Stray Voltage on T8 LED Heatsink

Did line and neutral get cross-connected somewhere in supplying the fixture(s)?

If not, I wonder how UL or other lab missed this.

I think there was a similar CR4 thread about non-isolated LED supplies.

How is the power connected to the heatsink? You need to trace to the source...

I have disassembled one tube completely. I cannot see how anything but phantom could be induced into the heatsink. When I tried to bond the heatsink, pop...no more light...smps fried.

Bonding out phantom/induced voltage shouldn't fry a power supply...right?

Did line and neutral get cross-connected somewhere in supplying the fixture(s)?

If not, I wonder how UL or other lab missed this.

I think there was a similar CR4 thread about non-isolated LED supplies.

No breakers are tripping, so I can be somewhat assured that L-N are not crossing.

Phillips had a recall for some consumer mogul based bulbs, but that was attributed to a loose wire internal to the bulb contacting the heat sink. http://www.usa.philips.com/c-e/us-recall/led-lamp.html

I didn't mean crossing to each other, but being crossed (i.e., interchanged) before being connected to the fixture.

oh, no. checked that.

Perhaps a heat sink insulator is missing, faulty or the isolation has been defeated (for example by an incorrectly located metal instead of nylon screw)?

I have come across live electronic switch cases electrically insulated from accessible heat sinks (sometimes earthed other times double insulated) in consumer and commercial products, but don't have any experience with this when it comes to modern LED light fittings with heat sinks.

Is it something as simple as this?

I have disassemble faulty and functional lamps. They all look identical. The LED circuit card is friction fit into the heatsink. The end connectors are screwed in, but there is no connection between the SMPS and the heatsink. The SMPS is inside of a insulating sleeve. The SMPS/insulation and the circuit card are held with silicone as well.

I'll disassemble another one tomorrow with pics, maybe. I might add some screenshots with the readings, too.

Lots of pictures would be great

So how does the heat get transferred to the heat sink? Is it possible that there's some direct contact between driver semis and the heatsink that isn't apparent when the power supply is slid out from the heatsink.

Even more innocuous would be a semi whose heatsink is specified to be galvanically isolated from the die, but somebody saved a penny per device by purchasing a functionally equivalent part number but without the isolation.

I did a little googling on the subject and I found this.

Yeah, I am trying to hit it up on three fronts:

1. Electricians

2. Electrical Engineers

3. The designer/manufacturer

Interesting. At first I thought it might be capacitive coupling voltage that a DMM might pick up due its high impedance, but when you grounded it and it popped...

Don't bother with the manufacturer, go to the certifying agencies found on the label. According to the manufacturer's website, UL and DLC list these tubes; I'm sure that UL will be very interested to hear and see your test results. The DLC doesn't seem to be interested in electrical properties, just that the light illuminates as advertised.

If there are no safety labels from a recognized listing agency present then the manufacturer is violating federal law by offering them for sale in the US.

I wouldn't go off half-cocked, I mean you must identify the problem first, and then I think verified by a second party would be minimal requirement to raise a justifiable stink....I would probably also find a fix that is simple and straight forward...

The problem has already been identified, even if not yet the cause.

No a condition has been found...the problem would be the cause...

That doesn't even pass muster as a semantic quibble.

I would rather go off half-cocked than fully unprotected. Imagine the field day the lawyers would have if one of your workers was working on a ladder installing those bulbs when he's twisting the bulb by grabbing it when the back of the hand touching the heatsink simultaneously contacts the properly grounded reflector. The surprise shock causes him to lose his balance, etc.

Plaintiff's Attorney: "Mr. Eagle, we have heard testimony from your consulting engineer Mr. Cuba, that his testing had turned up a shock hazard in the lightbulbs being installed by the Plaintiff, and he immediately notified you in writing of the hazard. You have also testified that rather than contact the manufacturer as recommend by Mr. Cuba, you continued to allow your workers to install the faulty lightbulbs, which in turn caused the shock that knocked the plaintiff off his ladder, and into the nearby vat of hydrochloric acid, subsequently causing his permanent disability..."

You can easily write the rest of this scenario. This is not raising a "justifiable stink", this is performing your fiduciary duty to protect your employees from a hazard about which you have prior knowledge. And your risk in going to UL and making them aware is...an angry supplier?

Well right off the bat you have lost your case....the instructions clearly state that the power must be disconnected before any handling of the light...

WARNINGS: • Risk of fire/ electrical shock • LED Hybrid Tube installation requires knowledge of luminaire electrical systems. (If you are not a qualified electrician, please do not attempt installation and contact a qualified electrician.) • Risk of electric shock - disconnect power at source before installation • Do not make or alter any open holes in an enclosure of wiring or electrical components during installation. • If lamp or luminaire exhibits abnormal operation (buzzing, flickering, minimal light, etc.), turn off power, remove lamp and contact us. For more information visit www.eurilighting.com or call our toll-free number 1-888-743-5766.

http://www.homedepot.com/catalog/pdfImages/c8/c8fdfbf7-8e2e-42f9-9c49-93edd5f520f6.pdf

Just alter the scenario to a person not working on the fixture, but merely adjacent to it, and accidentally contacting a live external part.

We're not talking about an external part....and I might add the light is in any common use, going to be covered with a lens of some type...

The OP said that the heat sink was exposed....

To OP: If these are "Eurilights" do they use blue wire for neutral and brown for line?

My apologies...

Since I didn't fill everyone in on this in this thread:

The light I found energized initially was already installed. I was taking readings with a clamp-on meter on the LN wires...simple power logging for an energy efficiency report. The tray cover had been removed, exposing only the ballasts and wiring.

The luminaire is an Ledalight/Ergolight model 101. The ballasts, wiring and tubes are accessed via the top. The tubes can be replaced from the top without removing anything.

Also, as a matter of fact, I can stand flat footed in my office and touch a heatsink easily. The fixture only has an eggcrate diffractor on the bottom side.

Well it seems you're using a dimmable fixture with non-dimmable LED bulbs...I don't think this plays any role, but just thought I'd make a note of it.....Is the control system disconnected, or non-existent....?

http://www.brite-lite.com/Pdf/ergolight.pdf

http://media.wix.com/ugd/979e78_4492539139524e9585756d6d8d291b5b.pdf

The control system was disconnected, as were the ballasts. The luminaire is more just a lamp holder at this point when compared to its former state.

The fixture and tombstones are rated for the wiring method.

Note that in your link, the Cautions include: "•Do not touch the product with wet hands."

That sounds to me like the manufacturer is aware of the problem, and is depending on them being installed in locations where ordinarily no one can reach them...

I have one better: "Don't touch the product at all, even while it is still in the package."

America does not win any more. We let a bunch of loser teachers, teach a bunch of mush to even losier kids, now they are out engineering a global short out.,

Yes, well...I didn't divulge the entirety of the discovery.

I was on a ladder, but wearing my electrical safety boots, anyway.

I was taking readings in the fixture (powered) to get accurate numbers for a report on our new found electrical efficiency.

I brushed one of my hands, quite inadvertently...not expecting anything at all like this, across the heatsink of the tube while simultaneously touching the case of (the now disconnected but still bonded) ballast.

I knew I had felt the 277 but was so "shocked", more incredulous...like...what the hell was that??? I grabbed my meter and found that the heatsink was energized...which led to further testing.

Man, I HATE getting shocked...so it really just pissed me off more than anything.

The scenario you posed may have been the outcome had I but touched in two different places without knowing what lurked there. I don't think it's that common to test every exposed metal part of everything we have at work.

If we buy UL products I would expect that had already been accomplished. How does one go about calling the UL to reference a product's testing?

You are lucky you didn't become a statistic of this foolish head-long rush to convert everything in the world to LED lighting.

I would be pissed too!

Bullet dead on the mark with that one.

So New Zealand doesn't recognize UL labels on anything imported? Better check with these guys.

The UL label doesn't guarantee compliance is what I am saying, especially with certified ATEX American gear people import and try to install down here which doesn't generally comply with local hazardous area standards.

As for those guys, I used to work at the test lab they brought out when they moved in to NZ (oh it is a small world). Fortunately I had already been gone a few years before it happened.

I try to not buy anything from Home Depot. I don't like their store, and I refuse to go there when I have other alternatives, anyway. Home Depot sux.

I hope they all grab one of these fixtures, and stick their finger in a light socket. They don't care what they market, and most of it is crap.

Furthermore, I think they left out the insulator, or the dielectric isolator between the semi and the heat sink - really bad idea.

OR they did something that turned the connection between the heat sink and the semiconductor casing and one of its leads to a capacitor - translation for those in Rio Linda: direct short.

I was able to remove the LED strip from the heatsink (without destroying it).

There is no direct connection.

This is looking to be a great capacitor.

Generally speaking: the requirements are, carbon for charge storage, dielectric, or electrolyte with high polarizability (phosphoric acid types), and metal contacts. The fact that it is essentially connected by a diode makes me wonder.

Are you sure the voltage you are seeing is AC? Could this be DC potential stored up, i.e. filtered by your "capacitor"?

I am using a Flue 43 power quality analyzer for readings, which works on ballasted and un-ballasted, giving the frequency, and a waveform image.

I'll post pix as soon as my browser stops screwing with CR4...I can't use any of the advanced features right now.

I contacted the Underwriters Laboratories, and after the nice lady on the phone helped me navigate their page, I found the product listing for the company. The file number on the sticker returned four hits.

*None* of those four hits contain the model number for the product (in my hands) which bears the UL sticker.

Now, maybe the page at UL needs updating. But, this does give me impetus to start a Market Surveillance Report Form with UL. Especially so, since the company has discontinued correspondence.

The plot thickens...? Comments...?

If the lights are not UL listed then you may need to sue for damages...I would contact the legal department and decide what documentation and supporting evidence you may need to support a legal action...just in case...

since the company has discontinued correspondence.

Do you mean UL?

When I have tried to get detailed information out of UL on specific products I have also been met with a lack of help.

If I remember correctly the response to my question was UL was a world wide company, they had no master product record and I had to contact the group that actually performed the testing.

Now, I think (although I may be corrected, it was based on a single phone call to UL about 8 years ago) that UL also mentioned they outsource (or similar) the testing to third parties making tracking of who actually performed the tests more difficult and I should contact the manufacturer directly.

Hardly inspires confidence when traceability and accountability is this poor. Plenty of other test labs have detailed records available online by way of a simple search engine.

The mfr. has not written since my last email with them.

UL was actually a great help.

The mfr. has not written since my last email with them.

That doesn't surprise me, it can be like pulling teeth trying to get detailed consumer product certification documentation out of them.

Are you really stating this is an EU product? The same folks who don't allow our products in because they don't meet EU standards. Shocked I say, Shocked I am

Static charging is possible and if there is no discharge path then change can hold on for long. Charge may also develop by leakage.

"Charge may also develop by leakag" - Depends

Depends works for thousands of seniors here in the States that have "leakage" problems.

Have they used some power switching MOSFET or Transistor which has no insulation or insulation is worn out and AC voltage is leaking into heat sink? There may be push-pull transistors on heat-sink for DC-HF AC conversion.

Yes, the SMPS uses the 5N65 Mosfet.

However, the internal SMPS is pretty well insulated and does not appear to make direct contact with the heatsink in any manner. I cannot see any physical contact with the heat sink anywhere.

As near as I can tell it is due to extremely efficient capacitance...enough to blow the power supply if the heatsink is bonded to ground.

If one touches the heatsink, one will receive a shock.

I am not able to post pictures to CR4 right now...my browser is acting weird. I will fix that as soon as I can to help clear up some of these questions.

Did you recheck the stray voltage after grounding the heat sink....was it possible that the discharge neutralized a built up potential? Have you tried powering up a light with the heat sink already grounded?

Grounding/bonding the exposed heatsink of a tube results in sparking, frying of the tube's SMPS and release of magic smoke.

They normally continue to sizzle and smoke until power is secured.

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

This, in my view, violates NEC article 410.5 where:

Live Parts. Luminaires, portable luminaires, lampholders, and lamps shall have no live parts normally exposed to contact. Exposed accessible terminals in lampholders and switches shall not be installed in metal luminaire canopies or in open bases of portable table or floor luminaires.

I agree this is an unacceptable safety hazard. End of story.

AC leak may not build up static DC potential but power drawn on grounding the heat sink may trip the MCB.

Similar to SE's comment: Have you checked the resistance and/or capacitance between line and the heat sink?

...and I might add a continuity test between the heat sink and line in voltage, disconnected of course.....

that is a valid point. Anything that connects to Earth through instrument heat-sink is worth checking for failure at operating voltage and not at multimeter battery level resistance check. If there is voltage controlled resistor for over voltage protection and if it fails and starts firing at lower voltages then leakage of current from excessive voltage is assured.

I believe that something which requires heatsink is leaking current from mains power source itself.

If one gets shock then there is enough voltage > 100V and enough current > 10mA perhaps up to 100mA through human body of about 1K-10k leak resistance.

If 10K resistance is placed between heatsink and ground then sure >100V across it due to leakage current must be seen in oscilloscope. You can also see if it is DC or AC as multimeter in AC mode can also measure DC if it is only having bridge rectifier diode array to turn AC-into DC and will work for both AC and DC in identical way.