Clive, just wanted to say thanks for keeping your video style consistent. With all sorts of RUclipsrs changing up once they get some popularity, yours has kept going as is. Thanks for the great content!
LED driver (and electronics as a whole) circuitry miniaturization these days is simply incredible. Due to the use ICs and SMDs the whole driver can fit inside the regular bulb socket cap...
You got me again. After almost 3 minutes of looking at the printout, my brain had already forgotten it was a printout and went WTF?! when you placed the notepad flat on all the "components".
The old voltage doubler rectifier! I know that one well. It supplied the power for my old coilgun project. It's actually more efficient for low voltages than the usual full bridge, as there's only one diode drop. Good for energy harvesting.
Fascinating Clive, a voltage multiplier circuit to power an LED string. It's great to be kept up to date with the latest. Thanks Clive! Your blood is worth bottling.
I was in my dentist's office the other day and noticed right away that he'd had all the drop-ceiling fluorescent lights converted to LED. Though, as I walked down the hall I could easily spot places in the arrays where individual LEDs had failed. I guess he went low-budget on the lights or this was just a poor design that didn't diffuse the light enough to hide the occasional dud. Obviously, the LEDs themselves should have an operating life of several years and not just a couple of months!
Sebas Eu, interesting observation. I've heard before that the capacitive dropper types are susceptible to peaks or switch bouncing. Those mean a high dv/dt, equivalent to higher frequency, which in theory can pass through the dropper capacitor and cause overcurrent through the LED chips..
Clive can i just say that i watch your videos just because of your voice. Im not even that interested in electronics but you make it interesting vocally
I start playing the new Clive video and as usual the cat comes running to the room, he must like Clive's voice. Clive, are you available to cat sit when I go on holiday/vacation?
11:07 "J W Here" - heard that in my head right about the time you said it. Have I watched too many of your videos, Big Clive? Damn shame l'm on the other side of the pond from you or I'd come to the Manx meet-up to hear your wonderful accent first hand.
Usually you'll see Greinacher style voltage doublers, not the Delon style one you find here. The advantage of the Greinacher is the input and output share a common lead, which matters in some situations, but not this one. When I built my first strobe light, since I live in the US with its 120 volt mains, I needed a doubler to come up with the voltage, and since I was young, I couldn't wrap my brain around how the Greinacher doubler worked (it seemed to me at the time that it was wasting half the power by dumping it to ground), so I went with the easier-to-comprehend Delon style seen here.
I'm amazed at how many LED bulbs seem to fail. I have bought all of my LED bulbs from Poundland or Screwfix when it was cheaper to do. I typically buy 5 to 9 watt bulbs. In the past 3-years I have only had one LED bulb in my house fail, and this had been heavily used. I have no dimmers in my house, just the normal 240v supply.
Since LED lamps have become popular, I've found myself wondering why they continue to make the bulbs out of glass (yes, there are some that don't), but seeing over 600 volts across the filaments in this one (and probably 300+ for 120v bulbs), I no longer wonder. "Yes, I'll have the glass bulb please...." Thanks for the video.
is there a technical reason why they're using a voltage doubler like this vs doing the capacitive dropper to do three pairs at 200V? I can't seem to think of a benefit to doing it this way, unless they got some of the parts for cheap and just kinda made the rest work.
It's a really simple and compact way to drive that many filaments with just a single current limiter. I would guess the main driving factors were cost and space.
You could use a single limiter. They may be trying to avoid mismatched current between parallel circuits and perhaps reduce dissipation in the limiter by keeping the current low.
Yep, that's what I figured. Also I suppose if one of the filaments failed open circuit like this the increased current on the other two probably wouldn't be good. Glad I'm not crazy, though. Thanks Clive, awesome videos as always.
I was expecting that 2KV insulation tester to come along. I guess it wouldn't deliver enough current though. I wonder about these lamps though. Not really knowledgeable with capacitor droppers (Needs more tutorials materials Clive...) or this design, but I wonder why they even bother with the chip. If they've got a "smooth" 670v DC why not just match the LED's close to that voltage and let the inrush resistor (or one at the LED's) limit the current? What advantage is this chip really giving? If the capacitor was sufficient to stop voltage droop on load then there's no real need for any regulation or active current control.. Ach, I dunno. More tutorials on driving LED's using droppers requested. Maybe including details on driving larger loads too. Maybe a 3w/10W 12V LED... How do we really calculate these things?
Without the margin for regulation the current through the LEDs would vary wildly with supply voltage. So the lamp could be dim in one house and very bright in another. The intensity would also jump up and down with the slight voltage variations caused by loads switching on and off.
220 - 240 is the official voltages. I'm in about 6 to 8 houses a day, part of the job is doing a Ze test at the incoming supply, the loop impedance meter displays voltage first before the earth test. Average is about 228 - 245, but i have found a few as low as 215 and as high as 287. Which is a bit naughty from Scottish Power.
I thought the current official spec was 230V +/- 10% (207-253V) - the "mid harmonisation" period. Eventually once all the substations in Europe have been renewed, the spec will tighten up to +/-5%
Clive, I have a power supply brick to a PoE network switch and the switched wall outlet it was connected to which exploded, effectively, last night. Would love to send them to you for a tear down so we can learn what happened. If it wasn't for fuses and RCDs, our house would have been rather smokey!!
There is a limited data sheet for the JW1691H online - 500v max and 5 -80mA. You can get the same constant current effect with a three terminal voltage regulator by connecting the LEDs in series with the Vin terminal and connecting a fixed resisitor that determines the current from Vout to Common.
The advantage of the dedicated chip is that it is rated for a higher voltage range and has internal thermal limiting that will lower the output progressively if it overheats. Many of these lamps do start of at higher intensity and then nudge back gently when they have warmed up.
I gave up trying with LED filaments (granted they were all direct-from-china, but still), the majority of the ones I have/had ended up with broken bond wires that turned them into seizure-inducing flashing lamps as the filaments blinked on and off like a bad Star Trek effect, the rest, well, they got broken due to butterfingers here... :P
I guess the trick is to find out how many are in the corrugated cardboard boxes they receive them in, and order in whole boxfuls. I've gone off buying super cheap LED bulbs though. I think it's cost effective to get Osram or Philips and actually get 20 000 hours, even at 5x the price.
hey BigClive I have become very interested in electronics because of you. I will picking up a soldering iron in a few days. But I was curious if you've ever "taken to bits" any older mice or computer game controllers. I have and it was fascinating (even though I didn't know what I was looking at. so many components and very few chips.) I was looking for a video where maybe you did with no luck. any way... If you're reading this thank you for making amazing content.
My son and I were at pizza hut, the other night. They had just remodeled the interior of the joint. I noticed all the lamp fixtures were these types of led lamps. I was like cool I've seen these taken apart by Bigclive.
Unfortunately, you can't even replace faulty ones. If it would be possible to do that, I'd be interested in this kind of solution, but no way I'm going to buy it like this.
The 22 Ohm inrush resistor is likely (well that's what I'd design in there for low cost) an NTC component with a 22 Ohm cold rating. Also, I'm surprised the 2 Meg resistors are not balancing the large caps?
Interesting. I've just pulled apart a dead Philips dimable LED filament bulb. The driver is all transistor (with possibly a CC IC) The led filaments (three in parallel)require about 90vdc to light . Should have sent it to you Clive, for an autopsy (still can if you want the bits???)
Interesting. I've once built an almost identical circuit, but with a darlington transistor and a current divider to set the output current. Except It used 12V AC from a halogen lamp transformer (so about 24V on the LED), not 240V AC (560V on the LED) haha omg that's so dangerous.
The problem with LED filaments is that you have a lot of tiny LED chips in series. If one chip fails in a filament, the entire filament will likely fail. In this lamp's case, one bad chip has caused the entire lamp to fail as they are all in series (perhaps as a means to reduce current).
Clive, I think you should mention the inherent potential for fatal shock with the cap. dropper circuits. Although of course this LED isn't using that method. I was all jazzed about using CD circuit for somethings until I saw another video wherein the YT'er pointed out how it could go wrong in so many ways. I guess its common knowledge in the industry that you really only want to use CD in devices that are completely enclosed, with no external signaling wires or connections do to the fact that many points within the circuit can be at mains or close to mains voltage with respect to User-ground. Just thought it might be worthwhile to mention occasionally in the event some watchers will be like myself and think its a neat, cheap shortcut to a mains powered LV PSU. Which it is, however the clear fatality potential is extremely risky to save a dollar or two.
Just wondering how a SBC connection would differ? There wouldn't be much room to cram those big capacitors in. I own several filament LEDS of that particular brand. A lovely 'warm' white but not very reliable due to the Black Dot of Death.
Ive just had one of those lamps fail (at least it is visually similar) & i'm not sure what caused it. I know it made a little hole in the e27 screw bit & smoked. I was using my electric oven at the time & the circuit breaker on that tripped as well as the lights. The oven has been fine since.
That's definitely a lot of voltage, so at what ma do each of the LED's then run at? Thanks for another good video Clive. Very interesting to learn all that.
bigclivedotcom I was referring to the lamp power supply Although I'm thinking that it probably wouldn't work cuz they need a higher voltage start/strike them right? Maybe give it a shot for science? Perhaps you could start the tube with your custom power supply then move over to the lamp psu?
These days most channels have a standard sentence with “click here to subscribe to my channel and here is the patreon page and please give the video a thumbs up blah blah”. I'm glad that you don't do that.
anlumo1 Totally agree with you. I simply hate those video creators that are reminding and begging for the likes and subscribes sometimes even multiple times during a few minute video - I mean, if I like the video, I will show my appreciation, but those f*cking reminders basically insults every viewers intelligence.
Surely it should be "always remember to put your meter leads back into the voltage sockets after measuring current"....or something like that....Always brilliant content however they end (smoke and/or flames are preferred but not mandatory).
All the electrical knowledge and impressions as well; first John Ward, then Homer Simpson, doing MmmmMeg Ohms (instead of doughnuts), finally I thought we were going to get Electro Boom, with the spicy voltage and getting the Argon Mercury driver our and 'we have ways of making you light'!
The filament LED style might well be the way mainstream LED bulbs are heading. I have noticed both Osram and Phillps are now actually making frosted versions of them and just marketing them as plain bulbs rather than the "heritage" tags they give to the clear envelope versions. Stands to reason as they can obviously be made in plants set up to make traditional tungsten filament bulbs with relatively little re-tooling. And easier for the general public to assimilate as the end product is once again the "familiar lightbulb" that they know. I can also see more sinister motives here too. Seems it's easy to make a bulb with so many LEDs in series (especially the one you have, but even the conventional series/parallel types) fail in a relatively short period of time. Maybe once again getting back to the 1000 hour planned obsolescence of the "Lightbulb Cartel"? Nice John Ward reference btw. :)
I picked up one of those frosted LED filament bulbs a while back from Screwfix. It's still working fine, and it's quite nice because it has basically the same radiation pattern as an old incandescent, unlike many other LED bulbs. They don't seem that common though. My experience with the Chinese eBay LED filament bulbs is more of a mixed bag. Some smaller ones (two 4W candles and a 4W GLS) are still going strong, but an 8W GLS didn't last long before it started flickering.
They seem to be a fairly new thing over here in NZ, although the clear ones have been around for quite some time. Offered in the standard glass envelope, the "vintage edison" pear style & a host of other silly alternatives. I have recently found another advantage of them seems to be that they radiate much less RFI than conventional LED bulbs. Which is important to me as I monitor a community radio station that I manage, which is quite some distance away. Of course my results here might be due to the particular ones I have, but I suspect that using the higher voltage and lower current is what mainly offers this benefit.
xanataph Maybe it's the kind of power supply they use - switchmode supplies can produce RF noise due to the high frequencies they run at, and the low current means the filament bulbs can get away without one.
Looks like the one that Clive took apart definitely had a SMPS in it though. But if they need less current it would be possible to use lower switching frequencies and that would help with reducing RFI. And also the lower current too I suppose.
The data sheet for that chip suggests it's a linear regulator, not an SMPS. The high voltage drop across all those series LEDs (about 85% of supply voltage) means that even a linear regulator will be reasonably efficient.
Peter Resch standard is the large bayonet fitting but we see a lot more ES fittings now as we get chinese/European fittings and shops sell bulbs for them. All Ikea’s stuff is ES and they’re a big driver for the changeover. We get small bayonet as well but it’s uncommon.
BC used to be dominant and is probably what a lot of us grew up with... ES used to be a speciality bulb that'd have a few inches of supermarket shelf space at best... now it's probably more like 50/50 with BC. I think Ikea is mainly responsible... everyone else in lighting retail saw they were getting away with selling ES fittings and decided to stop worrying about it.
Hey Clive, I've been having a difficult time trying to wrap my head around how comparators work, and what I've found online so far hasn't helped much. Do you think you could clarify how they work?
If you mean the simple op-amp style comparator then it compares the voltage at it's two inputs and the output goes high or low depending on which is higher. The plus and minus inputs determine whether the output goes high or low.
Hunter's Moon you’ll love this, I investigated a cheap 2 stroke generator that put out nearly a kv instead of the usual 240v. It was the subject of a complaint from a woman who was running a tv off it when camping. What I found was that her husband had fucked with the carburettor and set the idle speed too high. It was running at over 4500 rpm instead of 3000 and making crazy volts at low load (like when the tv was plugged in). Anyway it fired too many pixies up the poor old telly and the magic smoke got out, much fuss and bother etc. When I tested it with a heater it would run fine at about 500w or above because the governor could regulate the speed properly but at low loads, the throttle couldn’t close far enough to maintain that 3k RPM. Gotta love the bother you see in amateur hour 😆
Don't ask about the voltages used in microwave ovens then... several thousand volts, and low enough output resistance to easily push enough current to kill you.
dingo about 2.5kv usually, and they’ll supply around 400mA which will do a proper job on you. Avoid. Especially when it’s hooked up to a voltage doubler and some big ass capacitors.
I was hoping to get my bigclive fix today, you didn't disappoint, the board looks a real mess, both capacitors are 400v,arent they divisible in series?
Phonotical They would be if they were bipolar and connected conventionally but these are used in a kind of push - pull configuration, are polar and connected in parallel with the pair of mirrored diodes so each one charges only during one half of the incoming sine wave.
Im sure I remember some horrible maths involved with parallel resistance. Could you....measure two 100k resistors - solder the ends....predict the resistance with a calculator and then measure it !
Resistors in series-> resistance adds up Resistors in parallel-> conductance adds up conductivity is measured in "siemens", that's the reciprocal of "ohms" and vice versa. 1/siemens = ohm
It's not that horrible, you just need to add the inverses of the resistances (then invert back at the end). So overall resistance = 1 / ( 1/R1 + 1/R2 + 1/R3 + ...)
electronics certificate program DC Circuits semester 1 week 2 The stage in the course where you find out who did and did not actually pass gradeschool math.
I disagree about the filament lamps. Instead of relying on a few high quality high wattage LED chips these rely on many cheap LEDs in series not only making dead on arrival much more likely but also weakens the punishment the LEDs can take over time. The fact that it stayed on from the filter cap tells you a lot. There isn't much after the filters so if it dims out the supply is most likely working. The fact that the LEDs are all in series means even small ripples in the AC will kill the LEDs. So the lamps supply needs to make up for that with more regulation and filtering. So touch chance using a capcaitive dropper they must then use a linear regulator. It sounds weird but the higher voltage accomplishes this better because it now requires a doubling of a voltage spike to do the same damage to the LEDs.
The earlier ones were bad for earlier failures and they definitely don't like hot environments like fully enclosed fixtures. It pays to get lower power lamps and use more of them to create the same level of light.
definitely. I put CFLs in my cabin when I moved here 15 years ago. Just within the last year have some of them failed. ALL of the failures have been the result of the emissive coating on the filaments becoming depleted so that the discharge would not strike. This long life includes a 145 volt, about 5 seconds long power surge when a utility tech did the wrong thing at a primary voltage regulator. I'm just nerdy enough to keep a Dranetz line disturbance analyzer permanently connected to my breaker box. It recorded the surge to ms accuracy. I bought an Osram 100 watt equivalent LED bulb at Home Depot. It lasted about a month. I got a new one under warranty. I expect this "premium" LED to fail shortly. The conical shape between the plastic hemisphere and the Edison socket runs too hot to touch. And that's burning in an upright, non-enclosed configuration. The very best CFLs used a chip that ramped the operating frequency upward. Before resonance, the current lit the filaments red hot. When the frequency hit the resonance of the transformer primary and a capacitor, it generated enough voltage to strike the lamp. Then the chip settled into constant current mode. But people bitched about the half second delay so now most CFLs use an instant-on design. This type of lamp fires the discharge immediately upon power-up, bombarding the cold filaments with mercury ions, resulting in the emissive coating being blasted off the filaments much faster than had they been hot. All my long lived CFL use the ramp design. Sold in bulk under the GE trademark at Sam's club. Newer is seldom better, especially when the new stuff comes from Communist China.
Grossly unrelated. I once put a bare 5mm red old led across the rectified opìutput of a f/art neon transformer hoping for a short flash of glory. The led worked fine.
To be expected. Neon transformers are constant current devices. The voltage rating is only open-circuited. Typically, neon transformers come in 5, 10, 20, 30, 160 and 240 ma rating. I think Clive's is a 20 ma. That should make an old LED quite happy.
Bulb: "Do you really expect me to light?"
Clive: "No, Mister Bulb, I expect you to die!"
SigEpBlue This needs to be the top comment!
Clive, just wanted to say thanks for keeping your video style consistent. With all sorts of RUclipsrs changing up once they get some popularity, yours has kept going as is. Thanks for the great content!
LED driver (and electronics as a whole) circuitry miniaturization these days is simply incredible. Due to the use ICs and SMDs the whole driver can fit inside the regular bulb socket cap...
You got me again. After almost 3 minutes of looking at the printout, my brain had already forgotten it was a printout and went WTF?! when you placed the notepad flat on all the "components".
The old voltage doubler rectifier! I know that one well. It supplied the power for my old coilgun project. It's actually more efficient for low voltages than the usual full bridge, as there's only one diode drop. Good for energy harvesting.
Fascinating Clive, a voltage multiplier circuit to power an LED string. It's great to be kept up to date with the latest. Thanks Clive! Your blood is worth bottling.
Oh! I was hoping for the X-Ray treatment again!
The bigclive X-Ray is only for very special occasions :)
I could X-ray it and replace that filament.
What about that X-ray hammer?
I wish someone would give me a head x-ray lobotomy.
I was waiting for X-Ray but I think I just want to see it used again in general
I was in my dentist's office the other day and noticed right away that he'd had all the drop-ceiling fluorescent lights converted to LED. Though, as I walked down the hall I could easily spot places in the arrays where individual LEDs had failed. I guess he went low-budget on the lights or this was just a poor design that didn't diffuse the light enough to hide the occasional dud. Obviously, the LEDs themselves should have an operating life of several years and not just a couple of months!
Sebas Eu, interesting observation. I've heard before that the capacitive dropper types are susceptible to peaks or switch bouncing. Those mean a high dv/dt, equivalent to higher frequency, which in theory can pass through the dropper capacitor and cause overcurrent through the LED chips..
Interesting design, and clever use of voltage multiplication
Clive can i just say that i watch your videos just because of your voice. Im not even that interested in electronics but you make it interesting vocally
I wonder how you found his channel. Was it in some ASMR playlist :)
I start playing the new Clive video and as usual the cat comes running to the room, he must like Clive's voice.
Clive, are you available to cat sit when I go on holiday/vacation?
Clive has PO Box....just post on over your moggie when it comes to vacation time with a return address box packed also..........lmao.
don't forget the air holes :D
Don't make the mistake of creating the air holes with a knitting needle AFTER you package the cat - I've lost several pets that way! 😜
Ahahahaha omg....
Pete Allum omg :)
11:07 "J W Here" - heard that in my head right about the time you said it. Have I watched too many of your videos, Big Clive? Damn shame l'm on the other side of the pond from you or I'd come to the Manx meet-up to hear your wonderful accent first hand.
Usually you'll see Greinacher style voltage doublers, not the Delon style one you find here. The advantage of the Greinacher is the input and output share a common lead, which matters in some situations, but not this one. When I built my first strobe light, since I live in the US with its 120 volt mains, I needed a doubler to come up with the voltage, and since I was young, I couldn't wrap my brain around how the Greinacher doubler worked (it seemed to me at the time that it was wasting half the power by dumping it to ground), so I went with the easier-to-comprehend Delon style seen here.
I'm amazed at how many LED bulbs seem to fail. I have bought all of my LED bulbs from Poundland or Screwfix when it was cheaper to do. I typically buy 5 to 9 watt bulbs. In the past 3-years I have only had one LED bulb in my house fail, and this had been heavily used. I have no dimmers in my house, just the normal 240v supply.
Wow - an arking sparking brute force led driver circuit. The first one i've ever seen. That's amazing !!!
Since LED lamps have become popular, I've found myself wondering why they continue to make the bulbs out of glass (yes, there are some that don't), but seeing over 600 volts across the filaments in this one (and probably 300+ for 120v bulbs), I no longer wonder. "Yes, I'll have the glass bulb please...."
Thanks for the video.
is there a technical reason why they're using a voltage doubler like this vs doing the capacitive dropper to do three pairs at 200V? I can't seem to think of a benefit to doing it this way, unless they got some of the parts for cheap and just kinda made the rest work.
It's a really simple and compact way to drive that many filaments with just a single current limiter. I would guess the main driving factors were cost and space.
Hm. Would you need more than one current limiter using parallel strings?
You could use a single limiter. They may be trying to avoid mismatched current between parallel circuits and perhaps reduce dissipation in the limiter by keeping the current low.
Yep, that's what I figured. Also I suppose if one of the filaments failed open circuit like this the increased current on the other two probably wouldn't be good. Glad I'm not crazy, though.
Thanks Clive, awesome videos as always.
Nicole Cummings it also creates a nice income stream. A single point of failure in one of those 200 led generates a new bulb sale...
I was expecting that 2KV insulation tester to come along. I guess it wouldn't deliver enough current though.
I wonder about these lamps though. Not really knowledgeable with capacitor droppers (Needs more tutorials materials Clive...) or this design, but I wonder why they even bother with the chip. If they've got a "smooth" 670v DC why not just match the LED's close to that voltage and let the inrush resistor (or one at the LED's) limit the current? What advantage is this chip really giving? If the capacitor was sufficient to stop voltage droop on load then there's no real need for any regulation or active current control..
Ach, I dunno. More tutorials on driving LED's using droppers requested. Maybe including details on driving larger loads too. Maybe a 3w/10W 12V LED... How do we really calculate these things?
Without the margin for regulation the current through the LEDs would vary wildly with supply voltage. So the lamp could be dim in one house and very bright in another. The intensity would also jump up and down with the slight voltage variations caused by loads switching on and off.
Yeah. Suppose so. With line being all over from 220v-240v that makes total sense.
220 - 240 is the official voltages. I'm in about 6 to 8 houses a day, part of the job is doing a Ze test at the incoming supply, the loop impedance meter displays voltage first before the earth test. Average is about 228 - 245, but i have found a few as low as 215 and as high as 287. Which is a bit naughty from Scottish Power.
I thought the current official spec was 230V +/- 10% (207-253V) - the "mid harmonisation" period. Eventually once all the substations in Europe have been renewed, the spec will tighten up to +/-5%
MattOGormanSmith
Try telling THAT to PowerGen! (or whatever they're called these days)
Clive, I have a power supply brick to a PoE network switch and the switched wall outlet it was connected to which exploded, effectively, last night. Would love to send them to you for a tear down so we can learn what happened.
If it wasn't for fuses and RCDs, our house would have been rather smokey!!
You're great man. Really like your RUclips. I have so much fun when sitting home alone
rian ritzen it’s great bedtime watching. Very relaxing.
Clive, you rock sir. Thanks for another insightful video!
Clive is a genius, and we get to learn!!
There is a limited data sheet for the JW1691H online - 500v max and 5 -80mA. You can get the same constant current effect with a three terminal voltage regulator by connecting the LEDs in series with the Vin terminal and connecting a fixed resisitor that determines the current from Vout to Common.
The advantage of the dedicated chip is that it is rated for a higher voltage range and has internal thermal limiting that will lower the output progressively if it overheats. Many of these lamps do start of at higher intensity and then nudge back gently when they have warmed up.
Thanks so much for your channel !! I watch every episode. One of the most interesting on RUclips!!
What did they use to seal the glass to the board? Looks like the inside of a Crunchie bar :D
Whatever that is, seems like the same thing which is used for normal lightbulbs.
Mmmmm crunchy......
John Carr rOHS approved Crunchie bar
Seen the same gunk inside many a bulb
Robbie Edward's MASSIVE business SayersTM as China make most of these I’m sure the paper trail is solid. 🤔
I gave up trying with LED filaments (granted they were all direct-from-china, but still), the majority of the ones I have/had ended up with broken bond wires that turned them into seizure-inducing flashing lamps as the filaments blinked on and off like a bad Star Trek effect, the rest, well, they got broken due to butterfingers here... :P
The best part of LED filaments is if one LED fails making the whole lamp open circuit and making the smoothing capacitor go bang :D
I use Osram and "v-tac" filament lamps for my whole apartment, all of them running great so far... except for the one i took apart of course
just order 200 at once, they'll pack em right at higher volumes
it is just not worth their time caring about packaging when you make them two cents by ordering five or something
I guess the trick is to find out how many are in the corrugated cardboard boxes they receive them in, and order in whole boxfuls. I've gone off buying super cheap LED bulbs though. I think it's cost effective to get Osram or Philips and actually get 20 000 hours, even at 5x the price.
The only autopsy I enjoy watching.
"some spicy voltages" ... I think I'll use that highly inspirational technical euphemism now. :-)
hey BigClive I have become very interested in electronics because of you. I will picking up a soldering iron in a few days.
But I was curious if you've ever "taken to bits" any older mice or computer game controllers. I have and it was fascinating (even though I didn't know what I was looking at. so many components and very few chips.) I was looking for a video where maybe you did with no luck. any way... If you're reading this thank you for making amazing content.
My son and I were at pizza hut, the other night. They had just remodeled the interior of the joint. I noticed all the lamp fixtures were these types of led lamps. I was like cool I've seen these taken apart by Bigclive.
my local theatre has alot of them in wall panels along the inclined floors going up to the theatre room's entrance.
Okurka cuz our local pizza joint was closed.
All those LEDs in series. Reliable!
Unfortunately, you can't even replace faulty ones. If it would be possible to do that, I'd be interested in this kind of solution, but no way I'm going to buy it like this.
anlumo1 I believe that Mark was being sarcastic...
I'm aware of that. If you could replace faulty ones, it wouldn't be that big of a problem, but alas…
I'm slightly more worried about the reliability of high voltage electrolytics. Especially in that packaging style.
The LEDs are much more likely to die very long before the caps anyways.
Nice video! But.... i have followed your vids for a while and still no bang or slip... the anticipation is killing me!!! =) Keep up the good work sir!
Did you smash the bulb to harvest the 5 good filaments?
I probably will salvage them for re-use.
bigclivedotcom 1 week later... Hilarity ensues
"JW here!"
He does a good impression.
lol
It's actually ".. Hello I'm JW.." tickled me tho 🤣
The 22 Ohm inrush resistor is likely (well that's what I'd design in there for low cost) an NTC component with a 22 Ohm cold rating. Also, I'm surprised the 2 Meg resistors are not balancing the large caps?
Clive I think you need to come up with a super duper disassembly kit.
Interesting. I've just pulled apart a dead Philips dimable LED filament bulb. The driver is all transistor (with possibly a CC IC) The led filaments (three in parallel)require about 90vdc to light . Should have sent it to you Clive, for an autopsy (still can if you want the bits???)
Hah, for a minute there, I thought it was going to open like a Spam can.
Hmm... BANG! "Ooh, spicy soiled shorts. Oh, my."
I was waiting for you to short out the bad strip!
The bad strip's voltage would then be added to the voltage dropped in the current regulator chip, maybe causing it to overheat...
But if you then change the programming resistor to a suitable value, then surely it will work?
Interesting. I've once built an almost identical circuit, but with a darlington transistor and a current divider to set the output current. Except It used 12V AC from a halogen lamp transformer (so about 24V on the LED), not 240V AC (560V on the LED) haha omg that's so dangerous.
absolutely fascinating, thanks clive!
Quite spicy voltages! 😂
What voltage would be the equivalent to a vindaloo?
I am guessing tonight’s voltage is around a bhuna maybe a madras?
Vindaloo = Microwave oven transformer output!
we've had the exact same fault on two of these bulbs (same make), both lasted about 2 months.
I looked it up, the JW1691H now has a datasheet available in English - it's not rated for this application. It's only supposed to go up to 500 volts.
The current regulators usually only see the voltage they are dropping. The bulk will be across the LEDs.
The problem with LED filaments is that you have a lot of tiny LED chips in series. If one chip fails in a filament, the entire filament will likely fail. In this lamp's case, one bad chip has caused the entire lamp to fail as they are all in series (perhaps as a means to reduce current).
I'm a little confused, I don't understand how can the 230V supply charge a capacitor to 335V?
230V RMS (root mean square) is the average value. Multiply it by 1.41 for the peak value of the sinewave.
Ohh okay that makes so much more sense! Thanks for the reply, love the videos and really informative to an electronics student :P
Edison lamps, ribbed for your convenience.
It has a certain visual value in fault state. Could be quite good as Halloween porch light, bit spooky. :)
Clive, I think you should mention the inherent potential for fatal shock with the cap. dropper circuits. Although of course this LED isn't using that method. I was all jazzed about using CD circuit for somethings until I saw another video wherein the YT'er pointed out how it could go wrong in so many ways. I guess its common knowledge in the industry that you really only want to use CD in devices that are completely enclosed, with no external signaling wires or connections do to the fact that many points within the circuit can be at mains or close to mains voltage with respect to User-ground. Just thought it might be worthwhile to mention occasionally in the event some watchers will be like myself and think its a neat, cheap shortcut to a mains powered LV PSU. Which it is, however the clear fatality potential is extremely risky to save a dollar or two.
5:24
Is it me, or does the top e-cap have a burnt/melted spot on the shrinkwrap sleeve?
They had silicone rubber and a dab of the glass to metal cement on them.
Just wondering how a SBC connection would differ? There wouldn't be much room to cram those big capacitors in.
I own several filament LEDS of that particular brand. A lovely 'warm' white but not very reliable due to the Black Dot of Death.
13.33ish..670 volts..quite spicy...oooh how am I going to do this without blowing myself up....classic 🐻
30 seconds of excitement followed by a bang and a mild sense of disappointment.. seems oddly familiar
Ive just had one of those lamps fail (at least it is visually similar) & i'm not sure what caused it. I know it made a little hole in the e27 screw bit & smoked. I was using my electric oven at the time & the circuit breaker on that tripped as well as the lights. The oven has been fine since.
Sounds like the circuitry may have shorted onto the inside of the screw cap.
Hmm, how was the circuit able to charge a capacitor to 335V using a 240V input?
@9:28
240V is the average sinewave voltage. The peak is about 1.4 times that.
I see, I thought 120/240 was the peak voltage of the wave. Is that the difference between peak and RMS?
well more leds in series means same light output, less current and thus less power to dissipate when dropping the voltage through the regulat000r
I wondered if you'd smash it open and bridge out the dead line
Maybe keep a corned beef/spam can key handy? They are also useful as trolley keys.
Thumbs Up for the john ward reference :D
It's so funny hearing you with headphones. "OoOoOo" "One Megom" lol (I know it's Mega ohm)
14:06 I was waiting for an ElectroBOOM moment.
Thanks for sharing 😀👍
Err..... Stupid question. Why does 2*110k=55k?
When two identical value resistors are in parallel the value is halved. If in series the value is doubled.
Okay, thank you. I was unaware and now I am aware.
That's definitely a lot of voltage, so at what ma do each of the LED's then run at? Thanks for another good video Clive. Very interesting to learn all that.
i rekon its a boost converter to boost from 340ish to 670v dc
could that circuit drive argon/mercury tubes?
The circuit shown at the end is suited to short argon mercury tubes. Just a couple of feet though.
bigclivedotcom
I was referring to the lamp power supply
Although I'm thinking that it probably wouldn't work cuz they need a higher voltage start/strike them right?
Maybe give it a shot for science?
Perhaps you could start the tube with your custom power supply then move over to the lamp psu?
Best line:
Okay, we have ways of making you light.
What a strange circuit. LED's work on about 3v, so what does this circuit do? It doubles the already far too high mains input voltage.
how many of your videos end with "indeed"?
I'm not sure. Sometimes it's hard to find a suitable way to end a video. I usually try to end it with a very brief summary.
These days most channels have a standard sentence with “click here to subscribe to my channel and here is the patreon page and please give the video a thumbs up blah blah”. I'm glad that you don't do that.
anlumo1 Totally agree with you. I simply hate those video creators that are reminding and begging for the likes and subscribes sometimes even multiple times during a few minute video - I mean, if I like the video, I will show my appreciation, but those f*cking reminders basically insults every viewers intelligence.
"Keep your dick in a vice."
Surely it should be "always remember to put your meter leads back into the voltage sockets after measuring current"....or something like that....Always brilliant content however they end (smoke and/or flames are preferred but not mandatory).
All the electrical knowledge and impressions as well; first John Ward, then Homer Simpson, doing MmmmMeg Ohms (instead of doughnuts), finally I thought we were going to get Electro Boom, with the spicy voltage and getting the Argon Mercury driver our and 'we have ways of making you light'!
Is that device a glorified current mirror?
The filament LED style might well be the way mainstream LED bulbs are heading. I have noticed both Osram and Phillps are now actually making frosted versions of them and just marketing them as plain bulbs rather than the "heritage" tags they give to the clear envelope versions.
Stands to reason as they can obviously be made in plants set up to make traditional tungsten filament bulbs with relatively little re-tooling. And easier for the general public to assimilate as the end product is once again the "familiar lightbulb" that they know.
I can also see more sinister motives here too. Seems it's easy to make a bulb with so many LEDs in series (especially the one you have, but even the conventional series/parallel types) fail in a relatively short period of time. Maybe once again getting back to the 1000 hour planned obsolescence of the "Lightbulb Cartel"?
Nice John Ward reference btw. :)
I picked up one of those frosted LED filament bulbs a while back from Screwfix. It's still working fine, and it's quite nice because it has basically the same radiation pattern as an old incandescent, unlike many other LED bulbs. They don't seem that common though.
My experience with the Chinese eBay LED filament bulbs is more of a mixed bag. Some smaller ones (two 4W candles and a 4W GLS) are still going strong, but an 8W GLS didn't last long before it started flickering.
They seem to be a fairly new thing over here in NZ, although the clear ones have been around for quite some time. Offered in the standard glass envelope, the "vintage edison" pear style & a host of other silly alternatives.
I have recently found another advantage of them seems to be that they radiate much less RFI than conventional LED bulbs. Which is important to me as I monitor a community radio station that I manage, which is quite some distance away. Of course my results here might be due to the particular ones I have, but I suspect that using the higher voltage and lower current is what mainly offers this benefit.
xanataph Maybe it's the kind of power supply they use - switchmode supplies can produce RF noise due to the high frequencies they run at, and the low current means the filament bulbs can get away without one.
Looks like the one that Clive took apart definitely had a SMPS in it though. But if they need less current it would be possible to use lower switching frequencies and that would help with reducing RFI. And also the lower current too I suppose.
The data sheet for that chip suggests it's a linear regulator, not an SMPS. The high voltage drop across all those series LEDs (about 85% of supply voltage) means that even a linear regulator will be reasonably efficient.
Great and interesting video.
Does the UK use Edison Screw or Bayonet Cap sockets as standard in consumer light bulbs?
Never mind I googled it, it's BC like Australia. It seems though that be are both moving to include ES.
Peter Resch standard is the large bayonet fitting but we see a lot more ES fittings now as we get chinese/European fittings and shops sell bulbs for them. All Ikea’s stuff is ES and they’re a big driver for the changeover. We get small bayonet as well but it’s uncommon.
Thanks, yes that's exactly like what's happening in Australia. Though it seems us colonials are still under the control of the British empire ;)
Peter Resch
Well, you are still part of the Commonwealth.
BC used to be dominant and is probably what a lot of us grew up with... ES used to be a speciality bulb that'd have a few inches of supermarket shelf space at best... now it's probably more like 50/50 with BC. I think Ikea is mainly responsible... everyone else in lighting retail saw they were getting away with selling ES fittings and decided to stop worrying about it.
That is a rather pretty filament lamp though.
clive just a quick question does it matter which way round you fit resistors in a kit build ?
sparky allen...... they don't have a polarity so you are good either way except sideways because that would be uncomfortable.
hahaha thanks clive its all very new to me i built one of them transistor testing kits and it doesnt work so i wondered about polarity thanks bud
Thanks for showing us.u a clever man
Hey Clive, I've been having a difficult time trying to wrap my head around how comparators work, and what I've found online so far hasn't helped much. Do you think you could clarify how they work?
If you mean the simple op-amp style comparator then it compares the voltage at it's two inputs and the output goes high or low depending on which is higher. The plus and minus inputs determine whether the output goes high or low.
Why Bayonet style bulb sockets?
Common in the UK.
Very nice video
More Amperes ! we need more Amperes !
ooh is it the voltage doubler i wonder... watches to see
I've never seen anything run that high, 670 volts is crazy.
Hunter's Moon you’ll love this, I investigated a cheap 2 stroke generator that put out nearly a kv instead of the usual 240v. It was the subject of a complaint from a woman who was running a tv off it when camping. What I found was that her husband had fucked with the carburettor and set the idle speed too high. It was running at over 4500 rpm instead of 3000 and making crazy volts at low load (like when the tv was plugged in). Anyway it fired too many pixies up the poor old telly and the magic smoke got out, much fuss and bother etc. When I tested it with a heater it would run fine at about 500w or above because the governor could regulate the speed properly but at low loads, the throttle couldn’t close far enough to maintain that 3k RPM. Gotta love the bother you see in amateur hour 😆
Don't ask about the voltages used in microwave ovens then... several thousand volts, and low enough output resistance to easily push enough current to kill you.
dingo about 2.5kv usually, and they’ll supply around 400mA which will do a proper job on you. Avoid.
Especially when it’s hooked up to a voltage doubler and some big ass capacitors.
How does 230v charge a capacitor up to 330v?
230V average value of sinewave. 330V peak value of sinewave.
What camera and CCU do you use? You might have mentioned this before but you frequently adjust the settings without reaching up to the camera.
Moto G4, Open Camera adjusting on-screen controls.
I was hoping to get my bigclive fix today, you didn't disappoint, the board looks a real mess, both capacitors are 400v,arent they divisible in series?
Phonotical
They would be if they were bipolar and connected conventionally but these are used in a kind of push - pull configuration, are polar and connected in parallel with the pair of mirrored diodes so each one charges only during one half of the incoming sine wave.
Pete Allum ooooooooookayyyyy
Im sure I remember some horrible maths involved with parallel resistance.
Could you....measure two 100k resistors - solder the ends....predict the resistance with a calculator and then measure it !
It's easy with identical values. Harder with different values in parallel.
Yes. Those horrible memories are coming back now. 100k and 4k7 aaaaarrrrrgggh.
Resistors in series-> resistance adds up
Resistors in parallel-> conductance adds up
conductivity is measured in "siemens", that's the reciprocal of "ohms" and vice versa. 1/siemens = ohm
It's not that horrible, you just need to add the inverses of the resistances (then invert back at the end). So overall resistance = 1 / ( 1/R1 + 1/R2 + 1/R3 + ...)
electronics certificate program DC Circuits semester 1 week 2
The stage in the course where you find out who did and did not actually pass gradeschool math.
Is your neon design in the website?
www.bigclive.com/neon.htm
I disagree about the filament lamps. Instead of relying on a few high quality high wattage LED chips these rely on many cheap LEDs in series not only making dead on arrival much more likely but also weakens the punishment the LEDs can take over time. The fact that it stayed on from the filter cap tells you a lot. There isn't much after the filters so if it dims out the supply is most likely working. The fact that the LEDs are all in series means even small ripples in the AC will kill the LEDs. So the lamps supply needs to make up for that with more regulation and filtering. So touch chance using a capcaitive dropper they must then use a linear regulator. It sounds weird but the higher voltage accomplishes this better because it now requires a doubling of a voltage spike to do the same damage to the LEDs.
Use the neon FART transformer...
That´s AC though, and I think around 8kV of it. So he definitely would need a variac for that.
Not sure if he has one.
Same here. 12 year lifespan, did not last 12 months. And I paid £8 for it! Grrrrrrrrrrrrrrrrrrrrrrrrrrr
Ve haf vays of makink you light, Mr. Bond.
Why voltage double them?
Probably to get a higher duty cycle and decrease flicker.
not sure if circuit above would be full wave -
To allow a higher power lamp with a longer string of LEDs, but at low current.
I guess they did eliminate any type of voltage reducer circuit
Spicy voltages: Creole delights for all electricians.
Is it just me or do led lamps seem to burn out quicker than cfls?
The earlier ones were bad for earlier failures and they definitely don't like hot environments like fully enclosed fixtures. It pays to get lower power lamps and use more of them to create the same level of light.
definitely. I put CFLs in my cabin when I moved here 15 years ago. Just within the last year have some of them failed. ALL of the failures have been the result of the emissive coating on the filaments becoming depleted so that the discharge would not strike.
This long life includes a 145 volt, about 5 seconds long power surge when a utility tech did the wrong thing at a primary voltage regulator. I'm just nerdy enough to keep a Dranetz line disturbance analyzer permanently connected to my breaker box. It recorded the surge to ms accuracy.
I bought an Osram 100 watt equivalent LED bulb at Home Depot. It lasted about a month. I got a new one under warranty. I expect this "premium" LED to fail shortly. The conical shape between the plastic hemisphere and the Edison socket runs too hot to touch. And that's burning in an upright, non-enclosed configuration.
The very best CFLs used a chip that ramped the operating frequency upward. Before resonance, the current lit the filaments red hot. When the frequency hit the resonance of the transformer primary and a capacitor, it generated enough voltage to strike the lamp. Then the chip settled into constant current mode.
But people bitched about the half second delay so now most CFLs use an instant-on design. This type of lamp fires the discharge immediately upon power-up, bombarding the cold filaments with mercury ions, resulting in the emissive coating being blasted off the filaments much faster than had they been hot.
All my long lived CFL use the ramp design. Sold in bulk under the GE trademark at Sam's club.
Newer is seldom better, especially when the new stuff comes from Communist China.
Grossly unrelated. I once put a bare 5mm red old led across the rectified opìutput of a f/art neon transformer hoping for a short flash of glory. The led worked fine.
To be expected. Neon transformers are constant current devices. The voltage rating is only open-circuited. Typically, neon transformers come in 5, 10, 20, 30, 160 and 240 ma rating. I think Clive's is a 20 ma. That should make an old LED quite happy.
2 * 110K = 55K naughty.
It's, 2 resistors of 110K in parallel = 55K (i.e. 110 / 2 = 55)
A brand-new lamp that turns very dim after 30 seconds. Must be made of neodimium... (SCNR)
“These are the kind of lamps we like... DEAD ONES!”
Not heard “Bwahahaha” (evil mad scientist laugh)
Clive a nod to John Ward 11:06...lol
"This is quite spicy voltages"
The Hammer of Knowledge? The Vice of Knowledge?
his hammer has an identity issue and is called X-Ray
Not to be confused with the handymans toolbox consisting of a hammer and a flat blade screwdriver with the secret helpers ducktape and WD40