Hi with the recent well of video. I Thought this one would be quite appropriate to have a look at and it's a retro one. Take to mind back to 2005 when the original Xbox console reigned supreme ends up if you still got one and you're still using it Anyway, there was a problem with the Xbox console. Unfortunately, about one in every 10,000 give or take would release the magic smoke, so it was a big deal.

at the time, millions of boxes involved parently only like 30 of them caught fire or smoke, door melted down and did whatever. Microsoft actually seemed to do quite well at identifying these problems and owning up to them and fixing them. but in this particular case, they didn't recall the actual boxes and they didn't really specifically say what the issue was that was causing the problem. With these things, all they did was saying oh, we're going to for a voluntary recall of the power cord for this thing.

So what they did as they shipped out millions of these things, these replacement power cords to Xbox owners. You can go onto the website and you can request a new power cord and I did this at the time and sure enough, they shipped me one of these power cords. So the reason I'm doing this video is because I'm still cleaning up the old lab and I came across this original replacement. Xbox Power cord in quote marks it's actually an electronic fuse by the looks of it.

I Thought: be interesting to take a look at this issue and do a teardown which was designed to cut the power when your power supply developed or whatever the problem was inside the unit actually developed the fault and rather than burn down, it's like a fuse that will add to most of their products. But let's not go there again. Now here's the original power cord again in quote marks that Microsoft voluntarily replaced. There wasn't any like like official like government like safety recall order or anything like that, but they volunteer replace the power chords on 14.1 million.

Xboxes and I don't know if that's if that's how many they ship but that was the potential impact of this thing. So why did they replace a normal power cord like this with this electronic fuse that has a reset button and a test mode. You press test on it and it flushes a few times and then the electromechanical cut out. You can see that there's a green like mechanical indicator in there which will then switch off.

So why would you replace a normal cord with an electronic fuse like this? Well the only reason is that it wasn't a faulty power cord now. Microsoft Were very cagey about their actual wording for this thing saying I wear a place in the power cord and things like that I Think they kinda sorta admitted that there was an issue with the console, but anyway, um they decided I guess the console didn't have adequate or or any fuse in it. So we'll do a teardown of an actual Xbox console shortly, so stick around for that to have a look. But they decided that rather than recall the consoles which could have been one way to fix this problem that would have been expensive I Mean if you recall 14 point 1 million Xbox consoles like it's it's cheaper just to design and manufacture this and just ship it out to people and problem fixed.

and you know, guess there's nothing inherently wrong with that. You gotta weigh up the costs of doing this. I mean just imagine if the bomb cost and the shipping cost of this thing is ten dollars for example. Well that's a hundred and forty million bucks right there.

but it would be much more expensive to actually and much more damaging to the brand, etc. and the Xbox reputation, all that sort of stuff. actually recall units cuz apparently this all came about because only like 30 or 40 units actually caught on fire or smoked or melted down something like that Microsoft actually came out and said that you had a one in 10,000 chance of it happening. but the actual number of units that that the fault did actually hiren was actually much much smaller than that.

So anyway, if anyone has any details on the you know the background for all this like actually inside information or something like that, please do leave it in the comments or over on the Eevblog forum. Anyway, let's tear down this thing. nominal 610 milliamps. Not sure how accurate that is, but it's made in China it's got the C tick mark.

It's all happening anyway. catalog number qo1 so let's tear it down. So obviously designed a trip at 610 milliamps or there abouts. Thankfully, it's got some screws.

Let's get into it now. Unfortunately, you can see the pain in the ass screw down there. It's one of those pintle OB type things with the security pin in the middle and it's actually a large one. So I've only got like the smaller ones for like the newfangled iPhone Whatnot.

Ease and well. I Drilled out the screws and check this out. It is phenomenally complicated for an electronic fuse, which is basically all. It is unbelievable.

Look at all the control circuitry on here. I'm just absolutely stunned. We've got two eight pin jobbies on here. They're probably little micros, might have a closer look at those in a minute.

So that's your test switch. That's your reset switch and you can see that that physically pushes that green. Indicate that mechanical green indicator. Why are you go to all that effort for this mechanical green indicator and have all this molded plastic stuff? It's absolutely remarkable.

Got a massive move there. By the looks of it, you can just picture the design meeting for this and the poor scared little engineers in there. Or and they're going. don't screw this up.

Otherwise, bill billion selfs gonna come here and personally kick your ass. It's like hilarious. Yeah, they've just gone to so much effort and even like all these dedicated plastic mouths to mount the board on and everything else is just insane. I Mean you know they could have just put a fuse in a lead if they really wanted to look at this taking off that plastic cover look.

They're breaking both the active and the neutral. They're like. Unbelievable the amount of complexity they've gone into doing all that little solenoid in there and it looks like they got a low voltage transformer here for powering various stuff. They would have got that right.

I'm sure with all the requisite approvals, they wouldn't be very careful with everything in this thing because they didn't want to screw it up. Anyway, let's have a look at how they're actually doing the current tripping here. now. It looks like they've got two current transformers here.

this one here. Some windings down there inside that pot in down in there and they come out these brown wires down to this chip down here. We'll have a look at that in a minute and then they've got this other one. You can see the windings more clearly in that one.

that one's not actually potted in place. That one looks like it has much finer and much higher number of turns in there and that's just like a regular. PCB Mount our current transformer there got a Diode bridge rectifier. We'll look at that for the transformer in a minute.

but look at how the mains current goes here. Why do they need to if they're just doing simple current sensing? Now, this first current transformer here has both the primary, the brown wire and the black wire running through it. That does not make sense for measuring the current of the product under test because the if the currents flowing this way through the active Brown wire and then back this way through the black neutral wire, then they cancel each other out and you can't measure anything so. But this is a classic configuration for a Earth Leakage circuit breaker or a Current Balance transformer.

So if there's any imbalance between the current between the primary and the secondary here, then it will generate a magnetic field which can be sensed on the coil and then it trips at a certain current. So that's how an Earth Leakage circuit breaker works. Now we've got this extra winder, which we'll have a look at in a minute, but obviously they're doing some Earth leakage circuit breaker functionality so right than just the 610. Obviously, they wouldn't be doing 610 milliamps in terms of our Earth leakage current there.

That's way like above what the standard is like 20 milliamps or 30 milliamps. So obviously that's the primary current that they're tripping the electronic fuse at 610 Williams So they don't mention anything about earth leakage breaker inside here, but they're obviously taking no chances. So this thing has dual functionality. Now if you have a look at the second transformer, it now makes sense.

The black wire, the neutral avoids going through that and they've only got the active passing through. So obviously this current transformer is measuring the product consumption. Current from the active goes out to your Xbox and then if current comes back and then avoids that. So we're measuring the current taken by the Xbox so that one is doing your current sensing and the circuitry on the back.

it looks like and that's going to bugger off to your control board. Here there's a track on top and if you can see that, but there's a track that goes from the pin of that current transformer over to that board. So there you go, it's got your functionality. Wow Didn't expect that.

So what's going on with this white wire here that's all wrapped around and buggers off down here? Well, obviously this is the test functionality so that because it's going through the primary current measurement transformer here, they're obviously going to put a load on this thing and then test out both the product current, the electronic fuse, trip current, and also the Earth leakage circuit breaker as well because you can see there's a couple of turns wrapped around there like that and then it bugs off down to here and down to this relay down here which is a nice Omron job' haha. Spared no expense, really spectacular, spared no expense. and then it goes through these three 1206 resistors. That's to get the voltage requirement.

So three of those in series and here are the two contacts on the coil and then the contacts just go off to to the new chopping down the bottom. and of course the top side of that is connected to the active after this transformer here. So they're obviously just putting those three resistors directly across active and neutral with that relay. and that's designed to test both the Earth leakage circuit breaker via these couple of loops here which give you more current by the wayside like magnifies the current going through there, and also by putting a small amount of increased current into the main current measurement transformer.

But how small? Well, we're only talking these are three a 10k resistors in series. so we're only talking like four odd milliamps or something. So obviously they're not testing at the full load because if you wanted to test at the full load, you did a massive power resistor. especially like for the six hundred and ten million, some massive power resistor or a varistor or something like that.

Now I Know there's a lot of stuff on there, but you know, surely you would have as a first pass. You would have tried to integrate that onto the main board down there. Maybe you know I Know they've got to keep the size down, but geez, you think they could have some I mean it's double sided populated anyway though. although it's only through hole on the top I guess.

but still. it's almost. Maybe like you know two separate design teams. you know one worked on this is test control or the other one worked on the rest of it or something.

and then they had to integrate them together. But it's just seriously calm, complex, and unbelievable. Son of a medium microchip. What does that mean? I Is that like an in-joke? There's our lead and the microchip fanboys go wild.

That's a 12 F 675. They had some sort of marker pen on top of that, obviously to show that it's been programmed and then elsewhere. We just have a Lm35 eight Op-amp so little a pin micro Joby just controlling that test functionality. Oh, that's so8 that makes sense.

It's not his OB but maybe that's what they referred to it internally as I'm sure now. if I was designing this and wanted it to be reliable I wouldn't be doing the detection of that 610 milliamps in software in the micro wouldn't be relying on that micro to then like the ADC measure it, continually measure it and then TripIt Not only is it well, it's going to be slower, it's still going to be fast enough of course for the application, but then you add an extra layer of unreliability in there. As reliable as these are every extra process you add in there. So I'm wondering, is there any like adult land log functionality and all of this other? The micro and everything else is just doing the test functionality.

Perhaps. So the active stuff might be done analog wise. We certainly do have lots of discrete transistors on here and stuff. And of course we've got the Op amps and things like that so it could.

Certainly the thresholds could certainly be done in hardware, but then what do you do for different regions? You're going to need different test currents. so if you did it analog wise, you would have to change some of the values of the resistors. And you can see the microcontroller has the test pads up here so obviously this is programmed in circuit and it wouldn't surprise me if that had different programming for different regions for the different out required test currents. All right, let's just run this.

I'll push the tests which test it. you use my poker here Wow Hey I'm talking about belt and braces engineering they said in our series love in here Big ass trans orb and if that's not enough, coupled thermally to a thermal fuse 105 degrees C Thermal fuse. At what point in like the engineering cycle do you go? Well, you know. Yeah, we've got to move in there, but we want to have a thermal fuse in series with that just in case the moth eats up and gets too hot.

What? like I Got my kingdom to be a fly on the wall at the safety review meeting for this thing because well, you know, look, hundreds of millions of dollars are on the line for this thing. You've got to remember that. And you know, the Brant the Xbox brand that's worth billions and billions of dollars is all on the line. And if they don't get it right, you know.

So the pressure on the engineers to over engineer the out of this would have been yeah and pretty intense. And that's what they've done. I Mean this is just massively over engineered. So there's a fuse in series with the mob, and they've run a wire over a jumper link directly over here.

Just. it's not an afterthought, really. But you know why they couldn't have snaked the traces on their clearance? Why's they could have just slotted that out? I guess. But anyway, yeah, directly across the mains input before any of the switching happens in here.

So obviously our test button that goes in the back here pushes up these armatures, it latches in place, and then they activate this solenoid to actually switch it. You know to deactivate or trip the thing Basically, so that solenoids is connected under. here. you'll notice that's going over there.

it's going over to looks like is that a little solid-state relay job'? No, that's actually just a bridge rectifier down in there. Here's the neutral over here. going into the solenoid here. Then the solenoid goes into the AC side of the bridge rectifier and then the bridge rectifier comes out.

He should DC out. So they've got a little transistor in there and so they're what Cro3 am? No, it turns out that's a thyristors. They're using that to switch the solenoid. That's rather interesting, isn't it? And they've just got a very stir in series with that to protect it.

But yeah. um. thyristor control of the solenoid. What do you know? That's a dedicated earth leakage breaker chip.

But you'll notice that the coil goes the bridge rectifier between active and neutral there. and the fire Esther is on the other side of that bridge rectifier there. So it's got. it's.

not actually switching the coil like directly across that. That's not how it works. If you look at the typical application circuit for our Earth leakage circuit breaker chip. Here, we've got our bridge rectifier.

Okay, here's our contacts over here. Here's our solenoid to drive it. So there's our thyristor there as part of this latch circuit here. and that's solenoid, Of course.

dries the contacts, but that's that's not what we see here. Look, this is neutral. goes through the bridge rectifier, through the coil to active. It's just permanently enable.

That solenoid is permanently enabled across. That means how else does it get driven? It's bizarre. So I've traced this out here and this middle pin is the anode of the Thera stirrer we're talking about at this point up here. instead of it going to the solenoid, which makes perfect sense, it doesn't go through these four resistors here in series, going over to pin eight of the chip the power pen.

So those four resistors are obviously the 50k there. So of course we're getting now. you know our 240 volts. AC rectified across this full wave bridge rectifier here.

So we're gettin' orti volts on there. That's why we need those four resistors in series high-voltage doing a direct mains powering of this chip here. and it's got a built in regulator. It's designed to do this, so no problems whatsoever.

But where's the solenoid? Where's wally? It's directly across, active and neutral permanently. I Swear. I must be going nuts and I have confirmed that that is correct. The gate goes over to pin seven here as part of the latch circuit.

have confirmed that and the other. The cathode of course goes down to ground. so that's exactly the same. But instead of that coil there, this just connects directly to there.

There is no coil. it shorts out the output of the bridge rectifier. So instead of having our solenoid here there, it's installed directly in series with the AC side of the bridge rectifier there. That's it.

There doesn't seem to be any other way to activate that's solenoid. I'm going nuts. I think I'll go start at Jim's Mowing franchise. Okay, so the only way this thing can work.

look what happens. The solenoid okay is normally off like this because like, you've got 240 volts on here so comes through here. There's no way that it can go anywhere else, right? Because it's just it's just going through and powering the circuit here. So if the thyristors shorts out this point here and this point here, bingo, we've got a short across there.

Bingo. You suddenly have a path for the current to flow through that diode, through here, through there and back like that, activating the solenoid and disconnecting the circuit. So it's like they doing a completely opposite to what the application circuit for this chip says. But that certainly explains why they've got that varistor in there.

Ah, because like you know, you don't want the full mains across that you want a very stir in series to then go high resistance as it heats up, but you only need the solenoid to activate for you know, a split second and then Bingo cuts off the power. So you could say actually, that's a rather clever variation on this classic application circuit which no doubt the designers would have seen. They would have had the datasheet just like I've got for this chip which has this application note on the very first page and it looks like they're doing it almost are identically except for the fact that they haven't bothered to put the solenoid here. They've actually put it in series with the bridge rectifier and then once it trips shorting out, the bridge rectifier cuts its own power off because once they spread rectifier sorry, shorts out across here of course.

Then then there's no more power on the rail up here. and the switch chips and switches off the thyristors off. everything's off, so it just goes. BAM But why they decide to do that over the standard implementation with the solenoid coil here I Don't know.

Only you guys fight it out in the comments. Go for it. It's for the solenoid activation. Let's push this lever all the way with LBJ right down to the bottom.

Push that in like that and then that holds it in place until the solenoid releases and the whole thing flips up. And what does this transformer doing? well? It's just a little low power transformer just to power the electronics. It's connected directly across the mains. there.

it just goes over to the other side here and then that just powers the this board. So this is the AC mains. Input here goes into the AC side of the full way of bridge rectifier. there.

the other side of that buggers off to one pin of the module and the other ones over here go through here over to here so this can find its way back. Certainly find its way back into the main earth leakage circuit breaker tree. So obviously you know you've got to have two pars. one's got to come from the test board.

logic control has to be able to trip it, so it's obviously has that extra path through there to do that. So let's just have a quick look inside the Xbox shall we? I've got to get all this cage stuff out before we can access the mains down here because there's going to be something to do with the mains input. Take out the hard drive we're into the power supply. A totally separate board of course.

as you'd expect. Very nice construction inside the Xbox by the way. very well designed. Nice.

I use of the envelope with avid thermal oi or for all you avid fanboys, there you go. they've got those in the right direction for the fan here. If you had them in the other direction, they wouldn't work. You wouldn't get the airflow across on it.

It's made by Fox Link Technologies made in China Looks pretty good. Single sided board of course as you'd expect in all of these ones, but that looks so nicely laid out, nicely designed. It's got a tipo main cap there. geez, haven't seen T pay for what happened to tipo and probably anyway.

Nice common-mode choke there. there's your full wave bridge rectifier, there's your a suppression cap no worries whatsoever and the backside here. we've got our mains input directly across here, the clearances just fine. Everything's hunky-dory there.

no problems whatsoever. They've got decent amount of clearance between there I Can't fault that at all. Look like a standard 240 switch mode supply. so I got, but no issues with that.

But let's look at the protection. There's our protection 2.5 amps. It would be different for the hundred and ten. Looks like they've got a mob in there.

don't know. They've got a resistor across that to bleed any residual when you pull the cord out. so that's nice. Everything's hunky-dory There has been speculation that there was bad soldering joints on these and that's what would heat up and start.

You know they catch the fire and everything else, but of course those joints look fine and of course you'd expect them to be fine. My just the one random unit that I opened that dates from March 2002 by the way. Then if the random sample was out, then you'd expect a much higher fire rate than the expected one in 10,000 that Microsoft claimed or even. you know point.

But is it triple Oh 2% actual field failure rate. So of course you know the odds of us actually seen anything in here was zip. Yeah, the only real issue there is the strain relief is just these two little plastic clips. Like that.

There is no other way to hold that except for the two solder joints and these little plastic clips. So every time some kid comes along and goes wham look at that, you can see the plastic in there wiggle like that. The rest of that strain and stress is going to be taken on those solder joints. There's your problem.

But there could be two issues here. You know they do mention maybe some sort of component fire, but the component could include the PCB the solder in the assembly. What not? A 2.5 amp fuse on 240 volts? that's 600 watts? Yeah, I Think that was just a badly Specht fuse which would have been like it's okay for like gross overloads and gross shorts and stuff like that needs to be over 2.5 ants for a significant amount of time for it to blow. It just doesn't magically blow it instantly at 2.5 amps.

that's not how fuses work there. A you know, a there's a thermal delay mechanism plus some tolerance added in there. So so with hindsight in that there might have been some sort of a solder joint fire in here that was drawing power, causing these to heat up and even in the odd case, catch fire, that 2.5 amps wasn't Specht Well enough to do that. But of course, as I said, recalling all of these, all the Xbox units to replace that fuse may not.

It was much cheaper to simply supply an external one. So Microsoft just went. Ah, bugger it. Look, after extensive testing I'm sure they would have determined that 610 Milliamps was the optimal trip point for the you know, the 230 240 volt market unit they determined will protect these Xboxes and stop them melting down if that particular fault.

Whatever it is, which they're probably keeping very close to their chests happen like it's been 13 years at least now. Has the has the truth come out I Don't know if it has, Please leave it in the comments: Well, not only do we need a fuse in there, let's make it electronic and hey, while we're at it, let's let's add some earth leakage circuit breaker as well. just built and braces approach. Anyway, they came up with this, no doubt.

very expensive and massively over engineered response to this particular fault, which is very fascinating. It you know might be a classic example of there's a lot on the line. As I said, billions of dollars on the line and the engineers have to get it right. So they decided, well, belt and braces when we're taking no chances and we're going to over engineer the crap out of this.

And they did. And when I thought I'd do a quick teardown of this sorry, it's been like half an hour or something now. Dry joint crack joint on on the AC in which is pretty important, the buzzing noise we're hearing was probably the arcing of that to the actual board. That was fascinating.

It's so if you liked the video, please give it a big thumbs up. And as always you can discuss down below. If anyone actually worked on this for Microsoft at the time and now you can talk about it, then please let us know. But I Hope you enjoy that fascinating look at some retro safety engineering to avoid a product recall.

It still cost them hundreds of millions of dollars they caused. There might a couple of hundred million bucks, and that's not counting reputation. Everything else happens. Catch you next time.