Some LED lightbulbs on sale now in the UK are potentially LETHAL

Stuart Wright

AVForums Founder
Staff member
Joined
Jan 24, 2000
Messages
17,040
Solutions
1
Reaction score
14,135
Points
7,651
Location
Birmingham, UK
I want to preface this post by saying that I'm a big advocate of LED lights and I think we should all invest in more efficient, ultimately cheaper, greener LED light bulbs whenever we have the opportunity.
However, if someone gets killed by a dangerous LED, the publicity (specially if it is misreported in a typically hysterical, Daily Mail way) could put people off LED bulbs. So let's focus on the problem here - LEDs with SMDs (Surface Mounted Devices) - which have circuitry which is exposed.

This video shows a dangerous 'corn cob' LED which uses SMDs with exposed circuitry.
(Note that the speaker does use some 'emotional' language).

I have seen many GU10s and 'corn cob' bulb type LEDs with no protective cover over the actual LEDs and their exposed circuitry.
I'm not suggesting that all LED lights with exposed circuitry are dangerous, but how can you tell whether they are 100% safe or not?
You can buy SMD LEDs right now from various online stores including Amazon and ebay
B22 12W 60 LED 5630 SMD 960LM White Corn Spot Light Bulb Lamp AC 220-240V: Amazon.co.uk: Lighting
UK Stock E27 B22 SMD Corn Cob LED Bulbs Spot Lights Warm White Lamps | eBay
other UK retailers
LED Hut - 4.5 Watt - G9 High Power LED Bulb - LED Spotlights
Lamp Shop Online - 20w LED Corn Light MH/SON Replacement ES Cap - LampShopOnline
Energy Bulbs - G4 & G9 LED Capsule Bulbs - LED Light Bulbs
and the store referred to in the video
Eco-ERS

I actually have a corn cob LED above my head in my office here which I bought from Natsen_UK's ebay store in November 2012 (incidentally Natsen_UK no longer have items for sale on ebay).
corn.jpg

To think that it's potentially lethal to touch horrifies me!
An RCD in the lights would kill the supply before killing me (hopefully), but I don't know whether we have a separate RCD for our lights as not all houses do.

This corn cob LED here was cheap, though, and that is the issue. A quality LED bulb is still expensive to buy relative to the incandescent or halogen type bulb it replaces. So there is an understandable attraction to cheaper LED light bulbs.
Personally I would never buy another LED without proper protection between the LED circuitry and the outside world.
I've bought cheap and not-so-cheap LEDs and I think it's worth spending a few more quid to get better quality lights. You can read my guide to LED lights with a £££ saving calculator here.
 
Last edited:
-- As an eBay Associate, AVForums earns from qualifying purchases --
— As an Amazon Associate, AVForums earns from qualifying purchases —
I'm assuming this only applies to mains voltage LED bulbs rather than low voltage (12V) versions (which use an external transformer/driver). GU10 = mains, MR16 = 12V (usually). You can get 6V and 12V corn style LEDs too though they're a lot smaller than the one in the video.
 
I'm assuming this only applies to mains voltage LED bulbs rather than low voltage (12V) versions (which use an external transformer/driver). GU10 = mains, MR16 = 12V (usually). You can get 6V and 12V corn style LEDs too though they're a lot smaller than the one in the video.
I'm not an electrician, but yes, I would assume this only applies to 240 volt mains bulbs where there is enough current can be drawn to stop your heart.
Both the corn cob style bulb in the photo above and also a GU10 LED we owned (which has since failed way before the promised 25,000 hours) have exposed SMDs and are 240 volt.
 
A great posting, thanks for the heads up.

This is pretty scary. I've stayed away from buying corn cob LED bulbs because I think they are ugly. We have a fair few chandelier light fittings and corn cobs would look pretty disgusting in those , I will certainly be staying away from a safety perspective now. The CE and ROHS markings are obviously fake. Surely a plastic sleeve would add less than a quid to the overall cost?
Ikea have a good range of LED bulbs (either screw or GU fitting, no bayonet though :( ) I've got 6 of the 4 watt GU10's in my kitchen to replace 8 40watters, took a couple of days to get used to the change in light output but it is certainly good enough and a pretty good saving.
For bayonet bulbs I've been getting the cheapy halogens from the supermarket when they are on offer.
 
I'm no electrician or electronic engineer but don't all these bulbs run off 12V? The only difference being whether the control circuitry is onboard ie a GU10, bayonet or screw bulb or a separate transformer for your low voltage ones. When you power up those SMD strips you use a 12V power supply, so surely that is what the corn bulb is supplying to the LEDS.

I mean, did he at any time measure the voltage of the current drain? Because if it was at 12V then 80mA isn't going to kill you, or is it? And let's face it he wasn't measuring any potential leakage to earth, he was short circuiting it to the neutral terminal: would anyone be sticking their finger into a neutral socket and then touching a bulb in real life?. I'd like to see these things tested by a PAT engineer before getting too panicked by an hysterical Youtuber.

Not to say I'm not pleased for the heads up so we can look into whether there is an issue here, thanks for the link Stuart :thumbsup:

Edit: Having looked into it it does appear that depending which way around you put the bayonet in you have a 50% chance of there being live mains on the surface of the bulb :eek:

The same applies to GU10 bulbs with surface mounted LEDS where there are bare wires visible on the board...
 
Last edited:
I'm no electrician or electronic engineer but don't all these bulbs run off 12V? The only difference being whether the control circuitry is onboard ie a GU10, bayonet or screw bulb or a separate transformer for your low voltage ones. When you power up those SMD strips you use a 12V power supply, so surely that is what the corn bulb is supplying to the LEDS.

I mean, did he at any time measure the voltage of the current drain? Because if it was at 12V then 80mA isn't going to kill you, or is it? And let's face it he wasn't measuring any potential leakage to earth, he was short circuiting it to the neutral terminal: would anyone be sticking their finger into a neutral socket and then touching a bulb in real life?. I'd like to see these things tested by a PAT engineer before getting too panicked by an hysterical Youtuber.

Not to say I'm not pleased for the heads up so we can look into whether there is an issue here, thanks for the link Stuart :thumbsup:


It is possible for a few milliamps to kill.
 
Yes, I know it is, at 240V: that's why RCD's trip at 30mA. It wasn't clear from our Youtube friend whether it was mains voltage or not, I now know it probably was.
 
Last edited:
I'm really not up to speed with circuit diagrams etc, so don't really get why reversing the way the light is installed makes it so much worse. One question is what was the voltage across his little test circuit in each case? presumably not the same as the test lamp was much brighter in one case.

My other query refers to Johnny HUK2's comment about lower voltage LEDs with an external driver being safer. Why? The corncob bulb had an LED driver, albeit an internal one and that was dangerous, why does having an external driver make anything safer? Surely the quality of the driver is the key thing? In this case the LED driver was found wanting. I ask because I recently bought a 30V 50W LED on a chip* with an external driver supplied (just the circuit board and four wires, I had to take an inspired guess as to which was ac in and dc out). How do I test that to see if its dangerous. (It is dangerous, because I made the mistake of not looking away when I first set it up, my poor bloody retinas!)

The LEDs in the corn cob will undoubtedly run off low voltage, so how come they were surviving the 230V?
 
I'm no electrician or electronic engineer but don't all these bulbs run off 12V?
MR16s run off 12v, but all standard bulbs and GU10s are 240v. I knew someone when I was in my 20s who had a huge scar on the top of his hand. When asked about it, he told me that as a kid, he had poked the pins of a bulbless table lamp and it had blown a teaspoon sized scoop of the top of his hand off. That's 240v.
 
Yes, I know they have an input of 240V, I meant that the LEDs themselves all run on a low voltage regardless of whether they are 12V or 240V bulbs. In the mains bulbs the circuitry to drop the voltage to run the LEDs is part of the bulb. It's absolutely shocking that the 240V supply can run on the exposed circuit boards. Pun intended.

Another peril is that on the GU10 bulbs there is a risk that if they have a metal heatsink casing there is a risk that that could become live if the LED boards overheat, and they can do...
 
To put it into perspective, I'd rather have a made in China cheap and nasty GU10 LED Lamp in my house than a Halogen Lamp.

Not that I would ever buy them from China. I tend to stick with Megaman Lamps. £7.50 for a lamp is about right.
 
Yes, I know they have an input of 240V, I meant that the LEDs themselves all run on a low voltage regardless of whether they are 12V or 240V bulbs. In the mains bulbs the circuitry to drop the voltage to run the LEDs is part of the bulb. It's absolutely shocking that the 240V supply can run on the exposed circuit boards. Pun intended.

Another peril is that on the GU10 bulbs there is a risk that if they have a metal heatsink casing there is a risk that that could become live if the LED boards overheat, and they can do...
The problem is whether or not there is a good earth with a well isolated power supply.
With an ordinary lamp holder there isn't an earth.
 
Yes, the damn things are potentially lethal, I thought I said that ...

:)
 
nice post, i work as a lighting designer for an led manufacturer so will do my own testing tomorrow. Most smd ( surface mounted diodes) have a cover anyway and i have only seen them on the corn style lamps without covers. Be very careful on the corn style as its more the heat dissipation from internally in that could potentially melt the luminaire. Just been told to stay away from the corn style lamps in general, that and dimmable gu10s can be a bit of nightmare to pair correctly.
 
I'm really not up to speed with circuit diagrams etc, so don't really get why reversing the way the light is installed makes it so much worse.

If you look at the circuit diagram at around the 19:10 mark you'll see the difference. With it one way round (the "safer" way) you have between the live incoming supply and the LEDs:-

L1 (inductor)
|
V
R1, C2 and C3 (capacitors for current limiting, discharge resistor to drain the capacitors when switched off)
|
V
R2 (20 ohm current limiting resistor)
|
V
D1 or D2 (diodes for passing current in one direction; D1 carries the positive half of the AC waveform, D2 the negative half)
|
V
9x6 LED array

If connected the "nasty" way round, you have:-

F1 (10 ohm resistor with built-in fuse)
|
V
D3 or D4 (diodes for passing current in one direction; D4 carries the positive half of the AC waveform, D3 the negative half)
|
V
9x6 LED array

At 240 volts a 10 ohm resistor will quite happily pass 24 AMPS, obviously this is a fusible one so it won't pass 24 amps for long (not least because it'll be dissipating 5.7kW in the process) but needless to say it's not going to provide any protection from electric shock. Compare this with a 270 kilohm resistor in a mains test screwdriver (which at 240 volts will pass 0.9mA and dissipate 2.2mW).

The LEDs in the corn cob will undoubtedly run off low voltage, so how come they were surviving the 230V?
The LEDs will happily run at a higher voltage as long as the current is limited. A typical, cheap red LED (a standby light for example) has a forward voltage of about 2.5 volts, any higher than that and a resistor (R) is needed to limit the current to the desired amount, in the cheap LED case 20mA.
To run it at 12 volts:-
R=(Vin-Vled)/Iled), R=(12-2.5)/0.02, R=9.5/0.02, R=475 ohms (nearest standard resistor size would be 470 ohms so you'd use that)

The "dropper circuit" in the corn cob is a bit more sophisticated, using capacitors to limit current (don't ask me how, I've not looked into them in detail) but the circuit will be designed to limit the current for a 7W corn cob to 29mA, the task being divided between the inductor-capacitor-resistor on one leg of the supply and the fusible resistor on the other leg.
 
Last edited:
I mean, did he at any time measure the voltage of the current drain? Because if it was at 12V then 80mA isn't going to kill you, or is it? And let's face it he wasn't measuring any potential leakage to earth, he was short circuiting it to the neutral terminal: would anyone be sticking their finger into a neutral socket and then touching a bulb in real life?. I'd like to see these things tested by a PAT engineer before getting too panicked by an hysterical Youtuber.

Neutral and earth are at the same potential. In fact if your electricity supply is TN-C-S (a.k.a. PME or Protective Multiple Earthing) the earth connection is provided by the electricity company straight to the incoming neutral, which is spiked to earth at regular intervals all the way back to the 11kV supply transformer. I deliberately chose to measure to neutral so as not to trip the RCD that protects my sockets (although thinking about it now I do have a single socket in the attic that is non-RCD, I could've used that).

The video only applies to mains voltage bulbs. 12 volt bulbs (such as replacements for MR16 bulbs) should be perfectly safe as these are fed from a separate step-down transformer.

The 83mA was shown being drawn through a 25 watt pygmy bulb. Going into some technical sums (Ohm's Law):-
A 240 volt 25 watt pygmy bulb will draw 25/240 = 0.104 amps = 104 milliamps
The resistance of the bulb's filament is 240/0.104 = 2308 ohms

The voltage required to drive 83mA (0.083A) through the same bulb (wired as shown in the video) = 2308*0.083 = 191.6 volts

Therefore, under load, the LED contact under test was at 191.6 volts relative to neutral (and therefore earth).

(techies, please feel free to check these calculations in case I've messed them up).
 
You could try doing your sums with the right UK nominal voltage voltage, which has been 230 volts for a number of years :)

Some other things more general to the thread:-

1) Never rely on a RCD to save you - it's akin to driving without a seatbelt and using the airbag as a safety device.
2) Test your RCD's quarterly by using the test button - if it fails to trip FIRST TIME, get it replaced.
3) Mains 230v is low voltage, 12 volts DC is extra low voltage.
4) 80 milliamperes is generally accepted as being enough to kill you if your unfortunate enough to have a nice path across your chest.
5) in regards to saying would 80 miiliamps at 12 volts be enough to kill you, we'll it might, however even if a 12 volt circuit is running at 80 ma, when you touch it you create a path to earth with your resistance, which is nominally 1 k ohm, plus the resistance of whatever you are standing on, it is this value you plug into good old ohms law.
5) Never rely on a RCD......yep, repeating myself, but it can't be stressed enough!

It's the amps the kill and you calculate this with the given voltage and resistance, not by the current being drawn in the circuit you touch!


Neutral and earth are at the same potential. In fact if your electricity supply is TN-C-S (a.k.a. PME or Protective Multiple Earthing) the earth connection is provided by the electricity company straight to the incoming neutral, which is spiked to earth at regular intervals all the way back to the 11kV supply transformer. I deliberately chose to measure to neutral so as not to trip the RCD that protects my sockets (although thinking about it now I do have a single socket in the attic that is non-RCD, I could've used that).

The video only applies to mains voltage bulbs. 12 volt bulbs (such as replacements for MR16 bulbs) should be perfectly safe as these are fed from a separate step-down transformer.

The 83mA was shown being drawn through a 25 watt pygmy bulb. Going into some technical sums (Ohm's Law):-
A 240 volt 25 watt pygmy bulb will draw 25/240 = 0.104 amps = 104 milliamps
The resistance of the bulb's filament is 240/0.104 = 2308 ohms

The voltage required to drive 83mA (0.083A) through the same bulb (wired as shown in the video) = 2308*0.083 = 191.6 volts

Therefore, under load, the LED contact under test was at 191.6 volts relative to neutral (and therefore earth).

(techies, please feel free to check these calculations in case I've messed them up).
 
I'm pretty sure the voltage here's unchanged, only the tolerance. Supply used to test was in the region of 240 volts.

Sorely tempted to rerun the test this weekend with a separate, dedicated RCD and more meters (looks like my 3-phase power analyser's coming back out of hibernation :) ).

The plan:-
Set up RCD socket on non-RCD supply to provide power to the "test bench"
Use Metrel Instaltest to test the RCD to find its trip point
Use LEM Analyst 3QC to monitor the supply voltage coming to the test bench
Use Belkin power meter to monitor the power consumption (the LEM is better at measuring multi-kilowatt loads)
Use another old cable to provide an earth point rather than a neutral point
Use Fluke 79 III to measure current flowing through the pygmy test bulb
Use another meter to measure the voltage between the "live" side of the pygmy bulb and earth
Optionally use a 4kVA variac to drop the supply voltage to 230v or 220v to see how it affects the results
Throw in some calculations based on the test results

Possibly make a "test finger" (pork rib?) to probe the corn cob

Oh, and no swearing this time!

Any other suggestions before I do this will be greatly appreciated.
 
Last edited:
If you have the resources, get some corn cob bulbs from Amazon, eBay and other retailers as per the links in post #1.
It's one thing finding a small retailer Eco-ERS selling dangerous bulbs, but if it's possible to buy them through multiple retailers plus well known organisations, then it's a bigger deal.
 
That's a good call, something that was pointed out to me last night by a fellow YouTuber (Photonicinduction) over a few emails. It's payday, I'll do some shopping.

The bulb tested, incidentally, came from Eco Ers' eBay channel.

EDIT: 5 different bulbs from a couple of different sellers, on their way from Amazon UK, including one of the new COB (chip on board) designs, these are the ones with the yellow panels instead of discrete LEDs.
 
Last edited:
Hey AintBigAintClev, thanks for joining the thread and for the clarification(s) :thumbsup:

I've got some "bare" GU10s with exposed circuitry in my kitchen, which I'm now going to swap out for covered ones as a result of this thread.
 
5) Never rely on a RCD......yep, repeating myself, but it can't be stressed enough!
Amen to that, especially as in many houses the lighting circuit isn't RCD protected.
 
No worries, many thanks to everyone on the forum who's helping to highlight this issue. Constructive criticism welcome, because it results in defensive testing by me digging my heels in :)
 
Amen to that, especially as in many houses the lighting circuit isn't RCD protected.
Well the light I bought the corn cob for is (a) earthed brass, (b) non-RCD and (c) on a push-on-push-off button so, like a bathroom pullcord, you can't tell whether it's on or off if there's no bulb in it. Easiest way to fit a bulb to it is to have the holder in one hand and the bulb in the other. What could possibly go wrong...?
 

The latest video from AVForums

TV Buying Guide - Which TV Is Best For You?
Subscribe to our YouTube channel
Back
Top Bottom