Do you guys use an isolation transformer for your lab/workbench?

[bigfoot22]

deleted

[tooki]

Let’s put it this way: if you don’t thoroughly understand why and when you’d want an isolation transformer, and what it does and doesn’t protect you from, then you shouldn’t work on mains-powered stuff yet, specifically, things with a live chassis.

If you get one, make sure you don’t let it lull you into a sense of false security, such that you actually let your guard down around mains voltages.

As for the specific question of using one to isolate all of your test equipment: an emphatic no!!!!

If you want to add safety to your bench, install a sensitive GFCI (more sensitive than the one your home may already have installed).

[bdunham7]

Just asking because I'm fitting out my workbench and wondered if I shouldn't isolate all of the equipment from the mains with the aid of an isolation transformer then do noise filtering on it.

Why no grounding on isolation transformers?

There's probably not a good reason to do so and most test equipment will isolate their innards from the mains anyway via the power supply transformer or whatever.  Y-cap leakage can still be an issue in some cases though.

Isolation transformers typically pass through the ground connection but that is a separate issue.  You typically would not want to break that ground, certainly not by default.

[AVGresponding]

Let’s put it this way: if you don’t thoroughly understand why and when you’d want an isolation transformer, and what it does and doesn’t protect you from, then you shouldn’t work on mains-powered stuff yet, specifically, things with a live chassis.

If you get one, make sure you don’t let it lull you into a sense of false security, such that you actually let your guard down around mains voltages.

As for the specific question of using one to isolate all of your test equipment: an emphatic no!!!!

If you want to add safety to your bench, install a sensitive GFCI (more sensitive than the one your home may already have installed).

This  ^  100%

Simple quotepost because it bears repeating.

[AVGresponding]

I don't know how things go in the US, but here in the UK, that would NOT MEET electrical safety standards, and would be technically be illegal to sell.

@bigfoot22 sorry, forgot you're an Aussie. I can't imagine it would pass safety regs there either.

[bdunham7]

Not for this one apparently:
https://www.ebay.com.au/itm/254536293226

You don't want that for powering your bench unless you plan on modifying it with a ground somewhere.  It might be appropriate for powering a DUT in certain cases, but I don't like the fact that it has an earth socket that isn't connected to anything.  If there was a device to measure the voltage at that isolated earth socket to actual earth and disconnect the power if that potential rises above SELV voltage, then it might be acceptable. 

[AVGresponding]

Primarily the reason it's non-compliant is that it has the capability to plug in a class 1 (requires an earth connection) device, without having the required earth connection. If it had 2-pin outlets it would comply.

[nightfire]

As a proud owner of an old isolation transformer of the 70ies- has anyone some hint of the actual IEC (Or VDE/german) standard revision for those?

[TimFox]

An electrostatic shield between windings greatly reduces capacitive coupling of high-frequency common-mode garbage (on line and neutral) to the secondary, and such a transformer usually connects the PE ground to the output.
Topaz transformers used to be popular in critical laboratory applications, and can be found at surplus vendors and eBay, but seem to be hard to locate since Schneider bought SquareD. 
They look nice and have very low primary-secondary capacitance.
Here are some at a large US surplus vendor:  https://www.surplussales.com/class/inductor/isolation_1k-4k.html
I couldn't find anything Googling for Schneider Topaz.

[Gregg]

I have a medical grade toroidal isolation transformer dual voltage windings primary and secondary and an electrostatic shield that I rewired the secondary so that I can easily disconnect the output receptacle from the input mains ground via a jumper link on the front where it is readily visible as to the ground status.  This also allows connecting a jumper to an ammeter etc in the ground circuit or to connect to an isolated ground. The ground binding posts are connected to 4mm banana jacks on the side of the case.
I also installed a switch type circuit breaker and 4mm banana jacks on the output. 
It has come in handy in a few cases but it certainly doesn't reside on the bench top.

[nctnico]

Yeah I have no intentions of using it for my lab/workbench after reading your helpful replies. If it doesn't pass through earth then its dangerous anyway. I'll get an RCD/GFCI for my setup.

And get CAT rated high voltage differential probes. These are very affordable (from MicSig for example) and will provide safety while measuring in mains powered circuits.

Also think long and hard if you even need to provide mains power to a mains powered device. Switching power supplies always have a low voltage supply in the primary side. If you use a bench power supply to power the primary side (and a different channel to apply some power to the secondary side) you can test 99.9% of a power supply without even connecting it to mains.

[NiHaoMike]

I have one that does about 100W, made by taking apart two halogen lamp transformers and reassembling one with two primary windings. I use it more for the current limiting although the isolation is definitely a plus.

For more power, I run a 1000W power inverter from a surplus 1400W server PSU. It's rare that I have to do that, even working on fairly large PSUs the 100W isolation transformer is sufficient to see if it at least powers up correctly.

[Electro Detective]

Study isolation transformers in detail till you understand what type for what task,

how to test the actual level of input to output 'isolation', chassis relations,

and that it will do the task before using it,

but not before understanding the source of the incoming power from your mains switchboard first, the earth bond/grounding thing etc


It's actually SUPER SIMPLE even using generic isolation transformers once you have understood the above,

with the wiring option of various degrees of 'isolation' from the mains and or connected suspect equipment etc


My rig has the option of any type of isolation needed,
regulated, varied or limited AC, 
and options to use GFCI or RCD,
and be able to externally test trip them also to ensure they actually do work


i.e. If risky business    I simulate the snafu first, then hook up and power up

and there's a UPS rig too but that's another story as to how it's transformer works on backup disconnected from the mains


My intentions when learning about this stuff were really to be personally 'isolated' from hospital beds
and or a one time permanent earth/ground connection, after a few good overconfident idiot zaps and walkaways over the years. 

'Isolation' has way too many variables for people unfamiliar with all the voodoo and confusion,
to either zap their equipment and themselves,
or cop spark shrapnel in the eyes,

or worse still, send panic'd kitty and fido packing down the street 
into an oncoming beer swilling, bad hair day, 'friday tradie' ute, with cranked up mindless detuned metal mayhem trash. 

Isolation: No research? = No Do ! 

[Brumby]

Let’s put it this way: if you don’t thoroughly understand why and when you’d want an isolation transformer, and what it does and doesn’t protect you from, then you shouldn’t work on mains-powered stuff yet, specifically, things with a live chassis.

If you get one, make sure you don’t let it lull you into a sense of false security, such that you actually let your guard down around mains voltages.

As for the specific question of using one to isolate all of your test equipment: an emphatic no!!!!

If you want to add safety to your bench, install a sensitive GFCI (more sensitive than the one your home may already have installed).

This  ^  100%

Simple quotepost because it bears repeating.

I was going to do exactly the same - for the same reason.

[rstofer]

If you want to add safety to your bench, install a sensitive GFCI (more sensitive than the one your home may already have installed).

In the US, UL standard 943 covers GFCI devices and they trip at 4 to 6 mA of ground current.  I don't think you will find anything more sensitive.  These are used for individual branch circuits, we do not trip the panel main breaker with a GFCI nor do we use something like the UK RCD device.  Typical ratings are for 125/250V up to 50A

It is my understanding (from a quick read on Google) that UK RCD devices trip at 18..23 mA (30 mA max).

[tooki]

If you want to add safety to your bench, install a sensitive GFCI (more sensitive than the one your home may already have installed).

In the US, UL standard 943 covers GFCI devices and they trip at 4 to 6 mA of ground current.  I don't think you will find anything more sensitive.  These are used for individual branch circuits, we do not trip the panel main breaker with a GFCI nor do we use something like the UK RCD device.  Typical ratings are for 125/250V up to 50A

It is my understanding (from a quick read on Google) that UK RCD devices trip at 18..23 mA (30 mA max).

The US is a bit of an anomaly in this regard, in that it uses very low current GFCIs. Elsewhere in the world, 30mA is typical. At my previous job (vocational training center) here in Switzerland, they have 10mA GFCIs protecting each workbench, both so they trip sooner if there’s a fault, and so that it won’t take out the entire lab if there is.

[Infraviolet]

I've heard they can be useful for some forms of high (mains) voltage work, haven't really looked in to the accuracy of such rumours or reasons behind them, but for lower voltage stuff I'd expect they could be a practical menace. Lets put it this way, the things you're designing are probably going to end up being battery powered or run from wall wart supplies, either being imperfect. If you get used to producing designs in a "perfected" environment where everything is isolated from the ground to avoid some particular sort of noise then things you design might then not work in the real world where that noise is present and the isolation isn't.

That said, a lot of low voltage stuff will be isolated anyway, low voltage wall warts are floating and ungrounded relative to mains. And bench power supplies often have a phsyically grounded pin socket on them, but unless you use it the supplies will be floating relative to mains.

[timeandfrequency]

Hi bigfoot22,

tooki's first post sincerely provides you with good advice.

When I have to probe into a device (DUT) with mains/live exposed parts (like boards with in situ SMPS), I do not use an isolation transformer (even if I've got one).
My DUT is powered via a 10mA RCD/GFCI (trips at about 6-8 mA) and the sole allowed TE are first tier (bench) multimeters with IEC1010 compliant accessories, and a truly isolated oscilloscope (Scopemeter) with adequate fully isolated probes.

You may also check out these test leads with selectable safety category and double-insulated silicon wire.  Same stuff but with banana connexion.
Another set from Pomona.

If the DUT has a metallic enclosure, I always use a strong alligator clip to make an additional link from chassis to earth.
Even if it's not practical, wearing class 00 or class 0 insulated gloves provides an additional level of safety.

[tautech]

Hi bigfoot22,

tooki's first post sincerely provides you with good advice.

When I have to probe into a device (DUT) with mains/live exposed parts (like boards with in situ SMPS), I do not use an isolation transformer (even if I've got one).
My DUT is powered via a 10mA RCD/GFCI and the sole allowed TE are first tier (bench) multimeters, and a truly isolated oscilloscope (Scopemeter) with adequate fully isolated probes.

If the DUT has a metallic enclosure, I always use a strong alligator clip to make an additional link from chassis to earth.
Even if it's not practical, wearing /]class 00 or class 0 insulated gloves provides an additional level of safety.

Gloves implies 2 hands are used which for mains breaks the good old rule of keeping one hand in the back pocket when working on mains.

[nctnico]

Has anyone mentioned not to wear jewelry?

[timeandfrequency]

[...]

Gloves implies 2 hands are used which for mains breaks the good old rule of keeping one hand in the back pocket when working on mains.

The well-known rule is undoubtfully good advice when working with bare hands. If you need both hands to probe inside a DUT, wearing insulated gloves usually provides a sufficient level of safety. I couldn't find any seller that only sells one glove.

[tautech]

I couldn't find any seller that only sells one glove.

[themadhippy]

I couldn't find any seller that only sells one glove

Ya didny look very hard,although buying a pair might be cheaper
https://www.smoothradio.com/artists/michael-jackson/white-glove-sold-auction-value/

[Electro Detective]

I couldn't find any seller that only sells one glove.

Bob has you covered:     www.bobthewelder.com.au/products/xt-left-handed-welding-gloves-green-long

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

FWIW folks: 30ma RCDs rarely trip at 30ma

Anywhere from 15 to 28ma  is what usually happens, and a good unit will have about 2 to 3ma repeatability

i.e. test the suckers properly before trusting your life and gear with them,
better still, use two RCDs  piggy backed.
One of the two 'may' trip 'faster', or both will work if the trip current approaches zap city.

But they won't do you much good with an electrical fault or short in the chain, so..FUSE UP, just past the DUTs current rating,

and you won't need an imitation rabbits foot or garlic necklace 

[Electro Detective]

Welding gloves would be too thick I'm afraid.

I already use Nitrile gloves. A highly coveted stockpile of exactly 1 box of them. Got no idea where I'm going to get the second box from but that is a problem for the future.

Once your hands get sweaty in those, you better think of possible conductivity consequences,
especially if there's a pinhole or paper cut slice you can't see, especially if salt is a favorite in your diet. 

I usually have cheap thin cotton gloves on first, then work large nitrile or similar gloves over. You get a few good dry hours use from that.

Or if it has to be tight thin nitriles for finicky handling, then I swap them over for dry ones once I feel the hands getting sweaty

The sweaty gloves get 'flipped out' and put out to dry etc

= a bit hard to put on again, but the effort is great for the wallet