Hi, it's repair time. If you're following on Euv blog 2 and you should, You would have noticed that I had a problem with one of my uh, oyster lights in my kitchen and well, it was sending alien signals or something. It looked like it looks like an alien spaceship and it was flashing. Uh, a weird code.

So anyway, I've ripped it out. I've brought it into the lab here and I've hooked it up. and that's what it's doing. Let me turn the lights off.

so I've just got it hooked up to a lead string which I uh, got out of ripped out of an old uh Lcd, uh, dumpster Tv. It's coming very handy. I've used it for quite a few videos. Actually the Leds have got like a five volts drop per lead.

They're actually quite high. and um, yeah, I do like they're in string configuration. I use them for all sorts of things. I've got two strings of those.

yeah, does anyone know their morse code? What's that flash in? Yeah, it's hiccup in. So anyway, that is, obviously uh, hiccup in as they call it. Um, that's one of the names for the fault. I don't know if you call it something else, but um, yeah, in the industry it's often called uh, hiccup in And the mains power supply is hiccuping.

It just can't start up. It's trying to start up. Uh, the mains, um, side of it. but the main switching side.

But it just can't do it. So it keeps rested, retry and retry and retry. and no, it just comes guts or every time. So anyway, let's take this thing apart and uh, see if we can find the culprit.

All right, we're at the Tagano microscope, but I've got something a bit new for you today. I've now integrated my Ni National Instruments Virtual Bench scope. It's just sitting right down here next to me and I've got my probes. I can have my oscilloscope probes, I can have power supply function generator, logic analyzer, digital, I O everything integrated into my Uh video capture system which not only I can record with, but I can also stream as well so you'll see in my straight ahead camera.

I've also got uh, my Tagano microscope here. it is right under me and uh, there will be a B cam, but it's not hooked up at the moment. And also there's my national instrument screen. But one of the cool things I can do now is I've set it up so that I can now overlay this.

I can overlay the multimeter here. So here's here's here's the multimeter winner and all the power supplies as well. And there's some digital I O in there as well and I can stream that and um, yeah, that's really cool, huh? So anyway, uh, let's get to it. Let's have a look at.

sorry, let's have a look at this. Okay, so let's have a look at the driver here. It's a luchy thing. model number for those are playing along at home.

It's a it's a dimmable uh jobby and it's a two watt lead. Um, that's individual Led. It's actually up up to 20 or 22 uh watts output. so it's actually rated for up to 80 degrees.

But yeah. Anyway, let's have a look. What's failed here. Here's the output section.

Now, I wouldn't expect there's no bulges in those caps, of course. First thing you look for is bulges in the caps. But oh, I forgot to tell you that this actually comes with the heatsink over here and it's just screwed in there. And I've taken that out because we want to see some stuff under there, a couple of extra caps.

and anytime you see caps near or under heatsinks like this, yeah, you want to be suspicious of those, uh, puppies. So anyway, I've taken that off. Um, so yeah, it's just a plastic package. so that's just uh, like they didn't even bother with the heatsink compound so they just bit how you do it.

they just slapped it on there. But you know better than nothing I guess. Uh, these are a shy caps on the output. Now it's not going to be the output A because those caps looking good, nick and B there's multiple ones in parallel, so like you know, I wouldn't be concerned with that and that's not going to cause the power supply to hiccup.

Now Is this a secondary regulator? Yes, it is. This is a secondary regulator. Okay, because here's the primary secondary split. So this is all the secondary side.

This is the mains input over here like this. And so these are my initial thoughts. I I'm pretty sure it's not going to be this secondary side. I mean, you can go in there.

Obviously, you do a visual on the uh caps, but these have got the vents on the top. I can't see any bulges whatsoever. They'd be in parallel if you actually follow the money on there. Yeah.

Parallel, Parallel. parallel. Um, and another one there, is it. I don't know if there's four in parallel or whatnot.

No, there's another one that's not in parallel. So we've got three in parallel. So when you've got three caps in parallel like that, it like it's not going to be the output caps and really the output driver side of things. Um, you're generally not going to see like a hiccup.

hiccup, hiccup, hiccup. Um, as we call it, that's usually a primary side thing where this converter can't start up. So yeah, I would not be looking at the secondary side. I'd be looking at the primary side.

now. They've actually rubbed the number off this chip so we can't see what this is. Maybe I could get a knife on there and try and scrape it off, but I think that they've actually scraped the number off that chip. Anyway, you want to give a visual on here and one of the first things I notice is that there's discoloration around there.

That resistor in there? Is that my imagination or is that a bit discolored too that? that looks a bit how you doing doesn't it? So we'll get in there and measure that and you can see that on the bottom as well. Look, you can see the fiberglass and like you start seeing like the the the weave in the fiberglass pattern right? You don't see that anywhere else. That means this has been heated up to buggery. and uh yeah, it's just the fiberglass is not.

uh, you know your high temperature high quality stuff so it's just start to decolor and and yeah it's it's not looking good. Also, you can see the same thing effect happening over here as well. So that's under. Um, what's that under? That's these big resistors over here.

These are, just, um, input resistors. Are they? Yep, So yeah, they're just input. Um, limiting surge resistors? Whatever. Anyway, um yeah, there's no major caps on the input.

even though this is a full wave bridge rectifier. they just don't need a massive amount of capacitance on the input. Now of course, as I said, these caps are under and near the heatsinks here, so you would suspect these puppies, right? 50 volts and they would be to do with the startup as well. I don't know what that.

What's that in there? Is that a St25c? Is that just an external uh switching transistor? I have to look that one up. St25c: It's a Bjt. It's an Mpn Bjt. Yeah, it's just a jelly.

Looks like it's just a jelly bean. Bjt. No worries. And we've got a bunch of diodes in here as well.

Uh, you could go in and you could measure those as a matter of course. But I'm I'm really quite concerned that that resistor there is heated up a lot and it's right next to the caps as well And let alone the heat from the heatsink. But that that that resistor? Um, that's obviously gotten hot because it's just yeah. Um, so anyway, let's get turn on our multimeter shall we? and let's have a probe in there.

What? What value is that brown black something? right? Brown Black Burnt Brown Black burnt is the is the value of that resistor. Now when resistors get hot like this, obviously they're not going to be in the kilo ohms or the hundreds of kilohms because Ohm's law, you'd need like thousands of volts before you do anything actually heat them up. So this is going to be a low value low value resistor. And was that a gold band on there? Like you know, like a silver or something like this one over here? Anyway, so that's supposed to be a low value, so let's measure that.

shall we? 287 K: Okay, no, no, there's nope. Um, Houston. We've had a problem. Um, the resistor is that that is Gonski.

That is Gonski. I don't need to measure anything else. I mean, I could. I mean, I could whack.

Um, I could go uh D mode and I could go start checking my deities and stuff. But you want to fix anything you come across. 0.56 Nine Point Six Seven. I think we're good on the diodes.

Just check diodes because they're easy, you know? and you can generally check them in circuit Point seven. You know. no, no, no wackers. So yeah, I can't see anything visually.

anything else. really. I mean, these suckers. They like it.

It's it's discolored. I mean, it's it's discolored and it's 270k. Nah, that doesn't add up. That does not add up.

So we have Kamagatsa there. and uh, we could have found a culprit. Is it that easy? Is it one resistor? Let's hope so. Makes for a boring video.

But anyway. Oh wow. wow. That that other joint.

I, I haven't even desoldered. That and that other joints. Gonski. That's the other thing that I should have looked for in my visual was joints.

And let's actually go down here. Oh, hang on there. there's a pin sticking out. No, that can't That can't be it.

And oh yeah yeah. that one look. Yeah yeah yeah. there you go.

It's uh. is that just a that's just a transfer. Oh, that's just a transformer leg. Okay, that's right.

Yeah, there you go. This sucker has heated up so much that it's just delaminated. The copper's just delaminated from that Wow. Yeah, that's Gonski need to tongue at the right angle there.

Geez, There we go. Got that characteristic bernie smell. That Ernie Bernie smell And yeah, that that is one sick. oh look at that.

that's cracked a buggery. Wow. Wow. That suckers had a harsh life look at that.

So that's uh. well. no. I thought that was brown.

Um, black something. It's red. What is it? red? brown? something? I don't know. That is just that's terrible.

Muriel. Wow. Yeah, that is completely gone. No wonder.

I wonder if we can Just like, is it just gonna fall apart? I wonder if we just grabbed that with two pliers? Is that going to come apart? Oh no. It's still strong as a mallee ball. But uh, yep, it's It is completely cracked. It's gone and that of course is probably open.

The 270k was measuring in circuit and yep, yep, it's just it's it's gone ski. So it could just be the resistor that just heated up so much that it eventually went open. And I can guarantee it'll do exactly the same thing now if we power this up because that resistor is missing. It was open before it's still open.

So yeah, maybe that's it. Maybe did they have the wrong resistor value? So it's got to be like a piss-poor design Because like, you wouldn't design it so that resistor heated up so much that it failed like that. And it's not like because, and like an ambient thing because it's inside the oyster light. I mean, that doesn't help, right? the fact that it's got no airflow in those um, sealed oyster lights upside down on the roof.

But wow. You know that's just that. that's a harsh life. It's not good.

I don't know. It could have been brown. It could have been black. That's just faded.

It could be gold. That's just faded. I'm that third band. It does matter.

It matters a lot because that's the. um, that's that's the multiplier. So I'd be tempted to go with gold. Maybe.

so we're talking like one. Ohm, silver would be. um, you know, like it does look red on the screen here. But when I look at it, uh, with my eyes under my improper color, um.

lights here. it? uh, it looks brown. So I I think it's brown black. Um, and it's just yeah.

it. I know it does look red on camera. but when I look at it straight, it's actually it's It's more brown. So um.

anyway, that's what I'd expect. Yeah, I'd expect uh, like in the order of like, Ohmskeys. So where is it in circuit here? There It is there. That's one of the transformer taps.

Uh, these are the transformer taps along here. I don't really want to reverse engineer this, so no, that's a large value resistor in there. Yeah, so I'd say that's just a current shunt resistor, is it? Then I I'd have to look up a typical primary side switcher for that, because this, obviously it's not doing the regulation on the primary side. it's doing the regulation on the secondary side here.

That's the only reason why you'd have the S08 controller there, which I think they've rubbed the numbers off that puppy as well. I think the conformal coding doesn't help because you know, like, you can still see the diode marking on the diode bridge over here it? Um, yeah, I tried to scrape that away and it's just there's there's nothing under there. but uh, yeah, I don't know. And like, just throw in a couple of Ohms or something and see if it gets back on track.

So I'll actually, um, leave this one sticking up a bit. Why not? That's that's not going to touch the heatsink, just leave it. leave it up there like that rather than have it right down on the board. That gives it some extra heat sinking.

Um, due to the legs. Whoops. That pad's lifted. didn't go through.

That's a bit cleaner. Hopefully she'll go through. now. That's the problem with lifted pads.

Really rather annoying to try and get these back through. I think I got it. Yep, sweet. So I'll just leave that flapping around in the breeze up there.

And bob's your uncle. So here's our main primary switch in transistor over here. You can see that goes over to the coil there, and that's our resistor in there now. so I'm you know, it seems like that is not in the way of goes under there.

it goes over to here, which is then that small little to92 transistor we saw before. So I'm not seeing how that's part of the main line. so I'm not seeing how that resistor is in series with the main switching transistor there, so that shouldn't be a problem, even if we make that too low. I don't know.

Um, but yeah, if it was, if it was in series, um, then obviously the value is going to matter. But um, yeah, because it's in there. I don't know. There's this extra switching trending in here that's like it's it's probably doing some like power factor correction as well.

something like that. So it's obviously doing something different. So maybe as I said, like it could be some other sense resistor or something like that. but it obviously heats up so it's important to get that value right.

It's going to be critical and, well, I don't know. cross your fingers and hope, um, I've got the value right? But anyway, um, yeah, let's just power it up. Whatever. All right, let's give it a bowl, see if we get anything.

Whoa. Magic Smoke released. Holy crap. I guess that's the wrong value.

Whoa. What? What? What? Yep, I would say uh, that wasn't. uh, gold. that band.

That's what happens when you underestimate the value. The actual resistor's fine. but look at the input. it was obviously drawing like probably an order of magnitude more current.

And those input resistors from the mains. Nothing else seems fried. Um, which is really interesting. But those.

wow. Look at those input resistors there. They are just. ah, they're burnt.

They are horrid. That is unbelievable. Um, yeah, I've completely kamagatsa. I am.

I don't think I'm gonna bother to fix this now. I'll just probably upload this as a hilarious what not to do. Um, yeah, I got a bad. My spidey sense gave me a bad feeling.

Just before I was about to switch it on, I thought, oh no, I'm gonna come a guts or I think I've got. I should have gone up in value rather than down. Um, yeah, that was dumb wasn't it? Actually, I presume that's an input fuse there. That just goes to show how poorly that's rated that input.

I'm gonna cut that off because like that, that should have popped before all of those just absolutely cooked. Um, yeah, I think they've rated that wrong. Oh no. So much for that.

I thought that had to be a little axial fuse in there, but it's not. It's just a resistor. anyway. Oh boy.

Those puppies cooked. Sorry if you got that fan noise in the background. that's my, uh, filter going berserk here, trying to get rid of the wrenched, uh, smell from this sucker. Uh yeah.

Well, that's embarrassing. That's a double facepalm worthy. yeah, please leave it in the comments down below. I'm an absolute uh deal.

but I thought I had the right value. But then I like. I knew that I should have actually, like, at least attempted to reverse engineer this. See what was going on there? See exactly where the resistor was.

The lower value was always going to be a concern, but I I thought it was gold in there. But I have. It's yeah, that multiplier band is gonna Kamagatsa. obviously.

Yeah, I was off by at least an order of magnitude, but it seemed reasonable at the time. because if you have a look here, here's the resistor here and it goes under there and it goes to this small To92 Npn transistor that we looked at. So it's not like you know that's going to be carrying amps or anything. So it's not the like emitter resistor in the main switching transistor which will be in series with the transformer over here.

So it's like it's not that it's somewhere else in here doing something. but obviously given the size of that resistor and the fact that it's um, heating up yeah, it was, obviously, um, doing something more important and dissipating a fair bit of power. So with hindsight, yeah, that was just a dumb 1.8 irons was a dumb choice. Um, yeah, just didn't put the thinking cap on enough, Don't All right? I've decided to do a basic reverse engineer in here.

It's not complete, so please excuse the cruddy of the model. Didn't have time to build to scale or paint it, so I'm going to work out where I goofed up. My first move was not remembering that this is a dimmable lead controller. Which means that it has to start up at low voltages.

Which means it's going to need some sort of low voltage. uh, startup. You know, just in case the dimmer when you turn it on is like set to like a low voltage. like, you know, 50 volts.

You know, something like that or whatever it is. Anyway, um yeah, we've got 240 in. You know, there's some filtering and other stuff over here and uh, including the resistors and that burnt the hell out. And then we're going to breed a dire bridge rectifier.

And then we just got a main small filter cap which is, uh, here. And then um, here is our mystery resistor right here. and it looks like, yeah, we've got a Zener. A basic zener um, circuit here with a emitter follower transistor.

here. that's a little, uh, To92 jobby in there and um, and that just powers the eight pin chip. I still don't know what that uh chip is. Look up: uh, dimmable flyback? uh, lead controller or something like that and there's 10 million of them, but it's more likely to be one of like the lesser known, like asian variants or something like that.

Anyway, um yeah. and then of course, um, the Dc. The main. uh, couple hundred volts Dc.

um that powers the transformer over here just like uh, I suspected. um with of course the uh, main switch in mosfet here and then the other one. What resistor down in here, which I knew was uh, the series resistor for that and obviously they're like then they're tapping that off, uh, somewhere else? I didn't bother. Um, you know, going into details around the uh controller chip, but then you've got a tap coming off the primary of the transformer here which goes via a diode and that goes into by basically bypassing the 180k and the other resistor up here.

the 180k one. what? Uh, Jobby, that's the other one watt, uh, resistor up here which wasn't cooked at all so you wouldn't expect that 180k to be dissipating much at all because it's basically are in parallel with the coil over here and which of course can be low impedance to then uh, drive directly into our main zener dropper down here. Obviously this is going to be a Zener. I don't know the voltage, but I did there.

It is. tiny little jobby down there. It's a sad looking uh thing, but it has had the snot blown out of it. Um, it's measuring like three.

Ohms. so um yeah, that's no good at all. so that probably that's not one of the original faults. Um, that would have happened in uh, my goof up and we can see how it's blowing here.

This, I totally goofed up. I, you know, I thought it was some other, you know, current sense resistor or something dumb like that. I just, you know, it didn't have my brain engaged. And of course it's A.

If you did some basic reverse engineering, you would have seen that this was, uh, a zener dropper basically. So this needs to be in the order of tens of k's. Um, so that uh, resistor value probably is either um, 10k or uh, 20k. So that third band which I thought was gold was actually I think it was orange.

So uh yeah. so I think this was originally a 20k resistor and here I am going putting in a 1.8 Ohm. So I was only out by four orders of magnitude. so we can see now how it snotted itself.

Um yeah, we've got our high voltage Dc here basically going straight through the coil like this and then straight across through the diode. I'm surprised like the diode survived and a straight through basically a short circuit which then, um, snotted the uh, Zenna down here. So it it turned, it failed. Uh, short circuit.

So if either of those failed open then um yeah. we wouldn't have had the magic smoke escaped, it just would have went pop and pop goes a weasel. And yeah, it would have been fine. We wouldn't have burnt the um input protection resistors over here.

They wouldn't have turned completely black and charred like that. So yeah, we just, um, it looks like I haven't measured this other diet that could be uh, shorted as well. Um, likely. So yeah, it just boom.

Went straight down like that. So this should have been, you know, 10 or 20k, something like that. And yeah, complete goof-a-rama I chose Paulie. He chose poorly.

and what that Zeno voltage there would have been? I, you know, it's like in the order of maybe 20s. you know, the tens of volts. 20 volt. I mean, there's you know, a 50 volt rated cap.

Both of these caps are 50 volt uh rated here and I won't go into details about. you know how all this works. Doesn't really matter if anyone does know, um, they can, you know? guess uh, based on the basic, at least some of the pin out I've got here. Um, exactly what chip that is.

Please leave it in the comments down below. And if you do want me to do a complete reverse engine area, well, if we got that chip, we probably shouldn't have to. They've probably got an example uh circuit in there. It's probably almost identical to the example application circuit.

They usually are very few differences usually. Anyway, if you want me to do a full reverse engineering, then you know, leave it in the comments down below. But I that like we just needed to know where I goofed and that's what it is. It looks like it was a um, Xena dropper circuit doll.

So yeah, that was embarrassing, wasn't it? But anyway, um, I hope you found that valuable my goof is your gain I guess. Um, in terms of yeah, just you know, like a couple of more minutes. If I didn't rush this thing, a couple of more minutes, just figuring out exactly what that resistor was doing would have prevented this. Obviously, if I knew it was a Xena dropper like this, there's no way I would have made that like, you know.

Ohms. I would have made it tens of Ohms something like that and I would have probably guessed. You know, it's not going to be like in the hundreds of K region. it's going to be in the tens of K region.

If you do your uh, Xena, you know, dropper calculation and stuff like that, assuming like, you know, a milliamp or two draw in here and the minimum zener current you can work out. You know, a basic, uh, you know, ballpark resistor value and it's going to be in the tens of k. uh, regions. something like that.

So yeah, I definitely wouldn't have made that mistake if I simply went and spent a little bit more time doing the reverse engineering. So that. let that be a lesson to you. anyway.

I hope you found that valuable. If you did, please give it a big a thumbs up. As always, discuss down below: catch you next time you.