Pixel fires

[VirutalElectrician]

I was just reading the thread about the USB charger in the sofa fire....

I decorate a bit for the holidays, and I'm starting to look into using pixels for color changing animated displays. Pixels in case you don't know, are computer controlled 3 color LEDs operating on 5v or 12v. People typically have thousands if not 10's of thousands of these in a display. After all is said and done in buying wire, props, controllers, etc, you're looking at about a $1 per pixel to do a display.

So what does this have to do with electricity you ask? Well, these things are starting to cause some fires. Especially the 12v variety. All the pixels operate at 5v, so the 12v variety use a regulator or resistor to step down to 5v. So why use 12v then you ask? Because these have a fairly high current draw, so there's a fairly high voltage drop on runs that are fairly long with a large number of pixels. So 12v is used on longer runs (mega trees, house outlines, etc) to avoid voltage drop or havin to do mid-span power injection.

The power supplies used are UL listed, but the pixel strings themselves are of course not. There's been some discussion on whether or not an insurance company can deny a claim based on using non-UL listed pixel strings if they cause a fire.

Here's some pictures of the fires...The last image is of a house fire that the owner believes was started by his pixels. He hasn't posted a formal follow up since his fire 2 weeks ago.

 

[petersonra]

The power supplies used are UL listed, but the pixel strings themselves are of course not. There's been some discussion on whether or not an insurance company can deny a claim based on using non-UL listed pixel strings if they cause a fire.

Have you ever heard of an actual case of this happening?

There is no rule requiring such things to be UL listed.

Even if there was, the insurance company is required to cover hazards in accordance with the state insurance regulations, regardless of what the policy might say.

Insurance companies have to cover damage caused by drunk drivers they have insured, even though what they are doing is illegal, in some cases even felony illegal. You really think they can avoid coverage by claiming something "should" have been listed?

 

[hillbilly1]

I would say ocp is probably too high for effective fault protection. Was this design engineered? Looks like the pixel failed, but due to high ocp, the internals failed before tripping or blowing the ocp. Also looks like all of the current is going through a printed circuit board, in which it wouldn’t have the capacity the wire does. Pigtailing the pixel, if possible would remove that weak link. Fusing down lengths of pixels would also help. More expensive, but possibly help eliminate catastrophic failure.

 

[ramsy]

Here's some pictures of the fires...
The last image is of a house fire that the owner believes was started by his pixels. He hasn't posted a formal follow up since his fire 2 weeks ago.

If that crap was installed by an electrical contractor with commercial General Liability insurance, and property owner demanded a certificate of Additionally Insured, then the chances of recovering damages would be better.

Property insurance won't pay casualty claims caused by temporary lighting, or Christmas lighting decorations, much less for anyting involving extension cords, or DIY wiring.

Insurance companies pay employees to deny such claims, delay, and drag out legal any proceedings, which cost home owners $500.hr if they must hire an attorney to challenge it. Even if the equity in the home is work 1/2 a million, home owners can't afford the first year of legal proceedings, much less appeals, and collections. The best deal their going to get with insurance is to settle the claim out of court for peanuts.

 

[ggunn]

Insurance companies pay employees to deny such claims, delay, and drag out legal any proceedings, which cost home owners $500.hr if they must hire an attorney to challenge it. Even if the equity in the home is work 1/2 a million, home owners can't afford the first year of legal proceedings, much less appeals, and collections. The best deal their going to get with insurance is to settle the claim out of court for peanuts.

But of course you can drag it out as well; you don't have to accept their first lowball offer. Until you cash their check your claim is still on their books as a liability. I have had some luck with auto insurers by stonewalling them until they sweetened the deal.

 

[zbang]

Pixels in case you don't know, are computer controlled 3 color LEDs operating on 5v or 12v.

FWIW, unless you refer to them as "pixel LED units" or something like that, many of us are going to read that as "PICture ELement" (the meaning of pixel for, oh, 40+ years); I certainly did the first couple of times the thread appeared (and I thought "how can a pixel have a fire???").

Redefining common words and acronyms seldom ends well, especially when they're used without context.

 

[GoldDigger]

FWIW, unless you refer to them as "pixel LED units" or something like that, many of us are going to read that as "PICture ELement" (the meaning of pixel for, oh, 40+ years); I certainly did the first couple of times the thread appeared (and I thought "how can a pixel have a fire???").

Redefining common words and acronyms seldom ends well, especially when they're used without context.

I had not seen this usage before myself, but apparently there is an entire industry niche which has co-opted the word "pixel. It is not something the OP came up with. See https://learnchristmaslighting.com/what-are-christmas-light-pixels/

Among the distinguishing characteristics of these pixels is a protocol which allows individual control of the elements without the need for individual control wires or individual addresses. The most popular sends a serial stream of control values from the controller and each element (pixel) strips off the first 3 numbers and retransmits the rest to the next pixel down the line. Each controlled string of pixels is thus self-addressing and any individual unit can be replaced with no need to configure the replacement.

 

[Mr. Serious]

A few years ago when I was actively doing some experimentation with them, everyone called them WS2812s because that's the part number of one of them. But of course there are other brands, other part numbers, so people started calling them neopixels or just pixels.

 

[ELA]

I have worked with both the "Individually Addressable LED" strips and Arrays (Matrix). Fun to work with. They can be very inefficient depending upon the use.
Each Led section has a microcontroller in it. Even when the LED is off it still draws 1ma. at 5V. Does not seem like lot however, by example.

I use a 32 x 16 array to display a primitive graphic that changes shape, draws an outline, and may be just blank sometimes.
For this project's general use: the current draw at 5V is approx. = 0.65A ( a particular graphic is displayed , with few LEDS On, and at a very dim setting).

During initial testing of the entire array:
With no LED segments On current is = 0.52A ( 512 microcontrollers additive demand)
With all LED segments On (R+G+B) for a very bright cool white light, the demand is = 11.0A

For this project I added a power supply on/off driver MOSFET to feed the LEDS so that when the graphic is powered up, but has nothing to display, it would not be wasting 2.5W of power.

Perhaps some failures with smoke or fire are due to the system controller allowing all LEDS to be at full brightness, with the Cool White setting, for too long.
Current demand can add up to a large current for large displays. There may be a lot of designs that do not account for the "worst case" scenario.

 

[VirutalElectrician]

I had not seen this usage before myself, but apparently there is an entire industry niche which has co-opted the word "pixel. It is not something the OP came up with. See https://learnchristmaslighting.com/what-are-christmas-light-pixels/

Among the distinguishing characteristics of these pixels is a protocol which allows individual control of the elements without the need for individual control wires or individual addresses. The most popular sends a serial stream of control values from the controller and each element (pixel) strips off the first 3 numbers and retransmits the rest to the next pixel down the line. Each controlled string of pixels is thus self-addressing and any individual unit can be replaced with no need to configure the replacement.

Yup! Exactly right! Which is why when one of the LEDs fails, you have to replace it, you can't just patch over it, other wise your entire string is "off by 1" at that point.

 

[VirutalElectrician]

I have worked with both the "Individually Addressable LED" strips and Arrays (Matrix). Fun to work with. They can be very inefficient depending upon the use.
Each Led section has a microcontroller in it. Even when the LED is off it still draws 1ma. at 5V. Does not seem like lot however, by example.

I use a 32 x 16 array to display a primitive graphic that changes shape, draws an outline, and may be just blank sometimes.
For this project's general use: the current draw at 5V is approx. = 0.65A ( a particular graphic is displayed , with few LEDS On, and at a very dim setting).

During initial testing of the entire array:
With no LED segments On current is = 0.52A ( 512 microcontrollers additive demand)
With all LED segments On (R+G+B) for a very bright cool white light, the demand is = 11.0A

For this project I added a power supply on/off driver MOSFET to feed the LEDS so that when the graphic is powered up, but has nothing to display, it would not be wasting 2.5W of power.

Perhaps some failures with smoke or fire are due to the system controller allowing all LEDS to be at full brightness, with the Cool White setting, for too long.
Current demand can add up to a large current for large displays. There may be a lot of designs that do not account for the "worst case" scenario.

Current guidance is to run them at 30%. Both to lower current draw, and otherwise, they are just too bright.

It is believed most of the pixel failures are caused by poor manufacturing processes (a race to the bottom to get the cheapest possible price out the door). The epoxy around the pixel to seal it against the weather also seems to be a high failure point, allowing water to get in and causing a short.

 

[VirutalElectrician]

If that crap was installed by an electrical contractor with commercial General Liability insurance, and property owner demanded a certificate of Additionally Insured, then the chances of recovering damages would be better.

This is a hobbyist industry. Very little electrical engineering going on. Most people installing these have never heard of Ohm's Law, or don't have any clue that you can't pump 20amps of current down the "18awg" wires the Chinese manufacturer claims to use, but it's really undersized 20 or even 22awg for the really cheap sets. See my previous post about race to the bottom.

The first couple generations of the controllers didn't even have fused outputs. The newest controllers at least do now.

You should follow some of the Christmas lighting groups on Facebook. The stuff you read about people doing just makes you shake your head.

Property insurance won't pay casualty claims caused by temporary lighting, or Christmas lighting decorations, much less for anyting involving extension cords, or DIY wiring.

Really? Where is that clause?

Insurance companies pay employees to deny such claims, delay, and drag out legal any proceedings, which cost home owners $500.hr if they must hire an attorney to challenge it. Even if the equity in the home is work 1/2 a million, home owners can't afford the first year of legal proceedings, much less appeals, and collections. The best deal their going to get with insurance is to settle the claim out of court for peanuts.

American courts and state laws have already pretty much neutered insurance companies. That policy they send you doesn't mean much, what matters is written in state insurance laws and court cases.

 

[don_resqcapt19]

Have you ever heard of an actual case of this happening?

There is no rule requiring such things to be UL listed.

Even if there was, the insurance company is required to cover hazards in accordance with the state insurance regulations, regardless of what the policy might say.

Insurance companies have to cover damage caused by drunk drivers they have insured, even though what they are doing is illegal, in some cases even felony illegal. You really think they can avoid coverage by claiming something "should" have been listed?

They appear to be a type of decorative lighting.

590.5 Listing of Decorative Lighting.
Decorative lighting used for holiday lighting and similar purposes, in accordance with 590.3(B), shall be listed and shall be labeled on the product.

 

[VirutalElectrician]

They appear to be a type of decorative lighting.

The pixel strings themselves are 5v and 12v. Does that fall under the NEC? I use Mean Well power supplies for my personal use, which are UL listed.

My dad was an attorney, so he was always paranoid about legal things. I remember shopping for Christmas Lights with him when I was a kid in the 70s. He always insisted on UL listed light sets. Not all sets were UL listed back then. You could tell the difference between those that were and those that weren't. Those that weren't had smaller wires, flimsy sockets, and you could even see the bare copper going into the sockets as the insulation wasn't even secured to the socket.

 

[tortuga]

Really? Where is that clause?

If you like reading dense legal stuff Google
“efficient proximate cause”
and
"Anti-Concurrent Causation Doctrine (ACCD)"

Here is a example with public details, not with wiring but faulty venting;

https://www.millernash.com/industry-news/in-extraordinary-ruling-washington-supreme-court-applies-efficient-proximate-cause-rule-to-liability-coverage-dispute

 

[ramsy]

If you like reading dense legal stuff Google
“efficient proximate cause”
and
"Anti-Concurrent Causation Doctrine (ACCD)"

Here is a example with public details, not with wiring but faulty venting;

https://www.millernash.com/industry-news/in-extraordinary-ruling-washington-supreme-court-applies-efficient-proximate-cause-rule-to-liability-coverage-dispute

The discussion of "Anti-Concurrent Causation Doctrine" ACCD describes "Electrical Damage" as (an Excluded Peril), meaning not covered by insurances, so it seams homeowners are screwed when casualty claims are caused by electrical.

 

[ramsy]

In Xia vs ProBuilders, homeowner gets CO2 poisoning from faulty venting of gas water heater, installed with new home. Title insurance kicks the can to builders GL policy (ProBuilders), who deny claim due to policy "pollution" exclusion.

Xia takes ProBuilders to court and loses to a summary judgement, since policy "pollution" exclusion is successfully used to deny coverage for identical claims in higher courts in nearby Oregon State.

Homeowner pays for another legal battle, this time against Washington appellate court, which also deny Xia's claim on appeal. In spite of multi-State legal precedent, this homeowner finances appellate attorneys for yet another round, this time with Washington State supreme court, who finally overturns its own appellate court.

Xia's legal odyssey probably cost more than most reasonable homeowner's can be expected to afford, to prove Bad Faith insurance.

 

[Knuckle Dragger]

My guess it was the exposure to the outdoor elements/ environment that helped do it in...Besides probably just being cheap crap.
That's just my opinion.

 

[winnie]

Xia's legal odyssey probably cost more than most reasonable homeowner's can be expected to afford, to prove Bad Faith insurance.

That is an interesting story.

I only know what was in the link that Ramsy posted. But it seems that the builder was negligent, but the builder's insurance denied the claim.

Xia still has a valid claim against the builder, and the builder settles. Part of that settlement is assigning any payout from the insurance company to Xia, who then goes after the insurance company.

The insurance company is denying the claim on the basis of a pollution exclusion. But in this case the pollution was not the cause of the failure that caused the injury, but rather the pollution was the result of the negligence.

Ahh the webs we weave....

 

[ggunn]

In Xia vs ProBuilders, homeowner gets CO2 poisoning from faulty venting of gas water heater, installed with new home. Title insurance kicks the can to builders GL policy (ProBuilders), who deny claim due to policy "pollution" exclusion.

CO2 or CO? CO2 is not poisonous, and faulty venting can cause CH4 to burn less completely and produce CO instead of CO2. My brother and I got CO poisoned from a faulty wall heater in his house years ago.