skywave]

Originally Posted by ms660

I wonder if Al's transformer is centre tapped?

No, it isn't, But I have done some testing of it - and it seems that there is an intermittent low resistance insulation failure from the secondary to the transformer's earthed laminations. But all is not lost, since I do have a suitable spare which has a 115-v. - 0 - 115-v. secondary and is rated at 2 kVA.

So, thinking in terms of Mark H's recent posts about applications for such transformers and switchery, q.v., current thinking is to supply all bench outlets with bi-phase 115-v. plus earth, using sockets with DP switches (which are in existence anyway). The xformer sec. will feed a suitably rated RCBO.

Does that sound sensible ?

I will just add - for the sake of completeness - that I also have another isolation transformer unit (rated at 1 kVA) that does not have any supply earth fed through to the outlet on the secondary. (The metal case and xfmr. lams. are earthed, of course, and is appropriately fused). That xfmr. is used when working on a.c. / d.c. equipment - which is not very often.

Al.

[mhennessy]

Originally Posted by Skywave

So, thinking in terms of Mark H's recent posts about applications for such transformers and switchery, q.v., current thinking is to supply all bench outlets with bi-phase 115-v. plus earth, using sockets with DP switches (which are in existence anyway). The xformer sec. will feed a suitably rated RCBO.

Doing this creates Scenario #2. Make doubly-sure sure the earth is carried through.

You also have another option: you could re-wire the transformer to create L and N, as in Scenario #3. This is more "conventional", but like #2, it gives discrimination between the local RCD and an RCD further upstream. If I had your transformer, I'd lean towards this one. I say that because although a balanced supply reduces the magnitude of a shock before the RCD (hopefully) trips, it doubles the opportunities to receive one.

Although - and it's just a personal view - when you weigh up the pros and cons, it's debatable as to whether it's worth using the transformer:

Pros: RCD discrimination (which might or might not matter, depending on your house wiring; in my case, it would be a bonus). Some filtering and surge protection from the transformer.

Cons: Isolation transformers can be mechanically noisy, generate a fair amount of heat, and have large inrush currents that need to be managed. Over-current and over-temperature protection must be provided.

So overall, it might be better to stick to simple RCD protection for general bench sockets. But as I say, it's personal choice.

Hopefully these thoughts also help others in similar positions

Originally Posted by Skywave

I will just add - for the sake of completeness - that I also have another isolation transformer unit (rated at 1 kVA) that does not have any supply earth fed through to the outlet on the secondary. (The metal case and xfmr. lams. are earthed, of course, and is appropriately fused). That xfmr. is used when working on a.c. / d.c. equipment - which is not very often.

That takes care of Scenario #1

[Skywave]

Mark: again, interesting comments - thanks. You raise a few points -

House wiring, etc. This house is on a TT system; overhead incoming lines; earth rod near front door. The workshop has its own ring main, fed from a C-type RCD in the consumer unit. The workshop has a red 'panic button' which disconnects that entire ring. When I shut up shop at the end of the day, it gets pushed. Whenever I get visitors to the 'shop, that is the first thing I point out to them - and why.

The existing xformer ass'y.: its case has a fan fitted and suitable ventilation. It is fed from a conventional 13A plug with a 13A fuse. When the mains enters the unit, it goes through a DPDT switch with an indicator neon fitted. On the secondary, in addition to the RCBO, both outgoing lines are fused, again with a neon indicator. The entire unit sits on a thick piece of soft-ish rubber to absorb vibration; the xformer sits on a shelf. Those measures help keep the noise down - within the room and substantially reduces transmitted noise through the fabric of the building.

Picking up on an earlier comment of yours, a 115v. - 0 - 115v. arrangement does provide for half the voltage, phase to earth, compared to 0 - 230 v., in the event of accidental contact. To me, that is worth having. Now perhaps this bit is pseudo-psychological, but somehow I feel happier working with equipment in the knowledge that that equipment is not connected to the 'raw mains'.

Al.

[McMurdo]

worth noting that if you fit a 13A socket outlet to the case of your transformer that the earthed mounting screws of the socket don't inadvertently contact the metal transformer enclosure, since 13A sockets have the fixing screws connected to the earth pin receptacle by brass strapping.
On my transformer, which is in a typical hammer-grey transformer box, the 13A outlet is the type intended for mini trunking where the m3.5 tapped holes are double insulated along the length of the moulding.

[mhennessy]

Just one thing to clarify: when you say "C-type RCD", do you mean RCBO (MCB and RCD combined)?

In the case of an MCB, the letter refers to the ballistics of the cutout. I'm not aware of such a classification when talking about RCDs - those are specified in terms of imbalance current (usually 30mA) and response time (usually 30ms). Those numbers are nominal, of course. Lots more detail here: http://electrical.theiet.org/wiring-matters/50/rcd/

Too many acronyms

Anyway, I can see that the additional transformer does serve a useful function in terms of RCD discrimination. IIRC, your workshop is in the attic, so that's potentially a lot of stairs between you and the consumer unit

The choice between balanced or unbalanced mains at your bench is personal, providing it's safe. I think that either should be fine, provided the mains earth is firmly carried across to the earth pins of the sockets (I know I said that earlier, but for the benefit of anyone else considering such a scheme...). Labelling on the affected sockets would be a nice touch.

Also good that there are separate fuses (and indicators) for each "phase".

[Skywave]

Mark - "Too many acronyms": RCD, RCBO, MCB, etc.: yes, I know - and I get confused - which I think is quite common! But the answer is simple: for now, I'll just call it the 'trip'.
This trip is labelled Proteus Bayer C32 with a single lever to reset it. It only switches the L feed. I think that makes it an MCB.

This was fitted to replace what was there before: an 'ordinary' B type - and the in-rush current to the xformer on the top floor caused that one to trip with annoying regularity - as you would expect. From the top floor to the consumer unit is 39 steps: not exactly convenient! With the C32 fitted, the 'annoyance' has gone.

Al.

[Argus25]

With all the variations on the theme of shock hazards it could could be confusing for a newcomer.

Perhaps it could be boiled down to a few simple instructions and warnings:

1) Always use an isolation transformer to power an AC/DC "hot chassis or mains transformer-less" radio or TV set where the chassis is not earthed.

2) Be aware that working inside this set its possible to receive a shock if you get connected between parts of the sets circuitry and its chassis, between two points (this is the same for all sets) so use precautions such as rubber gloves and the one hand in the pocket technique to avoid current passing through both arms and across your chest and heart area.

3) Be aware that if you connect the chassis (of the set you are working on with its isolation transformer) to any earth, say with the earth clip of a scope probe, (which effectively earths one side of the isolation transformer's output) then you can receive a shock from just touching one point in the set's circuitry and the current can pass through your hand and arm and body to the nearest other earth. Other earths could include the surface you are standing on (so rubber soled shoes help), or another earthed appliance nearby you are holding with the other hand.

In other words, as pointed out, earthing the appliance powered by the isolation transformer defeats the isolation.

4) Be aware that the isolation transformer itself defeats the dwelling's fuse box RCD so that no longer provides any protection.

5) Adding an RCD arrangement to an isolation transformer's output is not helpful.

Does everyone agree with that, or can this be added to & modified ?

[stuarth]

I'm not sure I agree with your first point about always using an isolation transformer.

As Mark has pointed out, an RCD is safer than a transformer. With an RCD, if you touch any part of the circuit at a significant voltage above mains neutral, (and you are not well insulated from earth) then the RCD will trip. The isolating transformer is only needed when you are using earthed test equipment.

A newcomer may well have neither an isolation transformer (expensive) nor earthed test equipment, so ensuring there is an RCD is in the mains supply path (on modern houses there will be, and if not, get a plug top RCD) is the best option.

Stuart