Philips streetlight "ledgine" fault and repair

[@wimwiddershins]

If he doesn't want it back, you've scored a fairly good colour studio light.

[@jlucasound]

Yummy! As Clive might say.

[@FrontSideBus]

Or Clive could hack it with colour changing LED's and have it on his driveway 😂😂😂

[@FrontSideBus]

@Lyall Moore I don't think you can make popcorn with them either, unlike the good old CP60 😂😂😂

[@kevinwingfield2007]

@Lyall Moore nor the boys!

[@justinpatterson5291]

Or a "garden" light for the inside. if you catch my drift. Probably has a 6500k output.

[@nikoshi1988]

Philips and LED streetlights never were all that great. My time at Philips we had countless of customer complaints, mostly from city councils, that used these Ledgine type fixtures as well as it's successor the Roadstar and Speedstar. Both had lots of issues with arcing tracks, burned up LED's and not being able to dissipate heat quickly enough. The 10% mean time between failures was 7000 - 10.000 hours instead of the 100.000 hours that they advertised. Another fun thing Philips did was to discontinue a series of LED fixtures every 2 - 3 years (on paper) while the warranties were 5. Kind of glad that I'm not working for them anymore.

[@mattfleming86]

There is substantial conversation in this comment section about this subject. I wish Clive would pin this one at the top.

[@alexatkin]

Surely if they discontinue that model they are required to supply a replacement model? Would seem rather odd that commercial lighting would have less warranty than consumer products where they absolutely ARE required to honour the warranty AFAIK.

[@nikoshi1988]

@@alexatkin They would "honour" the claim though the contract would stipulate that only the same unit would be provided if available at no cost. If no exact same unit would be available then a suitable replacement would be offered at a discount price.

[@youtubasoarus]

Just another way they've engineered shit to fail, to ensure successive contracts and keep the revenue stream going hot. Clever bastards.

[@ehsnils]

LEDs expected to do 100k hours, but rarely are. They still have to beat the low pressure sodium lamps when it comes to being effective. The sodium lamps are very monochromatic, which is a blessing when it comes to light pollution since that light can be filtered out, but a curse since it's not possible to distinguish colors in that light. So a "grey cortina" (Tom Robinson Band reference here) could as well have been red or blue. However even the low pressure sodium lamps have a limited lifetime of about 18k hours.

[@DerMarkus1982]

Kink palculator ftw! Clive going at it with a surgical workflow... First, mark the parts to remove; then, a bypass is applied. Clive is our one and only true "electronics surgeon"! I *love* this channel!

[@DerMarkus1982]

@the kink palculator Okay, didn't expect that. LOL Flickery Hopi, anyone? :P

[@rushilkisoon]

Imagine being so well known in a community that even government type organisations turn to you for advice.

[@crackedemerald4930]

Clive is the wise bearded sage in the far away mystical island of man.

[@Blu3ManiC]

Well he’s an electrical engineer in what amounts to basically a small town/ island so certain jobs going to him are a given. The KZbin content just so happens to be a nice bonus for us.

[@nyetloki]

He works for local government iirc. I recall he used to do the christmas lighting for the city. There's a video for it if you look far enough back.

[@nyetloki]

kzbin.info/www/bejne/mGGXqqRqodJ3ZtE

[@Zveebo]

@@nyetloki That was a long time ago - KZbin is his main job now. The only significant contract work he does is the Tattoo in Edinburgh in August (Covid-permitting).

[@Jim-si7wz]

They could use them in parks and riverside walkways' like you said less critical areas or they could sell them to locals for garden lights, just a thought, but yes a very interesting video thank Clive, another good one.

[@craigr306]

unfortunately because its public money that buys the lights they have to dispose of them such a waste. my local council just destroyed thousands of pound of electrical hardware that could of been reused to help a new business when asking this is what i was told

[@tbelding]

@@craigr306 - Some places will put them up for auction, in bulk. They can't just 'give' them to someone, or 'sell' them to people, but they can auction it off.

[@DigBipper188]

I'd rather they sold off the old 35,55 and 90w SOX lanterns for use as outdoor lighting. At least then we would see them being recycled, as opposed to the current state of affairs where working lanterns are just scrapped and sent to landfill.

[@SlinkyStoney]

I know some company here won't sell their used stuff from electronics to vehicles. They just put those outside and let them rot. Want to get some? No, you can't, there is a security guarding the place.

[@qwertyasdf66]

@@craigr306 I assume they could give them away because that's effectively the same as disposing of them. I have a bunch of used 12v halogens because I once worked on a job at a museum here in NZ where they annually change all of the bulbs illuminating displays. Were were specifically told we were allowed to take the old bulbs home for personal use but could not sell them.

[@thewatchworks1372]

That has got to be in the most interesting failure modes of an LED fixture I’ve ever seen. Besides that, it definitely seems to be a rather quality unit, and I don’t think I have ever seen a fixture that “soft starts” the LEDs.

[@acmefixer1]

The soft start isn't for the LEDs; it's for the rectifier, filter capacitor and current regulator.

[@DjResR]

I've seen quite a few different style of fixtures that do soft-start in Estonia, although I'm not sure is any of it Philips._

[@thewatchworks1372]

@@DjResR Ah, I live in the United States, so when I have seen the street lighting flick on at night, it just instantly is at full brightness

[@thewatchworks1372]

@@acmefixer1 that is rather interesting, thank you for the clarification, it’s cool nonetheless though

[@NOWThatsRichy]

@@acmefixer1 Looking at the fairly new LED street lights here in Portsmouth, ( Philips Unistreet model) they do have a soft start, I'd say around half a second, more noticeable on the larger main road lanterns, than the smaller side road ones.

[@IncertusetNescio]

My guess: It got hot by way of substrate failure (perhaps the pad the LED soldered onto was problematic in the fiberglass layer but it didn't show until failure/several heat cycles) then tracked its way back across the pad as it high voltage burned in a current-limited way (like a Lichtenberg figure) until it hit the huge pad and completely shorted/ran out of hot spot to burn with all that contact area. Hmm.

[@theelmonk]

Seems reasonable. Not a led failure but a substrate failure. Maybe that's why they were discontinued. I don't think bridging the LEDs out is realistic for a commercial repair. Maybe make a replacement LED panel if there are a lot to keep going ?

[@tbelding]

@@theelmonk - I agree that it's not excessively realistic, but if you have some of them "fixed" in a truck, you can QUICKLY get a light back to functioning without desperately trying to find a completely replacement. It's not like bypassing the LED will be excessively difficult (if you can heat it). I've done similar work to get a server back online so I could take a few days to properly decommission it.

[@wombatillo]

The insulating layer in an aluminum core PCB is very thin. 75 um to 150 um is typical. If you use the street light as 230V (270V or so after rectifying) series chains, you're looking at a significant voltage being held back only by the thinnest of dielectrics.

[@HighestRank]

And professional electricians have one name for both Lichtenberg failure and its solution.

[@mahlapropyzm9180]

@@wombatillo I wondered if there had been water ingress from the edge, suddenly you have electrolysis...

[@ChodaBoyUSA]

Big Clive's genuine excitement over learning a products secrets is infectious.

[@derdwastaken]

Every day the "Kink Palculator" gets out is a good day

[@NinoJoel]

I had to rewind the video 3 times to be 100% sure he said that and it was not my brain messing with me xD

[@markiangooley]

I have an unusual fondness for comical spoonerisms. Though I suppose talking of making burgers from bound grief gets tedious…

[@HighestRank]

@@NinoJoel I had to reread the OP 3 times to be 100% sure that it didn't say "comes* out"

[@johngarritzcx6733]

😁😁😀😂😂

[@UndercoverFerret404]

Soldering these types of "heatsinking panels" (called metal clad PCB), is done fairly easily by heating the entire PCB on a frying pan to around 100ºC. It makes it quite easy to solder the LEDs with a regular soldering iron.

[@jimmychin8313]

Thinking outside the square, could you have linked the joins with a spot welder using a flat strip of nickel and not have 'heatsinking issues'? Would a flat join also scoot below the plastic lens?

[@chrishartley1210]

@@jimmychin8313 I think with spot welding the risk is that it would go straight through to the substrate as before.

[@rpavlik1]

Or using something like the mhp30 mini reflow hot plate/preheater

[@eDoc2020]

My 140 watt Weller gun doesn't have much trouble soldering to aluminum PCBs. Similar units go over 500 watts which can probably be used to solder even with the external heatsink attached. The big benefit of this over preheating the entire PCB is that it can be done in situ.

[@UndercoverFerret404]

@@eDoc2020 The problem with just dumping high temp into the solder joint, is that you risk delamination. Preheating the entire PCB at a safe low temp, helps to avoid that.

[@DrTune]

I can never get over how good your PCB photos+prints are. It's very, very impressive

[@RabidBadger_]

I have a feeling there's a link between those failures and the fact that they aren't making the lights anymore.

[@FrontSideBus]

And for the money these things cost it's scandalous that they should not be supporting them! Taxpayer money up in smoke. I recently googled the spec of some solar powered LED road sign lights that look interesting and the range starts at £700 and goes all the way up to £1200!!! I think replacing an 8w fluorescent lamp every few years would be cheaper!

[@garrettkajmowicz]

@@FrontSideBus In many cases, the biggest cost isn't the materials themselves, but the labor involved in replacement. Even for a trivial replacement like a standard consumer light bulb, the cost of getting a vehicle out to the location, blocking off traffic, plus the extra safety gear and precautions required might cost $100+. Now add specialized labor and having to fiddle with multiple components to be secured and the labor costs could dwarf the product costs.

[@WizardTim]

Straight from the first page of the Philips LEDGINE brochure *“The future-proof solution for outdoor luminaires”* yet they have been discontinued and cannot be serviced…? The LEDGINE warranty information says a LED panel is considered defective when > 5% of LEDs do not operate so I would imagine putting this back into service with 3 LEDs out (6%) would be a liability problem. I’m also interested to know if Philips has offered a newer replacement luminaire or if they have just absolved themselves of any fulfilling warranty on a discontinued product. And I suspect this was caused by a cracked solder joint that started arcing and burnt through the fibreglass and progressed along the track so I would expect it to fail again on a different LED if it’s a systemic problem in all of their installations.

[@marcosfeddersen5114]

It’s insane to me that Philips absolved themselves of this failure within the warranty period because the product is “discontinued”. Seems like a potential lawsuit.

[@hyperboloidofonesheet1036]

@@marcosfeddersen5114 I'm sure a multinational conglomerate like Philips has factored in the legal costs and decided it was cheaper to discontinue the product.

[@AAAyyyGGG]

@@marcosfeddersen5114 Before Phillips had a chance to reply lets not condemn them; the original comment implies 'they might have done'...

[@shepardpolska]

@@marcosfeddersen5114 not sure about the rest of the EU but in Poland I think that would atleast force them to refund the faulty lights

[@FrontSideBus]

Pretty much all LED luminaires cannon be serviced. Often they use bespoke drivers or LED modules and the whole thing needs to be skipped. Great way to save money...not!

[@rbleth51]

That appears to be quite a block of aluminum for the heat sink. In my machinist days that would have been a fun job to run on the CNC mill.

[@bigclivedotcom]

I'd guess it's extruded and then finished on a mill.

[@rbleth51]

@@bigclivedotcom Almost certainly, yes. For something of that volume extrusion would be the base geometry with finish work done after. I have cut a lot of aluminum in the past, including hear sinks from scratch. Very fast cutting, lots of fun with the proper setup!

[@grmlab4510]

Very neat fix. Clive is going to have a very bright backyard.

[@garrettkajmowicz]

It could be an interesting case in anti-theft technology. Old-and-busted: outdoor security lighting to make prowlers easy to see. New hotness: outdoor lighting so bright that intruders will require wearing welding goggles to approach.

[@joshfriesen9401]

I just love how he always says “I’ll bring in the pink calculator “ the pink calculator is my favourite

[@NiyaKouya]

I think you misspelled the magical kink palculator :P

[@JohnnyMotel99]

Nice toolkit from Wera, get the ratchet handle as wel, it's a godsend.

[@theR6969]

Just grind a little track in the lens to fit the pypass wire and the lens will fit perfectly, and it can be used as a parking lot or garage light.

[@Farm_fab]

Clive, a street lamp like that one having been modified, might be suitable for security lighting at maybe a park or around a farmhouse.

[@volkhen0]

The problem is that the cost of installing it in high cost labour country exceeds the gains. And you don’t know if the lamp will last. So then again you need to pay the cost of labour...

[@Farm_fab]

@@volkhen0 if it's installed on the side of a building, an extension ladder is the easy way to reach it, so the labor issue is negligible.

[@volkhen0]

@@Farm_fab and the electrician has to drive to the place etc. if it’s private property it’s a different story but in public places it’s not that simple.

[@Farm_fab]

@@volkhen0 for a homeowner, there's no requirement for him to have an electrician to do this job as it's a very simple job to make these changes. As for using it at a park, many maintenance workers are qualified to do this type of work, and may have this training to begin with. Where I work, I'm not an electrician, but I took a course in trade school called general trades, and it qualified me to work on a lot of equipment that is 220v and down. I am restricted in that I cannot install any new electric circuits as I'm not qualified to do so.

[@oldbatwit5102]

What do they normally do with the faulty units? These have to be around £300 a pop new. I'd like one of these in my fridge, for late-night foraging.

[@greendryerlint]

The sudden burst of actinic light from your fridge would certainly dazzle any late-night fridge raiders that didn't know about your upgrade.

[@hinspect]

Remember when we only had red LEDs on the consumer market? Technology sure has evolved in the past 40 years! Thanks for the video

[@markkrusemer526]

I so remember the RED only leds..... and the Demo of a "Speek and Spell" by Texas Instruments. The real expensive LEDs had an Orange'ish hue to them.

[@Monkeylum1]

Well I‘ve installed about 1000 of Philips streetlights in the last 9 years and they seems to be very reliable. Not any LEDgine failures so far encountered.

[@mikebashford8198]

Never taken much notice of street lamps, but (much to the puzzlement of the neighbours) have just taken a close look at the lamp post outside my house, and am pretty sure it is one of these modules.

[@curtw8827]

As a retired lighting engineer I used to warn my clients that LED fixtures would become throw aways because the technology would keep changing. No replacement parts would be stocked.

[@Fifury161]

My local council are replacing the old style orange (high pressure sodium?) street lights with the new LED style. Except they aren't just replacing the heads. They are installing the complete street light on the pole and all less than 1 foot away from the existing pole! Madness!

[@eddiereed5025]

They move them slightly to get slack on the cable , if they are near the kerb new regs move them further off the road Pesumeably to stop vehicles hitting them

[@mattfleming86]

Even more maddening when they are replacing components or even new fixtures in just a few months/years

[@HighestRank]

in historic districts, it's often legislated as mandatory to precisely duplicate or replicate the current placement and original esthetics of all existing luminaries, etc.

[@Fifury161]

@@HighestRank Yeah - the point being that instead of just swapping out the head (or bulb - I get why they may not be able to do that) - they instead decided to plat an new pole beside the existing one! So the existing one just gets turned off (and presumably disconnected) yet they plant a new pole less than 1 foot away from the original pole. Madness, I don't understand why they don't just swap the head on the existing pole?

[@fixitdude74]

We use quite a lot of the Pierlite brand ECO led highbay gen2 and gen5, gen6 are out now but we don’t have any of them, anyway the gen2’s had a habit of failing(stopped working) due to the neg spot welded terminal coming unstuck, had many of them fail in and out of warranty, great light especially the 200w model, never had any of the driver models fail., I resoldered the loose terminal and they came back to life.

[@UhrwerkKlockwerx]

Such a shame to see such promising, capable technology go to waste because manufacturers care more about profit than people. At least you might now have a better studio lighting setup with how nice that warm white is. It’s like a sunny day, quite good for illuminating your bench.

[@RODALCO2007]

Bridge the LED's out, I have done that in some SL led clusters. Those tiny leds are overdriven very hard. I like your colour coded security bits set.

[@leorigby3069]

I've had to deal with damages caused by exactly this problem and the acrylic on the front of the fitting was all black and melted! Granted they were running hot but had a proper heatsink and heat transfer tape. The major UK LED manufacturer brushed it off and were not even bothered with the returned light fitting.

[@acmefixer1]

Clive talked about using a dremel to cut away the lenses. Why not cut away the PC board? Try to preserve as many of the LEDs as possible. Maybe save 1 or 2. If only 1 LED was out, it would hardly be noticed. Another idea. If the copper pad can be saved, try to replace the led. 48 LEDs at about 100 watts is about 2 watts each, so probably 3 watt LEDs. They're most likely Philips Lumileds. 👍👍

[@tonyweavers4292]

Meanwhile, the sodium head it replaced would probably still be working. It's such a waste of resources.

[@zh84]

Yes, I was just thinking about sodium lamps. They ran happily (if hideously) for years. Which, I wonder, would cost more to replace?

[@grmlab4510]

I thought the same. I wonder if the rationale is the energy savings makes up for the increased cost of maintenance.

[@MadScientist267]

Sodium sucks but yeah more reliable than this crap... I fail to see the point in upgrading all this lighting to increase efficiency if all we're going to do is drive them to the brink of failure and make as much heat as what they're replacing. Absolutely stupid.

[@dewdop]

@@MadScientist267 I would blame the manufacturer and not the technology

[@MadScientist267]

@@dewdop Yeah LED technology is awesome... It's the idea that they think they have to push everything to the limits. This is a problem everywhere... It's suicidal design-wise to rate things so close to safety margins as they do. I have a rule of thumb... Take the [pertinent] AMRs from the datasheets and cut them in half... *That's* the operating point that will give you maximum reliability *and* efficiency... In the case of LED, need more light, use more dies. There is some exception to this... But at the moment, THAT is the exception. But I get it... If everything lasted forever, companies wouldn't make any money in the long haul. Oh damn... There's that word again...

[@OpticalMan]

I have done quite a few of these types of LED repairs and I would have only bypassed the one that was damaged (its not my main job, I actually design the LED street light lenses for a living). All you needed to do was put a short right in the middle of your diagram between the two negative pads. Because the link would be very small you would then have just needed a Dremel to take out a small area which would be in right in between two of the lenses. Heating the whole assembly does help if you are soldering links on these types of boards but not essential if it is done with a really big soldering iron. If you use a small iron it is easy to get dry joints, the solder will seem to stick but the clue that it is not right is that it is lumpy, so make sure the solder has flowed properly.

[@phils4634]

A rather nice light engine there. That heatsink was certainly "substantial" and probably more than adequate for the three panels even if operated over a long period.

[@vulgivagu]

The actual mount where the cable is plugged in seems to appear to have overheating damage judging from the yellowing. Just how hot do these lamps get as they have such large heat sink. As I am 75 years old, sodium lights have a nostalgic memory for me, although they were a pain for railway operators as they had to be shielded from the engine driver. The new electric signals had yellow displays and they could be confused from a distance.

[@McTroyd]

Amazing power factor for an LED driver! Then again, cheaping out on the driver probably means a bigger electric bill.

[@HighestRank]

I wouldn't put it past the supply to still have been muntzed to the point where it may indeed be throwing out big spikes.

[@SeanBZA]

@@HighestRank Not those, with a thousand on a single line you want the power draw to be as clean as possible, because your power line supplying it will be running right to the maximum, and any noise will be radiated out by what is effectively a pretty good antenna. Blot out phone reception, TV and radio on a street, along with land line phone, and the electric company will get a thousand complaints in a single night.

[@walttrowbridge7687]

Nice and easy repair. It's interesting sometimes what gets handed to me by friends and family to see if I can fix it. Many times it's a very simple item that's failed and usually cost a couple bucks to repair vs. A whole new item, that costs hundreds. I would probably find out the resistance each LED is though. And put the equivalent resistor inline within the jumper, since now that branch of LEDs is going to be under more voltage.

[@bigclivedotcom]

It's a wide range current regulated supply.

[@Megatog615]

been a while since we've seen the HOPI!

[@trcostan]

We have had several street lights where the phosphor falls off or something because they turn from white to the obnoxious blue color! I’m trying to get ahold of a dead one from a friend at the power company to see why!

[@HighestRank]

It's not phosphor it's a filter, and those aren't just blue they're ultraviolet.

[@bigclivedotcom]

The phosphor can degrade with the base blue LED showing through.

[@theskett]

It's most definitely phosphor and not a filter (a filter would just block UV, and so give you no visible light -- or just blue, maybe; unhelpful). Plus true-UV LEDs are still ridiculously expensive because hard to manufacture; hence most modern 'colored' LEDs (including 'white') use near-UV LEDs with appropriate phosphor.

[@trcostan]

@@bigclivedotcom what ever is happening their is zero white light, it’s all blue. If I can get one from a friend I will send you pics

[@alexanderlieret2827]

The LED driver module would be interesting to see. I was expected a good power factor from this unit because above a certain power level power factor correction has to be build in. I am not surprised that the very thin substrate layer could not handle the open circuit voltage of this driver. This can be as close as the rectified mains input (around 340V). A slight manufacturing error and it can ark through it, especially with the temperature of those leds. If the led circuit is not referenced to ground (mentioned in another comment) than there is another short in this circuit board.

[@ICountFrom0]

Love how they get out of "lifetime" by doing "discontinued" ... my phone was "discontinued" and then released exactly the same, just with a internal item number that's ONE HIGHER.

[@Airsoftforall]

Sad that they was discontinued during the warranty!

[@danhard8440]

easy to see why it was discontinued lol

[@minitoe]

I'm surprised Philips didn't offer a similar or better product. Isn't that customary if a product under warranty is discontinued?

[@probablynotabigtoe9407]

@@minitoe Phillips does a lot of sketchy business practices

[@_BangDroid_]

Couldn't get away with that in Australia

[@minxythemerciless]

We had the Phillips luminaires installed in our suburb but everyone complained they were too bright. The council was able to drop the brightness fairly easily so I guess the power supply is programmable?

[@dogsoldier1994]

Phillips or 'Signify' as they go by now, offer a kind of system called interact that works as a management system for jus that kind of solution, I think they call it 'smart city'.

[@bigclivedotcom]

Some systems allow remote intensity control. Philips Xitanium drivers can be programmed to different currents with either a programming puck or resistor.

[@100SteveB]

That appears to be a warmer white than a lot of the led street lamps in use these days. Some of them are very a very harsh white. I think they should add a slight amber tint to the lens to soften the light even more. I often stay the night around my sons house, and the bedroom i use faces the road with a street lamp pretty much outside the house. The old street lamp never used to infiltrate the bedroom with too much light - just a soft amber glow coming round the curtains. But when they switched to led's i was shocked at how much white light was lighting up the bedroom. And just take a look at some of the 24/7 video footage from the space stations earth viewing cameras, just about every town is now just a mass of bright white light just rebounding off into space. Whereas a few years back you could only see a much dimmer orangey amber glow. I am all for led lighting, but why do they have to use such a harsh white light? Lots more light pollution, plus it seems to cast much darker shadows than the orange lights used to. I think to get around the dark shadows problem they have simply turned the intensity up even more - obviously hoping the light will eventually bounce off everything and fill the shadows.

[@bigclivedotcom]

The first ones here were harsh cold white. They've softened the colour.

[@bikkiikun]

I miss the the colour of sodium-vapour street-lights.

[@jamesplotkin4674]

Unfortunate they get the old "sorry, you're sol on warranty because we don't have any spare devices stashed away in a box somewhere" But, more importantly, how sad in these days, one must worry (even think on) how there may be a claim because the lamp has been "tinkered with" and doesn't work 100%.

[@fouzaialaa7962]

there is also the option to cut the shorted bit of the pcb out with a Dremel and repair the rest with wires just like you did !! i've seen it done on proprietary computer motherboards where its multi layer and super expensive !!! so a technician will use a thermal imaging camera to locate the short and then cut the shorted side since he doesn't know which layers are shorted !! then expose each layer and bond the layers together !! its super interesting and complicated

[@theskett]

Go check Louis Rossman, he does similar almost-daily: kzbin.info/door/l2mFZoRqjw_ELax4Yisf6w

[@bigclivedotcom]

Not so easy with a metal core PCB. The fibreglass layer is ultra thin.

[@amorphuc]

Thanks Big Clive. Very interesting. I recently bought some heat sink tape from the usual Bang suppliers and can't help but wonder if that might not be a better solution between the LEDs and the heat sink for perhaps preventing a short between them. I'm guessing the thermal conductivity might not be as good but over such a broad area, it might not matter?

[@theskett]

The problem doesn't seem to be the heatsink compound between the MCPCB and heatsink, though; instead, is either the way the burned LED was soldered, or a manufacturing problem with the MCPCB (insulation too thin, some kinda copper whisker, etc.). And you don't have a means to get heatsink tape between the LEDs and the MCPCB, so I don't see how that tape could help you.

[@electroshed]

I picked up an LED streetlight head once a few days after a car wiped out a lamp post, quite a heavy chunky thing, they took away the car but left the lamp post and head, oh well I recycled it, had to spend £20 on a heatsink, it's now one rather bright 150w LED light in my garage!

[@rarbiart]

On which end did the arcing start, probably at the solder point of the LED and then burned back to the large pad. i assume the root cause on the initial soldering of the LED... perhaps it was soldered at a too high temperature during manufacturing? Or was it just an bad spot in the isolator, prone to fail? Perhaps let it run for several k hours and see if it develops further fails.... with a bit of luck we will have in the future livestreams one of those "let me pause the video for a moment" incidents...

[@HighestRank]

How about plain simple planned obsolescence?

[@rarbiart]

@@HighestRank I don't like conspiracy theories.

[@mahlapropyzm9180]

@@rarbiart Wiki the Phoebus cartel - conspiracy fact....

[@rarbiart]

@@mahlapropyzm9180that is your generalization fallacy

[@PF-gi9vv]

I’ve worked on something similar. If one LED fails then the power supply is switched on, the power supply’s output is over 1000v when not under load, enough to start arcing across the LED or to ground. I suppose this could have happened here.

[@DigBipper188]

Annoyingly, the phrase "Not quite lasted the rated lifespan" is a common occurrence with these LED street lights, Especially the Philips ones from what I've been seeing. An entire section of our street was done with 150w HPS back in 2011, and they decided "Screw it, let's waste taxpayer money on some new street lights even though *ALL* of them are working in this area" and replaced the entire stretch AGAIN but with 100w LEDs. Just the first start-up alone had 3 out, and another 2 fail during operation. compare and contrast to the 0 and 0 of the sodium lanterns that they were supposed to replace and that sets a really bad precedent for LED. Not to mention there is a fairly significant dip in percieved lumens, making the area look way darker than it did with the 150w SON lanterns. They really need to stop wasting money on these damn things if the technology isn't even gonna last more than 3 seconds in normal operation.

[@mattfleming86]

Most of it is forced up-selling with made-to-fail components. LEDs are capable of such a long life with proper weatherproofing, isolation, and primarily current supplied. This is an ideal time for a new company to drop in and make some true long-life lighting. My current consumer LEDs hold up about as long as an incandescent no matter the brand. It is infuriating.

[@DigBipper188]

@@mattfleming86 I can agree with you there. I have cheap chinese LED modules in some home-made porta floods that have lasted way longer. And one of them is a 1,000 watt monster that drives the LEDs quite hard because it's built for raw lumens over efficiency or lifespan!!

[@alexatkin]

@@DigBipper188 It really is insane. I have a £5 eBay 12v light strip just dangling loosely from the porch ceiling, its lasted longer than any commercial bulb I have bought. Yet I'm just running it via a 12v dusk/dawn sensor also from eBay, powered with a 12v PSU stolen from a BT Home Hub. I have another cheap warm white LED strip in the hallway around a large mirror powered in an equally janky manner. A block of LEDs did fail on it and the silicon coating has gone very orange, but otherwise it too has been remarkably reliable. Although I do have that on a 12v dimmer. On the other hand, I've had several LIFX bulbs fail on me, the potted PSUs give out. Perhaps the most amusing failure I've had was a bathroom LED fixture by LightingEver of Amazon. The junction box plastic perished causing the mains wiring to fall out. I was sure it would be the PSU that failed but nope, a couple of WAGO connectors later and that too is still going strong and its had a LOT of use as we leave the bathroom light on 24/7 most of the time so the extractor fan can run (very annoying, it can theoretically be manually bypassed with a fan control switch but the council put the PSU so close to the ceiling I cant reach to add the wiring).

[@guygordon2780]

Sodium and Metal Halide arc bulbs fail all the time. Then maintenance changes the bulb and you don't notice it. The problem isn't your govt. They were suckered in by claims of 100,000 hour lifetimes by the lighting manufacturer. And none of those led fixtures have replaceable bulbs. Blame the company.

[@DigBipper188]

​@@guygordon2780 I'm well aware of this as you can get early lamp failures on any lighting technology. Especially in large volume production it is bound to happen at some point. I've had brand new incandescent, MH and fluorescent lamps fail near instantly after being powered on for the first time before. It's less common with decent quality SON. MH or SOX gear though than it is with LED. I've seen more lanterns fail in their first months of use with LED than I have ever seen with SOX, MH or SON, and there is quite simply no excuse for it bar manufacturer cost-cutting measures. Granted, there have been some. There used to be a 35w Philips MI26 outside my house that once had a dodgy lamp installed that lasted about a month before the discharge tube ruptured at the cathodes, but this, again, is much, much more common with LED fixtures for whatever reason... Whether it's water ingress to the diode packs or just a bad batch of drivers, I ain't 100% sure, but whatever it is that makes them fail so frequently needs to be addressed because it's costing us, the taxpayer more in replacing these bloody things than it did to just keep the original fixtures in place. Heck, I bet we'd still be saving money running 300w incandescents!!!

[@jjab99]

Great video as always, Many Thanks Clive.

[@RomanoPRODUCTION]

Sometimes an honorable citizen gets the keys of the city but here it's better, BigClive has the Light of the City 😇😇😇

[@HighestRank]

'Lights of the District'

[@davenorth1265]

I have the same Wera set, it was from a Wera Advent Calendar a few years ago. lovely set.

[@phonotical]

I recently managed to get ahold of the pointy thing on top of Street lights, I thought they were light sensors, however close up, it looks to be an antenna, if you're interested? Otherwise I might give it a shot at some point😅

[@TheSpotify95]

Older ones were definitely just photocells - however, the new ones (as you rightly point out) are antennae which connect to a base station somewhere, and the base station turns on groups of lights all at once. I preferred each unit to have its own photocell - less problems doing it like that, whereas the failure of the photocell at the base stattion will negatively affect a large chunk of lights.

[@CTCTraining1]

You don’t mean the connection to the mobile network that the council uses to switch the lights on & off with?!

[@mahlapropyzm9180]

@@TheSpotify95 Interesting Two problems - presumably hackable, and has a single point of failure at the base station and it's software (which of course never fails.. to disappoint). The future of street lighting looks dark.

[@bigclivedotcom]

I have both types here. It's a low power communicator.

[@Asbjoern]

@@mahlapropyzm9180 why can't there be 2 base stations? Or just 1 with redundancy built in?

[@russellhltn1396]

This begs the question of what failed first? Was it defective insulation? Did an LED open and the resulting high open-circuit voltage cause the failure, or was there a high power-to-ground voltage spike (perhaps during a lighting storm) that damaged the insulation?

[@theskett]

I'd put a dollar on your #2; a dry LED joint and some sparking, lots of local heating. The heatsink is isolated from the driver (BC says), so I'd expect a voltage spike to blow the driver rather than puncture the PCB insulation.

[@cougarhunter33]

It would be interesting to see what the baseline RF being put off compared to what was happening during the failure. We have a lot of RF noise around here from arcing power lines and once triangulated and identified, the power companies were pretty slow to remedy the situation.

[@uK8cvPAq]

nonce

[@KuntalGhosh]

if the throw these out it's worth picking them up just for the heatsink. and these r pretty good quality lights u can use it as a studio lamp or outdoor lamp.

[@dragonrider4253]

Those might be "no longer manufactured" because this fault (or a similar one) came to someone's attention after some number of units were shipped and installed. Also, the kink palculator gets a chuckle every time.

[@Purple431]

Hello Clive! Let's bring in the "kink palculator".

[@boban250]

Is there a reason for soldering to the large pads instead of the existing solder pads left by the LEDs? Burning through the top coat and leaving the solder pads seems unnecessarily ugly

[@86abaile]

I work maintenance for a well known super market chain. I hate philips LED lighting; I've had so many of their products fail, especially their OEM fridge lighting. Where possible I avoid philips like the plague.

[@SedatedByLife]

Let it be a lesson - warranty doesn't always mean anything. Love the little asterisk companies put in their agreement. "oh but if we don't make it any more - you're SOL."

[@plainedgedsaw1694]

Could it be weak spot on insulation layer of PCB that made it arc between track and substrate? In that case lowering the amount of LEDs in series is gonna shunt the voltage lower and help with the reliability.

[@angst_]

It might be possible to preemptively upgrade the units to prevent arching. Removing the PCB and installing some type of thermal pad as an insulator. Something that will still transfer heat well but provide a better isolation.

[@HighestRank]

"arching"? so you did think it's a problem with mechanical stress.

[@angst_]

@@HighestRank Typo. electrical arcing*, like he mentioned in the video.

[@userPrehistoricman]

Isolation to the heatsink? Won't help because the PCB itself has an aluminium plate.

[@francoisleveille409]

People find these lights too blueish. I never understood why the companies wouldn't simply add a few green or red LEDs in the grid instead of lowering the white point of white LEDs. It would be much more energy efficient. Myself, back in 2009, I made an electronic ballast for 1kW high pressure sodium lamps so LEDs were the enemy. HPS lamps can churn out 140 lumens/watt for more than 20'000 hours. LED are now nibbling at this level of efficiency.

[@Dingbat217]

Can you lower the set-point of the PSU to perhaps lengthen the service life, at reduced output obviously, then that could be a better longevity fix I would think?

[@vivekchauhan7468]

Recently i teardown black box which is between power line 240(1phase) and street light. Which has 3 input terminals(live, neutral, earth) That is full of some kind of resin. And after removing resin it has 5 MOVs. It's connection is: 1MOV between live and neutral, 2 parallel MOVs between live to earth, other 2 parallel MOVs between neutral to earth. I didn't get connection with earth MOVs

[@GeorgeGeorge-xj2bc]

A giant heat sink,a led module prone to failure and a separate inverter for each lamp i am still wondering which are the benefits of led lighting technology over the previous lighting technologies.

[@NoblePineapples]

"Kink Palculator" reminds me of Kyle from BoostedBoiz saying he was going to "Caint the Page" instead of Paint the cage (in reference to the roll cage in his AWD 1,000HP 2001 Honda Odyssey)

[@will_doherty]

Only watched half of the video as of yet, but my immediate thought was to lift the "pcb" off the heatsink, fit a thermally conductive but electrically insulating shim between the two, cut out the burnt area and replace the missing track with a wire jumper. The pads that you noted as being compromised are only potentially compromised with regard to earth, since between adjacent pads you're only looking at 3V max potential difference, which in the scheme of things is negligible, relative to the separation distance. If you only warrantied your own repair work, there's a killing to be made from repairing these units in this fashion.

[@will_doherty]

Having now read your description that accompanied the video, I see that it wasn't tracking to the heatsink, so what I wrote above is unfortunately not relevant. Next time I'll check first :-).

[@ehsnils]

Was it the LED or the substrate that failed? My thoughts here is that it might be a moisture issue causing it to arc to the aluminium sink. Depends a bit on the PCB used if it's prone to absorb moisture.

[@JJ-kr6ky]

Love the cink palculator!

[@Michel-7.7.7]

A nice heatsink to build a nice growlight

[@argoneum]

Looks like moisture got in, got between the substrate-plate and the trace somehow (found a weak spot), and then started eating the trace electrolythically. Add some conformal coating maybe? Just a thought, seen some effects of water acting on DC things at my job, the "anode" is what turns from metal lump into metal ions and vanishes eventually :)

[@bigclivedotcom]

No obvious water marks.

[@tonythemadbrit9479]

It looks like the connector to the PCB also ran a little warm based on the discoloration of the plastic.

[@nooneyouknow9399]

Rather than a wire jumper, a copper foil tape would be only ~0.005". You could also cut existing traces to isolate the bad area and possibly relight one or two LEDs. I'd be curious to see the back side of the board to know how it was able to punch through to ground.

[@theskett]

There's apparently a requirement that 95% of LEDs are lit; three LEDs => 6%, so voids that. But if you can kill only one or two LEDs, you get to stay in spec :-)

[@FrontSideBus]

It's quite common to see streetlights with dozens of failed or really dim chips, sometimes even whole modules out. Maybe it's by design as if it doesn't work then it saves energy!

[@HighestRank]

then *

[@Basement-Science]

Since it failed somewhere in the middle of the string, wouldn't that mean the fault current would have run through all the LEDs before it as well? Surely that wouldn't have been too beneficial to them either. Kinda depends on the driver I suppose.

[@HighestRank]

Like Clive says, it doesn't appear to have been a failure of the LEDs themselves.

[@Basement-Science]

@@HighestRank Of course. But the FAULT in the PCB may have also damaged a lot of the LEDs, which could lead to them failing soon after this fix.

[@theskett]

The driver is constant-current, so we'd expect the "fault current" to be closely-similar to the "normal operation" current and hence not be harmful to any LED that didn't burn. There's no particular reason to think that "the FAULT in the PCB may have also damaged a lot of the LEDs"; it's not clear that the PCB was faulty before the tracking started, because the problem may have originated with a dry joint on the burned LED, for example, causing significant local heating. Of course, it's certainly possible that the soldering was botched (such that other LEDs are soldered imperfectly) or that the PCB insulation is inadequate -- but neither would be a result of this particular failure. The main problem is liability, both from doing a "hack job" and from reducing the light output.

[@Basement-Science]

@@theskett I'm just saying it's a possibility, and if true, other LEDs could have been damaged. It is a constant current driver *_when working normally._* The reason I suspect it might have exceeded its normal current is because any CC driver powered by mains first uses a Full Bridge Rectifier, and when you short some connection after that to Ground/Earth (which the Heatsink and back of the PCB is connected to) some of the diodes in the FBR are bypassed and even an AC current could have flowed from the driver through the LEDs to the point where it shorted to the Heatsink. (again depending on the circuitry in the driver) In that case it wouldn't matter why it started to fail, just that it ended up shorting to Earth in the middle of the LED string.

[@theskett]

BC checked that the CC driver connections were isolated from the heatsink, and the CC driver appears to be functioning correctly. If OTOH the CC driver had a connection to the mains, and the output was shorted to ground, there would likely be a loud bang; and probably a jump-cut in the video as the GFCI / RCD tripped.

[@blackcountryme]

I live in a maisonette, my bedroom is about 10 feet from the road, and one of these type of sodding lights. I preferred the orange ones tbh.

[@frankowalker4662]

Same here, I had to get blackout curtains for my bedroom. This is'nt a main road or anything, the council just rocks up and puts these all over the damm place.

[@blackcountryme]

@@frankowalker4662 now, I don't know if it's just me, but are they darker? I mean you get a pool of radiance and a pool of darkness between the lights.... The orange ones gave a all over glow.

[@K-o-R]

Sometimes you can ask the council to put a screen around the back of a light if it is shining in a particularly nuisance way. There's a few round here with little shield pieces fixed on them.

[@shepardpolska]

@@blackcountryme LEDs usually trade lumens for color rendering. I for one see better on darker LED lighted roads then with ones with sodium lamps. Better color rendering is safer for pedestrians, or so my country says

[@frankowalker4662]

@@blackcountryme They are not darker when they are 8 feet away from your bedroom window shining down with cold white LED's. :) The warm glow of the old sodium was soothing by comparison.

[@quantumbits]

I guess Once and arc starts there is something reactive supporting it. Or maybe that's some weirdly specific lightening damage .... Nice repair

[@tybofborg]

I would suspect spring storms and moisture/water ingress played a role, especially since the failure occured on one of the outermost lines.

[@Zadster]

Condensation under the LED? Or perhaps moisture in the substrate causing initial conduction, then arcing causes carbonisation and away it goes. Or perhaps a dry joint caused by thermal cycling (then arcing, carbonisation...). Nice panel and PSU anyway. Would be nice to see inside that PSU.

[@mahlapropyzm9180]

Ditto, possibly a crack in the substrate, allowing water ingress...

[@bigclivedotcom]

The PSU is potted in what looks like pitch.

[@Zadster]

@@bigclivedotcom A shame, but totally understandable.

[@1kreature]

Design fault or laminate production fault and they "discontinue" it and don't stand by warranty?

[@milksheihk]

Was there a refund? Under Australian consumer law warranties have to be honoured even if the item is discontinued, either repair, refund or replace with an item fit for purpose, even in cases where the manufacturer/service provider has actually closed down there is a mechanism in place by which consumers of products that fail in warranty period can be refunded.

[@draconightwalker4964]

gotta love our ACCC

[@milksheihk]

@@draconightwalker4964 They have a lot of good laws in writing but they can be a bit toothless in the enforcement at times.

[@Ni5ei]

They failed under warranty but Philips says "Sorry but these are discontinued"? I hope they gave them a newer model for replacement?

[@renowden2010]

Hi - I came across a similar panel floodlighting Harlech Castle. I’m curious why some of the clusters are twos and some fours at random across the surface.

[@bigclivedotcom]

Depending on the light it may be for beam shaping or illumination evenness depending on the throw of the light at that angle.

[@stickystinky]

Im so thankful for the work you put in to your videos. Your videos, are a joy. I have a trolling motor I'm trying to rig up. When the 48v thrusting (forward backward) motor is on the 48v motor that turns the unit won't work any suggestions? I just used a contactor for thrusting and relay for turning motor...

[@theskett]

Do you hear the contactor click, and can you confirm the 48V at the thrusting motor terminals using a multimeter? What happens if you connect the 48V direct to the thrusting motor, eliminating the contactor?

[@stickystinky]

Yes I do. Both motors are working. And they are 24v sorry, so two gell cell car sized batteries. How ever both thrusting motor and turning motor will not operate at the same time. I feel like the thrusting motor is stealing all the power not allowing the turning motor to turn. So I can move forward momentarily and use the turning motor while the thrusting motor is off. I actually have it wired so the thrusting motor has two speeds 12v and 24v. While in low speed the turning motor turns a little. A push in the right direction is greatly appreciated. Where do I start? Thanks a million

[@theskett]

> However both thrusting motor and turning motor will not operate at the same time. Is one or the other motor perhaps locked up, and so taking all the juice from the battery? Do the motors operate ok, independently (i.e,, one at a time) ? So you know they're not both broken, and will run? I'd go down to the local scrap-yard, and get a couple car batteries (12V, $15 each), and tap those across across the motors. The motors should either turn, or catch fire. If they catch fire, you replace them; they're dead already. If they spin nicely, the problem is (maybe) the contactor or the control circuitry. Do you have some kind of voltage regulator, isolating the controller from the motors? Often, turning on the motors will put enough voltage spike (up or down) that the controller crashes.

[@geoffturner1487]

So when will you be converting it from a capacitive dropper to resistive to make a nightlight?

[@the1spyderryder]

That track looks like a lightning burn track. I am thinking that it is on a ceramic board that the LEDs are mounted on.

[@jkobain]

I wonder why they discontinued them before their EOL, especially during the valid warranty period.

[@HighestRank]

Let's suppose that the last purchasers of a proposed EoL product would receive shorter and shorter warranties, then the logistics of printing and supplying the correct inserts becomes burdensome.

[@jkobain]

@@HighestRank this is why they should recall the goods or implement a proper full refund policy _if they discontinue a thing that's being actively sold._

[@tothemaxx1991]

Our street lights have a purple failure mode. It's trippy, like a vaporwave thing

[@davey6024]

Yes Clive nice to see u rocking the weras instead of that screwy out of a cracker 😉

[@Enoch001]

Is a streetlight module what you call the band members of streetlight manifesto?

[@richardkaz2336]

Better to identify if all lights are failing in the same LED group and in the same manner and if it's only on a particular part of the distribution grid then find exactly why they are failing and permanently fixing the problem. There appears to be discolouration on the plug suggesting heat due to over current and or over voltage with the weakest link acting like a fuse.

[@AutoUnder]

It appears as though LED street lights are pretty much made to fail like standard LED light bulbs; I’ve seen quite a lot of LED street lights in my area (made by Thorn and Urbis-Schreder) fail not long after being installed

[@gudenau]

I think we might know why these got discontinued. That's quite an interesting failure mode.

[@dancoulson6579]

I wonder if they could develop a special LED that utilizes a similar breakdown enamel junction used in the old christmas fairy lights? If an LED fails, then the voltage accross it would drastically rise. Then the enammeled wire would conduct and the remaining LED's should continue to light, even if in series.

[@bigclivedotcom]

I've seen that in other lights.

[@Hyratel]

I have a set of LED fairy lights, I had to do the Bodgy version of that solution when one of the diodes went out, darking half the string - twist the dead diode so the legs short together. net light is the same, despite the rest of that half-string being brighter due to it being a series resistor on mains

[@TheWellnessLawyer]

Love it when the local utility says help us, find why we have doggy equipment.

[@nicholasvalentine3907]

Would not a compromise be to just replace the whole circuit board, (one of three) or do Phillips not supply them individually? Repaired in any form though, they would be fine for a carpark or for private driveways, I don't think I would risk the liability on any public road. PS. For my own drive I would take out 2 of the 3 boards to bring it down to about 30W.

[@bigclivedotcom]

It's a single PCB.

[@nicholasvalentine3907]

Ah, it looked like 3 at a glance on my tablet, but I can see that does save on the manufacturing costs and Phillips was probably thinking of a municipal budget! Thank you for replying.

[@sirflimflam]

How can it be in the warranty period but Phillips say "no, it's not made anymore"? Shouldn't they have to handle the warranty one way or another?

[@garrettkajmowicz]

One way the company *might* handle this is to send out a replacement of a compatible new model. In that case they simply accept the claim as-is and ask the current owner to dispose of the broken unit. This saves on shipping and disposal costs. Unless the unit has a lot of intrinsic value (eg. scrap value of aluminum), there isn't a lot of point in getting the replacement back unless they can refurbish, repair, or analyze the root-cause of the failure. Unlike with consumer products, there isn't a lot of motivation for purchasers of these products to try and run warranty replacement scams against the manufacturer.