Safe way to play with transformerless supplies?

[brainwash]

A few times per year I have to play with transformerless supplies and having already fried one board and shocked myself a few times I'm looking for a safer way to test these. Of course, common sense prevails but there must be something out there or some way to safely work with those.
I know about the isolation transformers but they are quite expensive and take quite a lot of space. Also, it would be nice to be able to limit the current somehow to a few mA since I don't power up heavy loads (boilers and stuff). Maybe there are some small 1:1 transformers in some pieces of equipment (modems? network cards?), chokes or other techniques.

[IanB]

You could try to construct a "safe" current limiting isolation transformer?

For example, if you put two 12 V transformers back to back you will get an isolated mains voltage output. If you now put a 12 V 1 A bulb in series with the two secondaries it will limit the current available at the output transformer. For example if you short the output it will try to draw a lot of current in the 12 V windings, but the bulb will limit this to less than 1 A, which means you won't get more than 100 mA on the output (assuming 230 V mains). But before you even reach 100 mA the voltage will drop significantly due to the voltage drop across the bulb.

[G7PSK]

If they are just wall wart types, you could use an electric shaver isolation transformer, they can be purchased quite cheaply.

[brainwash]

Back to back transformers is a great idea and you also get a visual indication of the current consumption.
My retired shavers have normal transformer supply (6/12V), I'm not sure I've ever seen and isolation transformer in any of them.

[IanB]

Shaver isolation transformers are a UK thing. It is not permitted to have a live mains socket in a bathroom so there are isolation transformers available in the UK that feed a special two pin shaver socket with an isolated 240 V output.

[Monkeh]

It is not permitted to have a live mains socket in a bathroom

Yes it is..

[IanB]

It is not permitted to have a live mains socket in a bathroom

Yes it is..

...if you live in a mansion with a 1000 sq ft bathroom and the mains socket is 50 ft away from any running water?

[G7PSK]

This is what I meant.

http://www.ebay.co.uk/itm/NEW-British-General-Dual-Voltage-Shaver-Socket-White-115v-230v-Bathroom-Mains-/140990979426?pt=UK_DIY_Material_Electrical_Fittings_MJ&hash=item20d3b79d62

[Monkeh]

It is not permitted to have a live mains socket in a bathroom

Yes it is..

...if you live in a mansion with a 1000 sq ft bathroom and the mains socket is 50 ft away from any running water?

3m from the edge of zone 1.

[codeboy2k]

  why do they  mandate this?? We don't have it here, and I've not heard of any cases of shaver electrocutions in North America...
People were always killing themselves with hair-dryers, but not so much anymore since we all use GFI / GFCI outlets now (called RCDs elsewhere)

By the way... to answer the OP's question, use a GFI on your bench too and use that to power your DUT.  If you touch the AC side, it will trigger the GFI and trip the breaker long before you feel it. ( ok, you may feel it a little bit )

But the isolation transformer with current limiting is also a good solution.

[dr_p]

I built myself this: https://www.eevblog.com/forum/testgear/smps-troubleshooting-aid/

Despite people sugesting to replace the glass fuses with DIN breakers, I didn't blow any so far, so I guess I'll keep them.

[firewalker]

There will always be charged capacitor waiting for you... 

Alexander.

[Smokey]

A few times per year I have to play with transformerless supplies and having already fried one board and shocked myself a few times I'm looking for a safer way to test these. Of course, common sense prevails....

Frying a board and shocking yourself isn't really using "common sense".  Having an isolation transformer isn't going to save you from blindly jamming probes (or your fingers apparently) into some circuit.  Just saying.

[brainwash]

Common sense also says that's the wrong kind of reply.
I got shocked most of the times by bad house wiring than by poking into circuits. Blew up a multimeter by probing a fly zapper. Made a few holes into a housing because I rested some beamer lamp leads on the [invisible] metallic paint. I'm sure everyone who is a tinkerer has quite a few stories to tell.
Nevertheless, playing safe is playing smart, there's always going to be that charged capacitor bang that makes you jump and hit something or just some alligator lead snapping shut.

[G7PSK]

why do they  mandate this?? We don't have it here, and I've not heard of any cases of shaver electrocutions in North America...
People were always killing themselves with hair-dryers, but not so much anymore since we all use GFI / GFCI outlets now (called RCDs elsewhere)

By the way... to answer the OP's question, use a GFI on your bench too and use that to power your DUT.  If you touch the AC side, it will trigger the GFI and trip the breaker long before you feel it. ( ok, you may feel it a little bit )

But the isolation transformer with current limiting is also a good solution.

Shaver sockets with the transformer built in are allowed over the basin it's general purpose power sockets that are not allowed within 3 meters of the basin, bath or shower but the daft thing is that you then see power sockets all over the kitchen most of which are less than 3 meters from the kitchen sink.

[Monkeh]

By the way... to answer the OP's question, use a GFI on your bench too and use that to power your DUT.  If you touch the AC side, it will trigger the GFI and trip the breaker long before you feel it. ( ok, you may feel it a little bit )

RCD != magical safety net.

Shaver sockets with the transformer built in are allowed over the basin it's general purpose power sockets that are not allowed within 3 meters of the basin, bath or shower but the daft thing is that you then see power sockets all over the kitchen most of which are less than 3 meters from the kitchen sink.

The sink in a bathroom does not create a zone.

[brainwash]

Thanks for the suggestions, I'll try to find one GFI as soon as I figure out how it's called in German.

I agree with the naysayers, there will always be a way to kill yourself no matter how many precautions you take. The thing is, I'm not the only one using the workbench, I have a few less technical (electrical) inclined friends that come sometimes and use it as well. Might as well make things better for everybody if there is a way.

It's a hobby, not my job, I did not have safety courses for this, just common sense. Also, I've got shocked at least 100 times (220V), starting at age 5, when I was using the apartment building three-phase power distributor as a xylophone. Time really seemed to stand still for a few seconds

Oh, another thing regarding venting about safety issues: you should watch PhotonVids on youtube, it really kicks the whole thing up a notch: high voltage, high current, alcohol, fire, xrays and a bunch of other stuff all at once. I haven't managed to trip the neighborhood's safety switch, yet.

[Monkeh]

Thanks for the suggestions, I'll try to find one GFI as soon as I figure out how it's called in German.

RCD.

I haven't managed to trip the neighborhood's safety switch, yet.

Safety.. switch?

Oh, you mean the 600A fuse which wouldn't blow if you shorted your supply out?

[madires]

Thanks for the suggestions, I'll try to find one GFI as soon as I figure out how it's called in German.

That's simple, just remove the "G" :-) It's a FI or Fehlerstromschalter. But pay attention to the trigger current, choose 30mA.

[codeboy2k]

RCD != magical safety net.

Of course not. You can always find a way to kill yourself. However, I never work on open AC powered devices without either powering the DUT using an isolation transformer or powering through a GFI/RCD circuit.  I would go so far as to say that the GFI/RCD is better than an isolation transformer because with an isolation transformer you're only safe if you touch one side of the isolated AC line.  If you touch both sides, you're going to feel it and there's nothing to protect you from that. However, a GFI/RCD still only keeps one side hot (the other side is always neutral and still bonded to earth at the panel). If you touch the hot side or both sides at the same time it will always trip the breaker.

And they are *very* *safe* .  It *will* trigger.  Hold the shank of a screwdriver and stick it into the hot side of a GFI/RCD plug while standing in water, you'll be fine  

I've never tried a GFI/RCD outlet hanging off an isolation transformer, but now that I think about it, this will still work, and may offer additional protections, by both eliminating the ground reference to the DUT *AND* providing for a safe shut-off in the event of an imbalance of current flowing out of and back into the isolated windings.   Just like a standard isolation transformer, there would be no more HOT and NEUTRAL sides, since neither side is bonded to ground anymore (this is of course, a tech version of the isolation transformer, not the consumer version)

The earth ground pin of a GFI/RCD outlet that is hooked up to the output of an isolation transformer like this would still be connected to earth ground.  You don't want to remove the chassis ground/earth ground of the DUT (or the enclosure of the isolation transformer), but the goal is to remove just the reference to earth that the AC power usually has.  The GFI/RCD would still trip if there was leakage current (i.e. all the current flowing out does not come back in). I can only see this as a good thing.

See Toddfuns video here too.

http://www.toddfun.com/2011/04/30/isolation_transformers/

[Rufus]

The earth ground pin of a GFI/RCD outlet that is hooked up to the output of an isolation transformer like this would still be connected to earth ground.  You don't want to remove the chassis ground/earth ground of the DUT (or the enclosure of the isolation transformer), but the goal is to remove just the reference to earth that the AC power usually has.  The GFI/RCD would still trip if there was leakage current (i.e. all the current flowing out does not come back in). I can only see this as a good thing.

The current coming out one terminal of an isolation transformer only wants to go back in the other terminal. Apart from a bit of stray capacitance across the transformer there will never be any current difference for an RCD to detect.

[Monkeh]

And they are *very* *safe* .  It *will* trigger.

Not if it's faulty. And there's no guarantee it will trigger fast enough, or at the right current level, unless you test it. Oh, and have I mentioned you can't rely on one to save your life even if it does trigger properly?

I've never tried a GFI/RCD outlet hanging off an isolation transformer, but now that I think about it, this will still work, and may offer additional protections, by both eliminating the ground reference to the DUT *AND* providing for a safe shut-off in the event of an imbalance of current flowing out of and back into the isolated windings.   Just like a standard isolation transformer, there would be no more HOT and NEUTRAL sides, since neither side is bonded to ground anymore (this is of course, a tech version of the isolation transformer, not the consumer version)

The earth ground pin of a GFI/RCD outlet that is hooked up to the output of an isolation transformer like this would still be connected to earth ground.  You don't want to remove the chassis ground/earth ground of the DUT (or the enclosure of the isolation transformer), but the goal is to remove just the reference to earth that the AC power usually has.  The GFI/RCD would still trip if there was leakage current (i.e. all the current flowing out does not come back in). I can only see this as a good thing.

....

Try thinking that one through some more. And don't touch mains until you figure it out.

[IanB]

If you touch the hot side or both sides at the same time it will always trip the breaker.

Of course not. If you happen to touch both the hot side and the neutral side at the same time you will just become a load on the supply, a load which the supply will be happy to energize 

If the breaker tripped on normal loads between live and neutral nothing would ever work.

And they are *very* *safe* .  It *will* trigger.  Hold the shank of a screwdriver and stick it into the hot side of a GFI/RCD plug while standing in water, you'll be fine  

Have you tried this experiment? I have (safely), and I can only say it hurts. 30 ms of mains AC is a hell of a jolt. I would not want to trust my life to it under unsafe circumstances.

The GFI/RCD would still trip if there was leakage current (i.e. all the current flowing out does not come back in).

Have you heard of Kirchhoff's current law? All the current flowing out always comes back in.

[codeboy2k]

The current coming out one terminal of an isolation transformer only wants to go back in the other terminal. Apart from a bit of stray capacitance across the transformer there will never be any current difference for an RCD to detect.

Try thinking that one through some more. And don't touch mains until you figure it out.

Oh yeah, now that I have thought about that some more ... its true.  There is no path to ground anymore .. doh! 
So I was wrong in assuming that the RCD would help with leakage currents even after an isolation transformer. Now I see it's not possible, there's nowhere for current to leak to even if touching a live end after an RCD device.  Yeah, I get it now 

[codeboy2k]

And they are *very* *safe* .  It *will* trigger.

Not if it's faulty. And there's no guarantee it will trigger fast enough, or at the right current level, unless you test it.

ok. So it's clear to me now that the isolation transformer is better than the RCD device for use on the bench.  Although I (clearly) don't do much mains work, I'll be sticking to isolation transformers whenever I do now.  Yes, I guess one could still die if the RCD is faulty or not fast enough.

Oh, and have I mentioned you can't rely on one to save your life even if it does trigger properly?

nope you didn't mention it .. go on.... I want to know more about this ...

[Monkeh]

Oh, and have I mentioned you can't rely on one to save your life even if it does trigger properly?

nope you didn't mention it .. go on.... I want to know more about this ...

The current has to pass through you to trip the device. And it obviously isn't limited to 30mA. It's quite possible for bad timing or an unknown heart condition to lead to a fatal shock even when an RCD correctly operates. They merely reduce (okay, drastically reduce) the risk.

[codeboy2k]

Of course not. If you happen to touch both the hot side and the neutral side at the same time you will just become a load on the supply, a load which the supply will be happy to energize 

If the breaker tripped on normal loads between live and neutral nothing would ever work.

No I realize a person can be a load, but the human model I had in my head when touching both sides (after an isolation transformer and RCD) was that the human formed (basically) a Y shaped resistor network to ground two hands, and a body, and feet on the ground... there would be leakage to ground and the return current would not equal the source current.. but then I was shown the light when others pointed out that the current from an isolation transformer always wants to return to the transformer i.e. there can never be a path to ground (because it's floating).

And they are *very* *safe* .  It *will* trigger.  Hold the shank of a screwdriver and stick it into the hot side of a GFI/RCD plug while standing in water, you'll be fine  

Have you tried this experiment? I have (safely), and I can only say it hurts. 30 ms of mains AC is a hell of a jolt. I would not want to trust my life to it under unsafe circumstances.

not with a screwdriver, but I have touched a live wire at the end of a long RCD fed socket chain.  It tingled and tripped.

The GFI/RCD would still trip if there was leakage current (i.e. all the current flowing out does not come back in).

Have you heard of Kirchhoff's current law? All the current flowing out always comes back in.

I thought it was just an advisory

[IanB]

not with a screwdriver, but I have touched a live wire at the end of a long RCD fed socket chain.  It tingled and tripped.

The experience here depends on the contact resistance between your skin and the live wire, and the availability of a return path for the current. For me it has varied between a mild tingling "oh, that's live" at one extreme to "Ow! #*($@&(##!!!" at the other. It all depends... 

[codeboy2k]

The current has to pass through you to trip the device. And it obviously isn't limited to 30mA. It's quite possible for bad timing or an unknown heart condition to lead to a fatal shock even when an RCD correctly operates. They merely reduce (okay, drastically reduce) the risk.

they need to be faster then

joking aside, I think I get it now.. so after all this, the conclusion is that the GFI/RCD should not be relied upon to save your life, but having one there can signficantly reduce the risk of a catastrophic accident.  And they are no good on floating circuits such as after an isolation transformer, which would seem much safer for use on the bench than relying upon the RCD.

 

[codeboy2k]

not with a screwdriver, but I have touched a live wire at the end of a long RCD fed socket chain.  It tingled and tripped.

The experience here depends on the contact resistance between your skin and the live wire, and the availability of a return path for the current. For me it has varied between a mild tingling "oh, that's live" at one extreme to "Ow! #*($@&(##!!!" at the other. It all depends... 

I've touched mains probably a 100 times... always felt a tingle nothing more... certainly never Ow! #*($@&(##!!!" .... I guess I have  too much resistance

[Monkeh]

not with a screwdriver, but I have touched a live wire at the end of a long RCD fed socket chain.  It tingled and tripped.

The experience here depends on the contact resistance between your skin and the live wire, and the availability of a return path for the current. For me it has varied between a mild tingling "oh, that's live" at one extreme to "Ow! #*($@&(##!!!" at the other. It all depends... 

I've touched mains probably a 100 times... always felt a tingle nothing more... certainly never Ow! #*($@&(##!!!" .... I guess I have  too much resistance

Just a puny low voltage.

[brainwash]

For a real shocker you should have a solid return path to the neutral/ground. It can be either your other hand or normal shoes on concrete. It's still up to skin resistance though, some people are more 'gifted'; I personally know a few electricians that wire everything while live and just use the back of the hand to test for voltage. I don't have the necessary dry skin to try that.

A nice story (of what not to do) was told by a friend of mine: he was wiring up some country house doorbell at the gate. During this, he always felt small bites on his feet and was constantly checking for ticks, spiders, mosquitoes and such. In the end it proved to be the wet morning grass touching his feet over the rubber flip-flops (thongs?) providing the temporary return path to ground. Somehow he always has stories to tell about how he thought the safety was off but in reality it was not.

Oh, regarding the 'tripping' of the neighborhood fuse from that guy (PhotonVids), I could not find the video. He was drawing more than 100A from the network. Maybe he's here to tell the story, though that's normally something I would not publicly admit. It was caught on video though.

[Zero999]

The sink in a bathroom does not create a zone.

It's part of the wiring regulations which divide that bathroom up into "zones" where you're allowed to put sockets and switches.

The isolation transformer vs RCD debate is an interesting one. I think both RCDs and isolation transformers have their unique advantages and disadvantages when it comes to electrical safety. An RCD is good because it will trip if there's any leakage to earth but if you form a direct path between live and neutral and there isn't enough earth leakage it won't help, it's a mechanical component which could fail to trip and it's no good if you want to test your circuit with a oscilloscope. An isolation transformer doesn't rely on mechanical components which can fail and you can test the circuit with an oscilloscope  but it gives no protection if you form part of the circuit, even if you are a good path to earth.

On balance, I think an isolation transformer is probably most suited to this application because it ensures the whole circuit is isolated so you can connect it to other equipment such as an oscilloscope.

Using two back-to-back transformers provides a greater deal of isolation and a lower voltage supply but it occupies more room. is less efficiency, has a higher voltage loss and poorer regulation due to the impedance being higher.

You can use a shaver isolation transformer or buy a small isolation transformer quite cheaply. I made my own 100VA isolation transformer by rewinding a transformer I got from work because it was going to be scrapped.

[richard.cs]

I bought a 240 VA isolation transformer on ebay - just the transformer, no case connectors etc. It was about £35 including shipping from Germany.

I put it in an old battery charger case with a few extras. There's a neon connected between mains earth and the transformer centre tap, it doesn't pass enough current for a dangerous shock but it does light up as a warning if one side of the output gets grounded, either by a fault or intentionally through a scope or similar. There's also a lamp holder on the top in series with the output that provides variable current limiting, that can be shunted with a switch when not needed. Note that this current limiting is in the region of hundreds of mA depending on the lamp fitted. It doesn't make anything safer but it can protect components.

[Monkeh]

The sink in a bathroom does not create a zone.

It's part of the wiring regulations which divide that bathroom up into "zones" where you're allowed to put sockets and switches.

The sink isn't.

[Zero999]

The sink is zone 2.

[Monkeh]

The sink is zone 2.

Perhaps that non-official drawing claims it is, the regulations do not.

[brainwash]

I put a little thought into the process and came up with a safer way to do this.
One of my pending projects involves making a new uC board for a capsule coffee machine which works directly on the mains. The previous design used some Holtek chip powered at 5V through a zener diode. I already have the new board ready, based on MSP430 but I did not want to plug the thing up until I had a safe way to power it.
My idea is to have the board powered up (at least temporarily) from a 2xAA battery holder switched on by an optoisolator. The draw is less than 10mA so it should work, at least in theory. The old board with the SCR and everything is still in place, I'm just piggybacking on the uC pads.
If anyone has a better idea I'm open to any suggestions.
My reason for the thread was that last week I was trying to measure how the [hall] water sensor works and have gotten quite a shock. This was on an identical working machine but have no idea where I got the 220 voltage differential from. IMHO it's poor design especially for something working with/in water.

[Gigawatt]

Tun it off !

[SeanB]

I bought a 240 VA isolation transformer on ebay - just the transformer, no case connectors etc. It was about £35 including shipping from Germany.

I put it in an old battery charger case with a few extras. There's a neon connected between mains earth and the transformer centre tap, it doesn't pass enough current for a dangerous shock but it does light up as a warning if one side of the output gets grounded, either by a fault or intentionally through a scope or similar. There's also a lamp holder on the top in series with the output that provides variable current limiting, that can be shunted with a switch when not needed. Note that this current limiting is in the region of hundreds of mA depending on the lamp fitted. It doesn't make anything safer but it can protect components.

So basically you made an isolation transformer with a fault warning indicator.

[brainwash]

Tun it off !

What, the Senseo? It's probably one of the the most popular models in the world, so they probably got it right and I got it wrong.