As much as presentation software and screen captures are great, I just love the fact that you print out the macro board shots and thermal stuff. It works so well.
Hey Clive, thanks again for another informative video. I have been playing with those other kind of 'driverless' chips. The ones I purchased were from 'MB Lighting' and were pretty cheap. I tried the 20,50 and 120w versions which were $2 a piece for the 20 and 50w and $9 for the 120w. I liked the look of the 120w as it's surface area to power ratio was the best (thinking of heat transfer, not really sure how important that is in relation to other things) I have not had any issues with the smaller chips although those are not in use as much as the 120w chips which are on almost 24/7. Out of the 12 I have used, 10 have failed, the majority of them would still be emitting light from some of the leds but you could just barely see they were on. I was extremely careful in fitting them into the housings I had (which were easily suitable for power rating) however I cannot rule out user error. I noticed that my second order of the same drivers showed that they had redesigned the chips since my first purchase, the size of all were noticeably larger (the plate at least). Of the second order, I have had two failures. I feel some of the failures could have been from the wire connections to the chip itself. With such a large 'heatsink' as such, it does make for a challenging solder job. A lot of these new chips have been repopulating Casings from older 'failed' floodlights that I have accumulated from my adventures in the quest for the perfect LED. These new chips look great, The warm white is perfect color temp for my liking. And the price, well suspiciously cheap, but they look as good as any other I've seen. My greatest find over the years was a 100w (75w it turned out) Flood on amazon for $26. Now the housing was junk, too flimsy and you could tell the driver didn't stand a chance, so I swapped it out for a better casing before I installed. This was my first 'driverless' device I had come across, I believe it was a 16x8 grid of 0.5w Led's with 110v plumbed direct into it. How was a great find you may ask? After the initial reworking, The light itself has been amazing, Has been on for 22 hours each day at least, color temp is perfect, and looks as good now as it did on day 1. The biggest down side though has been the costly pursuit to find more of the same, and that where most of my housings have come from!
most youtubers let the success go to their head and they stop doing what they started out doing and get lazy, the money goes to their head too, clive makes the exact same video now as he did way back, thats how i like it
I'd be nice if you could do a video on a proper floodlight you can buy in Europe (like the Osram Ledvance or Geyer floods that are sold at quite attractive prices). The comparison could be enlightening.
From a cursory outside look at the electrical supply store, there is really no comparison between the quality of an Osram light and a no-name Chinese light. Even normally imported (not gray import) Chinese lights are really bad in many ways and QC tends to be terrible.
Good to know others are fascinated by lighting tech as much as me. Uneven cooling of the led array and poor cable handling is not a good sign. Keep up the good work Clive. 👍
The squeezing of the cable is obviously intended as a strain relief so the cable isn't pulled out. That aluminium plate was probably chosen to match the thickness of the squeezed cable (if the cable wasn't squeezed in there, the plate could just be omitted). Unfortunately, they didn't include the cable sheath in the squeezed zone. Anyway, a flat product like that should use a side entry to not waste its own thinness.
I'm guessing their trusting the silicon potting and conformal coat to keep moisture out, even if the seal around the cord or lens gasket isn't quite up to par. They put stuff on there to keep it from flowing out before setting, so it stays thick which may actually work. Thermal paste seems to be a QC thing, somebody being lazy there. The strain relief or apparent lack of it seems to be the worse problem. So with better QC and better cord mounting, it might not be that terrible.
You could maybe fit a switch somewhere on the case, have the terminals switched off normally. One of those old-fashioned metal toggle switches would look nice.
Typical, once you got your tool in there there was sticky white stuff everywhere, well, glad to see you’re a gentleman and don’t wipe it on the curtains afterwards, Clive, but have you never heard of a wet wipe?
Andrew Cooke Imho, it's that the screws have direct large surface area contact with heatsink compared to much smaller (possibly indirect) contact area with the led driver heatsink.
Hi Matt, Infrared cameras are fantastic diagnostic tools, but they have some severe limitations that it's important to understand, it's complicated, but in short, don't try and measure shiney stuff.. have a quick tread of this : support.fluke.com/find-sales/Download/Asset/2563251_6251_ENG_B_W.PDF?trck=emissivityexplanation
Andrew Cooke The stefan-Boltsman law isn't that complicated and I own and regularly operate a Fluke Ti110 on many materials including the 'shiny stuff' I still stick to my opinion from experience in the industry. Thanks for your input though.
Andrew Cooke that's for you to decide but my unit has saved me hundreds of hours. Your point is a factor no doubt about that but mine is also. The screws are most of the way into a colder heatsink and the head down onto a white and somewhat insulating layer over the led plate including under the screw head. Also Note the halo of cooler area just around the screw head in the image where it is just partially cooling the led plate from the screw head. If the screws weren't cooler then there wouldn't be a hotspot on the led array in the bottom right hand corner where a screw is absent as mentioned in the video.
error079 This is very poorly assembled. It would make a huge difference if mounted correctly. But the surfaces are no way flat enough and the cable input is a no go. If you put on too much thermal grease it's a thermal isolator compared to aluminum. That transistor needs an additional heatsink, too.
I guess one way to get around the squished wires issue, use an externally-mounted driver in a junction box near the light and fit a "traditional" DC LED module with a much smaller footprint, yes it adds a requirement for additional wiring, but, saves squashing mains powered wires to a potential short... :) And were you filming this at around 8pm-ish? The mains frequency went down a fair bit on the Hopi, indicating the end of a dreary TV soap and everyone powering up their kettles... :P
twocvbloke , ¿¡ !? ¡¿ That much measurable frequency drift ¿¡ !? Gonna have to look for that during an analagous event here in the US ! Ponderous... wonder how "smart power grids" technology is supposed to mitigate said affectations...
The voltage dropped by a couple of volts, but only when that 52 Watt load was added. The total resistance (impedance) of the cabling to that meter was probably around 0.075 Ohms, which would account for a ~2V drop, using the formula (V^2/W) = R. So ~4/52 = 0.077 Ohms.
I bought four COB chips after your first tear down video, 2 x 30 W and 2 x 50 W. I suspect the throttling does not work that well because the first three chips died rather quickly, approximately 3 weeks after install in different setups. 1. 50 W COB, ran on an old GPU-cooler with heat pipes and a 12 V fan in the tower. When it was cold it hit 53 W, then after a couple of minutes it throttled between 44 and 46 W. It managed to reach steady 70-100 C and all of a sudden only haft the LEDS worked faintly. Every second row worked as if it rode half a wave. 2. 30 W COB, put on another cooler only with ambient temperature to keep it cool. The surface area was a bit larger. But the same problem. It started at 34 W when cold then throttled back to 29-30. It also died three weeks later with half the LEDs flickering. 3. 50 W COB, put on a large sheet of aluminium with a CPU-cooler behind. Experimenting with the size of the surface area. No better luck here. 4. 30 W COB. This one is still running! I used the same CPU-cooler as in 3 but with a 12 V fan attached. This runs at 23-35 C, but without the fan it runs up to 75-105 C and throttles to 27 W. With the fan it does 33 W. I have a fifth chip from another supplier at 50 W waiting to be put on the CPU-cooler when 4 dies. The throttling therefore have only a marginal effect, because it cannot lower the power enough. I also found out you do not need that hot of an soldering iron to attach the cables if you only use old, 1987, Fluoridic based paste. It does however make you dizzy if you do not evacuate the air fast enough. Perhaps that is why it is no longer sold (lead based solder...) Enough surface area and fans are a must with these. One car battery and 12 V fans with a charger attached and an array with these chips. Or find a large industrial fan at 480 V and try to convince the wife it is for science. www.processenvironments.com/cartPics/4cb8c6797c051IRO_782_ARR_1.jpg
Clive. would you be able to do some more usb charger tear downs ? its something we take for granted everyday and would be nice to see different good/bad designs :)
I was thinking the same thing, but for an entirely different reason. I'm surprised he hasn't taken the HOPI apart and reworked the circuitry to stablize the flickery display for the camera.
It's earthed! Will wonders never cease? However, after having watched some of John Ward's videos about Chinese mains-powered stuff, the next question I have after "Is it earthed?" is "How good is the mains cable?" John's videos have exposed some really crappy Chinese mains cables: seriously under-rated for the task, crappy insulation with flame-enhancement properties, and conductors that are copper-plated to disguise the fact that they're made of damp string. Then again, you say the cable claims to be decent stuff rated for high temp (but it could be lying). So my big worry about that unit is the placement of the cable gland. It's in a *really* bad place where water will run onto it rather than away from it, so unless it makes a perfect seal you're going to get serious water ingress.
This cable have CCC Certification (Chinese standard) markings which should mean is tested in proper third party certified lab and with equal/higher standards than CE. But anyway CE is good standard if there was not ability that you do CE marking self-certification (you do not need to any third party testing and certification, you basically just need to give statement that product is compliant with CE standards).
European CE is not the same as Chinese CE, which means China Export; not to be sold in China. The two symbols look alike, when viewed quickly, but they are subtly different, as can be verified by a quick online search. And there is nothing to stop someone who is dishonest from printing the Chinese CE symbol on something which is substandard, or even lethal, and then making sure that it is not sold in China.
RWBHere There is no such thing as Chinese CE (China Export), it is just incorrect application of CE by some Chinese manufacturers, to be able to sell merchandise in China it must comply with CCC standard markings (so they do not need to have some marking that will say it is for export only, if it do not have CCC markings it can't be sold in China) and must be verify by independent laboratory (only certified testing centers can verify that product is compliant with CCC standard), contrary to CE (or how you call it European CE, but only that one exist) which is self-certify for most of products (all electronic products do not need to be tested and certified by a third party certification body you just need to fill EC declaration of conformity), only little product require to be certified by " Notified Body".
Excellent video, as always. Could you maybe do a test with the temperatures when you've made sure the thermal paste is actually gripping the whole board, and with that one corner screwed also? Like re-route the cables so that that is possible. Also if you could test it with one of the "driverless" leds you have, seeing if there's a difference and how it works. I do like the idea of smaller and "neater" light. I need to get one of those so that it'll help me fight against the darkness and the blues it brings.
Clive, the squishing of the wires allows a clamping effect to keep the mains cable from falling out. What was the lumen output of this thing? I'm recently seeing 50w LED floods marked (true 35w), so they're starting to become a bit more honest.
The screws look colder on the LED because of them being reflective, put some black tape on them and then recheck their temp, or paint them black, then you will get the correct tmeperature.
Unit prominently claims an IP67 rating on the back. IP67 means: Ingress of water in harmful quantity shall not be possible when the enclosure is immersed in water under defined conditions of pressure and time (up to 1 m of submersion). So stick it in 3 feet of water for 30 minutes and no water will leak in that gland. Maybe.
That's why he lives on The Isle...Puts 'em in a cardboard box, writes "love from Ireland" on the outside, sets them afloat when there's not a northerly breeze, and gives them a "Viking Funeral". With any luck, lands on a beach near Liverpool, starts a wildfire, burns out Liverpool, and the parts make it look like the Chinese had a sub burst into flames on the high seas. Everyone is a winner...Clive gets rid of the leftovers, the UK intelligence services get to spend more money looking for the wrong people, China gets blamed, and Liverpool gets to rebuild.
I had one of those integrated cobs fail recently and thought about sending it to you.... It first started flickering after running for around 8 months - prob 10-12 hrs/day. Finally it just died entirely. Was on a very large AL heat sync - IRC is was a 30w 120v version, but long rather than square.
Besides Grilling Food outside. Another thing us guys are great at. Sizing up a screw or nut and picking the correct size screwdriver, socket or wrench. :)
Grow lights have been available with that driver arrangement have been available on eBay for at least a year or so. Haven't seen a white LED like that until now though.
STM I wonder if that's ST Micro? If that's the case then it could be a custom PIC12 re programmed to contain all kinds of goodies.... perhaps pre-programmed failure so you have to buy another?
STM could be ST Microelectronics and 18C242 is a PIC. However, it's probably some bullshit Chinese part number to cover up the fact it's a clone of something else.
restcure When you have a customer who wants a chip that looks legit-ish but won't check it beyond the label. Or you have a supplier with a fake chip that looks legit-ish and they'll sell them cheap. Nobody asks too many questions.
17:30 Well...that seems like a cable fire just waiting to burn down your whole house, with best wishes from China. I was planning on buying 2 of those exact same lights (may be different name/manufacturer, I don't know, but looks EXACTLY like the one I was eyeballing on Amazon) to serve as grow lights for indoor hydroponics...am I glad you picked this thing apart, sir. I own a house and a forest, and I would rather the damn forest burned down 10 times in a row than have a fire start in my house.
They could of pressed a depression into the metal to allow for cable entry (this can be done sit punches the hole, in a single operation, doing this would of increased the cable space, then they could of put on 4 mounting screws instead of 3.
I had one of those 50 watt modules fail after I left it on a 12 hour timer while away for 3 days. It might have been a heat-sinking issue since the sink I used only contacted the back of the module in certain places. It also occurs to me that transients might damage that sort of module pretty easily, which is probably why the work light one had an MOV on it.
I'm wondering if anyone can answer a technical question about LED lightbulbs. I'm in the US. All of the bulbs in my house are GE Reveal. Originally it was the Reveal incandescent, then the Reveal CFL's now the Reveal LED's. It's the only bulbs whose color I can stand, it filters out the yellow without being overly cool like the fucking god awful cool white daylight bulbs. The LED's aren't quite as good as the incandescent were but still very close to the same color. ANYWAYS, I've noticed that with the LEDs GE has updated them a few times over the past few years. Now they're branded as GE HD Reveal. It looks like they've got the wattage down a bit for the amount of lumens they put out. The older LEDs had NO noticeable flicker. Their newest ones DO have a slight flicker which is only noticeable if you know it's happening or if you scan your eyes over them quickly. It's not bad but it still annoys the hell out of me. My question is WHY do they do this? Is this a case of GE cheaping out? Is it a byproduct of increasing the lumens while decreasing the wattage? The bulbs say they are dimmable, does that have anything to do with it? I have older Reveal LEDs which say dimmable but don't flicker. Why, why, why? It gives me a headache sometimes but I hate the soft white LEDs and the warm white LEDs so I feel stuck with these. The color output really is amazing. It makes colors (paint and such) really look nice and whites really look great. I've tried every other brand and none of them look right.
I've found that different lighting takes 1-2 weeks before I no longer notice the colors and shadows looking weird. Maybe you should try leaving different bulbs installed for awhile and see if you get used to it? Might free you from your reveal prison. Perhaps you're just used to reveal color temperatures so everything else looks bad? My guess is the driver circuitry has changed (certainly because its cheaper) and now there's visible flickering. If the driver has changed there could be small differences in efficiency. I wouldn't put up with flicker from expensive bulbs they would be going back to the store. I've always bought $1 warm white led bulbs from lowes (the whole house is full of them) and never had problems except for poor heat sinking eventually causing driver failure. This is a design issue, LED screw in bulbs have terrible heat sinking in general. After 2 or 3 years I put another $1 bulb in and tear the broken one apart for the LED board. Most people actually prefer some (or a lot) of yellow in their indoor lighting because its similar to tungsten bulb colors and is good for relaxing
BigClive, are you planning on rebuilding that lighting fixture ? id be interested to see what you would do to that light to have it back up and running safely
A very proffessional breakdown , although it would have been more relevant to know if it can run 5 hrs without blowing , and if the power is reduced as it gets hotter
Hey Big Clive! I love your videos and sometimes will go to sleep watching your videos - your voice is so soothing lol. I was wondering if you, or your subscribers, could recommend a book or website were I could learn the basics and advance from there? I finding that I really like this type of stuff and really want to dig in to the knowledge. Any recommendations of yours would be so much appreciated! Keep up the good work big guy!
All the information you need is online. If you Google any questions you have you'll usually find answers. There are also quite a few electronic hobby sites online that are worth hunting down and seeing which one you click with.
I've got one of these modules (not the dodgy housing). Mounted it on an old PC heatsink.. its been running happily for 6 months now... But it does get quite hot without the heatsink... And the main input pads are flimsy I'll have a check to see if I can see any chip ids on my one.
Do they give an expected lifespan for its illumination qualities like they do with the regular bulbs? I can’t imagine that running so hot it’ll be very long. 3 deduce’s in this video side tracked me a bit and I started thinking about Arthur Conan Doyle.
Just Google search for a random number such as '16744821674482167448216744821674482' & you get nothing. Having said that if Google's bots scan this page, the above random number should return just this video in it's search. Another way is just search using a single asterisk.
I had high hopes for this floodlight being somewhat decent inside, such a shame the heatsinking and lack of waterproofing let it down (especially when the box shows such a thing)
Where did you get that cool little line power connection block? I could definitely use one of those on my bench as opposed to the "Death Cord" I have that is just a line cord with a line mounted switch and insulated alligator clips.
Love your videos, particularly the ones where you uncover fake and dangerous unit. I bought the new Riddex Quad pest repeller which turned out to have the same circuit as the one you tested a while ago, a complete waste of money.
Well, unless you're having some sort of emissivity issue due to the conformal coating (the goop) going on there, if the mosfet/transistor is 108 centrigrade on the outside, it's going to be as near as makes no difference at critical junction temperature (125 centigrade) on the inside, which should see it fail fairly rapidly. At my college they (probably still have) had a little chamber to stress test circuits for heat. As one of our lab excercises was to design and execute an experiment with regard to reliability, I built a little 4520 counter circuit which I then intended to run at critical junction temperature, to see how long it lasted. I wrote a little assembler program for a Z80 computer board, which would clock each of the counters in turn and checked whether it had in fact increased by 1 or rolled over to 0 if the previous value was F. If an error occured, it would energize a relay which would, in turn, stop a timer. In theory, according to the most widely used model for a single semiconductor junction, it should have lasted about 72 days, but, probably because there a quite a lot of junctions in this IC and it's not really expected to run much above ambient (so the non-military ones don't undergo testing for it), it lasted only 4 days. In case anyone's wondering, there's an about x2 factor increase in lifetime expectancy for every 5 centigrade below critical junction temperature, so at a typical 45 degrees CPU temperature, you're looking at about 60 years for every individual junction (and there is prototype testing in this regard, although the production process is trusted afterwards).
You're correct about junction temperature failure times. A semiconductor manufacturer in Europe (sorry, no specifics here, they would lose contracts) made devices which the customer specified were to be 100% QA tested for about a year at a specified minimum temperature, and under specific electrical conditions, before they were acceptable. Temperature had to be controlled very precisely, to ensure a 100% pass rate. One failure meant rejection of a whole large batch of devices, so a 5 degree Celsius over-temperature could have cost a lot of money. And a 5 degree under-temperature would have entailed a re-run of the whole test, which was also expensive and could have resulted in the cancellation of the contract. The customer was happy to pay around 20 times the retail price for something with ostensibly the same capabilities, which had not been life-tested so thoroughly. Why? The operational design life of the devices was quite short, but lives depended upon total semiconductor reliability during that time. The customer needed that guarantee of reliability. No device ever failed during use, and lives were indeed saved. Knowledge of the behaviour of Silicon junctions at temperature was essential, even critical, during design and manufacture.
I have gone through approximately 5 of those led chips your comparing and they all failed after about 2 weeks of continuous use. When they fail they go very dim and/or nearly half the cob blacks out completely. I even had a grow led fail and i used such an efficient cooling system that it never got warm. Failed after 2 wks.
get one of those led light bars meant for vehicles and test it, I bought one of ebay that was advertised as 180W but when testing it with a 240W capable PSU I only got a draw of aprox. 30W. Would be interesting to get a deeper look on one of those.
@bigclivedotcom what light would you recommend for an indoor, home, floodlight? I just want shit-tonne of warm white light, like a sun without sun. Preferably chinese and led-based. Thank you so much for your videos!
I wonder the rating on that bridge rectifier...just found the same pkg in one of my 3D printer gear, only have an smd code. Clive d'y'think this 50W performs as well in 60c 110VAC N.A. I shall see if there are any avail in that option. Perhaps one might determine power factor without the HOPI ? Best, C
STM18 seems to indicate this is part of the STM18xxxx voltage controllers manufactured by ST Microelectronics in Switzerland, a well known manufacturer. I would suspect that the "C" designates a custom part made for a specific customer and that the rest of the part number designates something like the intended voltage and the design version.
I bought 6 of the 'Driverless' LED floodlights, 4x10w and 2x50w, all of the 10w ones failed within 20 hours of use and one of the 50w ones after about 60 hours
I know this has nothing with this video, but maybe you can answer a question. I have some of these power banks that hold one 18650 battery. Can you explain why if I charge one of them up, 2 days later (without using it), it does not stay fully charged? Does the little circuit that has the LEDs on it discharge it even if it has not had a load on it?
As a layman but big enthusiast of this topic, I would like to understand the electromagnetic properties that an inductor "induces" on electricity passing thru it so it is capable of setting off phase the voltage in relation to the current. Any hint?
The second type of driverless LED you show (that you did a video on earlier this year), I had two of three 30-watt ones fail on me; I bought them from same eBay seller. They ran about 30 minutes a day for 6 months, before failing (both failed within a week of one another). The actual cob LED is what failed; I could see several dead diodes within same row. I don't think it is heat related, they were mounted in recycled flood light cases with expensive compound and everything stayed under 70 Celsius when measured. My guess is they have same issue as first generation/cheap cobs where they're poorly matched. I've got one original still working and two new replacements...will see if I get anymore failures. If I do, I'll probably buy 20 watt version.
I love the fact it has a MOV to protect it against voltage spikes, but I didn't see what protection they had if it shorts out. Are the traces thin enough to act like a fuse?
I don't think that will help if the MOV shorts. I believe there's been a few fires started by power strips where the MOV overheated before they figured out they need to add thermal fuses touching the MOV.
i've used a few of those 50w integrated driver leds. put 3 of them to the back of an old 2x4 lay-in fixture. one of the 3 went dark after a while. not completely dark just real dim. whats the life on those things?
They're still selling these floodlights on aliexpress and other platforms. I wonder if zip tying a fan on the back would help to cool the chip sufficiently.
Clive, whats your opinion of this style of aluminium core PCB with LEDs on it? I've got some 100w ones from china and they light evenly! I've linked to them on patreon if you want to take a look.
As much as presentation software and screen captures are great, I just love the fact that you print out the macro board shots and thermal stuff. It works so well.
Agreed! I picked up that technique after seeing it here.
While I quite enjoyed this video, I feel it would have been better if you had used a bigger calculator.
dead hahaha
That's what I was thinking too 😂
Hey Clive, thanks again for another informative video. I have been playing with those other kind of 'driverless' chips. The ones I purchased were from 'MB Lighting' and were pretty cheap. I tried the 20,50 and 120w versions which were $2 a piece for the 20 and 50w and $9 for the 120w. I liked the look of the 120w as it's surface area to power ratio was the best (thinking of heat transfer, not really sure how important that is in relation to other things) I have not had any issues with the smaller chips although those are not in use as much as the 120w chips which are on almost 24/7. Out of the 12 I have used, 10 have failed, the majority of them would still be emitting light from some of the leds but you could just barely see they were on. I was extremely careful in fitting them into the housings I had (which were easily suitable for power rating) however I cannot rule out user error. I noticed that my second order of the same drivers showed that they had redesigned the chips since my first purchase, the size of all were noticeably larger (the plate at least). Of the second order, I have had two failures. I feel some of the failures could have been from the wire connections to the chip itself. With such a large 'heatsink' as such, it does make for a challenging solder job. A lot of these new chips have been repopulating Casings from older 'failed' floodlights that I have accumulated from my adventures in the quest for the perfect LED. These new chips look great, The warm white is perfect color temp for my liking. And the price, well suspiciously cheap, but they look as good as any other I've seen. My greatest find over the years was a 100w (75w it turned out) Flood on amazon for $26. Now the housing was junk, too flimsy and you could tell the driver didn't stand a chance, so I swapped it out for a better casing before I installed. This was my first 'driverless' device I had come across, I believe it was a 16x8 grid of 0.5w Led's with 110v plumbed direct into it. How was a great find you may ask? After the initial reworking, The light itself has been amazing, Has been on for 22 hours each day at least, color temp is perfect, and looks as good now as it did on day 1. The biggest down side though has been the costly pursuit to find more of the same, and that where most of my housings have come from!
It's grounded, i'm shocked!
James Leinen if it wasnt't grounded you would be REALLY shocked ;-)
If it was ungrounded, it'll have the _potential_ to _shock_ you, which of course will _Hertz_ very much!
Ohm my that's bad...
Only if you were Amp-scent minded.
"ooo, its Earthed! that's nice"
Congrats on 300k subscribers! 🎉
Clive deserves a million! (Especially when you compare this channel's content with the "RUclips average" . )
You're so cool man, thank yuo for taking time to share your skills with us Clive.
most youtubers let the success go to their head and they stop doing what they started out doing and get lazy, the money goes to their head too, clive makes the exact same video now as he did way back, thats how i like it
Very good strip downs of so many things and replies to things you need advice about. Really clever guy well done 👍 thanks for your help
I'd be nice if you could do a video on a proper floodlight you can buy in Europe (like the Osram Ledvance or Geyer floods that are sold at quite attractive prices). The comparison could be enlightening.
ferongr A chinese company bought that part of Osram.
From a cursory outside look at the electrical supply store, there is really no comparison between the quality of an Osram light and a no-name Chinese light. Even normally imported (not gray import) Chinese lights are really bad in many ways and QC tends to be terrible.
Or at least illuminating.
ferongr Will Clive ever see the light?😁
+Frank Gormans Yup - Osram spit off the Ledvance lighting division in 2016 and this year it was bought by a Chinese consortium.
Good to know others are fascinated by lighting tech as much as me. Uneven cooling of the led array and poor cable handling is not a good sign. Keep up the good work Clive. 👍
7:34 OwO what's this?
18:30 This is all covered in, oh ... and, you know, uhh... Very goopy.
19:00 This is making a huge mess.
Oh my.
Google used to bring up absolutely nothing. Now it brings up _one_ result. Your video. :)
There is something quite satisfying about watching you take electronics to pieces.
The squeezing of the cable is obviously intended as a strain relief so the cable isn't pulled out. That aluminium plate was probably chosen to match the thickness of the squeezed cable (if the cable wasn't squeezed in there, the plate could just be omitted). Unfortunately, they didn't include the cable sheath in the squeezed zone. Anyway, a flat product like that should use a side entry to not waste its own thinness.
“I should be cleaning off my fingers at some point... but let’s see if I stab myself first.. hmm oh well” haha... you’re so chill Clive.
I'm guessing their trusting the silicon potting and conformal coat to keep moisture out, even if the seal around the cord or lens gasket isn't quite up to par. They put stuff on there to keep it from flowing out before setting, so it stays thick which may actually work. Thermal paste seems to be a QC thing, somebody being lazy there. The strain relief or apparent lack of it seems to be the worse problem. So with better QC and better cord mounting, it might not be that terrible.
Those springy terminals on the HOPI meter seem like they're more trouble than they're worth. Why not remove it?
Because as soon as I do I'll find I need them.
David Robertson
How about relocating them to another location on the housing?
You could maybe fit a switch somewhere on the case, have the terminals switched off normally. One of those old-fashioned metal toggle switches would look nice.
greenaum that wouldn’t solve the issue of the terminals interfering with a UK plug being inserted
That feeling when you search for component id and only result is BigClive video :D
Typical, once you got your tool in there there was sticky white stuff everywhere, well, glad to see you’re a gentleman and don’t wipe it on the curtains afterwards, Clive, but have you never heard of a wet wipe?
I think it's more a case that the screws are low emissivity rather than being so much cooler. Try colouring them in with a black marker pen.
Andrew Cooke Imho, it's that the screws have direct large surface area contact with heatsink compared to much smaller (possibly indirect) contact area with the led driver heatsink.
Hi Matt, Infrared cameras are fantastic diagnostic tools, but they have some severe limitations that it's important to understand, it's complicated, but in short, don't try and measure shiney stuff.. have a quick tread of this : support.fluke.com/find-sales/Download/Asset/2563251_6251_ENG_B_W.PDF?trck=emissivityexplanation
Andrew Cooke The stefan-Boltsman law isn't that complicated and I own and regularly operate a Fluke Ti110 on many materials including the 'shiny stuff'
I still stick to my opinion from experience in the industry.
Thanks for your input though.
that's cool, my time designing them seems to have been wasted.
Andrew Cooke that's for you to decide but my unit has saved me hundreds of hours.
Your point is a factor no doubt about that but mine is also. The screws are most of the way into a colder heatsink and the head down onto a white and somewhat insulating layer over the led plate including under the screw head. Also Note the halo of cooler area just around the screw head in the image where it is just partially cooling the led plate from the screw head. If the screws weren't cooler then there wouldn't be a hotspot on the led array in the bottom right hand corner where a screw is absent as mentioned in the video.
You seem.... extra caffeinated today lol. Love it!
What would the thermal performance be like if the LED-board was properly mounted to the case?
error079 This is very poorly assembled. It would make a huge difference if mounted correctly. But the surfaces are no way flat enough and the cable input is a no go. If you put on too much thermal grease it's a thermal isolator compared to aluminum.
That transistor needs an additional heatsink, too.
Frank Gormanns The transistor is heatsinked in the same way the LEDs are. Through the board it's soldered to.
Just add another properly heatsync greased layer of aluminum between the light and the assembly to allow room for the wires and profit.
I would use that after replacing the heatsink compound and making sure that cables aren't punched through. Seems quite a reasonable design anyway
I guess one way to get around the squished wires issue, use an externally-mounted driver in a junction box near the light and fit a "traditional" DC LED module with a much smaller footprint, yes it adds a requirement for additional wiring, but, saves squashing mains powered wires to a potential short... :)
And were you filming this at around 8pm-ish? The mains frequency went down a fair bit on the Hopi, indicating the end of a dreary TV soap and everyone powering up their kettles... :P
twocvbloke ,
¿¡ !? ¡¿ That much measurable frequency drift ¿¡ !?
Gonna have to look for that during an analagous event here in the US !
Ponderous... wonder how "smart power grids" technology is supposed to mitigate said affectations...
The voltage dropped by a couple of volts, but only when that 52 Watt load was added. The total resistance (impedance) of the cabling to that meter was probably around 0.075 Ohms, which would account for a ~2V drop, using the formula (V^2/W) = R. So ~4/52 = 0.077 Ohms.
@john howard, sounds like a trump quote!
I bought four COB chips after your first tear down video, 2 x 30 W and 2 x 50 W.
I suspect the throttling does not work that well because the first three chips died rather quickly, approximately 3 weeks after install in different setups.
1. 50 W COB, ran on an old GPU-cooler with heat pipes and a 12 V fan in the tower. When it was cold it hit 53 W, then after a couple of minutes it throttled between 44 and 46 W. It managed to reach steady 70-100 C and all of a sudden only haft the LEDS worked faintly. Every second row worked as if it rode half a wave.
2. 30 W COB, put on another cooler only with ambient temperature to keep it cool. The surface area was a bit larger. But the same problem. It started at 34 W when cold then throttled back to 29-30. It also died three weeks later with half the LEDs flickering.
3. 50 W COB, put on a large sheet of aluminium with a CPU-cooler behind. Experimenting with the size of the surface area. No better luck here.
4. 30 W COB. This one is still running! I used the same CPU-cooler as in 3 but with a 12 V fan attached. This runs at 23-35 C, but without the fan it runs up to 75-105 C and throttles to 27 W. With the fan it does 33 W.
I have a fifth chip from another supplier at 50 W waiting to be put on the CPU-cooler when 4 dies.
The throttling therefore have only a marginal effect, because it cannot lower the power enough. I also found out you do not need that hot of an soldering iron to attach the cables if you only use old, 1987, Fluoridic based paste. It does however make you dizzy if you do not evacuate the air fast enough. Perhaps that is why it is no longer sold (lead based solder...)
Enough surface area and fans are a must with these. One car battery and 12 V fans with a charger attached and an array with these chips. Or find a large industrial fan at 480 V and try to convince the wife it is for science.
www.processenvironments.com/cartPics/4cb8c6797c051IRO_782_ARR_1.jpg
Clive. would you be able to do some more usb charger tear downs ? its something we take for granted everyday and would be nice to see different good/bad designs :)
If you like teardowns check out Ing. Max Koschuh's channel: ruclips.net/channel/UCLsp71mRNqoRSi6YC0jLmFA :-)
many ppl inlcuding clive had covered cheap usb charger
Have you though about taking the HOPI meter apart and turning the plug 90 degrees? So you would not have to use the death dapter
Ha. Genius
Reduces the chance of danger, no fun. he would never consider it.
floating earth damn
Or just remove the speaker terminals.
I was thinking the same thing, but for an entirely different reason. I'm surprised he hasn't taken the HOPI apart and reworked the circuitry to stablize the flickery display for the camera.
It's earthed! Will wonders never cease?
However, after having watched some of John Ward's videos about Chinese mains-powered stuff, the next question I have after "Is it earthed?" is "How good is the mains cable?" John's videos have exposed some really crappy Chinese mains cables: seriously under-rated for the task, crappy insulation with flame-enhancement properties, and conductors that are copper-plated to disguise the fact that they're made of damp string. Then again, you say the cable claims to be decent stuff rated for high temp (but it could be lying).
So my big worry about that unit is the placement of the cable gland. It's in a *really* bad place where water will run onto it rather than away from it, so unless it makes a perfect seal you're going to get serious water ingress.
This cable have CCC Certification (Chinese standard) markings which should mean is tested in proper third party certified lab and with equal/higher standards than CE. But anyway CE is good standard if there was not ability that you do CE marking self-certification (you do not need to any third party testing and certification, you basically just need to give statement that product is compliant with CE standards).
European CE is not the same as Chinese CE, which means China Export; not to be sold in China. The two symbols look alike, when viewed quickly, but they are subtly different, as can be verified by a quick online search. And there is nothing to stop someone who is dishonest from printing the Chinese CE symbol on something which is substandard, or even lethal, and then making sure that it is not sold in China.
RWBHere There is no such thing as Chinese CE (China Export), it is just incorrect application of CE by some Chinese manufacturers, to be able to sell merchandise in China it must comply with CCC standard markings (so they do not need to have some marking that will say it is for export only, if it do not have CCC markings it can't be sold in China) and must be verify by independent laboratory (only certified testing centers can verify that product is compliant with CCC standard), contrary to CE (or how you call it European CE, but only that one exist) which is self-certify for most of products (all electronic products do not need to be tested and certified by a third party certification body you just need to fill EC declaration of conformity), only little product require to be certified by " Notified Body".
Is the cable gland an issue? The insulation continues all the way under the conformal coating. Water will be getting in via the louvres anyway.
en.wikipedia.org/wiki/CE_marking#China_Export
Excellent video, as always. Could you maybe do a test with the temperatures when you've made sure the thermal paste is actually gripping the whole board, and with that one corner screwed also? Like re-route the cables so that that is possible. Also if you could test it with one of the "driverless" leds you have, seeing if there's a difference and how it works.
I do like the idea of smaller and "neater" light. I need to get one of those so that it'll help me fight against the darkness and the blues it brings.
ilaril It will cause headache because of the flicker.
Clive, the squishing of the wires allows a clamping effect to keep the mains cable from falling out. What was the lumen output of this thing? I'm recently seeing 50w LED floods marked (true 35w), so they're starting to become a bit more honest.
The screws look colder on the LED because of them being reflective, put some black tape on them and then recheck their temp, or paint them black, then you will get the correct tmeperature.
Nice sound Clive.
I am a simple I see big Clive post I see !
Unit prominently claims an IP67 rating on the back. IP67 means:
Ingress of water in harmful quantity shall not be possible when the enclosure is immersed in water under defined conditions of pressure and time (up to 1 m of submersion).
So stick it in 3 feet of water for 30 minutes and no water will leak in that gland. Maybe.
I also noticed the prominently IP67 ( en.wikipedia.org/wiki/IP_Code ) and hoped BigClive showed this submersion in the video.
How much space do you have to store all the things you tear down? You must have quite the collection by now.
That's why he lives on The Isle...Puts 'em in a cardboard box, writes "love from Ireland" on the outside, sets them afloat when there's not a northerly breeze, and gives them a "Viking Funeral". With any luck, lands on a beach near Liverpool, starts a wildfire, burns out Liverpool, and the parts make it look like the Chinese had a sub burst into flames on the high seas. Everyone is a winner...Clive gets rid of the leftovers, the UK intelligence services get to spend more money looking for the wrong people, China gets blamed, and Liverpool gets to rebuild.
@@carlwilson4425 genius.
I had one of those integrated cobs fail recently and thought about sending it to you.... It first started flickering after running for around 8 months - prob 10-12 hrs/day. Finally it just died entirely. Was on a very large AL heat sync - IRC is was a 30w 120v version, but long rather than square.
What a shame, a potentially very useful light wrecked by that cramped cable - a disaster waiting to happen! Great video as always.
I cant see that transistor lasting very long at that temperature either...
Great video as always and a BIG congratulation to 300k subscribers Clive.
Besides Grilling Food outside. Another thing us guys are great at. Sizing up a screw or nut and picking the correct size screwdriver, socket or wrench. :)
Would those louvres be a safety issue in the rain, or is the internal waterproofing sufficient to effectively just make it water cooled?
awmperry the vent holes don't go through to the interior, just from front to back of plate.
Ah, so it's a single plate? Hard to see on a small screen, looked like there could have been a cavity. Cheers!
Your videos are very reveling. Thanks for your time and information sharing.
Grow lights have been available with that driver arrangement have been available on eBay for at least a year or so. Haven't seen a white LED like that until now though.
STM I wonder if that's ST Micro? If that's the case then it could be a custom PIC12 re programmed to contain all kinds of goodies.... perhaps pre-programmed failure so you have to buy another?
STM could be ST Microelectronics and 18C242 is a PIC. However, it's probably some bullshit Chinese part number to cover up the fact it's a clone of something else.
Deception? why relabel when it's cheaper/faster to just remove the number?
restcure When you have a customer who wants a chip that looks legit-ish but won't check it beyond the label. Or you have a supplier with a fake chip that looks legit-ish and they'll sell them cheap. Nobody asks too many questions.
17:30 Well...that seems like a cable fire just waiting to burn down your whole house, with best wishes from China.
I was planning on buying 2 of those exact same lights (may be different name/manufacturer, I don't know, but looks EXACTLY like the one I was eyeballing on Amazon) to serve as grow lights for indoor hydroponics...am I glad you picked this thing apart, sir. I own a house and a forest, and I would rather the damn forest burned down 10 times in a row than have a fire start in my house.
They could of pressed a depression into the metal to allow for cable entry (this can be done sit punches the hole, in a single operation, doing this would of increased the cable space, then they could of put on 4 mounting screws instead of 3.
I am not electronically inclined buy find your videos very interesting to watch.
I had one of those 50 watt modules fail after I left it on a 12 hour timer while away for 3 days.
It might have been a heat-sinking issue since the sink I used only contacted the back of the module in certain places. It also occurs to me that transients might damage that sort of module pretty easily, which is probably why the work light one had an MOV on it.
Brilliant Clive!
I'm wondering if anyone can answer a technical question about LED lightbulbs.
I'm in the US. All of the bulbs in my house are GE Reveal. Originally it was the Reveal incandescent, then the Reveal CFL's now the Reveal LED's. It's the only bulbs whose color I can stand, it filters out the yellow without being overly cool like the fucking god awful cool white daylight bulbs. The LED's aren't quite as good as the incandescent were but still very close to the same color.
ANYWAYS, I've noticed that with the LEDs GE has updated them a few times over the past few years. Now they're branded as GE HD Reveal. It looks like they've got the wattage down a bit for the amount of lumens they put out. The older LEDs had NO noticeable flicker. Their newest ones DO have a slight flicker which is only noticeable if you know it's happening or if you scan your eyes over them quickly. It's not bad but it still annoys the hell out of me.
My question is WHY do they do this? Is this a case of GE cheaping out? Is it a byproduct of increasing the lumens while decreasing the wattage? The bulbs say they are dimmable, does that have anything to do with it? I have older Reveal LEDs which say dimmable but don't flicker. Why, why, why? It gives me a headache sometimes but I hate the soft white LEDs and the warm white LEDs so I feel stuck with these. The color output really is amazing. It makes colors (paint and such) really look nice and whites really look great. I've tried every other brand and none of them look right.
I've found that different lighting takes 1-2 weeks before I no longer notice the colors and shadows looking weird. Maybe you should try leaving different bulbs installed for awhile and see if you get used to it? Might free you from your reveal prison. Perhaps you're just used to reveal color temperatures so everything else looks bad? My guess is the driver circuitry has changed (certainly because its cheaper) and now there's visible flickering. If the driver has changed there could be small differences in efficiency. I wouldn't put up with flicker from expensive bulbs they would be going back to the store. I've always bought $1 warm white led bulbs from lowes (the whole house is full of them) and never had problems except for poor heat sinking eventually causing driver failure. This is a design issue, LED screw in bulbs have terrible heat sinking in general. After 2 or 3 years I put another $1 bulb in and tear the broken one apart for the LED board. Most people actually prefer some (or a lot) of yellow in their indoor lighting because its similar to tungsten bulb colors and is good for relaxing
BigClive, are you planning on rebuilding that lighting fixture ? id be interested to see what you would do to that light to have it back up and running safely
Do you have to pay extra to dispose of all the stuff you test and break apart? You must have a big container...
Glad you survived the storm ...
A very proffessional breakdown , although it would have been more relevant to know if it can run 5 hrs without blowing , and if the power is reduced as it gets hotter
+oz93666 Power did not reduce significantly after a few hours.
love your old MK 13 amp plug... i use to collect them to use on everything that plugged in..
Hey Big Clive! I love your videos and sometimes will go to sleep watching your videos - your voice is so soothing lol. I was wondering if you, or your subscribers, could recommend a book or website were I could learn the basics and advance from there? I finding that I really like this type of stuff and really want to dig in to the knowledge. Any recommendations of yours would be so much appreciated! Keep up the good work big guy!
All the information you need is online. If you Google any questions you have you'll usually find answers. There are also quite a few electronic hobby sites online that are worth hunting down and seeing which one you click with.
I've got one of these modules (not the dodgy housing). Mounted it on an old PC heatsink.. its been running happily for 6 months now... But it does get quite hot without the heatsink... And the main input pads are flimsy I'll have a check to see if I can see any chip ids on my one.
Do they give an expected lifespan for its illumination qualities like they do with the regular bulbs? I can’t imagine that running so hot it’ll be very long.
3 deduce’s in this video side tracked me a bit and I started thinking about Arthur Conan Doyle.
4:46 that cheerful * missed * on wild guess 😂
I have had failures of the driverless cob but it always seem to be the "Ming & Ben" brand others have not failed.
you ar the best youtuber ever its so true
about 8:08 i have been trying to find one of those web pages for a long time now. care to share one you find or how you find one?
Just Google search for a random number such as '16744821674482167448216744821674482' & you get nothing. Having said that if Google's bots scan this page, the above random number should return just this video in it's search. Another way is just search using a single asterisk.
Are those vents on the rear? Wouldn't it get rather wet?
I had high hopes for this floodlight being somewhat decent inside, such a shame the heatsinking and lack of waterproofing let it down (especially when the box shows such a thing)
How much power could you get out of a bag of spuds Clive ?
Is U2 an STMicroelectronics a/d converter? It's a really familiar number. Do they use a/d to make LED dimmers?
2:40 - does the black deathdapter have a sleeved earth pin?
Yes it does. But it does actually make contact in the Hopi unit's dodgy uni-socket.
The ST Micro 8C24 is a thermal sensor destined for computer usage, used on Intel chipsets. Does that sound down the right path?
Where did you get that cool little line power connection block? I could definitely use one of those on my bench as opposed to the "Death Cord" I have that is just a line cord with a line mounted switch and insulated alligator clips.
Search my videos for quick test. I took it apart in one of them.
Almost looks like a grill when it's laying flat. Could you use two of these as a grill aka secondary life saving feature. Haha
Love your videos, particularly the ones where you uncover fake and dangerous unit. I bought the new Riddex Quad pest repeller which turned out to have the same circuit as the one you tested a while ago, a complete waste of money.
Just wondering, did you ever open the Hopi ? would it be possible to rotate the 3 pin socket 90 deg left or right ?.
Well, unless you're having some sort of emissivity issue due to the conformal coating (the goop) going on there, if the mosfet/transistor is 108 centrigrade on the outside, it's going to be as near as makes no difference at critical junction temperature (125 centigrade) on the inside, which should see it fail fairly rapidly.
At my college they (probably still have) had a little chamber to stress test circuits for heat. As one of our lab excercises was to design and execute an experiment with regard to reliability, I built a little 4520 counter circuit which I then intended to run at critical junction temperature, to see how long it lasted. I wrote a little assembler program for a Z80 computer board, which would clock each of the counters in turn and checked whether it had in fact increased by 1 or rolled over to 0 if the previous value was F. If an error occured, it would energize a relay which would, in turn, stop a timer.
In theory, according to the most widely used model for a single semiconductor junction, it should have lasted about 72 days, but, probably because there a quite a lot of junctions in this IC and it's not really expected to run much above ambient (so the non-military ones don't undergo testing for it), it lasted only 4 days.
In case anyone's wondering, there's an about x2 factor increase in lifetime expectancy for every 5 centigrade below critical junction temperature, so at a typical 45 degrees CPU temperature, you're looking at about 60 years for every individual junction (and there is prototype testing in this regard, although the production process is trusted afterwards).
You're correct about junction temperature failure times. A semiconductor manufacturer in Europe (sorry, no specifics here, they would lose contracts) made devices which the customer specified were to be 100% QA tested for about a year at a specified minimum temperature, and under specific electrical conditions, before they were acceptable. Temperature had to be controlled very precisely, to ensure a 100% pass rate. One failure meant rejection of a whole large batch of devices, so a 5 degree Celsius over-temperature could have cost a lot of money. And a 5 degree under-temperature would have entailed a re-run of the whole test, which was also expensive and could have resulted in the cancellation of the contract. The customer was happy to pay around 20 times the retail price for something with ostensibly the same capabilities, which had not been life-tested so thoroughly. Why? The operational design life of the devices was quite short, but lives depended upon total semiconductor reliability during that time. The customer needed that guarantee of reliability. No device ever failed during use, and lives were indeed saved. Knowledge of the behaviour of Silicon junctions at temperature was essential, even critical, during design and manufacture.
I have gone through approximately 5 of those led chips your comparing and they all failed after about 2 weeks of continuous use. When they fail they go very dim and/or nearly half the cob blacks out completely. I even had a grow led fail and i used such an efficient cooling system that it never got warm. Failed after 2 wks.
Dose it has certificate ETL, UL listed?
that L shaped aluminium spacer should just have been made thicker to make room for the cable
thicker and NOTCHED for the cable to pass through
"You're going to need a bigger calculator". (Jaws; 1975).
get one of those led light bars meant for vehicles and test it, I bought one of ebay that was advertised as 180W but when testing it with a 240W capable PSU I only got a draw of aprox. 30W. Would be interesting to get a deeper look on one of those.
Can we have a link to where you got it please.
@bigclivedotcom what light would you recommend for an indoor, home, floodlight? I just want shit-tonne of warm white light, like a sun without sun. Preferably chinese and led-based. Thank you so much for your videos!
doesn't that code at the top read DOB -driver on board and AC220v mains voltage and V.1 means version 1 that says it all ya think?
would you have refrained from taking it to bits in a non reversible manner if it was a warm white flood light ?
No. It was bought to be taken apart and wasn't up to expectations. Technically speaking I could have fitted it with a warm white driverless array.
Nice video again. It would be nice to know the frequency that it is flickering at. Could you measure this and mention it in your next led vids?
It's flickering at mains frequency. 50Hz here so actually 100Hz when rectified.
I wonder the rating on that bridge rectifier...just found the same pkg in one of my 3D printer gear, only have an smd code. Clive d'y'think this 50W performs as well in 60c 110VAC N.A. I shall see if there are any avail in that option. Perhaps one might determine power factor without the HOPI ? Best, C
where did the 50w LED light come form ?
do you have link for these bigclivedotcom?
More Big Clive moth moments:-)
STM18 seems to indicate this is part of the STM18xxxx voltage controllers manufactured by ST Microelectronics in Switzerland, a well known manufacturer. I would suspect that the "C" designates a custom part made for a specific customer and that the rest of the part number designates something like the intended voltage and the design version.
I bought 6 of the 'Driverless' LED floodlights, 4x10w and 2x50w, all of the 10w ones failed within 20 hours of use and one of the 50w ones after about 60 hours
I know this has nothing with this video, but maybe you can answer a question. I have some of these power banks that hold one 18650 battery. Can you explain why if I charge one of them up, 2 days later (without using it), it does not stay fully charged? Does the little circuit that has the LEDs on it discharge it even if it has not had a load on it?
There's a small standby current, but it shouldn't discharge the cell too quickly.
As a layman but big enthusiast of this topic, I would like to understand the electromagnetic properties that an inductor "induces" on electricity passing thru it so it is capable of setting off phase the voltage in relation to the current. Any hint?
The second type of driverless LED you show (that you did a video on earlier this year), I had two of three 30-watt ones fail on me; I bought them from same eBay seller. They ran about 30 minutes a day for 6 months, before failing (both failed within a week of one another). The actual cob LED is what failed; I could see several dead diodes within same row. I don't think it is heat related, they were mounted in recycled flood light cases with expensive compound and everything stayed under 70 Celsius when measured. My guess is they have same issue as first generation/cheap cobs where they're poorly matched. I've got one original still working and two new replacements...will see if I get anymore failures. If I do, I'll probably buy 20 watt version.
I love the fact it has a MOV to protect it against voltage spikes, but I didn't see what protection they had if it shorts out. Are the traces thin enough to act like a fuse?
From destructive LED tests I found that the LED bond wires blow like fuses and the arc is contained by the phosphor loaded gel.
I don't think that will help if the MOV shorts. I believe there's been a few fires started by power strips where the MOV overheated before they figured out they need to add thermal fuses touching the MOV.
Ah, I misread your post. The track or supply fuse would fail.
Против Глобал, that's what I don't like about capacitive dropper - because spikes are high frequency, those go right though.
7:40 that orange circular object looks like a pptcd (polymeric positive temperature coefficient device) or... well, a resettable fuse
Thanks Clive, I enjoy your videos thoroughly. They actually help me learn alot about what to stay away from and how different circuits work!
What do you do with all those floodlights you buy?
Xmas tree ala Tim Allen? "the power went out at the airport, but we can see the runway thanks to this guys xmas lights!"
cool but is it safe outside in the rain ? or is foxy going to fry again
i've used a few of those 50w integrated driver leds. put 3 of them to the back of an old 2x4 lay-in fixture. one of the 3 went dark after a while. not completely dark just real dim. whats the life on those things?
not sure, how long did the one last?
They're still selling these floodlights on aliexpress and other platforms. I wonder if zip tying a fan on the back would help to cool the chip sufficiently.
For indoor uses it would help. You could carefully drill the case for fan support pillars.
So much for the IP67 rating on the back. BTW. Heat transfer compound stops bleeding quite well.
7:42 "I'm gonna have to explode this a little bit"
Clive, whats your opinion of this style of aluminium core PCB with LEDs on it? I've got some 100w ones from china and they light evenly! I've linked to them on patreon if you want to take a look.
It was looking surprisingly good until 16:24 😄
Should have drawn circuit diagram like rest videos.
I guess what driver are they using there can we find the datasheet.