Magnetic base LED machine light - WARNING MAY BE UNSAFE

I'd be interested to see the reverse of the PCB and what track separation they used. Also the markings and hopefully manufacturersname on the brown capacitor.

i would like to add to this thread as i bought one of these lights in the original thread and was using it on my mill. Within less than 5 minutes of use the flexible neck touched the quill and with a blue flash I received a mains current shock before the RCD cut in, Without even looking inside I threw the thing in the scrap. I reported the fault to the supplier and received a reply saying this has never happened before, but the fact is it did and anyone wearing a pacemaker could have been in serious trouble so I say to those praising this lamp BEWARE

Mike,

Can you dig it out of the bin and check what the fault was , if possible - was it mis-wired or a short to earth?

I'm changing the title of the thread.

Neil

Thanks for posting, Mike

Glad to know you lived to tell the tale.

MichaelG.

Mike P

Something does not add up in your description of the course of events: A blue flash would trip the breaker, indicating a short to the earthed quill, at this point of contact, in the miliseconds before the breaker opened, no voltage would be available to give you an electric shock. as all the metalwork would be at earth potential !

Malcolm

Some interesting comments about RCDs **LINK**

MichaelG.

OK, here are pics of both the component side and trace side of the (single-sided) PCB. For orientation, the PCB is flipped laterally (left to right) between these shots. Also I'm adding the schematic that I traced out from the PCB. If the diode across the output (zener?) is the wrong way round, blame me. I was never good at diode polarities/marking (so I became a Mechanical Engineer).

The PCB measures 0.95" across its flat sides.

As I mentioned earlier, all the markings on the brown capacitor are smeared and illegible.

Edited By Bandersnatch on 28/06/2016 22:46:34

Michael

There is an old saying... "It's the volts what jolts, but the mils that kills"

But the voltage is required to drive the current. A blue flash as the short occurs won't give enough voltage to jolt as a weld is most likely to occur between the components.

No blue flash: touch a live wire, circuit to earth through the body will give a jolt and trip the breaker.

The "let Go" phenomenom determines whether you live or die. Less than 6mA and you can let go, more than 22mA, with AC, results in Tetanic Contraction, which means the muscles permantly contract and you can't let go.

30K resistor looks like 3K with the photo colours.

Malcolm

Edited By Malcolm Parker-Lisberg on 28/06/2016 22:59:00

Is it possible that the short was from the 40V LED output to ground? I would think that would give a fair spark if it wasn't current-limited. It just seems from the construction of this thing that the most likely short to the metal conduit would be either damaged insulation on the LED wires or a short from the conduit to the wire connections on the LED PCB - particularly if the conduit on your particular sample protruded more than anticipated into the LED housing (see pic).

(Incidentally, I do have a pacemaker ..... or actually an ICD which is arguably worse)

Just to confirm, it is actually orange-black-orange.

.

Indeed there is.

However; a 'jolt' whilst working on a machine tool is something I prefer to mitigate against.

MichaelG.

Capacitive divider circuits as drawn by Bandersnatch are a respectable design. The safety critical component is the brown capacitor he marked with a question mark.

Whether or not the lamp is suitable for a workshop surely depends on the workshop. I'd be happy to use one in an indoor hobby room, less so in my garage workshop, and definitely not in a bathroom.

It's a pity Mike P hasn't been able to provide more information about his faulty unit. In the absence of evidence we can only guess what was wrong with it. I suspect an insulation fault put the metal sleeve at 40V DC. You would get quite a bang from a 100uF capacitor charged to 40V. And 40V is sufficient to give you a jolt if you have damp hands. Try licking the terminals of a PP3 9V battery!

You do have to be careful with this stuff. Many years ago I read of an unusual fatal accident resulting from a 30V DC shock. A colliery worker running to get out of the rain fell against a sheet of corrugated iron that in turn contacted an uninsulated low-voltage main. The poor chap was very unlucky: he was wet, sweaty and then the iron sheet maximised the flow of current due to the size of the surface area connection it provided.

Knowing that 30V can be fatal didn't stop me building a small Tesla Coil to make 1m sparks, but I only operated it after writing a risk assessment. 600W at 6000V; about 1MV out; x-rays; ozone; arcs; potential to interfere with a pacemaker etc. The high-voltage output isn't particularly dangerous because it's low current, but if the spark jumps to a mains connection like a lighting circuit, mains current can flow down the ionised path. That's very dangerous.

The fun part is the effect a Tesla Coil can have on a computer. From the patio, it turned on my daughter's Apple which was indoors 5 metres away and unplugged from the mains. I found the computer stalled on a blank white screen with the cooling fans full on. I had a bad few minutes waiting to see if would reboot successfully. Please don't tell her - it's a Secrecy in the Workshop issue.

Cheers,

Dave

I ask merely for interest, as mains electrical safety is out of my pay grade. As the supplied 2 pin plug isn't handed, you can get live on the side with the 100 ohm resistor in series, so the + output to the LEDs can be at mains potential? Or have I got it wrong?

If I'm right would we be better chopping off the supplied plug and connecting to a 3 pin, which is handed? Always supposing you can work out which is which.

Quite right Duncan. Imagine a poorly wired socket with live and neutral crossed. With this circuit plugged in but the socket switch off the positive terminal will rise to mains voltage.

Martin

Beyond my pay grade too Duncan so I risk being blasted by the professionals!

The circuit operates whichever way it is plugged in and you are quite right - it is possible to imagine a fault condition that would cause live mains to appear on the DC side. That sounds dangerous but a precondition of this type of circuit is that the wiring must be sealed inside the appliance.

Being sealed makes it rather less dangerous than a good old glass filament lamp. If you break the glass bulb of a filament lamp you expose a live conductor that's connected directly to the mains. Even a lighting circuit can deliver a good few amps. In comparison the capacitive divider always has components in circuit that tend to reduce the current that can flow. So it's rather safer than an ordinary light bulb.

That said, Mike P's lamp was unsafe because it wasn't "sealed" - that metal neck should never have had volts on it.

Cheers,

Dave

Hi all, I did not keep the remains of this lamp but from memory the item marked R2 on Bandersnatches picture had flashed over to the switch and the end of the flexible neck. Whilst appreciating all being said about RCDs and breakers I speak from live experience not theory and the shock I received was not theoretical but real that is why I posted to warn others of the danger. Mike p

The way I interpret Duncan's comment is that it wouldn't take a fault condition for the mains potential to appear at the output.

The plug (on mine at least) is 2-pin and not polarised in any way. The blades are identical. In fact there is no way to determine which is the "correct" (intended) way round to plug it in. Under those conditions the switch might be in either the live or neutral lead with the other lead unswitched.

Same disclaimer as Duncan, but it seems to me that the live side of the mains could then (in normal use) be present on the LED output terminals/wires (via the 100 ohm resistor and a diode). If they shorted somehow to the metallic flexible neck on Mike p's sample, he may indeed have had a mains short to ground.

You should always be wary when you're handling electricity of any sort, as it's always potentially lethal. Be sure everything is grounded and isolated properly and the right sort of wiring is used. And before trying anything on the mains try it on an RCD first.

Having said that, treating it like the monster under the bridge shouldn't stop you from learning about it.

Michael W

Edited By Michael Walters on 29/06/2016 22:30:35

Th whole of that PCB needs to be regarded as being 'Live' from a design point of view, none of it is isolated from the mains and the end of R2 is only 100 ohms away from the mains when plugged in so white is live. The bare tip could easily deliver a dangerous shock.

If the design is such that anything on that PCB isn't effectively isolated form metal parts I wouldn't touch one of these with a bargepole.

It is an acceptable design for sealed inside a non-conductive envelope like many LED lamps (as pointed out no worse than a filament lamp) but not suitable for inside an unearthed metal case.

Can anyone post a picture of how these PCBs fit in the case?

Neil

Am i reading this right? An AC current with no earthing terminal?

Ps. if anything from the pcb circuit was in contact with the casing it could become a conductor. 

Michael W

Edited By Michael Walters on 29/06/2016 22:48:25