Reece sent this plug-in thermostat and extension cable after an incident when he plugged it in and the cable immediately started burning in the vicinity of the sensor cable.
Quite interesting analysis of the cable damage with bonus teardown of the thermostat.
In hindsight, the fault may have been initiated by the core being partially cut through, as indicated by the outer damage on the cable.
If you enjoy these videos you can help support the channel with a dollar for coffee, cookies and random gadgets for disassembly at:-
http://www.bigclive.com/coffee.htm
This also keeps the channel independent of YouTube's advertising algorithms allowing it to be a bit more dangerous and naughty.
#ElectronicsCreators

This is an interesting puzzle. We have an amazon, plug-in thermostat with external probe. We have an extension lead and we have a hole burnt through the extension lead. There's a cover note with this hello.

My name is rhys. I send you an email explaining my bad experience with a cheap thermostat i purchased from amazon. It was getting chilly one night, so i plugged this thermostat into an outlet at which point the temperature probe wire shorted with a nearby extension lead. I froze staring at the ball of plasma, which would be you know when that happens.

It is quite scary. Even my friend on the phone ben heard the ordeal anyhow, i figured this would be right up your alley to tear down and diagnose the issue with this device. Assuming there is one. Preliminary investigation suggests that there's not actually an issue with the thermostat, but there's definitely an issue with the cable, and i think it's just sheer chance that the thermostat probe cable was lying in the vicinity of it.

Although i don't actually feel feeling along it uh he did mark, we stood mark. The uh cable put a bit of tape around it in the uh place that had arced, but i don't feel that having removed that, i don't see any damage to this. So, let's investigate the cable first and see if we can work out, what's happened when we go in close to this, are you going to be able to see this? Can i zoom in close, can you see that lovely, clear line there? Actually, what am i doing? I've got a picture of it hold on. I took a picture yeah.

That's that would have been helpful if i just uh cut straight to that here is the cable with a bit of a clean slice across it and then the burning now. Here's what i think has happened because it happens a lot in the entertainment industry when people don't separate cables in the correct way. Quite often, entertainment, you'll find people uh roll the cables up. I mean it's a fairly standard practice and put some tape around it, but sometimes they put too much tape around it and when they do that, let me just zoom out again here when they put too much around instead of just being able to pull the cable Apart, you have to actually have to cut it when that happens.

Sometimes people cut it a bit too deep. So let me show you what happens here. This is a cross section of the cable, a blade gets run across it and it hits the live and neutral. Copper, so this is the blade i'll just show us a standing up.

I'll show us a stand. The knife blade does, that is it called a stand, the knife blade elsewhere in the world, but it cuts through, and if you were to look at that cable afterwards, you'd see that lovely, clean white line where it's cut but you'd also potentially see the exposed, brown And blue, underneath or in summonses brown to earth and as you parted it further you'd actually see the bare copper as well, but that doesn't mean that it's instantly going to arc across here. However, as in the entertainment industry, what happens is that when it gets wet, say, for instance, the cable gets used outdoors or a drink gets spilled or you're spraying a spray or whatever and water creeps into that thin slot there, because a capillary action will always drag It in now you've got a conductive film in here between these and what happens is a little bit of current starts flowing and it's not really perceivable. It's invisible, but there's a tiny little crackling line there and it builds progressively.
It builds a little city, uh bridge. It's called tracking because it's creating a carbon track once a carbon track is enough, it suddenly becomes self-sustaining. It will start flaring out the side of the cable and burning and it coats that in the carbon, sometimes it'll blow itself clearly pop and sometimes tripping the breaker. In the process, or it will just burn and fart or uh in some instances it will uh just repeatedly do it every so often you'll get a pop and you won't if the cable's hidden from view the ziggurat's in there that somebody gets damp, sometimes dry, you'll Get an intermittent fault that it just occasionally just the breaker trips and you reset the breaker and it just lasts for ages, and then it doesn't uh later on it'll trip again the same happens inside connectors.

I reckon that's, what's happened here. So tell you what? Let's get a sharp knife and bear this back and take a look at the inside this cable. So i'm going to cut the cable completely the answer here if the cable isn't actually physically faulty. The answer here is just to cut the cable and re-terminate it into that, but also check the cable along its length for any sound of that in the entertainment industry.

We pull the cable through our hand, not while it's live and you can feel either bits of tape have been left on it or you can feel a nick in the cable a little notch. So let me ring this round. I'm not going to be too careful here, because i don't really care if i cut into the cores, because that's wrecked anyway, so if anything i'll just be super reckless, because i want to cut right in and bear it back like that and i'll cut this back Here i should zoom down for this shouldn't. I that would be better because then you can marvel as it all gets exposed.

So i'm just going to ring this cable around here and then i shall slit it long ways from the other side. So i've ringed it. I'm going to snap the insulation and then i'm going to try and slit it from the other side, notably the earth wire is on the bottom here as kind of predicted. If the earth wire got involved, it would hopefully have tripped the rcd if there was one, if there wasn't an rcd, it wouldn't have tripped it.

Oh into the fact: there's no rcd. So now i've saluted that. Can i peel this off where's my side, cutters. I've put in such a thin slot here that i can't even see it there.

It is let's peel this and expose the carnage inside carefully peering, that back, oh, that, actually the live wire has broken and it has it has actually punctured through to the neutral. I think this is a faulty cable. Either it's been stretched or it's been damaged in some way that it's actually so what i just said there about the nicking between the live and neutral. I mean it might be damaged i'll have to examine this more closely, but in this instance it looks as though uh i don't see, damage significant damage other than surface insulation damage.
It looks as though the actual live wire has broken suggesting. Maybe this cable has been stretched at some point: someone's actually pulled it too hard and that's resulted in the wire breaking, but it's possibly been hanged on by a whisker and just the disturbance of this being plugged in oh of course, it's close to the end of The especially if it's plugged into this uh just pulled that apart and then it started drawing an arc and when it drew apart, that's when it has actually started pulling because it's got a load, it started pulling an arc. I wonder what was plugged into that. Possibly a heater or a other high load device, and that's just finished this off right here now, we've seen that which is kind of unexciting, i could take a picture, but it's really just the brown wire has two broken ends again: a common problem in the entertainment Industry you'll get a bad, cable and uh when you go up to the plug end and you wiggle it like that or the socket end you wiggle it like that.

You can usually find it, but it usually fails at the point of the anchoring here. The restraint i'll tell you what let's take a look at this now, let's plug it in i just i think this is completely innocent. It doesn't mean it's well isolated. He said picking up the exposed metal bit, so this thing has a let mode, psycho, timer, say: d on cd off cd off okay, start time drawn start temperature off, oh and you can set two temperatures and you can choose which you know, i suppose ultimately uh, That means it can either be cooling or heating.

Well, that's quite useful um and how fast does it respond? Pretty fast yeah reasonable, reasonable enough right, let's open it, let's get it open and see if it uses a capacitive dropper or if it uses a proper isolate to supply it, because this is low voltage, cable, it's not something that should be run at 240, volts a Screwdriver one two screws do i have to take this one out, or is that just purely holding this together? Sometimes when you take those out in these assemblies the whole lot parts? Is it going to part? No, it's not going to part. Is there another screw under here? Will i have to spudge it sponge? Oh here we go. What do we have? We have a little switchboard power supply. We have the incoming supply, we have a bridge, rectifier i'll, take a picture of this and we can explore the circuitry one moment please and resume, and it's very modular, let's just zoom down a little tiny bit, not too much.
So what we have on here. We've got the relay, which is actually rated 60 amps at 250 volts. It's not uncommon in stuff like this to find it only rated for 120 volts, but it is really correctly it's a 12 volt relay. This looks a bit out of focus because the focus was down onto the circuit board the chip.

Well, let's start with the incoming supply. Okay, the incoming supply comes onto these connections. The neutral is just bringing the supply up for the switchboard chip, but the live is being switched. You get two lives and they come on here.

The two lives and go through the relay with a anti-tracking slot between those connections and the low voltage side of the relay so they're, making an effort for isolation here, because this is where the transformer is going across from one side to the other. Um. Not really much else to say on the back of the circuit board there, but there's lots to say in here this chip i'll show you the uh schematic for it. It is a pn8355 generic switch mode chip very simple: it's got the bridge, rectifier smoothing capacitor switches, the coil through the transistor in the chip.

It's got the snubber network that just i'll show you that on the thing it generates its own supply from a little feedback winding, and it has uh also a tap off that for the feedback, resistors and uh other stuff, like that, and the current sense resistor. For sensing the current through the windings, the output is just a diode and capacitor. It's just a generic switch mode chip, there's not a lot exciting to them once you really get to know them. So let's take a look at that.

There's the current censorship. You can tell that, because it's a very low value, here's the bootstrap resistor going to the bootstrap capacitor from its own little winding on there. The primary winding has that little what they call the snubber network is going through this diode. It just clips the spike when the coil turns off the transistor in this turns off, and it diverts the spike over to here this capacitor, which absorbs at that spike, but then there's also a resistor across that that gradually discharges it.

The power in the first place comes from the live: the neutrals coming on to one leg of the bridge: rectifier the live comes from the unswitched side through this resistor to the bridge rectifier, and then it's got the smoothing capacitor. The output is this diode to this capacitor. Also on this circuit board, the thermistor is connected to the negative with one connection it's under here. This is a little ribbon.

Cable, there's only four connections, uh. The four connections are 12 volt up in the two middle pins. The thermistor back because it references to the zero volt rail, the negative here and the other one goes through this uh really high value 10k resistor to our g6 npn transistor to switch the relay and there's a little spac emf: snubber diode there that's what's on this: Let's take a look, what it goes up to. It goes up to the lcd board, which has also got the buttons on it.
So this is the set of gold pads that go to the zebra strip and also, presumably with these round ones. It's for the uh. For probing it to test it, it's based on a little sc92fc7 chip. Let me see if i can just find that it's all in chinese, but not to worry.

The main thing is it's an 805 one. That's a really old microprocessor architecture. 805. One base chip uh with 256 bytes ram, eight kilobytes flash 120 base, bytes eprom analog to digital converter, that's ideal for the uh for the uh reading the thermistor and then things like pulses modulation not needed, and the uart the universal synchronous receiver transmitter uh, which is Abs of serial driver, which is maybe being used, i'm not sure what they're using for this, because this is the lcd driver that does all the sort of like the scanning of the lcd and the polarity swapping and all we've got here - are just a few data Lines coming across data and clock.

Probably so here are the four connections that come up here. We've got the 12 volts uh, the 12 volts comes up and it's got a decoding capacitor and then it goes through a 470 ohm resistor to the voltage regulator. The reason it goes through a resistor is just because then the resistor is disputing some of the power and the voltage regulators are dissipating some of the power it just shares the load. It means the regulator doesn't get too hot because it is dropping from uh the.

I mean it's not under really heavy load anyway, but it is dropped from 12 volts to 5 volts, which means you know it's dropping 7 volts across that, so that will just share the load loads, more recovering capacitors, one, zero ohm link uh, and this little bit Of circuitry here is a another decoding, capacitor for the thermistor coming back up and then the resistive divider that goes from the 5 volt rail to the thermistor. The thermistor goes to zero volt rail and then that way, this chip can actually measure the temperature from the voltage coming back from that, as the resistance of the thermistor goes up and down what else is there? The output from this pin goes to the this connection. In the ribbon, cable - and that goes out down to the relays, transistor we've got this 2.2 k resistor, which goes over to this led that is it. The buttons are just straight connected to the microcontroller, with internal pull up, because that's a common, zero volt rail there.

So when you push the button, it will just pull that input down low. That's it it's a very textbook design, all the secret stuff, all the secret sauces in the chip, as it always is so uh that is it. The fault was just a cable that may have been stretched at some point when it didn't get quite reached as far as it would was wanted, or it might have been damaged in a in the factory who knows but uh. Certainly it had split, and certainly certainly when all it took was a bit of disturbance to finish that off or the switching of a heavy load.
Maybe - and this is uh - presumably working then assuming that little ribbon cable has survived me, yanking everything out and presuming the the zebra strip actually makes connection again. You know what it's like when you work with these things you take it out and the zebra strip never connects again, but i shall try and put this together and i'll see if it works i'll. Do that right now. One moment please! Well, it took a couple of goes, but the zebra strip has now made connection.

It didn't make connection very well, the first time and the unit uh seems to be working. Okay, so uh yeah, that's it fixed, and that was a problem. So nothing really sort of significant. It certainly wasn't this unit that caused that it was just a damaged cable but quite interesting.

Taking this whole lot to bits anyway, and just seeing what makes it tick.

12 thoughts on “The case of the mysterious cable arcing”
  1. Avataaar/Circle Created with python_avatars M Bak says:

    Any chance of an investigation into zebra strips and ways to resurrect them? I have recently had to throw away a couple of kitchen appliances who lost their displays due to zebra strips giving up the ghost. Pressing on the display does not work when the item is a set of scales.

  2. Avataaar/Circle Created with python_avatars HiddenWindshield says:

    I was once trying to trace an audio (computer speaker) cable through a rat's nest of wires somebody had made behind their desk. As it turns out, "rat's nest" was more accurate that I'd have liked, because some rodent or other had chewed through the insulation of the computer's power cable. The copper was exposed but not cut through, so the cable still worked, until I pushed it aside and the copper brushed up against the bare metal desk leg. Loud bang, bright flash, breaker tripped. Fortunately, the only casualties were the cable itself, a scorch mark on the desk, and a… well, let's just say "mark" in my pants.

  3. Avataaar/Circle Created with python_avatars Roger Norton says:

    The Chinese Remote control version is worth examining. It works well although the transmitter (2xAAA) batteries need changing every three months. It has been known to migrate to the freezer mode (indicated by a frost symbol), from the heating mode (indicated by a flame). When it sends the temperature signal it flashes Green which can be annoying. I tend to keep the transmitter in a cupboard!

  4. Avataaar/Circle Created with python_avatars Scott Marshall says:

    I always marvel when it all gets exposed.
    I spent 40 years in industrial control, and more than once I got a roll of wire where they ran out of conductor and simply loaded more and kept extruding. The result is a roll of wire with a break or a simple contact connection somewhere in it. Maybe there's another explanation, but I've also found a "tape together" (not a splice, just the 2 wires taped together) in the middle of a 500' roll. None of them are good, especially when pulling into conduit.
    It could be this cable had a bad roll of wire that then went on to become a cable. Hard to believe wire manufacturers can get away with such practices, and well not common, it's not entirely rare either. To this day, I always ring wire on the roll when it's delivered, and keep a sharp eye (and hand) on the wire when pulling.

  5. Avataaar/Circle Created with python_avatars Chris's Garage says:

    I electrocuted myself on a building site because of a faulty cable. It was wet and I was coiling it up whilst it was plugged in, I got a shock, yelled out and tried to drop the cable as I let it go I got another shock yelled again in a higher pitch and tried to drop it again. This happened about 5 times with my voice getting successively higher untill I managed to drop the cable. Turned out I'd blown power to the entire site. Inspecting the cable I saw a very fine slice in the insulation. I was totally fine and I learnt a valuable lesson!

  6. Avataaar/Circle Created with python_avatars Greg H says:

    This happened to my hair dryer, spark came out like fireworks. I later found that the outer insulation was a normal beefy looking black sleeve with blue-brown inner insulation, but the actual conductor was so small an earbuds might even have more cross sectional area than those cable did..

  7. Avataaar/Circle Created with python_avatars Furrball Nirpaw says:

    I have an Amazon item you can tear down. It works just fine other than the angry pixies that tickle you from the usb metal exposed on the connector. Also tempted to upload a short video where there is a light switch blank on the shower wall. And on the other side of the shower wall the light switch to turn light on.

  8. Avataaar/Circle Created with python_avatars carlu bambi says:

    I don't know how many power cords I have seen go up in smoke ,slithering on the floor and melting and popping and crackling ,just like fireworks only with the smell of burning rubber .600 v can go off with a boom ,especially when run over by a lift or other machines ..we had a 12 /4 SOOW cabtire go off like a shotgun when someone with a towmotor decides to run it over .was on a 600v 3p 20a circuit .took out all 3fuses ,luckily

  9. Avataaar/Circle Created with python_avatars UpLateGeek says:

    I had a slightly different thing happen to an Ethernet cable when I was pulling through another Ethernet cable under the floor of a our data centre at work. The outer jacket was pretty aggressively conformed to the twists of the pairs inside, so it was particularly ribbed (For her pleasure? Because it wasn't for me!). When I pulled it through, it happened to run over a perpendicular cable, which evidently didn't have a very tough jacket, because it gouged right through it and then went through the insulation on the pairs.

    Only found out about it later when IT started complaining about a server becoming unreliable. Found the gouge when I was tracing out the connection from the server to the switch, so obviously I replaced it immediately. When I checked the port logs and stats, it was constantly going down and up, and lots of packet errors. The problems went away with the new cable.

    Needless to say we switched to a less aggressive brand of cable.

  10. Avataaar/Circle Created with python_avatars Joe Pollard says:

    A good example of where an AFDD would be useful. A series arc in an active won't trip the overcurrent or residual current protection (unless the ensuing damage eventually gets the neutral or earth conductors involved), but an AFDD can detect the arc by monitoring the AC waveform

  11. Avataaar/Circle Created with python_avatars Weißenschenkel says:

    People yanking off appliances' cables like my mother (a lady Born in the 1940's) will eventually face the same issues, unless it's a rugged coaxial cable. When I told her to do not yank the cables she started to pull the plugs with appliances like washing machines and other induction loads while draining current, melting not only the plugs but the power outlets themselves, due to arching.

    I asked her why she wasn't ever switching off the appliances. She had set one extension cord and one outlet on fire. Gladly the breaker switched off and avoided something worse to happen.

    She also plugged her vacuum cleaner in my no-break outlet and blew a fuse, rendering my file server unusable while I was trying to remotely connect from a client. Sigh…

  12. Avataaar/Circle Created with python_avatars Squelch's stuff 'n' things says:

    I think you were mostly correct in the first instance. The outer sheath definitely looks like it had been damaged at some point, although probably only partially cutting the live conductor. If the load was heavy, or has a high inrush, it may have just been enough to burn the remaining conductors as the current heated what is effectively a fuse. Fireworks then ensue.

    Interesting coincidence with the timer-stat though. Thanks for sharing Clive.

    [Edit to add] The V shaped cut on the outer sheath looks all the world like it has been trapped under the corner of something heavy at some point.

Leave a Reply to Weißenschenkel Cancel reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.