Isolating Transformer (UK)

[cprobertson1]

Hey folks!

   I'm in need of an isolating transformer for doing electrical repairs on live equipment - but I can't seem to find one (for the UK anyway!)

   Any ideas on where I can buy such a unit?

    Is it worth just building one? (and in that case, do I actually need to fork out for a full 3000kVA transformer (£165!) that would cover the max 13A from a plug socket, or do you reckon I'd get away with a lower power rating, perhaps the in 500-1000VA region?

   Many thanks folks!

[madires]

I got some questions too Do you need just an isolation transformer or an isolation transformer with a variac? What is the maximum power requirement of the devices you want to repair?

[cprobertson1]

Just regular isolation! (no variac).

Generally speaking it's small household items - off the top of my head, the most power-hungry thing I've repaired was a 400W television - I don't generally get things with motors or heating elements in them so the vast majority of stuff I repair is far below that!

I can get one built up from http://www.airlinktransformers.com/ with a 750VA toroidal transformer, MCB on the output with surge protection in an IP65 casing for £168ish (which is only a little more than the cost of a 3kVA transformer on it's own!)

I can shave £20 from it by dropping the output circuit breaker (I'll generally use a plug-in RCD anyway so it may not even be necessary)

[Richard Head]

If you are making a custom setup then also consider adding inrush limiting for the Variac. Variacs are notorious for tripping breakers on start-up due to their high remanent  flux caused by their tapewound structure. The output breaker is a good idea.

[cprobertson1]

If you are making a custom setup then also consider adding inrush limiting for the Variac. Variacs are notorious for tripping breakers on start-up due to their high remanent  flux caused by their tapewound structure. The output breaker is a good idea.

Oops! I didn't word my previous spec-list very well - inrush current / surge protection included on the input! - I think I'll keep the circuit breaker (as you said, it's a good idea!) - for the sake of £20 it's not much!

-750VA transformer (230V to 230V)
-UK plug w/ >=2m cable on input
-input single pole circuit breaker
-input surge protection
-UK socket on output
-2 pole circuit breaker on output
-IP60 or greater (dust proof) GRP case with handle

[madires]

Yes, a current inrush limiter for a toroidal transformer above 500VA is strongly recommended. Also spend the few extra bucks for the output circuit breaker. The RCD would only work for the primary winding, since the secondary is isolated 170 quid for the complete box seems fine to me.

[cprobertson1]

The RCD would only work for the primary winding, since the secondary is isolated

  - let's pretend I never said that

£170, delivered doesn't seem too bad tbh, I just like being stingy!

[nctnico]

I recommend against using an isolation transformer and get a good GFI/RCD instead. The chance you accidentally ground a piece of equipment by connecting test equipment (for example an oscilloscope) or something else to it is huge and this situation will make things worse where it comes to safety. Better get a differential probe with the right CAT ratings if you need to measure signals using an oscilloscope.

There have been several threads about isolation transformers and their inherent dangers so please read those. An isolation transformer is a stop-gap solution from the past!

[RGB255_0_0]

You have many more ways to kill the power from the mains on an RCD-protected circuit if you were to get into contact with a live wire.

Isolation transformers make some people think they are safe and forgo some precautions - such as using only one hand. If you were to get into circuit with an isolation transformer and pass enough current that exceeds your let-go threshold, there is almost no hope you can break the circuit on your own as there is no leakage.

At least with mains you could get pushed back onto a heating pipe or touch something else grounded to break the power.

Proper probes is the way forward.

[cprobertson1]

If you get one built you may as well get a 120V tap on it so you can quickly use it as a stepdown transformer. You also need to consider what ground/earth wiring you have present at the output.

That's a good shout! Having said that, I'm not sure how often I'd actually use the 110/120V output - but hey! The cost is negligible, so I might as well - never know when I might need it!

Would I be right in saying that the earth should be carried straight through from the input to the output side, with no direct connection to the neutral on the output (obviously with the TNC-S wiring in the house it will be bonded to neutral at the distirbution box of course)

You have many more ways to kill the power from the mains on an RCD-protected circuit if you were to get into contact with a live wire.

- But with an isolated transformer, you'd need to go out of your way to touch both live an neutral on the secondary side to get a shock - at which point the MCB should kick in (though you'll still get a jolt).

My understanding is that: yes, in principle there are more ways to cut the power in the event of a shock without the isolation; but the likelihood of a shock decreases dramatically with isolation in place; as long as you follow correct safety precautions (which you should be doing anyway!).

Obviously, if you're going in with both hands and touching things all willy-nilly then the risk of a lethal shock shoots up again - though if you're the sort of person to forego safety precautions then you probably shouldn't be playing with mains in the first place

Proper differential probes would of course be awesome, but I can't afford them - they'd actually cost more than my scope!

And even with an RCD in place, if you brush against a live wire or chassis while probing you'll make a path to earth via your body and you'll get a shock for at least 40 to 100 milliseconds before the RCD trips - which certainly gives you a fighting chance; but still, it'd be better to prevent a shock entirely

With isolation, the live is isolated from earth and its difficult to complete that pathway unless there is already an earth loop fault in the system (which you should have tested for before plugging it in! (i.e, as mentioned before, you need to be in contact with both live and neutral on the secondary - in this case we're using the earth loop to help us electrocute ourselves))

Feel free to correct me if I'm wrong though!

[RGB255_0_0]

Absolutely. As long as you know what you're doing you will always be fine, mains-isolated or not. Obviously if you've just drank 3 Red Bulls or 3 shots of Espresso you probably shouldn't be anywhere near live circuitry.

But "but the likelihood of a shock decreases dramatically with isolation [transformer] in place" I don't agree with. Wearing proper gear dramatically decreases shock hazard, not the isolation transformer. Test the RCD before working on the DUT. If you work on live equipment regularly, replace them regularly. Isolating yourself does the same job but if you happened to be in direct circuit with neutral and live, at least the circuit with RCD protection is there.

[RGB255_0_0]

http://www.hse.gov.uk/pubns/indg354.pdf

Supplies to equipment under test
Provide each item of equipment under test with its own test supply. These supplies
should be from designated sockets or terminals provided with covers, interlocked
with the supply isolator. The supplies should have suitable system protection
against overload and overcurrent in the event of faults, eg fuses. Note that:
? where an isolating transformer is used for the supply to the equipment under
test, it should comply with BS 61558 and a separate transformer used at every
test bench. If this is not reasonably practicable, the same isolating transformer
may be used for supplies to alternate benches, provided the risk of referencing
this supply to earth at any bench is properly controlled and the transformer
does not then have an unacceptably high leakage current;
? the supply from the isolating transformer should be provided from a single
socket outlet and clearly marked ‘only for use for making live equipment under
test’. No fixed wiring should be connected to the earth terminal of the outlet
socket. The face plate of the socket should be made of insulating material.
There must be no unnecessarily exposed live parts on equipment under test;
? in certain circumstances Class I equipment under test must be effectively earthed
unless supplied via an isolating transformer. This will bring with it an increased risk
of electric shock which may be minimised by using other precautions.
? when the equipment under test is Class I, any pre-existing earth fault must be
detected and corrected before making the equipment live. In the case of the
supply from an isolating transformer, failure to do this will mean that there may
be a hazardous shock risk, if there is a simultaneous contact between the
enclosure of the equipment and one or both poles of the isolated test supply;
? the integrity of the circuit protective conductor (earth) of all portable/
transportable Class I equipment must be re-tested after all test-bench work has
been completed, to make sure there are no earth faults before the equipment is
used again on a normal mains supply.

[MrSlack]

I don't use an isolation transformer when working on live equipment. I have a strict policy of doing the following though which is adhered to religiously:

1. Unplug, isolate.
2. Attach test probes/equipment.
3. Plug in, power up.
4. Observe measurement without touching the unit.
5. Goto 2.

If your measurement equipment is properly earthed and PAT tested it should be fine.

I don't own differential probes nor do I want to spend any cash on them, so I used an inexpensive AD8130 with two normal 10x scope probes and a secondary switchable attenuator. Works fine, CMRR drops off pretty quick at about 50MHz and noise creeps in so you need to use LP filter in trigger and it'd probably explode if you put more than 500v diff through it but it does the job well enough. I'll detail the design at some point. Cost about £10 in total and runs off a 9v battery. The worst possible current leakage from mains is about 500uA with the design.

[madires]

... or use the poor man's differential scope probe, i.e. two channels, ground leads removed, ch1 plus ch2 inverted and stay within the voltage limits of the scope.

[MrSlack]

That works fine for low frequency stuff, under about 20MHz. 3dB point on my 465B with some basic no brand probes is about 15MHz. Above that you need short probes; bring the amp to the signal rather than bringing the signal to the amp. 3dB point is about 70MHz now using some chopped down no brand probes and my diff probe amp.

[Ian.M]

Most of the personal safety benefits of an isolating transformer are moot assuming your bench supply is effectively RCD protected.  Also there is the issue that an isolating transformer doesn't detect a single fault to ground so any mistake, or even normal usage can lead to a false confidence that it is protecting the user.  A MCB on the output provides no protection against shock as the trip current is far far higher than that likely to cause electrocution, so its only useful to prevent equipment damage and fire.

However an isolating transformer is helpful when one is working on the primary side of SMPSUs,  non isolated PSUs etc. especially when working with more complex devices that require several steps to set up the desired operating conditions starting from fully powered down.   If you pre-attach suitable test points to clip the probe to so you do NOT have to hold the probe while making a measurement, you can even probe around the circuit with power on without excessive risk.

Differential probing is another matter - its rarely needed for working on PSUs, and usually Ch1-Ch2 mode is good enough.

[cprobertson1]

Most of the personal safety benefits of an isolating transformer are moot assuming your bench supply is effectively RCD protected.

See, my problem is that I'm repairing household goods: and getting all up inside them Depending on the product I can sometimes get away with just powering it from a bench supply and bypassing the mains side of things entirely - other times its not convenient to do so, or there are too many rails for me to power manually (rarely), or the problem is with the PSU itself - and of course, sometimes you just need more than the 60V my bench PSU can supply!

Its for cases like that I was thinking the isolated transformer would be invaluable - is that the case? Or would a good RCD a better option?

at which point the MCB should kick in (though you'll still get a jolt).

A MCB on the output provides no protection against shock as the trip current is far far higher than that likely to cause electrocution, so its only useful to prevent equipment damage and fire.

Aye! I realised that a little after I said it - a typical MCB trips at what, 3 Amps? 5? 13? Even with a 1-amp MCB, if you get to the point where you have an ampere flowing through you - cutting the power isn't going to help much (unless you happen to have a fondness for BBQ Engineers).

It doesn't take much to trigger fibrillation either - something meager like 30mA if I recall! (which, come to think of it, is below the tripping current for most RCDs - which, surprise surprise, is why you take precautions to avoid getting a shock in the first place )

[Ian.M]

I'm *NOT* talking about bench DC PSUs.  The AC mains supply to your bench should be protected by a 10mA RCBO.   Its a PITA if bench PCs, test & measurement equipment etc. and lighting are on the same RCBO as the D.U.T. and it also restricts the maximum safe current available to the D.U.T., so I'd suggest two 10mA RCD protected 13A 2.5mm^2 feeds to the bench, one for D.U.T. (with or without  the isolating transformer) and the other for all other bench equipment.

The only other practical alternative for more than 13A total bench load is to wire your bench as a 32A radial circuit from the main consumer unit, (possibly via a 32A socket so the licensed electrician doesn't have to sign off on your actual bench's wiring), to a sub-consumer unit at the bench, which lets you use 10mA DIN rail RCBOs for the individual bench circuits, + have a master switch to isolate the bench.

[Delta]

Most of the personal safety benefits of an isolating transformer are moot assuming your bench supply is effectively RCD protected.

See, my problem is that I'm repairing household goods: and getting all up inside them Depending on the product I can sometimes get away with just powering it from a bench supply and bypassing the mains side of things entirely - other times its not convenient to do so, or there are too many rails for me to power manually (rarely), or the problem is with the PSU itself - and of course, sometimes you just need more than the 60V my bench PSU can supply!

Its for cases like that I was thinking the isolated transformer would be invaluable - is that the case? Or would a good RCD a better option?

at which point the MCB should kick in (though you'll still get a jolt).

A MCB on the output provides no protection against shock as the trip current is far far higher than that likely to cause electrocution, so its only useful to prevent equipment damage and fire.

Aye! I realised that a little after I said it - a typical MCB trips at what, 3 Amps? 5? 13? Even with a 1-amp MCB, if you get to the point where you have an ampere flowing through you - cutting the power isn't going to help much (unless you happen to have a fondness for BBQ Engineers).

It doesn't take much to trigger fibrillation either - something meager like 30mA if I recall! (which, come to think of it, is below the tripping current for most RCDs - which, surprise surprise, is why you take precautions to avoid getting a shock in the first place )

Typical MCBs are 6A, 10A, 16A, 20A etc.

All RCDs used to protect against direct contact MUST trip by 30mA (for the reason you alluded to, even though you didn't realise).

Mate, no offence, but given your lack of understanding of mains protective devices etc, do you think you should be prodding around in live equipment?

[cprobertson1]

No offence taken

You're very right in your concern though! I've been thinking that a bit to myself as I've been replying to things on this thread if I'm being honest.

On the other hand - I generally power things from the bench supply if I can - it's a rare occurrence that I actually have something live exposed on the bench - the problem being that sometimes something dicky comes through that needs to be probed while either while powered up, or even while powering up - which was why I was inquiring about extra protection

--EDIT--
In fact, given the rarity of having to actually deal with live mains - I could probably get away with just refusing to repair anything I can't power safely from the bench PSU. That's probably a safer option overall.

[RGB255_0_0]

You're talking of uA to send your heart into fibrillation. Let-go threshold is way higher and depends on body mass though.

[Delta]

No offence taken

You're very right in your concern though! I've been thinking that a bit to myself as I've been replying to things on this thread if I'm being honest.

On the other hand - I generally power things from the bench supply if I can - it's a rare occurrence that I actually have something live exposed on the bench - the problem being that sometimes something dicky comes through that needs to be probed while either while powered up, or even while powering up - which was why I was inquiring about extra protection

--EDIT--
In fact, given the rarity of having to actually deal with live mains - I could probably get away with just refusing to repair anything I can't power safely from the bench PSU. That's probably a safer option overall.

The best piece of practical advice in this thread was to hook up your test probes with the DUT powered doon, then switch it on, take measurements, switch off, move probes etc...

I am no self-righteous "I ken better than you" type; I possibly owe my life to RCDs...     I spent years working on 24VDC instrumentation, and would probe around live panels will nilly, I then found myself being equally blasé when I had to go into a 110VAC relay logic panel, and a 415 or 690V TP starter panel.  I've had electricians grab "that daft ET" away from a panel and have a word. 
You just can't "get away with it" with mains.  This shit bites.

If you have to work on live mains equipment (and some times you do), do as mentioned above, and throwing a 5mA trip RCD (used on medical gear) on the supply wouldn't be a bad addition.

Stay safe, Brother.   

[cprobertson1]

The best piece of practical advice in this thread was to hook up your test probes with the DUT powered doon, then switch it on, take measurements, switch off, move probes etc...

Aye, that's basically what I've been doing - I'm not the kind to go taken chances if I can help it

This shit bites.

No kidding! Though I've met a surprisingly large number of electricians who say they've had a shock before.......

If you have to work on live mains equipment (and some times you do), do as mentioned above, and throwing a 5mA trip RCD (used on medical gear) on the supply wouldn't be a bad addition.

Stay safe, Brother.   

Any idea where the best place to purchase a low trip-current RCD like that? All the plug-in types I can find are 30mA, and I can only find RCDs as low as 10mA trip current (for DIN rail mounting)(but that's just checking Farnell and RS-components mind you)

PS - just out of curiosity, why are RCBOs so expensive?

[Ian.M]

5mA RCDs are mostly a US thing and are very rare in the UK, due to the UK's long history of an adequate PE on all socket outlets.  They are reported to be extremely prone to nuisance tripping.   10mA 230V RCDs are available - albeit a little difficult to find compared to 30mA RCDs.  e.g. https://www.amazon.co.uk/white-portable-adaptor-active-non-latching/dp/B00SVED44G should be suitable for field use
However, for a fixed bench installation, I would much prefer a reputable brand of DIN rail RCBO (or RCD if there is already MCB overload protection) + a suitable enclosure for them.

[SteveyG]

Most of the personal safety benefits of an isolating transformer are moot assuming your bench supply is effectively RCD protected.

RCDs are not suitable as primary protection against electric shock. Sure they help reduce the risk of death, but people have still gone into fibrillation from contact with an RCD protected circuit. A 'working' RCD only guarantees a maximum disconnection time not any kind of current limit.

Isolation transformers however are a very useful tool providing you understand what you are doing. I would never substitute working on some live gear with an RCD instead of an isolation transformer - they just aren't reliable enough.

RCBOs are not that expensive, but I would assume mainly because they are a bit more complex to manufacture and have quite a bit going on inside.