PROBLEM: Insulation Tests on Circuits with Electronic Loads

[@TeamSimpsonRacing]

Once again the people who write bs7671 don't live in the real world! The obvious answer to this is to put an off switch on the front of the socket, so the USB can be turned off before testing but instead of doing the obvious, lets just make an electricians life even harder 😪

[@johnaphone]

A socket outlet with 4 switches, not sure anyone would buy one. For initial verification its abit more work, but eicr i would say no more work cause you spot them when you unplugging equipment, and test at 250v

[@efixx]

BS 7671 isn't a product standard... 🤔

[@TeamSimpsonRacing]

@@efixx I know but that doesn't change the fact they don't live in the real world. For example, expecting us to IR test 100's of radial cables at first fix is absolutely ridiculous! Temporary linking it all is even more absurd. The USB's switched should be a requirement under british standards.

[@calmeilles]

​@@johnaphone Have the USB charger fed from one of the outlet's switched live - the then switch covers both.

[@5688gamble]

@@calmeilles Just have one switch for both- both on, both off, could perhaps be made into something people would find less intrusive, like a push switch by the USB plugs.

[@paulharris5409]

This has been a problem for years, way before integrated usb sockets. There has always been shaver sockets/lights, the tv booster in the loft you didn't know was there, the bathroom extractor that some twat installed without an isolator and countless other items.

[@johnlishman4946]

100% agree buddy wasted many hours looking for them

[@eric4709]

The bottom line here is... in the practical world .. there are really two scenarios: A new install, or.. an existing one where you are doing an EICR / fault find. In the case of a new install - there really shouldn't be an issue since surely it is plainly obvious ( you really don't need the "BS7671 bible" to tell you this) that you would test ALL cabling/wiring WITHOUT any active devices connected + circuits linked out !! - ( well....that is what WAGO connectors were invented for wasn't it?? - lol !!!) THAT way you PROVE the installation in terms wiring integrity ie - insulation etc ....... and of of course test continuity ie R1 +R2 etc while you are at it - because you should test continuity before testing IR anyway ) And of course - you should see "perfect" insulation - ie test meter out of range at >999 Meg ohm etc........ ......... BEFORE you connect any active devices such as usb sockets. And .... I would actually do a test on such a USB socket etc "on its own" to see exactly what insulation results ( ie leakage current) it gives - so that I KNOW what to expect when I connect it into my "new" circuit. Just blindly connecting it all up to a new circuit initially tested as above ... and then testing the "end result" is just going to add confusion "down the line" But... On an existing circuit - well ... all depends on how much you "know" about what you are going to test. In the REAL world - You wont necessarily know what is connected - and so what do you do? Doing insulation resistance (IR) tests "to the book" on old/existing circuits is always a challenge - and in my experience you will be lucky to get "good" values of IR on "first test" ( and what exactly do we mean by "good" - low values that might satisfy BS7671 might be ok from the BS7671 point of view - but in "my" book - low values - say in the 10's Megohms always raise suspicions. I actually always try to evaluate the circuit first as best as I can to see what is connected . ( where IS that socket behind the washing machine/cupboard/bed/wardrobe/wherever... and WHERE is that DIY pir light on an outside wall connected to ???) I like to start an IR test at 100V across line-neutral - to "see" what I get. IF you see silly values like 0 (which a IR tester will show for very low IR values) - or very low say 100K ohm - then you "know" something is likely to be connected somewhere. That is - it isnt a fault with the wiring insulation - but something "else" In fact - in the case of very low IR -I will then do a simple resistance test. In one case - I got a stable reading of 800ohms - obviously must be due to something connected which I hadn't yet found - -which turned out to be a WELL hidden TV signal amplifier in the loft that took ages to find. (it was a simple "transformer type" supply - and I was measuring in effect the "primary winding resistance" of that transformer in the TV amplifier supply - and 800 ohms was about right for such a small transformer. Once I'm confident that everything is "ok" - ill go to 250V - then 500. (or put another way - I NEVER EVER start with 500V - always 100V to start with ... then go from there. This isn't all about what BS7671 says. It is about the practical world and playing it out as a sensible "detective" or - as Forrest Gump would say / electrics is like a box of chocolates - you never know what you are going to get.

[@efixx]

Thanks for commenting! 👍

[@johnhoulihan4000]

Excellent post, thanks for taking the time.

[@yrification]

I’ll admit I was still doing to combined test on eicr because sometimes you can’t physically disconnect all loads or in-line stuff like that. Sometimes even doing the test at 250, you can get low readings. Something else to explain to property agents and landlords, that they won’t understand and will probably accept someone else’s lower price.

[@2Sorts]

The best solution would be to have a mechanism that disconnects the usb sockets when there’s nothing plugged into them

[@efixx]

Yup, this is a good solution till that becomes a Standard approach.

[@andycrask3531]

Yea like a longer plastic usb shaped plug that disconnects internally

[@jamesl1143]

I’m sure I’ve seen a brand that has switches on the back for this

[@garrywhiting8398]

...this rendering the installed charging outlet dependent upon another device. Not keep on that as a selling point. Device dead because the kids unplugged a lamp to save the planet ...

[@matthewgarlick2691]

A switch on the front would be a good idea so users could turn it off when not being used.

[@damianmcmullan7753]

Why is it not made compulsory for the manufactory to fit a dp switch on the usb points in the socket. This would address two problems problems first testing as removing sockets and refitting can lead to problems for the electrician and the questions we always get asked is it safe to leave a charge point on as all safety body's would advise users to disscount charger adapters. After 49 years in the electrical industry simply answer are always the best.

[@efixx]

Yup, maybe it'll get put into a standard in the future.

[@MervynPartin]

Regardless of the new double testing, on a personal level, I would not ever install outlets with combined USB sockets in my home. When I did my apprenticeship, it was drummed into us that the first thing we should learn about any apparatus was not how to turn it on, but how to switch it off. USB sockets are permanently on with no local means of isolation, so for me, they are taboo and will remain so. I have seen too many electronic devices fail catastrophically to trust anything that I can't switch off (easily) and the positioning of consumer units in houses often does not allow for easy access to switch off there.

[@efixx]

Pretty low risk I think.

[@Chris-hy6jy]

Agree totally. Absolutely no reason they can't power the USB from the switched side of the socket.

[@giacobbe1993]

What you on about 🥲, what they taught you? Usb is 5V so really low risk. In UK you are obsessed with isolation but yet neutral conductor in most cases is always "live". Eg. PME system

[@Ironic1950]

...and you have Nanny Gov telling you to switch off/unplug your phone charger, to eliminate tiny amounts of 'vampire' power, ignoring the wear involved in unplugging such things repeatedly...how does that work when your charger is built in, and are the old standard '5W' chargers, not the fancy 35W plug-ins...(The USB starts at 5v, then talks to the charger to see if it capable of higher powers, which means higher voltages, before going into high-power mode.)

[@Chris-hy6jy]

@@giacobbe1993 if you are saying "5V is really low risk" clearly you don't understand how SMPSs work and secondly, this video is about isolation testing between L and N. You can't do that with these things in circuit. A simple change by the manufacturers to include a switch would fix this problem but for some reason they don't do it.

[@Superkev001]

Wouldn't it be good for the sockets to have a switch for the usb part, so this could be turned off for testing. Simple solution instead of making life difficult

[@efixx]

Perhaps this will become an option in the future. 😃

[@paulbradbury4612]

That's a great clarification for new circuits but leaves questions over testing modified circuits. If an electrician makes a small modification to a circuit the two stage testing is not practical so it would be nice to know if just the second stage is adequate in this situation i.e. the live conductors connected to the earth.

[@tomcardale5596]

It's difficult to see how an IR test between (phase+neutral) and CPC at 250V DC could damage a device that wouldn't be damaged by 230V RMS AC. Whilst it's only for a fraction of a second, the line does get up above 300V with respect to earth 50 times in that second. Anyway. Wouldn't it be great if there was a little switch to turn off the USB elements of sockets? Would also mean you weren't in so much of a muddle if they let the magic smoke out later in their life.

[@Dog-whisperer7494]

So the boffins at the IET still can’t say with any degree of certainty that a 500v IR test will or will not damage such equipment. A L&N combined and tested to earth will not cause any damage or inaccurate readings as there is no potential difference between lines and neutral. Great video as always Joe 👍

[@efixx]

Thanks for commenting. 😃

[@MS-Patriot2]

The IET, despite all their convoluted fluff, can’t really be expected to specify whether a particular manufacturer’s circuit will have been designed to withstand a voltage in excess of 250V. L/N, L/cpc and N/cpc at 250V with a result >1Mohm is happy days. Any less and we need to hunt down anything still in circuit or isolate the fault. Why is it made so complicated ?

[@philipholme9911]

used to be known as test method two.

[@james.telfer]

Non-sparkys; have you ever been trying to commission your new installation on a site and wonder why something has just gone pop...? 🤔🤨🙄

[@efixx]

😬

[@avfczoff]

The regs might as well be in a different language. Make things more simple! Explain how you’re meant to do an insulation resistance test on a lighting circuit with 50 downlights & a ring final circuit with appliances plugged in that you can’t access to unplug (cooker hood & dishwasher as an example).

[@chrishorne3185]

Thanks for clearing that one up Joe, and especially the relevant regulations from BS 7671. 👍

[@efixx]

My pleasure! 😃

[@christastic100]

I not responsible for the thinking behind electrical testing but I have a question. If in normal use the circuit is single phase 230 volt (not a three phase circuit ) why would it need a stress test over twice the voltage between L and N anyway? I would love to know the actual practical reasons from those who make these recommendations and how this historical thing became a regulation for decades.

[@MartinE63]

The best solution by far is to ban USB socket outlets.

[@ElliottVeares]

Anyone doing a EICR on my place would be dreading the IR tests, finding not only USB sockets all through out my house but also Z-Wave automation modules behind all light switches.

[@frederikjacobs552]

I feel you 🤣 Luckily I have a shitload of individual breakers and differentials so it gets narrowed down pretty quickly.

[@smartchip]

I start with 250 volts, then click up to 500, Yeah yeah either earth -live conductors, (Update)Ok, he said it, but whatever,

[@G1ZQCArtwork]

During Portable Appliance Testing, it is important to remember to tie Live and neutral lines together in order to carry out the insulation tests. Particularly on Arcade Machines, or anything with CMOS chips that has an Earth. 500v DC is fine providing you are aware of this.! I always test Kettles, Toasters Irons and Gas Hob infighters this way, using either 500 or 1,000 v DC. These Class 1 items get the most abuse.

[@deang5622]

Utter rubbish. The voltage used for insulation testing does not get onto the CMOS devices.

[@G1ZQCArtwork]

@@deang5622 In that case, I hope you don't get any jobs insulation testing expensive electronic equipment.! If there is a break anywhere on the earth bonding, 500v can and will track across other places, best tie line and neutral together..... So they are at the same potential. Look it up, then come back and say sorry.

[@gillywild]

@@G1ZQCArtwork completely agree. I was involved in safety testing in the film industry for many years and I can confirm from reading circuit diagrams that the testing voltage can and does appear on the low voltage supply rails in many cases using switched mode power supplies.

[@G1ZQCArtwork]

@@gillywild On the subject of testing (or fault finding) of which I am a retired specialist, CPC continuity and insulation testing, has to be fully understood. E.G. Gas Hob / Oven Igniters are a good example. When plugged in, the CPC and connected Gas supply lines are bonded. I have come across times when the Earth connection on the appliance is broken, resistive or just not connected. The Igniter produces thousands of volts to light the flame, it needs to be properly grounded by the CPC, not the bonding, as it has Gas in it, leaky unions are perfectly possible, so stray sparks might make it go boom. Point being, if a CPC is faulty, and you jolt it with 500v DC, it can and will find a home path elsewhere eventually. CMOS devices get obliterated in an instant. So @deang5622 needs to apologise to all CMOS devices for putting them in danger.! The moral is, tie the Line and Neutral lines together BEFORE testing sensitive equipment.! David. 🥸

[@wiedapp]

Recently I came across two indoor cooling units in the company I work for that must have gotten an isolation test between L and N: The MOV was blown and the primary winding of the little transformer was blown as well. Yes, in both units. Needed a replacement PCB, as the transformer is so proprietary, you can't get it anywhere.

[@stevenhopkins2925]

I've been telling people this for ages and no-one believes me, I'll be sharing this with a few people to rub it in!

[@efixx]

Noice! Always happy to back you up Steven! 💪

[@gbelectricks]

Great bit of clarification there Joe. So to summarise, on initial verification 2stage testing @500V before equipment is connected and again at 250V after equipment is installed. And on periodic, L & N together tested against the MET with all earthing in place @ 250V only. So… how about when you make an alteration to a circuit? Do you just test the new bit initially at 500V then again at 250V once equipment is in place? Or.. test the whole circuit at @250V??? 🤔

[@Weaselspleen75]

Ok how about ir testing existing circuits after installing a new board for the eic? Do we go round disconnecting everything i.e 30 plus downlights/Boilers/spurs ect to test L-N/ L-E/N/E or do we test L&N to E The way i read it is initial verification was done when the cables were installed, could be 30+ years ago.L-N/ L-E/N/E carried out. So after a board change now this goes down as the indent 2 i,e after equipment was installed so test L&N to E.

[@TestGearJunkie.]

Personally I can't stand sockets with USB incorporated, they'res drawing power all the time. Not a lot admittedly, but if you've got a houseful of the things it'll add up over time. Why can't they be made so that either (a) a switch for the USB power is fitted, or (b) arranged so that the supply to the USB sockets only comes on when a plug is inserted..? I'll stick with my separate chargers thanks, or alternatively I'll charge stuff off the PC or laptop.

[@trevormorris1281]

Stop trying and frightening 6 month wonder sparks ? If you are qualified and competant you should know all of this!

[@darrenbone588]

Worse part when carrying out an eicr. Who is to know what sensitive equipment is plugged in where. Could have a loft with cctv nvr plugged in. I've even found things plugged in under the floorboards. Would i be liable to replace any items that get fried in the event of a 500v test? Not had one yet but never say never. Guess i just list it on my limitations?

[@StamateViorel]

Why not switch off the soket and usb at the same time inside the socket...

[@allsearpw3829]

AS per usual ? no one thinks of this on manufacturing products . just an extra switch on the front .

[@AmanaAbdelhamid]

i am first hhhh

[@gadgetman36]

Actually efixx were first! 😂

[@efixx]

Well in! 😃

[@AmanaAbdelhamid]

@@gadgetman36 hhhhhhh

[@rowles13]

Biggest flaws in the electrical industry is testing at 500.

[@StuartPearson-y8z]

If we have to do two IR tests & temporary connect vulnerable accessories, we are basically 2nd fixing twice? Who pays for the extra on labour?

[@efixx]

The customer...

[@jonhunter8737]

Simple solution. Are the rcd's/rcbo's tripping? Then the circuit is fine. Don't worry about insulation testing. It's all fine!!!

[@Ironic1950]

So when you have lots of USB faceplates, you would have to disconnect all of them...also, some 13A faceplates have LED indicators in place of neons, so make sure they are switched 'off'...

[@Chris-hy6jy]

So you don't have to test for electrical isolation between L & N on an established circuit?? An RCD will trip at 30mA when the isolation fails between live conductors and earth. What do we have to protect against leakage between L & N? A 32A MCB? Oh wonderful!

[@chiliphil64]

Maybe it is time to move to a 48v dc system as there is no need for 240vac for almost all devices in the home. I mean most! May need it for heating/cooking. However not required for vacuum cleaners TV, Computers etc:

[@guffermeister]

And then we introduce more potential problems, especially in ring finals as we can't easily test the condinuity of the ring once we've had to disturb the connections and can easily pinch insulation.... USB sockets are great but a pain for testing - there should be a means of isolating them. Once again the IET demonstrating that they are 10 years behind as per one of their vids the other week, As for "manufacturers instructions"? They go in the bin once it's signed off. No sparky is going to try and look up every device from 20 years ago to see if it can be tested at 500V. Also consider that 250AC is the RMS value. Peak is approx 350. So should testing be done at the upper limits of that? I get the need for a margin etc.

[@ciaranhughes8705]

On a job yesterday and the supply to their garage was a swa cable that consisted of a 6mm brown 6 x1mm blue with 6mm earth. So neutral is connected via 6, 1mm cables. I have never seen this cable before. Anyone here know if this is safe. The load should be shared but it's bizzare. I don't want to have to change the cable it's it's safe and sits correct with the regs but I can't find where it's stated otherwise. Cheers to anyone that replies

[@allsystemsdownagain9553]

If there is a fault the RCD will trip. This test is absolutely pointless. Most Electricians don't disconnect shit.

[@efixx]

L to N fault? Arc fault ?

[@tullgutten]

Not an actual electrician here but hobby electrician. The meggers is usually heavily limited output current around 3 to 10 mA so just a LED bulb would be just fine directly over it, ant the voltage never reaches 500 as it gets current limited. If something fails from using this then it isn't really a good design anyway on the pcb side. I wouldn't be too scared of it beeing a electrical load still connected when megging especially between live/neutral and earth. Good idea to try not charging up small digital appliances but if it has a indicator diode in it then that should be enough to save it and most high grade should have over voltage protection that will shunt it to earth

[@barrynee]

Why don't the manufacturers add a switch to disconnect the USBs when not in use. Better for testing and saves a small amount of electricity. It can't be just who thought of this, I'm sure.

[@newpuppy21]

Yeah on every job it would seem we’ll just remove everything, jump it out and wreck the paint finish and not get payed 😂👌

[@lifter98]

Could do with a video on how to use a voltage/continuity tester.. Explanation of the warbling sound when you probe an isolated circuit and pick up trace? How to test phase rotation, etc..

[@30mAkills]

Meggering is an art, not science.

[@whoisme678]

Assume nothing in a property.

[@Mike_5]

It seems like USB enabled Socket Outlets are becoming the new sort of in house SPD devices?

[@sbmorris2k6]

If L+N are connected together and tested to earth, would that not show a short circuit between L+N which would be where the most danger lies ???

[@efixx]

Other tests in the sequence would identify this.

[@SBTRIS]

How does this work with recessed integrated led lighting? You'll end up damaging half the ceil8ng plaster in the process

[@joshayres3211]

Why cant usb sockets have a switch on the back to disconnect the usb driver from the circuit this would be much easier then disconnecting all usb sockets?

[@benmarks911]

There's also smoke detector bases that communicate using radio frequencies that are permanently connected. So even if you've removed the detector itself there remains sensitive components at risk.

[@gabor.nadudvari]

I wonder that if in UK the sockets are switched why a dedicated switch in the center is not installed for the USB part.

[@abdulseaforth6930]

Thanks Joe. You evoked good comments too.

[@efixx]

Thank you! Lots to think about on this one for sure. 👍

[@vincenthigginbotham8729]

What about a Eicr when a system is in use and these tests like in hospitals where you are told to take a hike

[@gino2465]

I am confused now , its a sunday and i will revist the video lol

[@efixx]

😂 Come back to it on a Monday. 👍

[@gino2465]

Good morning. For the purpose of AM2 should we not do a verification calculation ie fraction, to make sure that each circuit result is acceptable . IE below 20m ohm from the osg brown book page 119.

[@retrozmachine1189]

From an electronics perspective, anything that is attached to the mains should be designed to cope with 500VDC across it. It already has to cope with 330V-ish [edit to fix typo] peak in normal circumstances and transients can easily reach 600V. It's not unusual to see transistors in switching power supplies with 600V ratings on them. Of course, this is the real world we live in and plenty of stuff is sub-par and will fail or be stressed by placing 500V across it, especially the cheap crap that comes out of certain countries in Asia. Personally I hate those wall plates with integrated USB power supplies. They are usually junk, and if the power supply fails you have to buy a new plate and have it fitted (well hopefully a pro would be engaged to fit it). Buy a quality separate plug in USB power supply and you'll probably still be using it in 10 years time. Same concept for those hard wired LED light fittings. Bleah. Standard bulb fittings for me, thanks.

[@TestGearJunkie.]

Standard bulb fittings with standard bulbs. I won't have an LED light in the house, horrible things. And don't get me started on the abominations that are called 'street lights' these days 🤬

[@Chris-hy6jy]

All of this could have been avoided if the manufacturers of USB sockets did the commonsense thing and fed the DC PSU off the switched side of the socket! Crazy!!

[@2Sorts]

On a new build (or a rewire) the single plates (up to 4 usb) or a double plate (up to 8) can be wired onto a 1mm 6A circuit. This could then be tested before the usb plate(s) are actually fitted.

[@Chris-hy6jy]

@@2Sorts but you still can't isolate them from the circuit after they're installed unless manufacturers stop cutting corners and start providing switches for them.

[@Hipyon]

I’m thinking the USB power supply is rectified with a diode and then to switch mode power supply for the 5 V so the reading would depend on the polarity of the test voltage If the test voltage reverse biased the diode you should not affect your reading

[@tomcardale5596]

it's most likely to be a full bridge rectifier so wouldn't make much difference.

[@Hipyon]

@@tomcardale5596 Actually most likely half wave because this half the voltage to start off with but I haven’t looked at many of these circuits the ones I have are halfway Full wave would use 3 more diodes I do find your videos very Informative and well presented

[@jix177]

One of your best video thumbnails! 😄

[@TestGearJunkie.]

Yep, that face says it all 🤣

[@JasonElson-e2q]

Love my kt66DL

[@TestGearJunkie.]

Don't like 'em. Megger 1741 for me, or Fluke 1664FC if I'm feeling contrary 😋

[@blow0me]

Clearly they need to design better equipment that does not have live and permanent connections/loads across them. Do these use electricity in operation when no USB is plugged in ?

[@procrastinator1842]

A while back I had to find a fault on a circuit with about 24 high bay leds on it....🫠🔫