12V LED filament lamp video disaster.

an LED filament lamp those of you who watch Julian ayelet may actually recognize this of layer two the filaments this of conical shape with strange supports than each filament has a support that has a connection at the base and a connection going up to the top but they both sort of clip into the circuit board and Julian air was going to use his in his air outdoor 12-volt system but I don't have an outdoor 12-volt system so i i'm going to open mine but let's try it first so with 12 volts and the polarity doesn't matter it's a non polarity sensitive at 12 volts it draws approximately 450 milliamps so this thing has rooted six watts that's close enough interestingly if i then turn the voltage down let's say turn it down to the point it starts dipping okay it starts to dip around about the earth which is four volts so if i turn that up to let's say five volts it's drawing about point seven amps let's say what's that working at as paralyzed five volts times one point seven amps equals oh it's come up to about eight point five watts on to how much of that is loss isn't the power supply versus the actual and the LEDs themselves interesting it does suggest that you could use this lamp in a USB power supply know that top is plastic so I don't stand too much of a risk of injury if I do this that is a very odd structure or the quite spring as well that almost gives it a resilience against being dropped or bumped about that's that's quite intriguing and it appears to be staked in or is that hold on let's say gas screwdriver and pry out the art of heat stick to actually jam it in position so here's the circuit board oh right I'm wondering is that a are these parallel arrays of LEDs we can test that here's the circuit board at the bottom it's hot melt i say it's hot melt glue think it's silly can you could be glued in oh or it's a tiny low power supply this is clear designed to go into the base of something like gu10 lamp it's got the inductor it's got a capacitor it's got a diode and there'll be a little trip in the back oh there's a fuller direct fire in the back and then there'll be the switching regulator this a buck regulator okay that's quite interesting it's unfortunately goo has covered the air the chip itself let's say I try and get some of that goo out the way and see if we can read number in the chip I may not be able to read number H at the moment that's okay if I can't read number in the chip now our pause and our find the number because there definitely is a number it doesn't look terribly clear I'm going to have a goal with the little magnifying glass oh it's Excel semi Excel semi and the silicon goo Excel I think it's 6 or 6/1 Yi the silicon goo has actually gone in and it's laser etched and it's kind of I don't know if you can see this let's say try a bigger lens I don't know if you can see that but you can see part of the number but not the whole number it might be that you can see it easier as better than me if you've got a big screen because there you may have more detail that's the parallel of High Definition and you actually get better vision than me yes better picture that is kind of cleared off a bit but it's still not cleared off that well I shall explore that later in the meantime let's just chop this off and get the bench power supply on again and we'll just strip these at all gooey silicon leads the very soft flexible silicon is stuff and let's see what voltage the LED array runs out I reckon it's a parallel array I reckon that as soon as I go above about 2.5 volts and I could be wrong here yeah they should start later maybe completely barking up the wrong tree it made actually boost the voltage so let's set the power supply let's bring in the meter actually let's bring in the media meter because it's got the alligator clips on it a nice contrast from the fluke so I'm guessing this is negative and getting this is positive and let's bring in the power supply which is set quite low and we'll stick that on there and we'll stick that on there and the voltage at the moment is why am I getting voltage there what fired done here Wow that's impressive oh there we go ah right let's turn this up let's zoom down a little but this is the curse when I zoomed down because then you know I'm going to screw up and I'm gonna forget sumed in we're currently at point nine of a vault I would reckon that if these are all in parallel then they're going to start conducting there is something grossly wrong here because I am NOT getting the crack reading for this meter as its battery flat or something is this just a terrible advert for these shitty meters what in Paris is going on there Wow oh there we go that's a unimpressive rate it visits the dial the rotary dial that's exciting that just goes to show you that spending an extra pound does make a difference the voltage is creeping up I'm not getting in late from those LEDs yet I would have thought they'd have been left there in parallel unless the polarities other way around negative is shown negative and positive so I let's keep pumping the voltage up then maybe that is a little boost circuit the voltage is rising the voltage is rising higher and higher and higher and still nothing's lit up and suddenly I'm full of doubt that's quite a high voltage air right okay as a precaution before I go any further I'm gonna swap the part in that in case everything is marked wrong because you just never know let's say I put positive on that I support another maybe to look at the air capacitor reaching the power supply output capacitor rating and see if it's rated quite high or not 25 volts it's not that high let's make a half-assed attempt at gripping this kibble I will start again so voltage going up I would have expect them to be lit at that if it's a right okay so it's not it's not a parallel array of LEDs at all I reckon that these all connected in series they maybe hold on let's say I get there let's get the dadima trim this is probably better idea let's say zoom back again so you don't miss any of this let's do some continuity checks with the daddy meter the Dhamma meter so this one is connecting to there but not to there and that's connecting these six filaments maybe in parallel but at right okay let's see if I can make this gloom gonna have to take this off-camera Maureen Charlie yeah it's glueing I don't know if you're gonna see this the meter is making it glow those are roughly a 3 volt filaments each and it looks like there may be six in series right let's start the experiment again so now let's say get the meter back here stick it and positive because I think that is the correct polarity and the negative and this time we're going to go for gold I'm going to change the arrangement in vitro let's do that while the power is off shall we and turn it on okay so we've got about 3 volts of them all ah let's zoom back down in this and when the voltage up optimistically until I see if current deflection happening here so there's six times two point five or don't let's see what voltage should have expecting here six times two point five so I don't get overzealous equals I reckon at fifteen volts which is very close it's going to start gluing but can I can I screw this video up anymore can I even can I do more wrong with this video ah right what the heck's going on here then let's get that air down a bit there's the polarity correct hold on zoom back Oh again yeah I'm screwing everything up here this is sometimes what happens nothing ever actually happens as you're expecting at least just out of air just out of interest let's do a quick the polarity check here in the mid her so let's put this to continuity I'm just off shot here I'm just double checking that air oh okay well that is not gonna show that so completely the things go wrong that's just what happens sometimes that was the wrong type of LED or tried testing right okay I validate adapt positive should activator should be positive for the LED so let's go onto here and again it right here I'm gonna have to come off short moment charlie uh-huh just bear with me one moment that does appear to if one bath is going on here did I have a bad connection before let's try that again this is just this is all gone horribly wrong and because I've no strip amount of it so I can't actually stop the video I'm just going to have to play one and what is turning into an absolute apocalypse it's the video of where everything went wrong sometimes that happens best lead best made plans and Mason men and August pear-shaped the final piece of icing the cake would be to connect up to their supply here and just blow up all the LEDs accidentally so 10 volts say I'm going to crank the voltage up higher I thought we were enlightened by now that's odd all right okay I think a rather net make it thicker myself anymore I think I'll just pause momentarily and then work out what's happening here and just like that all the problems are resolved and as Eric oh it's so it's mean automate see there's your problem lady two of the filaments in my beer leg force that I used to get this cover off actually snapped these are tops off the filaments where the connections go in the top so they're quite fragile that way just be aware of this if you ever want to open one of your own lamps it can break them if you do you have to just flex top sight to see and it'll part and then shut them out there soldier underneath so we're back and I also took the opportunity to find the data sheet which is the Excel 6 0 0 1 and is to quote a wide a 3 point 6 to 24 volt input a constant current sense driver for driving between three to eight series one more LEDs and that makes sense it's got a fix to 400 kilohertz switching frequency and it's effectively treating each of these filaments these are parallel it's treating them as a single LED but there's actually a parallel array and that placement of LEDs in them is very random I don't know if you can see that bear if we zoom down in that azumed I'm frog was a bit weird you'll see at the air the spacing in them is a bit sort of it looks at barcode hmm it's a bit strange but it's treating each of them as a 1 watt LED and why naman series and then runs them at a current this program will buy a resistor so I'm going to wrap this up now as you can see that the voltage across it is typical of it's just gone current limiting at eleven point three volts three hundred milliamps to set the current limiter so that is yeah just treating them as a 1 watt LED running them both hundred 50 milliamps or so that's quite interesting let's turn that off let's turn this cildren off that doesn't make a loud buzzing noise in the background and let's take a look at the circuit here that is very clearly the sort of power supply you'd find in the base force of mr16 e-type lamp I probably said gu10 over run yeah mr16 a table lamp which will driver an array of well it can drive from a 3 watt to the heat what I've arranged array of LEDs the schematic we don't have we don't have to worry too much about the block diagram here it's a very simple chip well it's a it's very simple all the work has done by been done by Excel semi it's the supply comes in that's good there it actually has a rectifier before this it's got a couple of smoothing capacitors its powers of the chip at pin 2 and it's also a reference and seven and eight to ground the reason it's got two pins for the current switch and the connection to ground is just to actually share the current between them it's got the inductor in that case this chunky little inductor here watch it rated 470 micro na or is that 47 micro Henry don't have something to test that for seven zero let's just presume that it may to actually just be a three digit code Zuba 47 micro Henry which could kind of make sense it does show a 22 microhenry here that makes more sense it's got the diode which is next inductor here and then the output capacitor and that's fundamentally at what it's doing here is it's pulsing this a head inductor to put a build-up a magnetic field in it and then when it turns off that field collapse isn't it sort of adds to the input voltage and go through that diode and push the voltage up to drive the LED the LED has a current sense resistor whereas that where's the current sense resistor I would expect to be near the output could it be these resistors here one more while I just magnify point naome and 1 ohm 10 ohm oh that would be quite accurate it was a raid like so let's scrub those it's a bit tear it's got more goop on it yeah it looks like a 10 ohm and let's say show you that and then lower value so I'm guessing they might just be fine-tuning the value there to get what they want and the idea is that when the voltage exceeds a certain threshold across this resistor the cent resistor it then starts cutting back and regulating the current a bit so that way it can actually sense current flowing through the LEDs it's an interesting variation of I've never seen this particular structure there's a mounting arrangement of the filaments before where they're parallel array on a ceramic base with the hole through for that disorder through they've got their layer of the force from the back as well so any of the blue lake roofing roofing LEDs through the back through the mm the ceramic layer does also contribute the sort of warm white layer as a push to the cold white or the dude Cree if it was just a bare blue net so on but most the light will come out one side you can actually see a slight ripple on the gel here where it's got the LED chips underneath it it's an interesting structure it's quite neat it's an interesting way of driving them just treating each one what led and why in the mall and series that takes me by surprise because I I really thought there on parallel as you saw at the beginning the video which had dragged for me by the way because everything was going wrong that's what happens when they're when videos go wrong it's almost like time stands still actually it's not like time stands still it's like it just seems like it's taken ages every screw up actually seems like it took minutes I'm not sure how long it did take hmm I'll find out and look at the video but there we go that is a that's what's inside these lamps don't take them to bits they're very very fragile but you could theoretically just scavenge if you can pop one open carefully you can then scavenge the filament so I guess you could carefully disorder these a wire support form then just put a couple of wires and if you wanted or create your own sort of frame and they are just effectively three volts a filament each them up to about 350 milliamps so quite neat useful as a source just for those filaments ah this video has just redeemed itself in a very interesting way because I took a closer look at one of the filaments I cut in new and off and I well initially I got one of the broken ones in a scrape the phosphor off and I was thinking it would be lots of little chips as bond wires on someone's it's not so I actually and screaming the phosphor off all the LEDs came off too and left no sign other than little bus bars with solder on them so took earlier another one and carefully picked away the the phosphor loaded gel around one the LEDs and here's the construction so we got ceramic and now how many LEDs would be half in here we've got one two three four five six seven eight nine 10 11 12 13 14 15 16 17 18 19 20 about 22 LEDs in there and each LED is mounted onto a little bus bar down in the middle so if this is the ceramic airstrip we've got two little sort of silver colored bus bars coming down the full length of LEDs and parlo and everywhere in LEDs mounted on it there's this really neat little perfect package at least even blue I've powered it up and looked at it through the microscope at low level and it's physically just soldered there's no bond wires no jumping wires anywhere it's just physically soldered straight across those bus bars so whatever machine is placed them is just basically sold him like tiny little LEDs along that bus bar lots and lots them with a rough spacing to give you an idea of size here glasses off I have to take your glasses off to get the microscope down and then a hold on just give me a second here bear with me Oh blame it's really wide space compared to LEDs right here so it's equivalent to about that spacing that's the size they are so if you imagine that you know this strip is about the 22 LEDs along it this is a spacing you can imagine what size they are and as I say I was kind of expecting it to have those air you know like a chip a pad with a chip and then a bond wire jump across then another bond wire green top and this is just so streamlined from mass-production this miss must be also how they're probably making the large aircar berries as well that is a very interesting evolution strange I wonder why that when you look at the LED if I drew a bit bigger it's got these of electrode dots like that and then a couple of other end as well and from this one there's a little stripe which is presumably a surface electrode and these ones too so I wonder if these are a black edge just connecting through from oh there's a thing that's probable is these ones will be connecting through to one substrate that right no how would it can I then connect no I'm actually talking crap no I was trying to work out how it connects from the bus bar down below up to that but without actually going in like super microscope stuff and and really sort of sort of reverse engineering it with like lab grade equipment and not sure how you could actually determine that but I'm guessing that these beer darts the room will be one and then these will be in the surface these are the other layer to actually spread the current over the surface for even illumination that's very interesting it's a new twist in the serve cob chip on board type a really DS I've just never seen before that's very interesting

Channel: bigclivedotcom

Views: 142,230

Rating: 4.862833 out of 5

Keywords: LED, lamp, low, voltage, filament, cob, ac, dc, 12v, 5v

Id: pu6usLoz2GM

Length: 22min 58sec (1378 seconds)

Published: Wed Dec 12 2018