When I was a teen I learned about capacitors. I built a beautifully constructed 300VDC tube power supply. Then, I turned it off and reached in to make some change. It shocked me hard and I pulled my hand out so quickly that my wristwatch stayed stationary and dropped to the concrete floor where it broke. Very instructive it was.
Ah yes, capacitors. Those little innocent looking cans and discs, sometimes packs of foil and paper, thin plastic film, or maybe a bit of foil and some oily stuff inside. Nothing dangerous until you put a bit of voltage on them. I learned about these little critters myself at around the same age but never got attacked until much later when I was a newbie tech. My job was to test power supplies of various sizes. Some were analog power supplies with big regulators and some big cannister-sized capacitors on them. I was always a bit on edge when I powered those up and one day, I had one explode on me because someone in manufacturing put the big capacitors in backwards. The explosion was so intense that it sent the capacitors up into the drop ceiling above me. It was a good thing I always ducked when I powered those up! The last encounter was with some switchers. These switching power supplies, I think 65W units used in some "portable computers", with portable being a relative term for machines weighing around 22 lbs. 10 kg. There was a batch of these small switching power supplies that had leaky caps that needed replacing. Field service collected the units and brought boxes of them down to the test department for repairs. The units were to be returned to field service for replacement in customer returned units, or sent to the outside depot houses where they came from for field replacement. I reached into the box of power supplies and got zapped instantly! The units, while sitting in the boxes for ages, still held quite a charge. After that, I got out a screw driver and would rub that across the leads sticking out of the boards before I even put my hand near them. The zap was enough to leave my arm achy and hand sore for the rest of the day.
i think i learned this from a tv originally... not to not wear a wrist watch but how to smack something in reflexive pain to the shock and accidentally break something, i forget what i knocked over... haha
Number of things about this don't even make sense, why 300vdc for tube technology, how if your hand was in something did anything come off and drop to the ground...
I've learned so much from your videos. I'm never going to use them in a technical sense, but they are hella helpful to look out for scams and poorly constructed equipment.
You mentioned the resistor to counteract the 'ghost glow' - the other approach is to add a resistor between the switch lines (1M or 2M maybe) to increase the effect for a useful night light. The switches then choose between 'full' and 'night light'.
He just inspired me, with this video, to stop halfway through and clean my electronics desk down through the layers to real desk. Knock on wood. I also mixed up some liquid flux to put in a syringe.
Possibly the other thing they were going for was a power factor requirement? The smoothing capacitor drops the power factor significantly, and using circuitry that is both flicker free and high power factor would be quite a bit more expensive.
I like the idea of re-grading a downgraded product, but the fact that the vendor is probably charging the same price or more compared to what it was before it got "cost-optimized" is irritating enough that I would think twice before buying the product. Of course it is happening constantly without my awareness; I just happen to know better in this case.
The wiring on those regulators in current regulator mode, looks like a LM317 wired as a current regulator. Interesting product. Thanks for the teardown!
Anytime you add a capacitor I smile, because I always want to add more capacitors to projects. I never been a fan of how many resistors companies add to products. I prefer 1 or 2 resistors over their dozen of resistors.
Adding caps is fun - until one day I added too many µFarads to the selenium rectifier of one of my valve radios... Selenium is one period lower than sulfur, but worse. The smell lingered for weeks
Main problem with the panels I bought is absolutely terrible CRI of about 60. They are bright but the room feels dull. Would not recommend wasting time and effort on these.
Kreosan was just scraping down one of the paired resistors for longer life with a meter hooked-up to monitor the drop till it was what he wanted. He was doing it to drop the input when one led burned and he shorted it with black silicone (yes lamp black and charcoal conducts electricity, I've elecroformed with both when I had problems with "pure" graphite that was more like silicon dioxide powder with graphite but visually you could not tell) to relight the "string".
Since those are linear regulator chips, I would put the capacitor in parallel with the LEDs instead of directly across rectifier output. It would still greatly reduce flickering without destroying the power factor, subjecting the capacitor to high crest current nor increasing the DC input voltage across the linear regulators beyond the initial power-on cycle.
Unfortunately that puts a lot of stress on the regulator at power up, as it sees a very high voltage differential as the capacitor can start at zero volts.
@@bigclivedotcom which is exactly what I meant by "initial cycle." 100mA at 350V crest voltage is only 35W instantaneous power for the sub-1ms it will last. As long as the regulator is designed to drop the full 350V, 35mJ per on-off cycle should be fine. Raising the DC voltage with an input capacitor increases energy dissipation in the chips by a much greater amount on a continuous basis.
Who would want lights that flicker? I wouldn't. I assumed it was because of your camera, but no, I guess not. Interesting video! Great improvement with the caps and resistors.
in normal day to day use, you will most likely never notice the flickering but there are those that are more sensitive to it, like i notice the backlighting of my keyboard flickering when im not looking directly at the keyboard, said flickering on my keyboard was even worse when i used it on a usb polling rate of 100hz, now at 250hz polling rate its far less noticeable flickering
Very interesting hacking experiments. These linear regulators seem fairly flexible, either using a capacitive dropper or just adjusting the value of the resistor. The smoothing capacitor is definitely necessary, I can usually tell when lights don't have one, even without the aid of a video camera.
Some Chinese teardown of a similar device suggests that the MLS3535A should be similar to the SM2082EGS. There is a Chinese datasheet available for this chip. The pinout matches, a lot of pins are marked as NC. There should be 0.6V present between pin 2 (Rext) and pin 1 (GND). The output current is specified to be 0.6V/Rext. The datasheet also allows smoothing at the input, and the chip is specified to work on DC as well. The seemingly strange track routing on the PCB with GND under the chip is also recommended in the datasheet, since there seems to be a thermal pad (or "IC pin") on the underside of the chip.
My take on the lack of an electrolytic capacitor is meant to reduce cost and prevent the explosive mode of failure that is caused by the capacitor being dried out due to the high temperatures produced by a largeish array of LEDs surrounding the cap.
I think that all of us who are Big Clive fans have managed to zap ourselves at one point in our past. Luckily, we're still alive! BTW, those old B/W CRT televisions were a bit of a killer if you tried to modify it to be an oscilloscope :(
I'd be inclined to buy better. Failing that, I'd improve the design with a smoothing capacitor and anti-ghosting resistor then reduce the wattage to 80% of the resultant value by changing both sense resistors. I've no desire to dim everything to night light levels nor make it outlast me.
I miss the smell of smoldering resistors. Electronics Lab always had a pall in the rafters and an odor of ozone and fried ceramic. EL was a recycled hangar with fans in the eave ridges, no real modern ventilation system. it was an Olde School. wow, that was a fuzzy olde memory.. thanks Clive.
Hey Clive, There’s a bit of a controversy brewing on the interwebs regarding UV led street lamps. I was wondering if you could perhaps do a little video to address what’s going on and why they create the purple light that is absolutely horrendous to drive under at night. I don’t know if you’re experiencing this issue across the pond, but I believe you probably have the knowledge to explain just what sort of failure is occurring, whether a coating issue, or failure of a particular color led, or whatever causes my windshield to momentarily flash a blinding purple glow as I drive home each evening. There is a lot of hyperbole out there about what I’m sure is simply dodgy engineering. I appeal to you, the beard of electronics knowledge for enlightenment. Thanks for the informative and entertaining videos.
Sounds like the phosphor coating of these LEDs got damaged in some fashion. The LEDs in those lamps normally put out light in the UV range. To turn that into nice white light, there is a coating on it that absorbs UV light and emits longer wavelength light. Kind of like how bleached clothing shines in dark light (UV). I'm guessing they screwed up the coating, or maybe didn't test if it could take the temperature cycles of an outdoors environment.
Sometimes blue chips can emit purple waves (spectrum), without being technical. More to the point they can be edging so slightly into purple rather blue. Also, it should be noted that car windshields, as you prefer, do have UV coatings or stuff baked into the glass it self so you don't turn into a tumbleweed or a baked potato. Crispy pork loin maybe... Either way, the glass if the refraction of light and the beam of light and the correct colour of light and the photon direction all hit correctly at the right junctions, you get glare. This purplish light you mention I see from LED headlights going down the road but because the angle of light is so that it doesn't refract accross the whole screen, I don't get blinded by that particular spectrum of light the led light is emitting. UV is a specific wavelength of light and while it does emit a purple colour, it's not the same as purple spectral light Note ozone generators and germ eradication devices that say they use UV light and they are just purple coloured LEDs. Fakery fake fakeness Pretty sure now I blabber that purple from LEDs is blue chip with a form of coating. I think the base answer is because there are so many factors that could reasonably say why you experiencing what you are it gives rise for the doomsayers to emit from the free to speak minds things that are not necessarily true while trying convince one that the egg most definitely came first...
I say all this however while not being aware of a campaign of rhetoric about the subject you speak. Sounds a bit intimidating, having the idea that your being cooked while driving down the road. Alas, your screens do have some protection, but I think it's only one, uvb or uva, some might have both, but it's a case of 'how fancy is your car, how fancy is your glass' kind of buck. I'm not an expert, so don't take my word for all of what I said but I believe what I said to be more or less correct.
Considering the voltagedrop over the leds as a constant. The power dissipation would not change due to using only one regulator because the current through each regulator does not change. The increase in power is caused by adding the smoothing capacitor because that increases the duty cycle.
I remember back to the last time we looked at a "driverless" LED module and experimented with fixing the flicker The capacitor works best after the drivers, that way they don't have to dissipate the extra voltage from the smoothed mains
Clive, I apologize in advance for this ... but ... I started watching and then you did the thing with the meter and it instantly came to me, "There's Clive, pluggin' things into the Hopi again." lol. Thanks for continuing to "take stuff to bits," Uncle Clive.
Excellent, informative, video explanation as always Clive. I have just fitted two of these cheapo Chinesium lights in place of ancient incandescent fittings. We shall see how long they last...
I have literally a dozen of 1000 lm LED lamps *per room* in my house, can't be bright enough for me. Watching these videos where you reduce the brightness or say "It is awfully bright" always hurts watching ;) (But I still love them)
This "product" is a bonus if it is inexpensive and can be transformed into an improved functioning lamp or an emergency light. Emergency lighting is a continuing sphere of interest for me, where I live, where tropical storm outages are a part of life.
....... That thumbnail brought back a nostalgic memory! 🤣 It has a made up word on it that is very similar to one a buddy, who I met by playing on his Quake II Deathmatch server, had come up with one day; his was "dildomagic". Though it deserves context... We would chat over ICQ, and I had found a plugin for it that would do Text-to-Speech, which being Win98 meant it was only barely better than Stephen Hawking's synthesizer.... and us being teens, meant juvenile fun. OH and it had an on screen avatar to read it, but that might've been part of Winderz, a classical Wizard (default) or a brown dog were choices. That's when he came up with this gem to have the wizard say, of which I'll never forget: _Westside fo' life, cube. Dildomagic, bonita ramsey_ _chunktastic, fo' digga di-gazzi madre._ Honestly, what made it so funny back then -aside from the fact that "dildo" was the only crude word needed- was the lack of inflection the TTS engine had, as it said it so monotonely in that classic robotic way, and the punctuation had more impact on the flow (greater pause than modern engines). Remembering how it hung off that first coma before _finally_ saying "cube" *still* gives me a good chuckle! ☺️ Anyways... Now to watch your video about hacking these "boobymeisters"! 😁
You are a gent Sir, I forgot to order a 2d LED replacement 2 pin for my bathroom light as fed up with it having to buy those stupid light bulbs for it and I had earmarked a LED panel replacement and your vid reminded me to buy it hehe
Let's not forget that there are extremely few vehicles that are not computerized. It is not unrealistic to assume that impacts on transportation systems may be regionally catastrophic. So not only will we be faced with months long blackouts, which will impact food storage, medical care delivery, and many other things ,but the logistics of responding to the emergency will be impacted as well.
I found a Chinese site with a similar teardown and a comparison of the MLS3535A chip with what appears to be pin compatible SM2082EGS, and a circuit diagram which had the smoothing capacitor (and quick shut-off discharge resistor) already in place. I found a datasheet for the SM2082EGS, unfortunately in Chinese, which showed the circuit diagram. Only 3 pins are used, 1=gnd, 2=Rext, 7=out. Presumably the 8 pin package rather than 3-pin is to allow for extra heat dissipation. How long until that changes I wonder. Their initial diagram suggests 22uF for the smoothing capacitor but a later diagram shows 6.8uF and I think suggests that this allows the discharge resistor to be omitted. They also suggest 15 ohms for the programming resistor giving output of 12W per chip but the teardown site suggests this settles down to about 11W (for the MLS3535A) which corresponds to what Clive saw.
Looks like nothing will beat (At least in the long run) a good ol' 50 Hz mains transformer with bridge rectifier and smoothing cap powering 5 mm white LEDs with resistors to limit the current to somewhere between 15 and 20 mA. Needless to say that a *proper* setup like this could stay illuminated for DECADES uninterrupted without flickering or blinking, unlike all those poor quality newer LED fixtures both in the home but also in commercial settings (Public streetlights, supermarkets car parks floodlights and so on...) that always have some sort of defect.
Well, I could see a few applications where a cheapy LED panel would be enough to give off a good amount of light. The ideal setup was with the 1uF capacitor and the smoothing capacitor on the board. With no modifications = LEDs may bake themselves, flickery With smoothing cap only = LEDs may bake themselves With 470nF cap and smoothing cap = power rating/light output a bit low With 1uF cap and smoothing cap = ideal power rating for optimal lamp life With smoothing cap and 11R resistor removed = regulator too hot, will shorten its lifespan
I bought a similar job about 3 years ago (came from Germany, so is actually well-made with a proper driver) to go into a dome light I bought to replace the bathroom fly-catching glass globe fitting, still not put it up yet though, the CFL in the bathroom light has refused to die... :P
i love the amount of wire these so called manufacturer's give you about 2 inch of white twin un earthed back plate am keeping out of this one though nice fix Clive
i sadly don't like the lighting heading in this direction in general but neat to mess with for sure. i like my replacement traditional sockets much more to ever replace them with something like this in general.
I bought a few similar ones to keep in the van for service calls thinking the magnet system was a clever and kinda quick, dirty and cheap repair but I haven't used one yet because every old light I've touched that required an upgrade wasn't in good enough condition and just crumbled at the first touch...
Adding the capacitor does make it generally better, but the power factor became worse. I'm amused the capacitor goes on the "R1" pads instead of the "C1" ones.
Hi Clive-Thanks for your video(always look forward to a good watch !!) Can I ask you a question? Would it be possible to make a video about adapting a set of Xmas lights to work on 12 volts. With the price of mains generated gold plated electrons this year, do you think this would be a good idea so they could be run off a car battery, charged by solar etc ? !! Thanks again
These lights are actually being sold as kits. The board and LEDs are sold pre-assembled. Everything else is you making them actually work the way the manufacturer claims they do.
„not power, but photones“ Most of the energy on this planet have reached us as photones. Either today or millions of years ago and someone (the old man?) stored it in smelly liquids for us to find.
I wonder how light output changes in each case. is there a sweet spot with minimum power, maximum light output and maximum life expectancy? thank you for your inspirational videos.
If the wite out isn't working in tape form, I know you can get the bottle of it, though its not as clean, it can be more forgiving on things that arent specifically text related
I have a few of this panels in use bur with a little more advanced driver. Some are running a lot of years. Longer than the cheap ceiling lights from the local dump store who died after almost 1 year. So i put a cheap panel in from arround €6,-. But i need to say that i screw them directly to the metal plate with some cooling paste between. And yes i also earth the light. But the lights i use are not flickering at all. CRI should be above 80. Ordered from a store called ´Health forces" on Aliexpress...
I'm curious as to why the PF went down when you added the smoothing cap and increased the resistor size. Was 0.9 originally, but once you added the capacitor to stop flicker and changed the resistors it went down to around the same as using the capacitive dropper at about 0.6.
Because with the smoothing cap charged there is only a very short time in the mains sine voltage where that voltage is actually higher than the caps voltage. And only in that period current will actually flow from the mains. So you have 100 current spikes (50 on the positive and 50 on the negative peaks of the sine for 50Hz mains). That's bad power factor.
Because the load changed from a resistive-type (from the regulators mainly) to a primarily capacitive load, caused by the pulsed charge current at the voltage peaks of the mains.
Sorry, a bit of a (probably) stupid question here... Wasn't there a dedicated placement for a capacitor (C1) between the two regulator chips; what's with that?
They basically designed the board to be able to use all the components, but in hundreds of thousands of units you could save quite a bit of money on capacitors, I guess most buyers don't care about the flicker that is also present in discharge lamps (fluorescent, HID, mercury vapor, sodium vapor) that use electromagnetic ballasts...
This LED system looks good because it has the magnets, easy to install. But it could have been nice to be waterproof as I would like to install it in my ceiling light in bathroom, so this clearly looks like it will be damaged by the steam, even though the ceiling lights have a cover too attached to the metal body that's screwed into the ceiling. The bulb is a GR8 type but I took it out as it used 24W bulb which would last for only a year. And I added a small bathroom LED ceiling lamp but it's too big inside the cover 😅 Maybe I can hack it with that magnet system which seems to get rid of the 2 screws. Easy.
There is another module somewhat like this that DiodeGoneWild tested. The leds are in 3 strings and are turned on at different points in the cycle. Are you sure there is only one connection between the leds and the driver ics? Check the waveform with a scope. Look at his video, it is interesting.
I was able to find some smd LEDs online, plus a few smd components for LED lighting. In theory, we could make one of these light panels ourselves, but I don't see any markers for the LED orientation, nor was I able to find any general resources. Would it be possible for you to do a crash course on these components?
Since the power (17W) in the last bonus segment matched the power in the initial test, I am wondering if the (2) 22 ohm resistors ended up in parallel with both chips sensing across the resulting 11 ohms, the same resistance as originally.
If only Congress' 'Flicker' could be Regulated with a Snip here and Swop there. OH WAIT, it's *Election* Day! _hmmm_ Temperature check: Let's See them Light up - _We wait for the _*_Bang._* 8 ) *Thanks Clive!*
In fairness to the chinese manufacturer, most people buy lights based upon the rated power consumption. Old habits from the incandescent era die hard, and if that means running the chips hotter so people think the light is brighter then that's what they'll do. Cheaping out on the smoothing cap is a mistake, however.
I will just stick to DC fixtures and my very nice SMPS with all the bells and whistles filtering- and smoothing-wise. I mean.. if you put the work hours into it is it still economical to buy them ..?! (One gets the feeling It really isn't a good idea in the first place. (thanks for the video)) (I would guess) The input resistor (right at the hopi) is probably not doing anything else than dissipating some heat (since there is current regulation right behind it) ??? There is a similar Part on mouser : NUD4011 (110V only)
when looking at the thumbnail i was amused to read that they apparently called themselves (or the product line) dildomatic ... got a huge laugh outta me
When I was a teen I learned about capacitors. I built a beautifully constructed 300VDC tube power supply. Then, I turned it off and reached in to make some change. It shocked me hard and I pulled my hand out so quickly that my wristwatch stayed stationary and dropped to the concrete floor where it broke. Very instructive it was.
First lesson in removing your wristwatch before working on electrical stuff as well!!
sneeze!
Ah yes, capacitors. Those little innocent looking cans and discs, sometimes packs of foil and paper, thin plastic film, or maybe a bit of foil and some oily stuff inside. Nothing dangerous until you put a bit of voltage on them. I learned about these little critters myself at around the same age but never got attacked until much later when I was a newbie tech.
My job was to test power supplies of various sizes. Some were analog power supplies with big regulators and some big cannister-sized capacitors on them. I was always a bit on edge when I powered those up and one day, I had one explode on me because someone in manufacturing put the big capacitors in backwards. The explosion was so intense that it sent the capacitors up into the drop ceiling above me. It was a good thing I always ducked when I powered those up!
The last encounter was with some switchers. These switching power supplies, I think 65W units used in some "portable computers", with portable being a relative term for machines weighing around 22 lbs. 10 kg. There was a batch of these small switching power supplies that had leaky caps that needed replacing. Field service collected the units and brought boxes of them down to the test department for repairs. The units were to be returned to field service for replacement in customer returned units, or sent to the outside depot houses where they came from for field replacement.
I reached into the box of power supplies and got zapped instantly! The units, while sitting in the boxes for ages, still held quite a charge. After that, I got out a screw driver and would rub that across the leads sticking out of the boards before I even put my hand near them. The zap was enough to leave my arm achy and hand sore for the rest of the day.
i think i learned this from a tv originally... not to not wear a wrist watch but how to smack something in reflexive pain to the shock and accidentally break something, i forget what i knocked over... haha
Number of things about this don't even make sense, why 300vdc for tube technology, how if your hand was in something did anything come off and drop to the ground...
I've learned so much from your videos. I'm never going to use them in a technical sense, but they are hella helpful to look out for scams and poorly constructed equipment.
You mentioned the resistor to counteract the 'ghost glow' - the other approach is to add a resistor between the switch lines (1M or 2M maybe) to increase the effect for a useful night light. The switches then choose between 'full' and 'night light'.
You keep giving me new ideas, or rather you point my mind in a new direction often. I appreciate you
He just inspired me, with this video, to stop halfway through and clean my electronics desk down through the layers to real desk. Knock on wood. I also mixed up some liquid flux to put in a syringe.
@@johnpossum556 I should follow your lead and clean my own lol
Possibly the other thing they were going for was a power factor requirement? The smoothing capacitor drops the power factor significantly, and using circuitry that is both flicker free and high power factor would be quite a bit more expensive.
I suppose you could say the flicker was working as PWM and giving the LEDs reduced duty cycle.
I like the idea of re-grading a downgraded product, but the fact that the vendor is probably charging the same price or more compared to what it was before it got "cost-optimized" is irritating enough that I would think twice before buying the product. Of course it is happening constantly without my awareness; I just happen to know better in this case.
If they were smart, they'd sell a non-cheap-out version with a higher markup and call it the "gold edition" or some such nonsense
I looked at iced tea, and at some point they basically added more water without warning and keeping the same price
The wiring on those regulators in current regulator mode, looks like a LM317 wired as a current regulator.
Interesting product. Thanks for the teardown!
They are essentially the same thing.
Anytime you add a capacitor I smile, because I always want to add more capacitors to projects. I never been a fan of how many resistors companies add to products. I prefer 1 or 2 resistors over their dozen of resistors.
sorry, but why not just use a single bug enough cap instead? Legit question. I'm a noob
Adding caps is fun - until one day I added too many µFarads to the selenium rectifier of one of my valve radios... Selenium is one period lower than sulfur, but worse. The smell lingered for weeks
Clive, this video was darn near perfect - your craft is in top form.
I guess we're not talking about that thumbnail!
I love you big Clive Your soothing voice helps me sleep whenever I feel depressed
A lot of people say that. I'm glad it helps.
Main problem with the panels I bought is absolutely terrible CRI of about 60. They are bright but the room feels dull. Would not recommend wasting time and effort on these.
That is a very important point with respect to cheap "white" LEDs
Kreosan was just scraping down one of the paired resistors for longer life with a meter hooked-up to monitor the drop till it was what he wanted. He was doing it to drop the input when one led burned and he shorted it with black silicone (yes lamp black and charcoal conducts electricity, I've elecroformed with both when I had problems with "pure" graphite that was more like silicon dioxide powder with graphite but visually you could not tell) to relight the "string".
Dude, you need a more violent name.
@@johnpossum556 😅
The insurance investigator will have a good laugh after the fire is out.
Since those are linear regulator chips, I would put the capacitor in parallel with the LEDs instead of directly across rectifier output. It would still greatly reduce flickering without destroying the power factor, subjecting the capacitor to high crest current nor increasing the DC input voltage across the linear regulators beyond the initial power-on cycle.
Unfortunately that puts a lot of stress on the regulator at power up, as it sees a very high voltage differential as the capacitor can start at zero volts.
@@bigclivedotcom which is exactly what I meant by "initial cycle." 100mA at 350V crest voltage is only 35W instantaneous power for the sub-1ms it will last. As long as the regulator is designed to drop the full 350V, 35mJ per on-off cycle should be fine. Raising the DC voltage with an input capacitor increases energy dissipation in the chips by a much greater amount on a continuous basis.
The pop from the capacitor discharging made me flinch...I bet that would have hurt...
Yes it does.
Who would want lights that flicker? I wouldn't. I assumed it was because of your camera, but no, I guess not. Interesting video! Great improvement with the caps and resistors.
in normal day to day use, you will most likely never notice the flickering but there are those that are more sensitive to it, like i notice the backlighting of my keyboard flickering when im not looking directly at the keyboard, said flickering on my keyboard was even worse when i used it on a usb polling rate of 100hz, now at 250hz polling rate its far less noticeable flickering
I've watched a bunch of your videos and I feel like I'm a starting to understand what my teachers meant
Very interesting hacking experiments. These linear regulators seem fairly flexible, either using a capacitive dropper or just adjusting the value of the resistor. The smoothing capacitor is definitely necessary, I can usually tell when lights don't have one, even without the aid of a video camera.
Some Chinese teardown of a similar device suggests that the MLS3535A should be similar to the SM2082EGS. There is a Chinese datasheet available for this chip. The pinout matches, a lot of pins are marked as NC. There should be 0.6V present between pin 2 (Rext) and pin 1 (GND). The output current is specified to be 0.6V/Rext. The datasheet also allows smoothing at the input, and the chip is specified to work on DC as well. The seemingly strange track routing on the PCB with GND under the chip is also recommended in the datasheet, since there seems to be a thermal pad (or "IC pin") on the underside of the chip.
Is it my imagination, or is Clive less willing to test caps with his fingers since EEVblog's "zappy zap" experience.
In this case I knew they were likely to have a charge.
Clive also isn't Medhi (Electroboom)...
My take on the lack of an electrolytic capacitor is meant to reduce cost and prevent the explosive mode of failure that is caused by the capacitor being dried out due to the high temperatures produced by a largeish array of LEDs surrounding the cap.
I really really love this type of quick hack videos!
I think that all of us who are Big Clive fans have managed to zap ourselves at one point in our past.
Luckily, we're still alive!
BTW, those old B/W CRT televisions were a bit of a killer if you tried to modify it to be an oscilloscope :(
I'd be inclined to buy better. Failing that, I'd improve the design with a smoothing capacitor and anti-ghosting resistor then reduce the wattage to 80% of the resultant value by changing both sense resistors. I've no desire to dim everything to night light levels nor make it outlast me.
Removed flicker but PF from 0.9 to 0.55. Do we want efficiency or
Flickering? Lmao 😂. You have the dream job Clive.
I miss the smell of smoldering resistors. Electronics Lab always had a pall in the rafters and an odor of ozone and fried ceramic. EL was a recycled hangar with fans in the eave ridges, no real modern ventilation system. it was an Olde School. wow, that was a fuzzy olde memory.. thanks Clive.
Hey Clive, There’s a bit of a controversy brewing on the interwebs regarding UV led street lamps. I was wondering if you could perhaps do a little video to address what’s going on and why they create the purple light that is absolutely horrendous to drive under at night. I don’t know if you’re experiencing this issue across the pond, but I believe you probably have the knowledge to explain just what sort of failure is occurring, whether a coating issue, or failure of a particular color led, or whatever causes my windshield to momentarily flash a blinding purple glow as I drive home each evening. There is a lot of hyperbole out there about what I’m sure is simply dodgy engineering. I appeal to you, the beard of electronics knowledge for enlightenment. Thanks for the informative and entertaining videos.
Sounds like the phosphor coating of these LEDs got damaged in some fashion. The LEDs in those lamps normally put out light in the UV range. To turn that into nice white light, there is a coating on it that absorbs UV light and emits longer wavelength light. Kind of like how bleached clothing shines in dark light (UV).
I'm guessing they screwed up the coating, or maybe didn't test if it could take the temperature cycles of an outdoors environment.
Sometimes blue chips can emit purple waves (spectrum), without being technical. More to the point they can be edging so slightly into purple rather blue. Also, it should be noted that car windshields, as you prefer, do have UV coatings or stuff baked into the glass it self so you don't turn into a tumbleweed or a baked potato. Crispy pork loin maybe...
Either way, the glass if the refraction of light and the beam of light and the correct colour of light and the photon direction all hit correctly at the right junctions, you get glare.
This purplish light you mention I see from LED headlights going down the road but because the angle of light is so that it doesn't refract accross the whole screen, I don't get blinded by that particular spectrum of light the led light is emitting.
UV is a specific wavelength of light and while it does emit a purple colour, it's not the same as purple spectral light
Note ozone generators and germ eradication devices that say they use UV light and they are just purple coloured LEDs. Fakery fake fakeness
Pretty sure now I blabber that purple from LEDs is blue chip with a form of coating.
I think the base answer is because there are so many factors that could reasonably say why you experiencing what you are it gives rise for the doomsayers to emit from the free to speak minds things that are not necessarily true while trying convince one that the egg most definitely came first...
I say all this however while not being aware of a campaign of rhetoric about the subject you speak.
Sounds a bit intimidating, having the idea that your being cooked while driving down the road.
Alas, your screens do have some protection, but I think it's only one, uvb or uva, some might have both, but it's a case of 'how fancy is your car, how fancy is your glass' kind of buck.
I'm not an expert, so don't take my word for all of what I said but I believe what I said to be more or less correct.
The violet chips stimulate phosphor which is detaching in this instance, resulting in a glorious (but safe) deep violet colour.
Search for "technology connections" streetlight vid.
Considering the voltagedrop over the leds as a constant. The power dissipation would not change due to using only one regulator because the current through each regulator does not change. The increase in power is caused by adding the smoothing capacitor because that increases the duty cycle.
I remember back to the last time we looked at a "driverless" LED module and experimented with fixing the flicker
The capacitor works best after the drivers, that way they don't have to dissipate the extra voltage from the smoothed mains
Would be interesting to see an efficiency and power factor test with the capacitor before and after the drivers. Clive, make it happen!
Although the capacitor itself is also an added load then, it's worth testing._
Clive, I apologize in advance for this ... but ... I started watching and then you did the thing with the meter and it instantly came to me, "There's Clive, pluggin' things into the Hopi again." lol. Thanks for continuing to "take stuff to bits," Uncle Clive.
Excellent, informative, video explanation as always Clive. I have just fitted two of these cheapo Chinesium lights in place of ancient incandescent fittings. We shall see how long they last...
1 Meg+ resistor across the cap is required.
Caps do recover considerable voltage after they are shorted briefly.
Hey, i would really appreciate it if you explained how does that happen, even if i short it for like a few seconds it still holds some charge!
@@theoneohmresistor dielectric absorption.
I had to click on this... Just because of the "Boobie Meister" label.
I'm mildly disappointed that they didn't really have that on the package...
I have literally a dozen of 1000 lm LED lamps *per room* in my house, can't be bright enough for me. Watching these videos where you reduce the brightness or say "It is awfully bright" always hurts watching ;) (But I still love them)
This "product" is a bonus if it is inexpensive and can be transformed into an improved functioning lamp or an emergency light.
Emergency lighting is a continuing sphere of interest for me, where I live, where tropical storm outages are a part of life.
So far, only two people have noticed the dodgy wording on the package in the thumbnail. Naughty Clive!
I'm happy to see HOPI is doing fine. ^_^
This view brought to you by the Boobymeister clickbait. Well played sir, well played.
Getting rid of flicker is very worthwile
Much as I like these videos (and I really do like them), I miss the random ebay kit building vids narrated by chaotic stream of consciousness.
....... That thumbnail brought back a nostalgic memory! 🤣
It has a made up word on it that is very similar to one a buddy, who I met by playing on his Quake II Deathmatch server, had come up with one day; his was "dildomagic".
Though it deserves context...
We would chat over ICQ, and I had found a plugin for it that would do Text-to-Speech, which being Win98 meant it was only barely better than Stephen Hawking's synthesizer.... and us being teens, meant juvenile fun. OH and it had an on screen avatar to read it, but that might've been part of Winderz, a classical Wizard (default) or a brown dog were choices.
That's when he came up with this gem to have the wizard say, of which I'll never forget:
_Westside fo' life, cube. Dildomagic, bonita ramsey_
_chunktastic, fo' digga di-gazzi madre._
Honestly, what made it so funny back then -aside from the fact that "dildo" was the only crude word needed- was the lack of inflection the TTS engine had, as it said it so monotonely in that classic robotic way, and the punctuation had more impact on the flow (greater pause than modern engines). Remembering how it hung off that first coma before _finally_ saying "cube" *still* gives me a good chuckle! ☺️
Anyways... Now to watch your video about hacking these "boobymeisters"! 😁
"It's not that great, but still worth playing about with and experimenting with." --- Big Clive. Words of wisdom: Life lessons.
Absolutely brilliant!
That was a bit of fun. I enjoyed watching it
You are a gent Sir, I forgot to order a 2d LED replacement 2 pin for my bathroom light as fed up with it having to buy those stupid light bulbs for it and I had earmarked a LED panel replacement and your vid reminded me to buy it hehe
this video is magic, there are mouths old comments, but it was posted 13h ago
I thought the same right now
It was first recorded and released on Patreon a while ago.
_snap_ *bang* _pop_ Never change, Clive. Live forever.
Buck Regulator you say?
I'd buy that for a Dollar! 💸
I saw Dave Get a nice belt of a cap the other day on one of his videos
This particular led configuration was initially designed for under water lighting.
Again with the manual dexterity. Impressive.
Let's not forget that there are extremely few vehicles that are not computerized. It is not unrealistic to assume that impacts on transportation systems may be regionally catastrophic. So not only will we be faced with months long blackouts, which will impact food storage, medical care delivery, and many other things ,but the logistics of responding to the emergency will be impacted as well.
Got me there… clicked the video based on the thumbnail alone… "This can't be real" combined with "Clive would totally buy it"… ah, well…
I found a Chinese site with a similar teardown and a comparison of the MLS3535A chip with what appears to be pin compatible SM2082EGS, and a circuit diagram which had the smoothing capacitor (and quick shut-off discharge resistor) already in place.
I found a datasheet for the SM2082EGS, unfortunately in Chinese, which showed the circuit diagram. Only 3 pins are used, 1=gnd, 2=Rext, 7=out. Presumably the 8 pin package rather than 3-pin is to allow for extra heat dissipation. How long until that changes I wonder.
Their initial diagram suggests 22uF for the smoothing capacitor but a later diagram shows 6.8uF and I think suggests that this allows the discharge resistor to be omitted. They also suggest 15 ohms for the programming resistor giving output of 12W per chip but the teardown site suggests this settles down to about 11W (for the MLS3535A) which corresponds to what Clive saw.
Looks like nothing will beat (At least in the long run) a good ol' 50 Hz mains transformer with bridge rectifier and smoothing cap powering 5 mm white LEDs with resistors to limit the current to somewhere between 15 and 20 mA.
Needless to say that a *proper* setup like this could stay illuminated for DECADES uninterrupted without flickering or blinking, unlike all those poor quality newer LED fixtures both in the home but also in commercial settings (Public streetlights, supermarkets car parks floodlights and so on...) that always have some sort of defect.
Very cool. Thanks Big Clive.
Well, I could see a few applications where a cheapy LED panel would be enough to give off a good amount of light.
The ideal setup was with the 1uF capacitor and the smoothing capacitor on the board.
With no modifications = LEDs may bake themselves, flickery
With smoothing cap only = LEDs may bake themselves
With 470nF cap and smoothing cap = power rating/light output a bit low
With 1uF cap and smoothing cap = ideal power rating for optimal lamp life
With smoothing cap and 11R resistor removed = regulator too hot, will shorten its lifespan
I bought a similar job about 3 years ago (came from Germany, so is actually well-made with a proper driver) to go into a dome light I bought to replace the bathroom fly-catching glass globe fitting, still not put it up yet though, the CFL in the bathroom light has refused to die... :P
So many opportunities to go all Electroboom in this video!
i love the amount of wire these so called manufacturer's give you about 2 inch of white twin un earthed back plate am keeping out of this one though nice fix Clive
Anyone else think Clive was moving into "electroBOOM" territory when he pointed to the HOPI display around 6 mins in?
Another very cool project!
i sadly don't like the lighting heading in this direction in general but neat to mess with for sure. i like my replacement traditional sockets much more to ever replace them with something like this in general.
I bought a few similar ones to keep in the van for service calls thinking the magnet system was a clever and kinda quick, dirty and cheap repair but I haven't used one yet because every old light I've touched that required an upgrade wasn't in good enough condition and just crumbled at the first touch...
omg i was SHOCKED to see the power factor so high, i was guessing it would be way lower
@bigclivedotcom you ever put a pot on a led panel?? i have some bigger(150w) led drivers that have 0-10v almost dmx like control
Adding the capacitor does make it generally better, but the power factor became worse. I'm amused the capacitor goes on the "R1" pads instead of the "C1" ones.
Hi Clive-Thanks for your video(always look forward to a good watch !!) Can I ask you a question? Would it be possible to make a video about adapting a set of Xmas lights to work on 12 volts. With the price of mains generated gold plated electrons this year, do you think this would be a good idea so they could be run off a car battery, charged by solar etc ? !!
Thanks again
You can run them for a very long time on a USB power bank. I've made videos about that.
Thanks Clive- I will look though your archives !!
Dildomatic Boobymeister 🤣🤣... you are a legend Clive ✌️👍❤️
12:30 "How cheap can we make it?"
"Yes."
These lights are actually being sold as kits. The board and LEDs are sold pre-assembled. Everything else is you making them actually work the way the manufacturer claims they do.
Awesome job! Thanks for the bonus!
„not power, but photones“
Most of the energy on this planet have reached us as photones. Either today or millions of years ago and someone (the old man?) stored it in smelly liquids for us to find.
This was great. Thanks!
I wonder how light output changes in each case. is there a sweet spot with minimum power, maximum light output and maximum life expectancy? thank you for your inspirational videos.
Lower current results in greater LED efficiency.
Great. Big Clive Industries has a rework department. Good luck! 👍
If the wite out isn't working in tape form, I know you can get the bottle of it, though its not as clean, it can be more forgiving on things that arent specifically text related
Did you notice that after you changed the resistors to 22 ohms the power factor dropped to .5 from the original .9? I wonder why,
It changed when he added the smoothing capacitor not when he changed the resistors. It's what capacitors do.
@@chrishartley1210 Okay that makes perfect sense.
LMFAO Boobymeister
I have a few of this panels in use bur with a little more advanced driver. Some are running a lot of years. Longer than the cheap ceiling lights from the local dump store who died after almost 1 year. So i put a cheap panel in from arround €6,-. But i need to say that i screw them directly to the metal plate with some cooling paste between. And yes i also earth the light. But the lights i use are not flickering at all. CRI should be above 80. Ordered from a store called ´Health forces" on Aliexpress...
I'm curious as to why the PF went down when you added the smoothing cap and increased the resistor size. Was 0.9 originally, but once you added the capacitor to stop flicker and changed the resistors it went down to around the same as using the capacitive dropper at about 0.6.
Because with the smoothing cap charged there is only a very short time in the mains sine voltage where that voltage is actually higher than the caps voltage. And only in that period current will actually flow from the mains. So you have 100 current spikes (50 on the positive and 50 on the negative peaks of the sine for 50Hz mains). That's bad power factor.
Because the load changed from a resistive-type (from the regulators mainly) to a primarily capacitive load, caused by the pulsed charge current at the voltage peaks of the mains.
So any linear PSU will show this effect? Interesting, never really thought about it like that before.
yeah, any non APFC SMPS will show that :(
Liking the hacks various! 👍😎
Sorry, a bit of a (probably) stupid question here... Wasn't there a dedicated placement for a capacitor (C1) between the two regulator chips; what's with that?
They basically designed the board to be able to use all the components, but in hundreds of thousands of units you could save quite a bit of money on capacitors, I guess most buyers don't care about the flicker that is also present in discharge lamps (fluorescent, HID, mercury vapor, sodium vapor) that use electromagnetic ballasts...
@@ruben_balea Yeah... that's what I thought too, they've chosen the path of the "cheepo" 😂
This LED system looks good because it has the magnets, easy to install. But it could have been nice to be waterproof as I would like to install it in my ceiling light in bathroom, so this clearly looks like it will be damaged by the steam, even though the ceiling lights have a cover too attached to the metal body that's screwed into the ceiling. The bulb is a GR8 type but I took it out as it used 24W bulb which would last for only a year. And I added a small bathroom LED ceiling lamp but it's too big inside the cover 😅 Maybe I can hack it with that magnet system which seems to get rid of the 2 screws. Easy.
There is another module somewhat like this that DiodeGoneWild tested.
The leds are in 3 strings and are turned on at different points in the cycle.
Are you sure there is only one connection between the leds and the driver ics?
Check the waveform with a scope.
Look at his video, it is interesting.
Just one. It's not the multiple tap style that rides the sinewave.
You just mentioned losing monetization. Could you package have any impact?
Not sure.
I was able to find some smd LEDs online, plus a few smd components for LED lighting. In theory, we could make one of these light panels ourselves, but I don't see any markers for the LED orientation, nor was I able to find any general resources. Would it be possible for you to do a crash course on these components?
Since the power (17W) in the last bonus segment matched the power in the initial test, I am wondering if the (2) 22 ohm resistors ended up in parallel with both chips sensing across the resulting 11 ohms, the same resistance as originally.
If only Congress' 'Flicker' could be Regulated with a Snip here and Swop there. OH WAIT, it's *Election* Day!
_hmmm_ Temperature check: Let's See them Light up - _We wait for the _*_Bang._* 8 ) *Thanks Clive!*
Nice tinkering, but you could have swapped the cap round at the end just to satisfy our need for BOOOOOOM 🤣 nice video👍👍
In fairness to the chinese manufacturer, most people buy lights based upon the rated power consumption. Old habits from the incandescent era die hard, and if that means running the chips hotter so people think the light is brighter then that's what they'll do.
Cheaping out on the smoothing cap is a mistake, however.
Maybe it was downgraded because they died faster when they were, possibly, better
What kind of capacitor would i need to smooth a 12v 10w'ish COB LED?
Is that your mic amplifying the EMF from the cap shorts?
I'm guessing it sounds louder on the video than in person...
I will just stick to DC fixtures and my very nice SMPS with all the bells and whistles filtering- and smoothing-wise.
I mean.. if you put the work hours into it is it still economical to buy them ..?!
(One gets the feeling It really isn't a good idea in the first place. (thanks for the video))
(I would guess) The input resistor (right at the hopi) is probably not doing anything else than dissipating some heat (since there is current regulation right behind it) ???
There is a similar Part on mouser : NUD4011 (110V only)
Nice one!
Makes you wish they'd just put a capacitor in by default, they're dirt cheep and make a world of difference.
excellent idea to increase the resistor values to drop the heat in this thing because at 17W its on a highway to hell
thanks for the bonus!
when looking at the thumbnail i was amused to read that they apparently called themselves (or the product line) dildomatic ... got a huge laugh outta me