Summary
a true horror show unfolded.
The best guess here is that someone took an originally cheap NiMH cell-based solar light circuit and simply replaced said NiMH cell with an – unprotected – Li-ion 18650 cell, as there is absolutely no charging management or cell protection. This is important as while NiMH will throttle charging current by rejecting it as heat the more it is in a charged state, Li-ion is not as tolerant.
Even ignoring that the ‘IP44’ enclosure has more gaps and holes in it than the average slice of Swiss cheese, the 5.5 V-rated PV solar panel happily tried charging the Li-ion cell at around 6 V. This raises the potential that on a particularly sunny day with an already charged Li-ion cell sitting at 4.2+ V, it’ll effectively grenade this cell.
During an outdoors test the 18650 cell was charged to 4.8 V before the sun vanished behind the building, which probably isn’t an amazing sign if you are expecting for this light to keep working for more than a few weeks. Or not violently explode/catch on fire on the side of your house for that matter.
Of course, treating a NiMH cell better than in this particular circuit will extend its lifespan as well, but the failure mode of that battery chemistry tends to be a bit less aggressive. Hence why people keep getting away with dodgy circuit designs like this.
I’m surprised. Not by the degree to which it is cheap and nasty; but by li-ion getting the nod for a product being cost optimized straight into dangerous dishonesty.
Maybe I’m just old and out of touch; but I still think of this as an application where a vendor who doesn’t really care about the fact that they are probably also lying about how long the light will run from a day’s charge will just shove a nickel cadmium cell in there and call it good: tepid energy density and the cadmium issue; but cheap and takes abuse pretty well. NiMH if they care slightly more, since the self discharge won’t be a huge deal on a 12-24 hour timescale and the energy density is better.
Has Li-ion just come down that much across the board; or is it a case where proper Li-ion would still be more expensive; but the volume manufactured for more demanding applications with actual QA means that irregular lots turn up for a pittance in quantities large enough to bodge in but not necessarily large enough to re-engineer around?
I’d guess rejects/recycled/reclaimed vape cells, which makes this doubly dangerous.
China may have run out of cadmium to dispose of in cheap yard lights and now has dodgy Li cells to dispose of.
They use second-hand cells salvaged from old laptop packs etc., hence capacity is somewhat random.
Adding protection isn’t that hard, those little 1-cell BMS circuits are cheap. They won’t regulate current (a little solar cell can’t cause too much havoc), but at least it’ll shut off at 4.2v and again at 3v-ish.
I once bought a knock-off laptop pack to salvage the cells, I wasn’t expecting much and the pack certainly delivered. I think the cells were supposed to be 2200mah based on the pack label, when opened they said 1800mah, peeling off the wrapper showed 1300mah, and they tested at about 1000mah. That was fine but the cascade of lies was amusing.
LOL nope. Any “entrepreneur” can import crap from China, slap a “CE” sticker on it (which is just “declaration of conformity”, not “tested for conformity”) and call it a day.
Sometimes consumer protection agency (like our Office of Competition and Consumer Protection, UOKiK) will take action and declare something as “dangerous good” which must be recalled (and in some cases existing stocks must be destroyed). They may also issue fines, but those are so insignificant that one may factor it as part of doing business costs. In the end it’s like using a flyswatter to destroy a single wasp on the wall when there’s multiple nests inside your house anyway.
Take for example those really small chinese “300A inverter stick welders” flooding the market. They (UOKiK) may issue a recall on one specific model because it: uses too thin welding cables which would catch on fire with current of less than 100 amps; crocodile clamp is the same one as in cheapest jumpstart cables; declared welding current is bullshit; inverter board has insufficient insulation between primary and secondary creating risk of electric shock, PE connection to the case doesn’t use locknut and may will fail, and finally, despite being rated at 3600W of power, the device is using 3×0.5mm² mains cable.
That’s fine but it takes time and money to test it, write a report and issue a proper, legal recall. Meanwhile, the same crap inverter welder will be sold under yet another “MAX MASTER 300XZ-PRO SOMETHING SOMETHING” brand within 6 months and the whole cat-and-mouse game starts again.
Only EU-wide bans are working really. There was a time when weedwacker heads with chainsaw flail chain were commonly available and became popular among DIY homeowners. Their (mis-)use led to some pretty horrific injuries. I mean this stuff: https://www.amazon.com/Universal-Weed-Eater-Head-Accessories/dp/B0D6N4BSQJ
They were banned a few years ago under some “flail mower” EU-wide regulation and they’re gone from regular stores. Nowdays you may sometimes find one only on a bazaar, imported directly from China in small batch, but that’s a niche and not worth the fight.
Honestly, I have no idea how to effectively prevent such bad products from entering EU-wide markets without also creating a massive shortage economy reminiscent of the 1980s.
One thing I forgot to mention in the list: included welding helmet is only a skeumorph which is dark but freely lets UV light pass through the glass. Mr Sum Ting Wong from Shenzhen who manufactured it is not concerned because no one will be able to sue him for injuries in the state capitalism of PRC.
Examples, please?
Please do keep in mind than polycarbonate (used for the clear cover) and acrylic (AKA perspex, also sometimes used) are both excellent UV blockers. 85%+ UV-A- 95 to 99% at most practical thicknesses- and 99%+ UV-B and UV-C, unless, of course, the significant extra expense is paid for Acrylic that transmits into UV-B. Standard silica glass is ok in UV-A (25 to 75% blocking, depending on the type), and almost as good and the polymers in UV-B and UV-C.
Anything used in a welding lend to block visible will Greatly improve any of these across all UV.
Plain, uncoated acrylic, polycarbonate, or glass may not provide the full shade 14 minimum in UV-A or B (optical density 5.5 or higher- blocking 99.9998% of UV), but to not at least come close to meeting this with cheap materials that are optically clear would be tough. The lens and clear covers may not stay optically clear for long (UV-B and C yellow or brown polycarb and acrylic fairly quickly), but they will likely provide UV protection.
Infrared, on the other hand, is a real potential problem, as many materials transparent in visible are also transparent in IR, at least near-IR. But, again, if the visible light is attenuated cheaply, likely IR is as well, unless a specifically near-IR transmitting material is used, like the window on your television remote. But that isn’t cheap. Typical IR is OD3 to OD4 (99.9 to 99.99% attenuation) for a welding lens, but ANSI
If it’s only 95-98% effective, you’re still burning your eyes out – just slower. You may not get “sand in the eyes” from a bit of hobby welding, but the damage is still there.
Another type of fraud is the auto-dimming glasses that don’t dim fast enough and/or block UV very well, because they’re just regular cheap LCD panels not suited for the job.
Again, UV is generally not the issue. Slow lenses lead to visible, and maybe IR, flash. regular, cheap LCD are slow, but the materials are still very good at blocking UV.
SLow response is problematic, as visible light flash is fatiguing, can lead to mild feel of arc flash, and can also (potentially, though I have seen no reports. Really, if you have any references, beyond “some guy on a forum”, tell me) cause primarily retinal damage (Blue visible and IR are hell on the cornea). Trust me when I say genuine arc flash injury is brutally painful.
There are a number of go-to guides, but the US military (I think it is NavPers, but I am not on my work computer to confirm) manual covering welding safety have the weld hazard index data, equation, and time-process-current charts. Also, the ANSI standard for welding lenses and the AWS references can be found online with minimal effort (I had to pay for them…)
Fun fact: this is a piece of the safety side of my professional work. We like our crew to go home with the same number of limbs they showed up with, in the same condition. Ditto for eyes, lungs, and all other organs. DO NOT take the guard off of your grinder unless the mounted tool specifically does not require it. Our foremen WILL boot you from the job to unpaid leave.
Maybe making marketplaces responsible for the crap they “totally don’t sell because they’re just a marketplace” would be a start? How the U.K. is making them responsible for collecting VAT.
The current economy operates on the premise of customer making a conscious decisions whether he wants a product or imitation of a product.
But does that really work when people are not really educated while conmen are everywhere?
i agree it’s dangerous but this story — “i found something sketchy. and i didn’t use it. and it didn’t start a fire. and it did not fail prematurely” — is specifically not telling ‘why’ unregulated lithium chargers are dangerous. it’s just asserting that they are, which, well, yes.
just sayin…headlines etc.
What about a solar light with anything less than ip66 rating. Crazy, if you need to put it under cover then how can the solar be effective?
Something seemingly well sealed but in fact permeable to moisture will tend to collect moisture. Something full of holes will collect moisture but will also vent out the moisture.
The IP rating isn’t anything to go by, because it’s so easy to follow the rules and cheat to claim that your stuff is “waterproof” when it just isn’t.
If you have an IP66 rated device and you leave it outside in the rain, water will find its way in – usually by wicking in through a seam or a crack, which takes longer than the 3 minute test. I’ve had IP67 rated devices destroyed by water in conditions where the water didn’t immediately drip or evaporate away from the gaskets and seams. In other words, outdoors in the weather. I no longer trust the IP rating at all.
Even IP68 doesn’t guarantee anything, since the manufacturer can specify how the device is dunked in water and for how long, This means that e.g. a cellphone can get an IP68 rating because the manufacturer says that it must be submerged with the screen side up, so the case traps an air bubble inside that keeps the water from leaking in through the back cover seals. The same trick they use for cheap “waterproof” watches.
Link to some video of fire of lion 18650 overcharged (by such small pv cell)? Nevermind, i will try to find one for me:)
I have three similar for 2 years. Functionality intermittent but in expectations accordingly to cheap price.