It's our favorite type of electronics, it's dead electronics, particularly with hot melty bits and a real electronic stink. So this is a passive infrared detector, typical type you might find the garden and it comes the story: hey clive, this pir operated a water feature. The water feature stopped found the mains three amp fuse blue. In the connection unit.

I put a new fuse in looked at pir and i'd lit up in the top left corner. I marked in purple, so it was actually glowing in here arcing. Presumably, it lasted about six seconds in total, obviously, something inside burning uh, which can't see, but on the top, has a melted. Okay, let's explore.

So this is a typical detector. It's got three wires: brown, blue and uh red, usually brown blue, are live and neutral and red is normally switched feed. Let's get just lop these right off and get them out of the way. I shall reduce the size of the unit a bit so after those because i don't think this is getting used again.

It looks as though it's gone smokey. It's also fetching shade of black, which is never terribly convenient for uh, visible things. Oh so the first thing i'm noticing is that normally there's a couple of screws holding the bottom here on. Oh, it is stinking so much you if you this was fitted with smell-o-vision you'd, be getting this right now, because this is stinking.

Oh, this is glued together. Let's get the vise of knowledge in and we'll squeeze it until it pops, so i shall try squeezing the base, not that i'm trying to like keep anything intact because, let's face it, it's dead, oh, that is, that is actually quite tightly together right. I'm going to have to i'm going to. I know what tool to use here.

One moment please pipe grips they're so destructive. Let's put it in here and do a squeeze that's making crunchy noises. These will open most things destructively. I mean: let's face it's not going back together again.

Is it oh that sounded reassuring in a sort of destructive way, let's stab a screwdriver into it, without stabbing a screwdriver into myself. This is not always guaranteed. Try and open this up a little tad more. Oh, this looks promising, so the inside is coated with oh, the smell, the smell is absolutely terrible.

It is stinking no okay, it's the really right. Okay, i have to get you a closer look at this relay one moment. Please right. This is complex uh because the fault could have started in three places, but i think it started in the relay down to the fact that there wasn't enough clearance in this contact.

This is a side view of the relay it's very, very burnt. The side is physically burnt off from the little fluffy fibrous stuff here, i'd say it's probably glass, reinforced plastic. The really has a little contact bounces up and down here, and that is this bit here. It goes up touches the contact here which wasn't in use uh, and it also has another contact which was the normally closed con.

The normally open contact, the one that would actually bridge when it comes down, and it's here and i get the feeling that it's arced between the common contact and the normally closed contact and that arcing has been sustained. Even the relay has possibly been off uh because it's just not been given enough clearance to break the load. There is a snubber network, uh i'll show you the the remnants the snubber network as well. But let's take a look, that's the capacitor.

It was right next to the relay they've, it's very hard to say, which went first, but this one is a little a little adorable little hole in the side where it's obviously shot a flame out. Whether it was this that set fire to the relay, or vice versa, is very hard to say it's one of these things that if the i'd got a verb, one that had just started flickering or going wrong, i would have been able to tell a lot more Uh, but this is uh - this is toast, as you may see, i shall put this out the way. It's also extremely stinky, because that uh plastic is very burnt and crusty right. The circuit board, quick, summary of the circuit board and we'll marvel at their token gesture.

Um, andy tracking slot, there's the token gesture anti-tracking slot. It doesn't really do an awful lot. Let's zoom down this a bit. So here's the front of the circuit board - and here is the back of the circuit board just cause.

You want to trace things out yourself. It's all a bit black down here, things worth of know before we go any further. There is the token gesture and the tracking slot between live and uh. The coil in this instance here plus also the track wise.

It's the live and neutral, but look here, they've actually just run it right. Next to the the neutrals just pass right next to the live, with a very thin track that has done its best to blow up, but the neutral continues along and uh. It's quite interesting. Their use here why they've continued neutral down here, because in reality they could have done it differently, but it's actually complicated by led loads.

That's not clearing things up. Is it let us uh move that out the way, but there oh there's the snubber network by the way, a resistor of unknown value, because it's cremated and the capacitor, which is 22 nanofarad, which is quite low for us, never network. Here's the circuitry a quick summary of the circuitry itself, a capacitive dropper, a bridge racked fire underneath a zener dart, probably 24 volt clamp, smoothing capacitor voltage regulator, feed a biss-0001 generic passive infrared chip, which has the light sensing facility uh and timing with these capacitors and Also, the uh, the actual pass infrared sensor, which just puts a low level signal output that gets amplified by this chip, and it looks for undulations as people walk by and then triggers when it reaches the threshold. It triggers a timer.

Everything is built into this chip by changing component values. You can tweak and adjust your heart's content. The unit turns on the chip turns on what the smoking remnants here with this transistor a 10k resistor in series of the the base of this transistor, and then it switches the coil. This little diode here is across the coil to protect that transistor.

The relay here's the other bit where it could have failed. It's possibly the heat that came down this track. They've er, the other, normally uh closed contact, they've just cut it off the bottom of the relay switch just sat on top of the other side of the circuit board, but this area here has a very obvious sign of tracking. Let me just zoom into this bit, because this is a bit we're looking at now.

I shall make sure i'm still focused on that. This is where all goes wrong with it. Suddenly, it's not in focus, so here is the uh feed it to the middle pin. But that's the common contact of this really and here is the normally open one that then feeds over to the output and there's also that little snubber network with a capacitor from here to here and then the smoke remains a resistor over to there with tracks blown Off and everything so here's the other possibilities that could have the possibilities that could have failed.

We've already had the relay kind of burned up inside. I think that's a prime suspect uh we've already had the uh, the snubber network capacitor could have failed, but it since sears this resistor and that should theoretically have kind of blown clear like a fuse. But now we've got the uh the neutral that was brought over for that snubber network is right next to the switch live, and there is very obvious signs of the pitting and etching into the circuit board. It's been tracking so that, while that was on uh, it may have actually just continued to burn, particularly if humidity got involved, um but then finally, just actually flashed over and when it flashes over.

That's when it will blow the fuse, sometimes it'll just arc and fizz. For ages before doing that, and sometimes you put the fuse in, it will have blown itself clear and it will just start fizzing and arc in the background again until it flashes over later. On that happens, a lot um, so let the snubber network. Let me just zoom back out again and we'll talk about the snubber network.

I shall bring in a notepad. Quite often when you have a relay contact. Let's uh just draw it like this. Quite often, you will have a snubber network which is capacitor and a resistor across that, and the point of that is that uh this would typically be 100 ohm and this would be 100.

Nanofarad is very typical. You actually get dedicated snubber networks. I was just looking for one here lying next to the bench: no, it's not going to be there. It's not going to be there when i need it.

I kind of want it, but right one moment please a snubber network. This one is 220. 9500. 100 ohm.

It's just a capacitor like this inside in series with a resistor all just built into one package, and this is normally wired across contacts and contactors, and here's the idea when this contactor, when this contact opens. Sometimes you get a little bit of an arc occurs between the two contacts and because the capacitor uh has it happily, the higher frequencies it'll actually pass them easier. It effectively provides an alternative path for that, and it takes a lot of the current away and lets that arc extinguish the resistors here to limit the amount of current can flow through the capacitor effect in that way. Also it's to limit having a capacitor across the contacts, because otherwise it can cause pitting.

So it limits the current. What they've actually done? Supposing this is live and it was going out to the load and there was neutral going straight out of the load load, neutral load live. What they've done is they've put that capacitor across there instead and the reason they've done. That is because, if you have a standard snubber network across a contact, switching an led lamp that 100 nano far 195 capacitor will easily pass enough current to keep that lamp glowing or make it start pulsing.

So that's why they can't really do that with a with very sensitive loads like compact, fluorescents and leds. So that's why they've put that there other things worthy of note. Quite often, if you have a circuit board, that's designed properly and it's got the relay on the board to switch a load to an output. You'll often find that there's a little thermal fuse mounted right up.

Next, the relay sometimes actually stuck onto the side of the case. The reason for that is that, when something like this happens, let me just grab the charred remnants of that when something like this happens, and the case starts burning up, because the the contacts are arcing inside or just make a bad connection result in the arcing. This fuse will break the circuit and uh, if that's say, for instance, leading to the common connection in here, then it just kills that part of the circuitry, sometimes it just kills the whole unit completely. When that thermal fuse goes that's when things are properly designed, this design is so perplexing, because it's got it looks like someone's copied a design or just based on a standard design, but their grasp of the vicinity of things like neutral and live is just not there.

They've not had that understanding. Likewise, the really just uh, it just seems so small, a lot of stuff you get from the grey imports from ebay, uh you'll find these relays are not rated even for in the case the uk 240 volts they're only rated for about 125 volts um, and they The current ratings are always ambitious. You know they say things like this. This can switch 10 amps.

In reality, it's not going to switch 10 amps in the case of some of the modules, but interesting, so without actually catching it earlier. I can't tell if the fault originated in the relay, i'm going to say that i think it is arcing the relay or if it's a little snubber capacitor, which has kind of blown out the side here and cracked. So it's obviously been pretty unhappy. But then it's been next to something that was on fire or this tracking down here, but this tracking could have been initiated by the heat from this pin getting red hot uh, causing heat to travel across and burning the actual circuit board itself.

The underside of that coil is pretty well, it's pretty black. The whole lot's been going up, but that's fascinating, interesting stuff. It's all always worth looking at designs like this and how they've failed, because it shows you what can go wrong in this case. Just for that tiny bit of extra space that little thermal fuse, even if they butted the thermal, fuse up the end here and moved the snubber elsewhere.

That could have actually saved this whole thing from happening. But there we go an autopsy on a completely smoked: um password infrared detector.