Because of you, today i was able to fix my bulb with exact same issue, Thank you sir !! , Randomly RUclips recommended this video for me and the idea came to my mind to check, nothing was planned
Wonderful, interesting video! If all videos were as clear, concise, void of annoying background music as this one is, RUclips would be a FAR better place. Well done, sir! 👍
Very informative video!! Especially showing the use of epoxy to stabilise the part so you were able to expose the interior structure of the inductor. I'll use that technique myself in the future. Well done, Sir.
I'm starting to think we ought to run 12VDC to our light sockets and have one big converter for the whole house. They already make 12V lighting for houseboats and caravans/motorhomes.
Oh I thought I was the only one who liked to do that kind of component analysis, somehow find it so satisfying to sand down the components and look at the cross sections with a microscope. You can use higher magnifications say 100x and stitch a larger picture, they are great as nerdy high resolution posters. Maybe its dumb but I was giggling at the sounds of the sanding fastforward. Subscribed!
I'm currently working on 8 more videos, 1 of which is almost done, unfortunately they take a lot of time to make, especially seeing as I only work on them on the occasional weekend or after my day job.
It’s very interesting to me, because I’m a retired submarine mechanic who worked on humongous parts from sonar dome to rudder & stern diving plans. But, in my senior years it’s the miniature/micro things that are of interest.
I just had a 9w Bayonet Smart bulb fail on me. It suddenly started flickering like it was trying to come on, but wasn't able to. Thought I'd take a look and see if I could fix it. I wasn't sure how to open them and came across this video. Thanks to this video, I was able to take it apart with relative ease. Turns out the issue was a bulging cap on the power supply board. It was a 10V 470uF cap. It was immediately noticeable. I have a bunch of power supply boards (some from TVs and and several ATX ones) laying around. Looked through the ATX PSU 1st and saw a 16V 470uF cap. Looked up if you can put a higher voltage one in and seems like everyone said it's fine as long as it fits. The 16V cap is larger than the one I took off, but the extra size was not a problem as there was more than enough space for it. So I de-soldered that cap from the ATX PSU and soldered it to the bulb PSU board. Put everything back together, fitted the bulb and it now works fine. No more flickering and it appears in the app again (it was showing offline before). All colours work and there's no flickering. So seems like it did the job. I'll have to see how long this cap lasts as it was taken from a faulty ATX PSU. I never knew these bulbs were fixable (thought you couldn't take them apart without damaging them). Nice to know these are fixable.
This video is exactly why I trawl youtube for interesting content. Smart guy, fantastic from start to finish. I'm subscribing hope it helps you make more great content.
I've learned from you. I have an LED light bulb on my Electronics Laboratory Bench. It's been flickering. This is the third video I've watched on LED light bulb repair so I can learn. Thank you 😊.
Awesome work, this should be in public schools under technology and sustained development, not some whiny kids and junkies sticking themself to art. Real electric magic!
Very interesting and creative approach. I often wondered when things just get tossed away in other videos - what about an "autopsy" to see what went wrong in that faulty component? And that's just what we get here. Thanks a lot for the vid., we can't always get the "smoking gun", but the process in attempting, and the theories, are still very interesting!
First time I've seen inside of one of those resistors. Also if you want to reduce the amount of heat remove the plastic diffuser or cut the end off of it. Another thing you can do is drill vent holes into the section below the LEDs so they've vent out. Apparently the LEDs can last basically forever if they can cool off sufficiently during operation.
nice tips - found a black spot of LED death on two bulbs with 12 LEDs each. Scraped off the dead LED down to bare metal, a blob of solder later - good to go :) Since they are in a decent shaded enclosure I left the diffusers off so they cool more efficiently now
You'll be driving the remaining LED's harder due to the missing one I think. Better to replace if you can. Although with the extra ventilation it probably won't matter. I'd be a little cautious running it without the diffuser.
What I did with one of these that had the same problem is take a few thin wires from a cable and wrap them around the negative positive Wnd it worked if it blows again try putting in a few more wires and it should last and work. great video thx.
What a deft presentation it is ! Wonderful video. Thanks Sir !! In the present LED bulbs there are no capacitors,condensers etc. All the parts are snugly mounted on an aluminium plate. So kindly teach how the modern LED bulb working pl.
waw masyaAllah, what a deep analysis. Thanks for the explanation! I really enjoy watching the video for 12 minutes. also, I got the other methods for checking the component failures. Thanks Wizard Team!❤
So, I found one of those large LED Spot Light with large heat sink in the back that they put on truck on the side of a road and I wonder how you would proceed to retrieve and reuse the LEDs and other parts.
Thank You Tim. Very informative and clear. We need such good quality videos to reduce waste and get most of Earth's resources. Lesson for those littering the U-Tube with Rubbish.
The failure analysis reminds me of the company I worked for. They would do this to the chips they manufactured but they used diamond paste to get through the individual layers of the chip and a SEM with an Xray analysis to determine the contaminates. At home I just throw away the bulbs when they fail. The local grocery store just had 60 Watt dimmable LED's for $0.50 USD each so besides the learning it is not worth the time to mess around with.
HAHAHA were same brother . i worked on engineering department in vishay, at wirebonding do like this and coss-sectioned on it after that do SEM (scanning electorn microscope) and an x-ray.
Some things that give it away it's an inductor: - PCB silkscreen has it marked as "L1" ("L" in honor of the physicist Heinrich Lenz) and an inductor symbol between the pins (easiest way to know it's an inductor) - Axial through hole inductors are often that light blue colour and are much wider than resistors with a slightly lumpy mid-section - The colour bands decode to 3 mH which is a relatively large inductance value, but makes sense in the circuit (if it were a resistor it would be 3k Ohms 10% which would limit the current too much and dissipate about 1.5 W of the 5.5 W of this LED bulb) - It would measure almost 0 ohms but about 3 mH on an LCR meter (if it wasn't broken) - The cross section shows it has a ferrite core and low resistance copper wire wound around it, a wire wound resistor would use a nichrome alloy wire wound around a ceramic core.
That was awesome, it would've been nice if you showed your multimeter on the screen so for people with basic knowledge understand how you measured the LED and the inductor. Btw, what is the difference between the inductor and resistance? can we use resistance instead? and how can we tell since both look similar?
Agreed, I even have a special bench-top multimeter than I can screen capture from, but completely forgot when recording this video to do that! In the new videos on this channel I have thankfully remembered (so far) to record the multimeter and put it in the video. The inductor smooths out the noise and voltage spikes on the AC mains to protect the LED circuit in the bulb but also prevent noise from the LED circuit going back out into the AC mains and interfering with other stuff in your house, it does this in combination with the capacitor, this is called an 'LC filter'. It does this by absorbing the voltage spikes in it's magnetic field. A resistor often used for a similar job in combination with a capacitor this is called a 'RC filter', but they have no magnetic field so instead convert energy into heat, thus it would get extremely hot and make the bulb very inefficient. It's often difficult to tell the difference between those axial inductors and resistors but typically you can tell based on the shape, inductors are usually shorter and wider than most resistors. Also ones like the brown one the coating is thin enough you can just make out the windings. In this example it's easier because there's a little inductor symbol on the PCB between the pins and it's labeled 'L1'.
@@WizardTim Thanks a lot, that was a very good detailed explanation. I'm surprised how no video I have watched for the past week trying to learn electronics explained or even mentioned inductors. I even googled inductors vs resistance and I didn't get a clear answer or picture comparison. I just bought a new multimeter so I'm excited to try and test these components myself.
@@mhnoni Inductors aren't used anywhere near as much as resistors or capacitors so they're usually not taught until later in electronics. You'll probably understand it better if you think of inductors as the compliment to a capacitor rather than a resistor. - Capacitors try to smooth *voltage* by storing energy in their *electric field* - Inductors try to smooth *current* by storing energy in their *magnetic field* - Resistors *limit current* by dissipating energy as *heat* Also you'll find almost all multimeters can't measure inductance, usually you're measuring resistors, capacitors or diodes, you rarely need to measure the inductance of an inductor so they don't include that feature most of the time, in this video I was just measuring the resistance of the inductor to check it wasn't broken internally which of course it was broken. I have an LCR meter to precisely measure inductors, capacitors, resistors and all their characteristics, but you only really need that when designing more advanced circuits.
Some things that give it away it's an inductor: - PCB overlay has it marked as "L1" and an inductor symbol between the pins - Axial through hole inductors are often that light blue colour and the colour bands decode to a sensible inductance value - It doesn't make sense for their to be a resistor in that place in the circuit but an inductor makes sense - The cross section shows it has a ferrite core and low resistance copper wire wound around it, a wire wound resistor would use a nichrome alloy wire wound around a ceramic core.
Some things that give it away it's an inductor: - PCB overlay has it marked as "L1" and an inductor symbol between the pins - Axial through hole inductors are often that light blue colour and the colour bands decode to a sensible inductance value - The cross section shows it has a ferrite core and low resistance copper wire wound around it, a wire wound resistor would use a nichrome alloy wire wound around a ceramic core. The colour bands are very similar to resistors, the multiplier is the only difference on inductors, there's plenty of online colour band decoders that make it very easy.
Very nice analisys of the damaged part. Would be nice that would give a detail on the voltage applied to the leds as you go through. Wich Is It? Thanks!
I had two Osram GU10 style lamps that failed after a while also with a similar open circuit inductor. They had a silicone encapsulated PCB, that didn't seem to improve reliability.
Interesting, I've heard the silicone encapsulation is meant to improve thermal conductivity to the outer shell and eliminate joint vibration failures, not sure about high thermal cycling. I'm tempted to buy a new one from 2022 to see if they've changed anything since the 2019 design.
Thank u for the informative video What did u use to test the individual led ? I use 2 1.5v batteries in series ( 3v ) to test but was not able to get any led to turn on.
The LEDs in this light bulb are special, they have three LED dies in one package so have a forward voltage of between 8.4 and 9.8 V, typically 9.5 V. To test them I used a Keithley DMM6500 bench multimeter in diode mode, unlike a typical handheld multimeter it has a relatively high diode test voltage of 7 to 12 V and can do higher test currents. But you can do the same test with a lab PSU at a similar voltage (>6V for those 3 die LEDs) in constant current mode set to just a couple of milliamperes. You could also do this with a 12 V battery, just make sure you use a resistor to limit the current otherwise you'll break the LED.
I used a Keithley DMM6500 bench multimeter in diode mode, unlike a typical handheld multimeter it has a relatively high diode test voltage of 7 to 12 V and can do higher test currents. But you can do the same test with a lab PSU at a similar voltage (>6V for those 3 die LEDs) in constant current mode set to just a couple of milliamperes.
Is anyone knows what's LED installed in a DLP LED projectors? They eats 0.6-3V and 1-3A. There are two wires connected-- + and -. When I tried to connect a 1.5V (AC\DC adapter with 3A) to one of them it does nothing. When I tried to check it by multimeter--then nothing. Regular diodes show some Ohms normally. But these one not.
I used to repair the old filament bulbs by holding the bulb at at angle to allow the broken wires to touch then apply power. Often it would fuse and you could get a lot more use
VERY interesting. especially on the inductor. likely scratching away the paint might have exposed the wire enough to add flux and resolder. as a first point of approach. But of course spending many dollars on a 50c component is not worth it. but saving the LED lamp was worth it; IF the hours are not more valuable elsewhere.
I would imagine the prestress placed on component during chinsy assembly, where they just handbodge it in place, will do them in with thermal cycling. Normally you are not allowed to bend the leads close to package and especially don't do so without proper tools. Then they bend or stretch legs to fit the hole distance, putting a constant stress in it, then they solder at 0mm distance from package end, always a no no typically for axial components. Well, they break every rule in the books. The board designer has magnificent magical stand off by the power of a fixed coordinate system, the assembly guys face gravity, time constraints and corner cuts. In todays outsourced manufacturing the left hemisphere isn't connecting to the right hemisphere, usually profit will sever that vital connection.
The DURIS E 2835 series LED chip that's in this bulb is rated to have a forward voltage of between 8.4 and 9.8 V, typically 9.5 V. To test them I used a Keithley DMM6500 bench multimeter in diode mode, unlike a typical handheld multimeter it has a relatively high diode test voltage of 7 to 12 V and can do higher test currents. But you can do the same test with a lab PSU at a similar voltage (>6V for those 3 die LEDs) in constant current mode set to just a couple of milliamperes.
Great idea to epoxy it. I wonder if the leads were thinner or more flexible if it would help, or its pure thermal cycling of the component itself as you said.
This inductor most likely got damaged due to heat and frequency. Ceramic capacitors get damaged similarly, especially in high frequency circuits. To prevent this, can be used e.g. two connected in parallel. They simply crack. They look functional but under a microscope you can see that they are cracked.
Nice repair and effort to spot the fault. I was a bit surprised you just didn't chip off the coating where the creak was to maybe see the broken wire. If it was because it got too hot, I think you would have seen burnt or discolored coating. So it had to be a broken wire... Just my opinion.
I've tried to chip off the coating on other components before, it's very difficult to do so without damaging what's under it so I wouldn't be confident if I found a broken wire to say if it was the failure point or if I had just broken it. If I were to do this again I would use an acid to dissolve the coating.
Im trying to fix mine, 24 led,s on a disc, resistance test shows continuity. The circuit board doesn't have a resistor like yours. it has 2 things the same as yours, a capacitor (no bulging) and the square redish parcel. it has 7 small black squares . it has a small glass fuse which is also good. Frustratingly weird cos I cant see the fault. Kanlux GU10 3.6 w
I assume you've tested the 24 LEDs individually like I showed in the video to make sure they all still work, if one doesn't work and they're all wired in series the bulb won't light at all, I would imagine this is the most likely cause of it not working. The 'square redish parcel' in your bulb is most likely a polyester X class capacitor being used in a 'capacitive dropper' power supply which is a very different design to the LED bulb in this video, if this is the case it should be a bit bigger than the one in my LED bulb (guessing it will be around 470 nF for a 3.6W bulb). If you have a multi-meter that measures capacitance you can test to see if the capacitor is still good based on the capacitance marking on the side of it (you may have to remove it from the PCB to test it 'out of circuit' to get an accurate reading), those types of film capacitors often degrade over time losing their capacitance which causes the bulb to be dimmer until it stops working completely. The 7 small black squares are likely a combination of diodes and resistors, those are unlikely to be the fault but you can test those with a multi-meter.
@WizardTim Hi. All led,s tested including on the back of the disk, the 2 solderd points. The x capacitor says cbb22 684J 400v. On 2000k test it says 550. I can't get any reading on the ų value. The other capacitor is a 4.7 ųf 400v and reads on 2000k 460 and again no reading on the ų value, trying from 20ų to 200m. The wires from the pcb to the bayonet prongs are good. Am i right in thinking if i connected a 3.6w power source to the back of the led disk they should light up if they are good? Thanks, John
@@knaptonmawson The 684J marking means the capacitor is "68 with 4 zeros" 680000 pf (680 nF). Your multi-meter may not be able to measure the capacitor as it's too small for it, it would appear as "0000.68" in the 20u range, assuming you have 2 decimal places on the LCD. I think your options are to de-solder the X class capacitor to measure it out of circuit on a better multi-meter or just assume it's bad and see if replacing it with an exactly equivalent capacitor will make it work again. I would recommend visiting an electronics hobby shop or even a makerspace if you have one nearby, you might be able to buy the right capacitor there or someone there might be able to test the capacitor for you but also help you test it safely. Here's a video that explains this type of capacitive dropper power supply: ruclips.net/video/-n22cqZkxNQ/видео.html To light up the LED disk you'll need the correct voltage and enough current rather than just any power source with enough watts, for 24 LEDs in series you'd need something between 40 and 100 V DC but you will also need to limit the current otherwise it will draw too much current and blow up instantly.
@WizardTim Thanks a bunch, i will consider the possibility of replacing the x capacitor. I hate throwing things away that looks simple enough to repair .(if you know what you are doing) I do understand that you need resistors when using leds, so you dont just instantly blow them. John.
Some things that give it away it's an inductor: - PCB overlay has it marked as "L1" and an inductor symbol between the pins - Axial through hole inductors are often that light blue colour and the colour bands decode to a sensible inductance value - It doesn't make sense for their to be a resistor in that place in the circuit but an inductor makes sense - The cross section shows it has a ferrite core and low resistance copper wire wound around it, a wire wound resistor would use a nichrome alloy wire wound around a ceramic core.
You might be able to carefully insert a thin pry tool between the lens and outer shell and then push it back and forth getting deeper every so often to try and cut any silicone adhesive, gentle heat may also help. Although you might find the lens is glued on with a stronger glue or ultrasonically welded which makes opening them without damage impossible.
In this video I used a Keithley DMM6500 bench multimeter in diode mode, unlike a typical handheld multimeter it has a relatively high diode test voltage of 7 to 12 V and can do higher test currents. But you can do the same test with a lab PSU at a similar voltage (>6V for those 3 die LEDs) in constant current mode set to just a couple of milliamperes, just be careful of the PSU's output capacitance damaging the LEDs if you set the voltage too high, adding a resistor will help prevent that.
Hi, the multimeter is in diode mode when lighting up the LEDs, continuity mode usually won't supply enough current or voltage to make white LEDs light up. However additionally those LEDs aren't the typical single die white LEDs, they're OSRAM DURIS E 2835 series LEDs so they have a rated forward voltage of between 8.4 and 9.8 V, typically 9.5 V. So your multimeter has to have a rather high diode test voltage. In this video I used a Keithley DMM6500 bench multimeter in diode mode, unlike a typical handheld multimeter it has a relatively high diode test voltage of 7 to 12 V and can do higher test currents. But you can do the same test with a lab PSU at a similar voltage (>6V for those 3 die LEDs) in constant current mode set to just a couple of milliamperes, just be careful of the PSU's output capacitance damaging the LEDs if you set the voltage too high.
Yeah, it's pretty interesting that they look hexagonal, however looking at them under a 100x microscope they are much more circular, it's difficult to tell but I think they may be deformed as they're wound onto the inductor core.
The inductor resists sudden changes in current by storing energy in it's magnetic field, this is useful for EMI/noise suppression in this LED bulb as it dampens the current spikes from the switching power supply from being conducted back into the house mains cables and possibly interfering with other devices or radiating as a radio signal. It can also help smooth the AC ripple on the DC bulk capacitor as well as adjust the power factor but it's main use is probably for basic EMI suppression (higher power devices will have proper "common and differential mode" EMI suppression filters).
Had no earthly idea where you were heading with the epoxy trick. That's cool! I'm impressed.
Fantastic how deep you went into details, it's a nice lesson of how things are really inside. Congrats!
Because of you, today i was able to fix my bulb with exact same issue, Thank you sir !! , Randomly RUclips recommended this video for me and the idea came to my mind to check, nothing was planned
Wonderful, interesting video! If all videos were as clear, concise, void of annoying background music as this one is, RUclips would be a FAR better place. Well done, sir! 👍
I agree whole heartedly. First class content & production..
Had to subscribe. Love the way you put this all together. No annoying back ground noise or music either
Very informative video!! Especially showing the use of epoxy to stabilise the part so you were able to expose the interior structure of the inductor. I'll use that technique myself in the future. Well done, Sir.
That epoxy trick is clever! Fantastic analysis all the way through. Those AD/DC converter boards almost always fail before the LEDs.
I'm starting to think we ought to run 12VDC to our light sockets and have one big converter for the whole house. They already make 12V lighting for houseboats and caravans/motorhomes.
@@charleslambert3368 Exactly what I have been thinking for a while now. I believe we will start seeing these in near future.
@@charleslambert3368, people do that already. Sometimes 12V, sometimes 24V (or maybe even 48V).
Came for the bulb, stayed for the inductor and the nice pictures! Awesome and very interesting video, thank you!
It's not an inductor, it's a resistor.
Wow... This 2K epoxy + sanding is a great idea to analyze a part, thanks for sharing! :-)
Im absolutely impressed. The mulfunction analysis, mainly. Good job. Thank you.
Oh I thought I was the only one who liked to do that kind of component analysis, somehow find it so satisfying to sand down the components and look at the cross sections with a microscope. You can use higher magnifications say 100x and stitch a larger picture, they are great as nerdy high resolution posters. Maybe its dumb but I was giggling at the sounds of the sanding fastforward. Subscribed!
So nice to find one of these videos with speech, and also well worth watching. Thanks for sharing.
Wow this went way deeper than I expected. Subbed. I never thought of using resin for something like this, it's genius.
Hey man, great stuff, you need to make more videos, your analysis is awesome!
I'm currently working on 8 more videos, 1 of which is almost done, unfortunately they take a lot of time to make, especially seeing as I only work on them on the occasional weekend or after my day job.
It’s very interesting to me, because I’m a retired submarine mechanic who worked on humongous parts from sonar dome to rudder & stern diving plans. But, in my senior years it’s the miniature/micro things that are of interest.
I just had a 9w Bayonet Smart bulb fail on me. It suddenly started flickering like it was trying to come on, but wasn't able to. Thought I'd take a look and see if I could fix it. I wasn't sure how to open them and came across this video. Thanks to this video, I was able to take it apart with relative ease. Turns out the issue was a bulging cap on the power supply board. It was a 10V 470uF cap. It was immediately noticeable. I have a bunch of power supply boards (some from TVs and and several ATX ones) laying around. Looked through the ATX PSU 1st and saw a 16V 470uF cap. Looked up if you can put a higher voltage one in and seems like everyone said it's fine as long as it fits. The 16V cap is larger than the one I took off, but the extra size was not a problem as there was more than enough space for it.
So I de-soldered that cap from the ATX PSU and soldered it to the bulb PSU board. Put everything back together, fitted the bulb and it now works fine. No more flickering and it appears in the app again (it was showing offline before). All colours work and there's no flickering. So seems like it did the job. I'll have to see how long this cap lasts as it was taken from a faulty ATX PSU.
I never knew these bulbs were fixable (thought you couldn't take them apart without damaging them). Nice to know these are fixable.
This video is exactly why I trawl youtube for interesting content. Smart guy, fantastic from start to finish. I'm subscribing hope it helps you make more great content.
Ohh my God, I just blew away how deep you went into the fixing and investigating the component. It felt like watching a crime mystery movie..😅😅
I've learned from you. I have an LED light bulb on my Electronics Laboratory Bench. It's been flickering. This is the third video I've watched on LED light bulb repair so I can learn. Thank you 😊.
Awesome work, this should be in public schools under technology and sustained development, not some whiny kids and junkies sticking themself to art. Real electric magic!
Wow , amazing video. Thanks. Please make such Quality videos again and show us how to repair different types of issues in LED light bulbs
Man, what a mystery solving of tech video this is! Definitely a like and subscription on my side
Fantastico mai visto ottimi dettagli, bravo
Very interesting and creative approach. I often wondered when things just get tossed away in other videos - what about an "autopsy" to see what went wrong in that faulty component?
And that's just what we get here.
Thanks a lot for the vid., we can't always get the "smoking gun", but the process in attempting, and the theories, are still very interesting!
Thanks for the teardown of the inductor -- that alone would have warranted a full video!
It's not an inductor, it's a resistor
subscribed from this. it is genuine content that I'm looking for.
Very cool, component analysis was awesome, Thanks
First time I've seen inside of one of those resistors. Also if you want to reduce the amount of heat remove the plastic diffuser or cut the end off of it. Another thing you can do is drill vent holes into the section below the LEDs so they've vent out. Apparently the LEDs can last basically forever if they can cool off sufficiently during operation.
it is not a resistor though
Love the detailed teardown
Keep up posting. Quality content!
nice tips - found a black spot of LED death on two bulbs with 12 LEDs each. Scraped off the dead LED down to bare metal, a blob of solder later - good to go :) Since they are in a decent shaded enclosure I left the diffusers off so they cool more efficiently now
You'll be driving the remaining LED's harder due to the missing one I think. Better to replace if you can. Although with the extra ventilation it probably won't matter. I'd be a little cautious running it without the diffuser.
That's a bad idea, the bulb is gonna fry in very few uses
What I did with one of these that had the same problem is take a few thin wires from a cable and wrap them around the negative positive
Wnd it worked if it blows again try putting in a few more wires and it should last and work. great video thx.
Thanks Tim for this informative video on faulty Led drivers. I am very interested in Led troubleshooting explanation type videos many thanks again.
What a deft presentation it is ! Wonderful video. Thanks Sir !!
In the present LED bulbs there are no capacitors,condensers etc.
All the parts are snugly mounted on an aluminium plate.
So kindly teach how the modern LED bulb working pl.
Don’t ever let this guy investigate the crime you committed.
waw masyaAllah, what a deep analysis.
Thanks for the explanation! I really enjoy watching the video for 12 minutes. also, I got the other methods for checking the component failures. Thanks Wizard Team!❤
Thanks for your hard work why resistor fail and good job led repair explained , thumbs up ^^
the final close up shot was great.
I watched the whole video. I can't believe you only got 4xx followers only! This channel is going to shoot up soon.
Thanks! I had 335 subscribers yesterday before posting my last video!
WizardTim I am telling you it's going to be big. Great job brother
Hay I jist became # 737 !
So passionate and patient.
Excellent analysis! Thank you!
Brilliant video. Very informative. Thanks
Bravo. Can’t believe you got that cross section. I would have just picked it apart and looked for a break or short…
this is youtube at it´s best!! Thanks!
really good failure analysis
Fantastic effort. Thank you
When I fix these Bulbs I normally increase the value of the current sensing resistor. So the bulb runs a little cooler.
Damn it I loved the video ❤ bravo , quite a fine job
really quality content! will definitely steal your epoxy technique
So, I found one of those large LED Spot Light with large heat sink in the back that they put on truck on the side of a road and I wonder how you would proceed to retrieve and reuse the LEDs and other parts.
Thank You Tim. Very informative and clear. We need such good quality videos to reduce waste and get most of Earth's resources. Lesson for those littering the U-Tube with Rubbish.
Extremely interesting analysis
Thanks for the info
The failure analysis reminds me of the company I worked for. They would do this to the chips they manufactured but they used diamond paste to get through the individual layers of the chip and a SEM with an Xray analysis to determine the contaminates. At home I just throw away the bulbs when they fail. The local grocery store just had 60 Watt dimmable LED's for $0.50 USD each so besides the learning it is not worth the time to mess around with.
HAHAHA were same brother . i worked on engineering department in vishay, at wirebonding do like this and coss-sectioned on it after that do SEM (scanning electorn microscope) and an x-ray.
As a semi-noob, how did you know that was an inductor and not a resistor? Really amazed seeing the internals of it.
Some things that give it away it's an inductor:
- PCB silkscreen has it marked as "L1" ("L" in honor of the physicist Heinrich Lenz) and an inductor symbol between the pins (easiest way to know it's an inductor)
- Axial through hole inductors are often that light blue colour and are much wider than resistors with a slightly lumpy mid-section
- The colour bands decode to 3 mH which is a relatively large inductance value, but makes sense in the circuit (if it were a resistor it would be 3k Ohms 10% which would limit the current too much and dissipate about 1.5 W of the 5.5 W of this LED bulb)
- It would measure almost 0 ohms but about 3 mH on an LCR meter (if it wasn't broken)
- The cross section shows it has a ferrite core and low resistance copper wire wound around it, a wire wound resistor would use a nichrome alloy wire wound around a ceramic core.
That was awesome, it would've been nice if you showed your multimeter on the screen so for people with basic knowledge understand how you measured the LED and the inductor.
Btw, what is the difference between the inductor and resistance? can we use resistance instead? and how can we tell since both look similar?
Agreed, I even have a special bench-top multimeter than I can screen capture from, but completely forgot when recording this video to do that! In the new videos on this channel I have thankfully remembered (so far) to record the multimeter and put it in the video.
The inductor smooths out the noise and voltage spikes on the AC mains to protect the LED circuit in the bulb but also prevent noise from the LED circuit going back out into the AC mains and interfering with other stuff in your house, it does this in combination with the capacitor, this is called an 'LC filter'. It does this by absorbing the voltage spikes in it's magnetic field. A resistor often used for a similar job in combination with a capacitor this is called a 'RC filter', but they have no magnetic field so instead convert energy into heat, thus it would get extremely hot and make the bulb very inefficient.
It's often difficult to tell the difference between those axial inductors and resistors but typically you can tell based on the shape, inductors are usually shorter and wider than most resistors. Also ones like the brown one the coating is thin enough you can just make out the windings. In this example it's easier because there's a little inductor symbol on the PCB between the pins and it's labeled 'L1'.
@@WizardTim Thanks a lot, that was a very good detailed explanation. I'm surprised how no video I have watched for the past week trying to learn electronics explained or even mentioned inductors. I even googled inductors vs resistance and I didn't get a clear answer or picture comparison.
I just bought a new multimeter so I'm excited to try and test these components myself.
@@mhnoni Inductors aren't used anywhere near as much as resistors or capacitors so they're usually not taught until later in electronics.
You'll probably understand it better if you think of inductors as the compliment to a capacitor rather than a resistor.
- Capacitors try to smooth *voltage* by storing energy in their *electric field*
- Inductors try to smooth *current* by storing energy in their *magnetic field*
- Resistors *limit current* by dissipating energy as *heat*
Also you'll find almost all multimeters can't measure inductance, usually you're measuring resistors, capacitors or diodes, you rarely need to measure the inductance of an inductor so they don't include that feature most of the time, in this video I was just measuring the resistance of the inductor to check it wasn't broken internally which of course it was broken. I have an LCR meter to precisely measure inductors, capacitors, resistors and all their characteristics, but you only really need that when designing more advanced circuits.
It's looks like a resistor. How can you tell from the first sight It's an inductor ???
Some things that give it away it's an inductor:
- PCB overlay has it marked as "L1" and an inductor symbol between the pins
- Axial through hole inductors are often that light blue colour and the colour bands decode to a sensible inductance value
- It doesn't make sense for their to be a resistor in that place in the circuit but an inductor makes sense
- The cross section shows it has a ferrite core and low resistance copper wire wound around it, a wire wound resistor would use a nichrome alloy wire wound around a ceramic core.
@WizardTim thank you God bless you
Thank you so much for teaching.
Thanks, how do you tell the difference between a resistor and an inductor with that colour coding? I was sure you meant a resistor BUT ….
Some things that give it away it's an inductor:
- PCB overlay has it marked as "L1" and an inductor symbol between the pins
- Axial through hole inductors are often that light blue colour and the colour bands decode to a sensible inductance value
- The cross section shows it has a ferrite core and low resistance copper wire wound around it, a wire wound resistor would use a nichrome alloy wire wound around a ceramic core.
The colour bands are very similar to resistors, the multiplier is the only difference on inductors, there's plenty of online colour band decoders that make it very easy.
Very nice analisys of the damaged part. Would be nice that would give a detail on the voltage applied to the leds as you go through. Wich Is It?
Thanks!
The DURIS E 2835 series LED chip that's in this bulb is rated to have a forward voltage of between 8.4 and 9.8 V, typically 9.5 V.
Thak you!
I had two Osram GU10 style lamps that failed after a while also with a similar open circuit inductor. They had a silicone encapsulated PCB, that didn't seem to improve reliability.
Interesting, I've heard the silicone encapsulation is meant to improve thermal conductivity to the outer shell and eliminate joint vibration failures, not sure about high thermal cycling. I'm tempted to buy a new one from 2022 to see if they've changed anything since the 2019 design.
Please did you use dry cell in testing the individual LED bulbs?
Thank u for the informative video
What did u use to test the individual led ?
I use 2 1.5v batteries in series ( 3v ) to test but was not able to get any led to turn on.
The LEDs in this light bulb are special, they have three LED dies in one package so have a forward voltage of between 8.4 and 9.8 V, typically 9.5 V.
To test them I used a Keithley DMM6500 bench multimeter in diode mode, unlike a typical handheld multimeter it has a relatively high diode test voltage of 7 to 12 V and can do higher test currents. But you can do the same test with a lab PSU at a similar voltage (>6V for those 3 die LEDs) in constant current mode set to just a couple of milliamperes. You could also do this with a 12 V battery, just make sure you use a resistor to limit the current otherwise you'll break the LED.
@@WizardTim Thank u very much for replying. 👍👍
Thumbs up for coil autopsy😊
cool, nice detective work, you know you can use devices to do the sanding for you, they are called, wait for it, sanders.
How do you light themup one by one, how many volts to light them how to do it?
Thank you for your video.
I used a Keithley DMM6500 bench multimeter in diode mode, unlike a typical handheld multimeter it has a relatively high diode test voltage of 7 to 12 V and can do higher test currents. But you can do the same test with a lab PSU at a similar voltage (>6V for those 3 die LEDs) in constant current mode set to just a couple of milliamperes.
Is anyone knows what's LED installed in a DLP LED projectors? They eats 0.6-3V and 1-3A. There are two wires connected-- + and -. When I tried to connect a 1.5V (AC\DC adapter with 3A) to one of them it does nothing. When I tried to check it by multimeter--then nothing. Regular diodes show some Ohms normally. But these one not.
I used to repair the old filament bulbs by holding the bulb at at angle to allow the broken wires to touch then apply power. Often it would fuse and you could get a lot more use
and the epoxy trick was fabulous...
VERY interesting. especially on the inductor. likely scratching away the paint might have exposed the wire enough to add flux and resolder. as a first point of approach. But of course spending many dollars on a 50c component is not worth it.
but saving the LED lamp was worth it; IF the hours are not more valuable elsewhere.
I did not expect so many windings inside that small inductor.
I would imagine the prestress placed on component during chinsy assembly, where they just handbodge it in place, will do them in with thermal cycling. Normally you are not allowed to bend the leads close to package and especially don't do so without proper tools. Then they bend or stretch legs to fit the hole distance, putting a constant stress in it, then they solder at 0mm distance from package end, always a no no typically for axial components. Well, they break every rule in the books. The board designer has magnificent magical stand off by the power of a fixed coordinate system, the assembly guys face gravity, time constraints and corner cuts. In todays outsourced manufacturing the left hemisphere isn't connecting to the right hemisphere, usually profit will sever that vital connection.
Excellent video, thanks! What voltage do You use to test LED "cells"?
The DURIS E 2835 series LED chip that's in this bulb is rated to have a forward voltage of between 8.4 and 9.8 V, typically 9.5 V.
To test them I used a Keithley DMM6500 bench multimeter in diode mode, unlike a typical handheld multimeter it has a relatively high diode test voltage of 7 to 12 V and can do higher test currents. But you can do the same test with a lab PSU at a similar voltage (>6V for those 3 die LEDs) in constant current mode set to just a couple of milliamperes.
What a So very good explanation......Many Thanks ! ! ! I learned a little more as always in our lives !
A great embrace of: carlitos
Great idea to epoxy it. I wonder if the leads were thinner or more flexible if it would help, or its pure thermal cycling of the component itself as you said.
Thank you for the video 😁👍
This inductor most likely got damaged due to heat and frequency. Ceramic capacitors get damaged similarly, especially in high frequency circuits. To prevent this, can be used e.g. two connected in parallel. They simply crack. They look functional but under a microscope you can see that they are cracked.
very interesting & useful
Very clever👍👍👍
I always think that was a resistance . Not an inductor. Good to know jaja. Thanka
6:34 Looking to fix a light? Why not Zoidbulb?
Thanks for this video. What is the function of the inductor in this circuit ?
Likely a combination of EMI/noise suppression, power factor correction, inrush limiting and ripple smoothing on the main DC capacitor.
@@WizardTim Thank you very much.
Well done!👏
Verry interesting n nice video
Nice repair and effort to spot the fault. I was a bit surprised you just didn't chip off the coating where the creak was to maybe see the broken wire. If it was because it got too hot, I think you would have seen burnt or discolored coating. So it had to be a broken wire... Just my opinion.
I've tried to chip off the coating on other components before, it's very difficult to do so without damaging what's under it so I wouldn't be confident if I found a broken wire to say if it was the failure point or if I had just broken it. If I were to do this again I would use an acid to dissolve the coating.
Excelent video !!!!!!!
Creative video, thanks :)
Failure should not be an issue. The manufacturers over run these lamps to make them fail.
Im trying to fix mine, 24 led,s on a disc, resistance test shows continuity. The circuit board doesn't have a resistor like yours.
it has 2 things the same as yours, a capacitor (no bulging) and the square redish parcel. it has 7 small black squares . it has a small glass fuse which is also good. Frustratingly weird cos I cant see the fault.
Kanlux GU10 3.6 w
I assume you've tested the 24 LEDs individually like I showed in the video to make sure they all still work, if one doesn't work and they're all wired in series the bulb won't light at all, I would imagine this is the most likely cause of it not working.
The 'square redish parcel' in your bulb is most likely a polyester X class capacitor being used in a 'capacitive dropper' power supply which is a very different design to the LED bulb in this video, if this is the case it should be a bit bigger than the one in my LED bulb (guessing it will be around 470 nF for a 3.6W bulb). If you have a multi-meter that measures capacitance you can test to see if the capacitor is still good based on the capacitance marking on the side of it (you may have to remove it from the PCB to test it 'out of circuit' to get an accurate reading), those types of film capacitors often degrade over time losing their capacitance which causes the bulb to be dimmer until it stops working completely.
The 7 small black squares are likely a combination of diodes and resistors, those are unlikely to be the fault but you can test those with a multi-meter.
@WizardTim Hi.
All led,s tested including on the back of the disk, the 2 solderd points.
The x capacitor says cbb22 684J 400v.
On 2000k test it says 550. I can't get any reading on the ų value.
The other capacitor is a 4.7 ųf 400v and reads on 2000k 460 and again no reading on the ų value, trying from 20ų to 200m.
The wires from the pcb to the bayonet prongs are good.
Am i right in thinking if i connected a 3.6w power source to the back of the led disk they should light up if they are good?
Thanks, John
@@knaptonmawson The 684J marking means the capacitor is "68 with 4 zeros" 680000 pf (680 nF). Your multi-meter may not be able to measure the capacitor as it's too small for it, it would appear as "0000.68" in the 20u range, assuming you have 2 decimal places on the LCD. I think your options are to de-solder the X class capacitor to measure it out of circuit on a better multi-meter or just assume it's bad and see if replacing it with an exactly equivalent capacitor will make it work again. I would recommend visiting an electronics hobby shop or even a makerspace if you have one nearby, you might be able to buy the right capacitor there or someone there might be able to test the capacitor for you but also help you test it safely.
Here's a video that explains this type of capacitive dropper power supply:
ruclips.net/video/-n22cqZkxNQ/видео.html
To light up the LED disk you'll need the correct voltage and enough current rather than just any power source with enough watts, for 24 LEDs in series you'd need something between 40 and 100 V DC but you will also need to limit the current otherwise it will draw too much current and blow up instantly.
@WizardTim Thanks a bunch, i will consider the possibility of replacing the x capacitor. I hate throwing things away that looks simple enough to repair .(if you know what you are doing)
I do understand that you need resistors when using leds, so you dont just instantly blow them.
John.
Very cool video
I would've thought that was a resistor... How can we differenciate them from inductors?
Some things that give it away it's an inductor:
- PCB overlay has it marked as "L1" and an inductor symbol between the pins
- Axial through hole inductors are often that light blue colour and the colour bands decode to a sensible inductance value
- It doesn't make sense for their to be a resistor in that place in the circuit but an inductor makes sense
- The cross section shows it has a ferrite core and low resistance copper wire wound around it, a wire wound resistor would use a nichrome alloy wire wound around a ceramic core.
Amazing work.
Very Nice !!
hello, can you tell me how to open the cover of a GU 10-5W bulb?
You might be able to carefully insert a thin pry tool between the lens and outer shell and then push it back and forth getting deeper every so often to try and cut any silicone adhesive, gentle heat may also help. Although you might find the lens is glued on with a stronger glue or ultrasonically welded which makes opening them without damage impossible.
Does you know how check Leds over 3V? My multimeter can't
In this video I used a Keithley DMM6500 bench multimeter in diode mode, unlike a typical handheld multimeter it has a relatively high diode test voltage of 7 to 12 V and can do higher test currents.
But you can do the same test with a lab PSU at a similar voltage (>6V for those 3 die LEDs) in constant current mode set to just a couple of milliamperes, just be careful of the PSU's output capacitance damaging the LEDs if you set the voltage too high, adding a resistor will help prevent that.
Hi, when checking each smd led using the multimeter till it lights up, do you set it to continuity mode?
Hi, the multimeter is in diode mode when lighting up the LEDs, continuity mode usually won't supply enough current or voltage to make white LEDs light up. However additionally those LEDs aren't the typical single die white LEDs, they're OSRAM DURIS E 2835 series LEDs so they have a rated forward voltage of between 8.4 and 9.8 V, typically 9.5 V. So your multimeter has to have a rather high diode test voltage.
In this video I used a Keithley DMM6500 bench multimeter in diode mode, unlike a typical handheld multimeter it has a relatively high diode test voltage of 7 to 12 V and can do higher test currents. But you can do the same test with a lab PSU at a similar voltage (>6V for those 3 die LEDs) in constant current mode set to just a couple of milliamperes, just be careful of the PSU's output capacitance damaging the LEDs if you set the voltage too high.
Very good thanks
Old led bulbs seem like a good source of transformers, what is the transformer in these bulbs? 110 / 12 vac?
It appeared to be a 3-pin ferrite E core non-isolated transformer for 240 V AC to about 24 V DC.
It's amazing that those tiny wire is Hexagonal.
Yeah, it's pretty interesting that they look hexagonal, however looking at them under a 100x microscope they are much more circular, it's difficult to tell but I think they may be deformed as they're wound onto the inductor core.
@@WizardTim Thanks for the additional information.
Very nice video i like this 🎉
I’m brand new to electronics so not really sure what I’m saying, but is the inductor not a suppressor?
The inductor resists sudden changes in current by storing energy in it's magnetic field, this is useful for EMI/noise suppression in this LED bulb as it dampens the current spikes from the switching power supply from being conducted back into the house mains cables and possibly interfering with other devices or radiating as a radio signal. It can also help smooth the AC ripple on the DC bulk capacitor as well as adjust the power factor but it's main use is probably for basic EMI suppression (higher power devices will have proper "common and differential mode" EMI suppression filters).