It was interesting to note your comments on these relays. I bought a voltage controlled relay for charging a second battery on my van, supposedly rated at 20A per circuit(2). The cover was just a simple sleeve and despite the attempt to scrape the markings off the relays it was possible to make out that they were only 10A rated and probably cheap ones actually only capable of carrying at most 5A. He was extremely rude when I started an eBay return. Even said most people just contact him and send things back (although he hadn't responded to me contacting him). My bet is that he claims he's never received the item back.
Ooof. I hate dodgy sellers. I used 12v planet and got a basic Victron unit. So pleased with it. I haven't bothered wiring in the emergency jump start feature but the charging works brilliantly (as one would hope). I did put breakers either side, rather than fuses. Thought process was that I could isolate eithef battery and charge with solar individually, otherwise the split charge relay could send the puny amount of charge from my small solar power to both batteries at once and I doubt I'd see any benefit.
A lot of the time the ebayers have no idea what they're selling. They'll just look at what they have and try to look up similar items and copy their stats. I purchased a LiPo battery charge manager that stated it could work as a "UPS" (uninterruptable power supply), only to find out that the module, when input power fails, would lose its output for 2 seconds before the battery would pick up the load. They tried all sorts of BS to get out of it, even offering to send me more of the modules for free so long as I'd give good feedback.
@@ConstantlyDamaged Yes, I've had that too with Chinese suppliers but in the case of the relay they were clearly built by the UK seller and he just didn't care that he could potentially set someone's vehicle on fire.
If the electronic item you are purchasing is important (i.e. outside of just playing around with Hobby stuff) I always find it better to purchase from a traceable source such as Farnell / RS, who will usually be buying directly from the factory
Splendid! I like it when they use those "Switches 10 amps, at least once!" relays. Being able to enjoy the audio/visual indicators that signal their operation, slash demise, is always a treat. Cheers!
C13 also got a hole burned into it from the arcing. Replaced plenty of those little Songle relays from them cooking the contacts, the 10A rating is very optimistic indeed on them. Years ago I needed some snubbers for a contactor, so looked up the price of the manufacturers one, sat down, then looked up the price of generic ones from RS. Also sat down, then went to the one local supplier and got 3 68n 4kV polypropylene flyback capacitors, normally used in TV set flyback use, and some 33R 5W resistors. A little overkill, but guaranteed never to fail, as those capacitors normally run 2A of 15kHz current through them in the TV set, so 50Hz mains would never even bother them. Resistors 5W because they were not much more than the 1W versions, plus had longer leads, as this had to go across the pins of the 11 bin base, so long leads were good. Put some supply filtering in as well, 3 4uF 450VAC motor run capacitors in star, so as to handle the current spikes from the contactor as it was rather rapidly switching, which is why the snubber was there in the first place, wanted plug in relay to last longer than 6 months. Old Brown Boveri took 40 years to fail, but seeing as the new ones were Omron plug in, much better than needing to do wire surgery, just plug in the spare. Old Brown Boveri failed because the steel of the core mushroomed out, jamming the core up in the housing, not from the contacts failing. When they said 16A there, they meant it, not will switch 16A, or will do 100k cycles. Brown Boveri meant that, at 16A inductive load, and 100k cycles, they will not fail, and would still look almost like new. They probably were well over the 50 million cycle count. Not made like that any more.
This looks to me like a case of over current…. Perhaps a seized pump ? And a 13A nail in the fcu ? These pir things usually power up in the ‘on’ state …….. awesome videos Clive, keep up the good work!!
I think the relay contact got hot, burnt up the PCB over time and compomised the isolation properties of the PCB up to the point when the arc jumped the carbonized PCB... The PCB design is dreadful.
Most of the relay failures I've saw were due to arcing pitting the contacts, which go resistive and overheat. That's typically due to a shitty or worse, absent snubber circuit. Motors are inductive devices, so they'll generate counter-EMF voltages, causing the arc. Whoever specified the device and whoever designed the unit need to be dropped into a solder pot.
This has brought back some very vague memories of working on an unit that had a plastic housing identical to this including the trim adjustments. I do however remember it having a through-hole multiple opamp, can't remember if dual or quad, but it was a chip I had used in university in the analog labs so I recognized it at the time.
On the subject of snubber networks, old rotary dial phones had what we called the spark quench circuit across the impulse contacts. They were also found all over equipment that used relays, which is why it was so rare to find burnt contacts causing faults. Of course, failure was usually in the form of what we called "low insulation" or LI faults, and was never spectacular, just noise (crackling) over speech, or where no speech CCT was involved, strange behaviour of equipment.
I have on of those on my garage light. Installed ages ago. I have this urge to take it apart and compare it to the circuit board design of the 2018 one in your video.
I recently replaced one I put up over 20 years ago. I think I got my money's worth. To a rough approximation (I didn't open up the new one) the design hasn't changed. Most of the time, these units can't set anything on fire in the first place.
I must comment, I was just watching you tear apart this unfortunate PIR unit, and my mind was wandering. I thought "Well, dissecting a HUMAN is called an Autopsy, dissecting an ANIMAL is called a Necropsy, so what do you call dissecting an ELECTRONIC corpse? Why, a Techropsy, of COURSE! lol LOVE your channel, you're always a hoot!
It's fascinating to me that someone would be able to design an operational circuit without the understanding of mains clearances. I'm just a little hobbyist, and any time I've built something for mains voltage, I lean toward the ridiculously cautious. Sure, why not have 10mm of clearance, and add slots between just in case.
Very interesting and odorous item! Do you think that the relay's demise could have been hastened by having to switch on the locked-rotor current of the "water feature" pump motor; as opposed to a light which this was surely designed for? Or; at turn-off; the inductive nature of that same motor might have overwhelmed the voltage rating of the snubber network capacitor?
Many pir's are designed for a couple of 100w lamps at most, but people use these on 300w (or more) halogens, i've seen several that the contacts have simply overheated (or the circuit track) It's much better to use a pir to control an external relay, and let the relay switch the load Incidentally, i just found a similar one that's advertised as 1200w, but zooming into the label it says 300w (on amazon)
You could have a look at some of the Shelly products. Especially the Shelly 1 or the Shelly 1 PM. The actual relays inside seem so miniscule for what they are (aparrently) rated for
I try to assume that any relays that are not from reputable sellers are going to be no more than 12.5% of the rated value...I used to assume 25% but to my detriment :(
Going to guess Songle or one of the many clones, though Songle generally is a pretty good relay, provided you use correctly, which means 1A at mains, with a snubber across the contacts, and a MOV as well on input and output. Or 10A at 12V, which they actually will do well. Not Omron well, where you actually have a warranty of 100k cycles plus, but at least should be able to get to 10k cycles before something fails elsewhere, most likely the capacitor dropper power supply.
Most cheap knockoff relays do be grand, I've used fake knock off omron clones, it comes down to the design mostly not nessarly the quality, it's all cheap shit from China, when you pay for quality what you're really paying for is design
I've been using FINDER relays recently...seem to perform as they are rated...a 120VAC@10A easily runs on my 8.3A PSW inverter used for my fail-safe computer power supply. It is set at NO on MAINS and NC on INVERTER...so if the mains power fails it goes to NC and the inverter takes over...the computers never even notice the switching.
@@gregorythomas333 Computer power supplies must stay online for 17 milliseconds before their output destabilizes, according to the ATX standard. I've never had a normal consumer UPS fail to successfully switch to battery power in that time. But over-engineering is fun, I suppose
@@gregorythomas333 Problem is most small relays are not really rated for mains switching on both NC and NC contacts, especially if there is a chance the 2 contacts will be out of phase, as there is a good chance the arc will be able to conduct enough to short the 2 contacts together in the arc.
In our 5th grade classroom, a kid next to me said "I smell burning raisins!" moments before a florescent fixture above popped and smoked. Every time I smell burning raisins, I recall that kid's face.
We Merkins in 120 VAC land have it easier than the rest of the 240 VAC world. Things are not quite as catastrophic and 'fun' in the fireworks sense. We do have failures but with half the voltage, the results are not as spectacular! My torture toy when I was growing up was an old TV transformer not much smaller than a microwave oven transformer. It had a HV winding that put out about 700 VAC - I didn't bother with the center tap. 😂😂 I tortured a lot of 1 watt carbon comp resistors with it. Thanks, Clive for bringing back the memories of the smells.
Good explanation of the use and reason for snubber networks. You'll find them across semiconductor switches too (mosfets, bipolar transistors, IGBTs) too with inductive loads to limit the rate of rise of voltage dV/dt from the flyback effect when the load is switched off. Before they were as well understood it was not uncommon just to find a capacitor with no current limiting resistor across switch contacts. I had an amplifier like this with a C across the mains snitch directly and on an occasion I was doing some servicing, found that this capacitor had obviously been subjected to a large current that had blown it into 2 pieces and the foil inside had unwound like some Christmas decoration. It's now in my rouge's gallery along with other remarkable instances of total destruction like the KT88 power output valve that dissipated so much power that the anode melted and fractured its glass envelope due to thermal shock and an Altec compression driver diaphragm that was so overloaded by amplifier clipping that the aluminium dome ended up like confetti. Curiously the driver still made sound since the voice coil hadn't itself burnt out but it was fabulously distorted.
I'm curious if this PIR matched the power of the pump that ran that "water feature" and whether its inrush current ablated the relay contacts, since those things can be triggered repeatedly by birds and falling leaves etc..?
curious about a PIR on a water feature - haven't seen that myself yet. Normally they're on a timer, to get plenty of aeration. So this would only operate pump if someone approaches - or maybe it's a cunning anti-cat device!
What an interesting experience. For a moment I thought I got smell-o-vision. Turned out some dump truck next door somehow started burning up a belt or whatever it was, while I had the window open. Have to stop watching the video, have to get some coffee and snacks and watch them try to get the thing back running again. Molten belt seems to have fused quite nicely with a melted pulley and they don´t even get the thing started again.
The leakage current from that snubber circuit is more than enough to drive a solid state relay, but not a mechanical one. Leads to some fun problems when "upgrading" equipment.
There's nothing like a good catastrophic failure. The more spectacular, the better. "Sniff... Ah I Love The Smell Of Burnt Circuit Board In The Morning"
The cause was probably the water feature (likely pump, inductive load) overloading the relay. Such mains automation things usually have two figures: for resistive loads and for inductive.
Near the beginning the live connection in the choc block looks like it might have gotten rather hot, and the insulation on the live wire going into it looks as if the cable has gotten hot as well
Thanks Clive.... I've been filling up the comments sections of my favourite electronics channels with "can you explain snubber networks please" for ages.... :)
I raided a couple of PIR modules from my dad's outdoor light assemblies when he bought self contained LEDs. One has a mechanical relay, the other a solid-state relay capable of half brightness. The first one, I'm going to "upgrade" it by having the line output push a relay or contactor. I've got enough of those laying around that I've got options. Maybe that 4PDT, or that NEMA 0 3ph...
Just a thought, but how about cutting away the charred embers of the PCB, and see if the rest of the circuitry still works, or could be made to work? Purely for the fun of it. We know it's not going back together again, but pure interest/curiosity makes me want to know how much actually survived, since we already know what didn't.
The first thing I would check for is a short or overload in the water feature. Excessive amperage could easily start the failure cascade you see on the board. Of course, an underrated relay would produce the same effect.
My mental image of "The Clamp Of Persuasion" has changed for the worse since what may or may not be the start of World War III. As if the Scots have not had enough to put up with already!
I work with security systems and a lot of access control equipment comes with something similar to the encapsulated resistor and capacitor which you showed (or it's recommended that they are fitted). Most of our kit is 12 or 24 volts DC and those the snubber devices are fitted across the relay terminals and across positve and negati0ve near to the lock, which tends to be an inductive load like a solenoid bolt or a maglock (electromagnetic lock).
The real problem with having such inferior products in the market is they create a demand for cheap goods, which creates pressure on otherwise safe product manufacturers to cut corners. On top of that is the fact that its impossible to tell whether something is safe or unsafe without doing an expert teardown. (Price is a good indicator but insufficient, because nothing stops a disreputable vendor from marking up the unsafe cheap products to the same level as the safe expensive ones.) We are going to have a number tragedies reminiscent of 19th and early 20th century West before this phenomen get fixed. Which is why BigClive's channel is a service to humanity, if you think about it!
This is actually where the standarts should theoretically come in, but as you can see virtually no one certifies their devices for anything. It is sad, i think it should be a more widespread concerne, but i doubt that it is gonna go anywhere without anyone pushing it, like with right-to-repair. If not for Lois Rossman i doubt it would go anywhere.
I try to look for certification marks (Intertek, UL, GS etc) when shopping for anything that plugs into the mains, but I don't really know how much that helps. Edit: I know that CE means absolutely nothing in the real world. Same with RoHS.
I think the clue is in the original covering letter "Water feature" to me indicates bl00dy big inductive pump, I suspect the relay contacts draw an arc when switching off creating so much heat the PCB turned into a resistor. If I was Toni Peters I would be checking that the new PIR is rated correctly for the water feature Load.
I was confused by the water feature aspect. Home made cat scarer maybe? Or just wanting to impress the posty with a triumphant fountain but not use too much electricity? We need more on the backstory for sure.
Street light collector Mike Barford from streetlightonline gave me a Royce Thompson Monostar 1000 NEMA photocell that had water ingress, this had caused tracking across the Common and NC relay solder joints, in turn burning the PCB under the relay to a crisp and melting the relay, it also burnt a hole in the casing of the cell. I managed to get the thing working just by replacing the relay and bridging one trace, those things are bulletproof!
I've taken several of these sensors apart to modify them for low voltage operation, so I have a sense of how they are constructed. It would be helpful to know the operational history of this particular unit. Did it work properly when it was first installed or did it fail immediately? I suspect that it had been working OK in the beginning and failed sometime later due to water infiltration due to rain/fog or other sources. This enclosure is not hermetically sealed and there are multiple paths for water/moisture to enter. That could have initiated some arcing on the PCB under the relay and which lead to the meltdown. Also, switching the reactive load of the water feature pump motor (without proper snubbing) is also a likely contributor to this failure.
And suddenly I vividly remember that foreman during my apprenticeship that would rip screwdrivers out of our hands if the size was like one size off for a screw… also he told us screw drivers aren't prying tools. We really learned a lot.
Those relays don't like capacitive or inductive loads. That's why finding a PIR that allows more than 100 W LED or 5-8 light sources is hard to find. You will need a contactor to take up the load. I don't know if the astro clock I put in will survive since it controls a lot more light than I like. But the customer is always right.
I don't think these devices were ever designed to be used with an inductive load (water feature pump as mentioned). Have used many of these with a great deal of success but only with low powered lighting. Yes the tracking dose look a bit skimpy but I don't think that's it. Remember these things are built to budget and once in a while a defective component will show its head. Great video by the way 👏
I'm wondering whether the failure was caused by the diode across the relay coil. When the relay is de-energised, the diode clamps the back EMF and prevents the magnetic field from collapsing quickly. This causes the relay contacts to open slowly, which can result in lots of arcing and sparking, and degradation of the contacts, increasing their resistance. With time, this can cause overheating and even fire. It's so much better to protect the driver transistor with a zener diode between collector and emitter, with value chosen to clamp the back EMF below the Vce(max) of the transistor (rather than 0.6V with a regular diode). If you choose a high voltage transistor with suitable zener, the relay will last much much longer.
2:15 that's usually my response when somebody asks me if I could open something for them. "Its not a matter of being able to open it but a matter of being able to close it again." 🤭
This is almost exactly the same PIR module that i used to revive an old halogen motion detecting lamp - and due to the magic of having bought two, i have one open here. Mine has a date code of 2019-12-29 and is thusly slightly newer than the burnt up one here, but yes the non-existent spacing between the relay output and neutral is still here in this one. (Now feeling that if i ever use this spare module, it will be with some extra safety mods... like cut the neutral near output, use a seperate snubber, add a fuse..) The relay model in mine claims to be a DAOHENG DHT78-A-24D-SS, rated (or "rated") 5A/250VAC, 7A/125VAC, 20A/14VDC. (Hopefully it is better built than that burnt up one lol). Looking at the PCBs, the differences between the one in the video (2018) and mine (2019) are mostly PCB component options, mine has one extra hole for different length resistors for R1, and the removal of the through-hole option for C9 (that one is a bit odd, why remove unused pads? just to drill less holes? hm). Also my CDS light sensor has been mounted up way higher (next to the PIR, not down low) than the one in the video.
20 years ago one of those relays stuck closed in a bread machine, had the heating element on for some time and smoked out the apartment pretty bad. The bread machine never caught fire but it turned the bread into a small cube of carbon carbon. Since then little tiny 10 + amp budget (or now even grey market) relays always scare me when they switch any actual current (over 2 amps). If I could I'd like to rig up a good test setup with lots of amps and try a few.
Interesting internal construction. I'm using a bunch of eBay 12V PIRs to control external LV LED lighting. Ours use two PCBs, the IR module and the 12V PS and relay PCB. So far none of ours have let out the magic smoke but we have had a couple of dead ones over the years.
Head on the nail there. That relay wasn't up to the task (under dimensioned). It started to arc, and the rest was just a slow chain reaction. I've seen these very exact devices myself, and they've failed the very same way.
That looks like it didn't enjoy an inductive load. I don''t know about over there but here in Australia a lot of those water feature pumps are low voltage with a transformer to supply them. In any other application than a single lamp I use a real relay to switch the load.
I realize I'm a little late to the party, but you missed the charred bits on the output plug that you snipped off at the start. The red wire's plug end was charred, so I'm guessing this unit wasn't rated for the current that the attached water pump(s) were drawing.
very likely, whizz poor contact materials within the relay failed, overheating and melting the solder at the board, then the board starts charring and ZAP arc over. this happens on tons of appliances, (fridge, freezer, washer, dryer, dishwasher, oven, microwave, hvac boards, thermostats, garage door openers) yeah I could go on and on, but that's plenty to get the point across. alternate ways, thunderstorms and lighting arced across leaving a carbon track. it got water inside and fizzled away.
I don't trust those dubious relays. I recently bought a temperature controller and replaced the unbranded 250V 10A relay with an Omron. The relay footprint looks the same on your pcb here, the Omron G5LE series matches it.
Is it possible the end user had the load wired between Load live and an Earth? Also you mentioned it was running a water feature, was it running a pump and as such an inductive load? Combining wrong wiring with an inductive load you could get huge spikes as the fields collapse across the windings of load or coil.
Could it have been a bad solder connection for the relay that caused all the heat? Looking at the photo the most damage seems to be where the connection is to the board.
That cap had a problem I've seen a bunch. You see, at the factory where electronic components are manufactured, they hermetically seal them with a little smoke inside. So, if they get a little hole in them, the smoke gets out and they don't work any more... A trainer at a new job many years ago told that tale to a customer who was an auto mechanic. The mechanic's high tension probe had quit (with a mysterious fan blade mark on the side of it). He'd connected the primary test leads directly to the high tension coil wire (should work in a pinch, right?). The 741 Op-amp that was straining to 'hear' the 100mV or so coming through the filter board received all 30KV that jumped over said filter right between the eyes. After that, when I was confronted with both primary and secondary wave traces missing, I'd look at the hall-effect secondary probe for high speed impacts and then tear down the scope to look at those 741's. I even saw one where all there was in its place were 8 pins standing there in testament to its existence and a tiny scorch mark... I suspect use on a C3I (computer controlled coil pack) at closer to 100KV... Poor little bastard never stood a chance. :) We often referred to this phenomenon as 'A Failure of the BDU to RTFB' (I was often tempted to put that on the repair invoice, but somehow refrained... :)
I have heard that you can buy the equipment to put the magic smoke back into components, but it's very expensive, and on a long lead time. Oh, and I also heard that the magic smoke is currently on back order due to a world-wide shortage of Unobtainium.
@@wiseoldfool Yeah, I've heard that. I myself have run up against this supply chain issue... :) Thanks Barry, didn't realize it was that damn Unobtanium being so dang hard to get... again. BTW, do you feel like Clive is an old friend after watching a few of his videos? One of the friendliest 'chaps' around, wouldn't you agree?
I bought the 12v variety of this PIR sensor and found it massively unreliable on my off grid 12v outdoor set up. Watching this has made me think otherwise on messing around with it further
Hi Clive, It's not quite as dramatic but have you ever tried Super glue debonder. It's dead cheap lasts forever and is instant Not in this case because it was destroyed before you touched it but if you wanted to save the item you're getting into, it might help. Brian Wharf
I'm supprised to see the ldr, I thought there was an issue with them or something with containing a heavy metal, so they don't produce them in products anymore, I guess this item is quite old, or do I just have bad info
![Blox Fruits Dragon Rework Update [Full Stream]](http://i.ytimg.com/vi/EqDAp8udhm0/mqdefault.jpg)
Maybe we should introduce a "BCA" marking on units that are "Big Clive Approved" !
I'd trust that!
That would be perfect!
Maybe together with a DGW rating (DiodeGoneWild)
Like the sound of that, much better than the much abused CE mark.
BCC: Big Clive Certified.
It was interesting to note your comments on these relays. I bought a voltage controlled relay for charging a second battery on my van, supposedly rated at 20A per circuit(2). The cover was just a simple sleeve and despite the attempt to scrape the markings off the relays it was possible to make out that they were only 10A rated and probably cheap ones actually only capable of carrying at most 5A.
He was extremely rude when I started an eBay return. Even said most people just contact him and send things back (although he hadn't responded to me contacting him). My bet is that he claims he's never received the item back.
😡
Ooof. I hate dodgy sellers.
I used 12v planet and got a basic Victron unit. So pleased with it. I haven't bothered wiring in the emergency jump start feature but the charging works brilliantly (as one would hope).
I did put breakers either side, rather than fuses. Thought process was that I could isolate eithef battery and charge with solar individually, otherwise the split charge relay could send the puny amount of charge from my small solar power to both batteries at once and I doubt I'd see any benefit.
A lot of the time the ebayers have no idea what they're selling. They'll just look at what they have and try to look up similar items and copy their stats.
I purchased a LiPo battery charge manager that stated it could work as a "UPS" (uninterruptable power supply), only to find out that the module, when input power fails, would lose its output for 2 seconds before the battery would pick up the load. They tried all sorts of BS to get out of it, even offering to send me more of the modules for free so long as I'd give good feedback.
@@ConstantlyDamaged Yes, I've had that too with Chinese suppliers but in the case of the relay they were clearly built by the UK seller and he just didn't care that he could potentially set someone's vehicle on fire.
If the electronic item you are purchasing is important (i.e. outside of just playing around with Hobby stuff) I always find it better to purchase from a traceable source such as Farnell / RS, who will usually be buying directly from the factory
Splendid! I like it when they use those "Switches 10 amps, at least once!" relays. Being able to enjoy the audio/visual indicators that signal their operation, slash demise, is always a treat. Cheers!
It's always very instructive to see failures and sleuth the cause of their demise. Thanks!
Once again I watch one of your videos and I'm amazed how interesting you make a subject, that in normal conversation would make people fall asleep.
C13 also got a hole burned into it from the arcing. Replaced plenty of those little Songle relays from them cooking the contacts, the 10A rating is very optimistic indeed on them. Years ago I needed some snubbers for a contactor, so looked up the price of the manufacturers one, sat down, then looked up the price of generic ones from RS. Also sat down, then went to the one local supplier and got 3 68n 4kV polypropylene flyback capacitors, normally used in TV set flyback use, and some 33R 5W resistors. A little overkill, but guaranteed never to fail, as those capacitors normally run 2A of 15kHz current through them in the TV set, so 50Hz mains would never even bother them. Resistors 5W because they were not much more than the 1W versions, plus had longer leads, as this had to go across the pins of the 11 bin base, so long leads were good.
Put some supply filtering in as well, 3 4uF 450VAC motor run capacitors in star, so as to handle the current spikes from the contactor as it was rather rapidly switching, which is why the snubber was there in the first place, wanted plug in relay to last longer than 6 months.
Old Brown Boveri took 40 years to fail, but seeing as the new ones were Omron plug in, much better than needing to do wire surgery, just plug in the spare. Old Brown Boveri failed because the steel of the core mushroomed out, jamming the core up in the housing, not from the contacts failing. When they said 16A there, they meant it, not will switch 16A, or will do 100k cycles. Brown Boveri meant that, at 16A inductive load, and 100k cycles, they will not fail, and would still look almost like new. They probably were well over the 50 million cycle count. Not made like that any more.
This looks to me like a case of over current…. Perhaps a seized pump ? And a 13A nail in the fcu ? These pir things usually power up in the ‘on’ state …….. awesome videos Clive, keep up the good work!!
I think the relay contact got hot, burnt up the PCB over time and compomised the isolation properties of the PCB up to the point when the arc jumped the carbonized PCB...
The PCB design is dreadful.
Minimise board area, but use the cheapest components as well, so you need the board area, and then you get this.
Most of the relay failures I've saw were due to arcing pitting the contacts, which go resistive and overheat. That's typically due to a shitty or worse, absent snubber circuit. Motors are inductive devices, so they'll generate counter-EMF voltages, causing the arc.
Whoever specified the device and whoever designed the unit need to be dropped into a solder pot.
This has brought back some very vague memories of working on an unit that had a plastic housing identical to this including the trim adjustments. I do however remember it having a through-hole multiple opamp, can't remember if dual or quad, but it was a chip I had used in university in the analog labs so I recognized it at the time.
On the subject of snubber networks, old rotary dial phones had what we called the spark quench circuit across the impulse contacts. They were also found all over equipment that used relays, which is why it was so rare to find burnt contacts causing faults. Of course, failure was usually in the form of what we called "low insulation" or LI faults, and was never spectacular, just noise (crackling) over speech, or where no speech CCT was involved, strange behaviour of equipment.
I have on of those on my garage light. Installed ages ago. I have this urge to take it apart and compare it to the circuit board design of the 2018 one in your video.
Get the pipe wrench!
I recently replaced one I put up over 20 years ago. I think I got my money's worth. To a rough approximation (I didn't open up the new one) the design hasn't changed. Most of the time, these units can't set anything on fire in the first place.
I must comment, I was just watching you tear apart this unfortunate PIR unit, and my mind was wandering. I thought "Well, dissecting a HUMAN is called an Autopsy, dissecting an ANIMAL is called a Necropsy, so what do you call dissecting an ELECTRONIC corpse?
Why, a Techropsy, of COURSE! lol
LOVE your channel, you're always a hoot!
It's fascinating to me that someone would be able to design an operational circuit without the understanding of mains clearances. I'm just a little hobbyist, and any time I've built something for mains voltage, I lean toward the ridiculously cautious. Sure, why not have 10mm of clearance, and add slots between just in case.
Agree, minimum spacing is critical on anything mains related, especially 220v used externally where humidity is unavoidably involved.
Sometimes our own Frankenstein projects are safer than original Chineseium products.
Although they always look like a very dangerous construction.
Then paint some liquid rubber on it, just for the halibut.
Electric current smells. Since my first AC transformer experiments in the early 196ties I found this wisdom proved a plethora of times.
Very interesting and odorous item! Do you think that the relay's demise could have been hastened by having to switch on the locked-rotor current of the "water feature" pump motor; as opposed to a light which this was surely designed for? Or; at turn-off; the inductive nature of that same motor might have overwhelmed the voltage rating of the snubber network capacitor?
Most small water feature pumps have a fairly high impedance even when stalled.
So close to a decent product. Interesting teardown and analysis, as always
The Slo Mo Guys have been Blowing up Capacitors at 187,000FPS of late. Perhaps a mashup would be in order?
I love when one of your videos hits on a problem that I've been having with arcing contacts. And I have the components it fix it.
Many pir's are designed for a couple of 100w lamps at most, but people use these on 300w (or more) halogens, i've seen several that the contacts have simply overheated (or the circuit track)
It's much better to use a pir to control an external relay, and let the relay switch the load
Incidentally, i just found a similar one that's advertised as 1200w, but zooming into the label it says 300w (on amazon)
Always grateful for someone who suffers that terrible smell and shows us what has happened - thankfully *Smellivision* is still a long way off!!
YES! The Vise of Knowledge! We don't get to see that tool often enough.
You could have a look at some of the Shelly products. Especially the Shelly 1 or the Shelly 1 PM. The actual relays inside seem so miniscule for what they are (aparrently) rated for
"a resistor of unknown value because it's cremated"
The bands were black, black, black and the tolerance band was black.
Love your videos Clive. Haven't watched is a little while, not your fault of course. Thanks for all you do. Your one of my top 5 youtubers.
I try to assume that any relays that are not from reputable sellers are going to be no more than 12.5% of the rated value...I used to assume 25% but to my detriment :(
Going to guess Songle or one of the many clones, though Songle generally is a pretty good relay, provided you use correctly, which means 1A at mains, with a snubber across the contacts, and a MOV as well on input and output. Or 10A at 12V, which they actually will do well. Not Omron well, where you actually have a warranty of 100k cycles plus, but at least should be able to get to 10k cycles before something fails elsewhere, most likely the capacitor dropper power supply.
Most cheap knockoff relays do be grand, I've used fake knock off omron clones, it comes down to the design mostly not nessarly the quality, it's all cheap shit from China, when you pay for quality what you're really paying for is design
I've been using FINDER relays recently...seem to perform as they are rated...a 120VAC@10A easily runs on my 8.3A PSW inverter used for my fail-safe computer power supply.
It is set at NO on MAINS and NC on INVERTER...so if the mains power fails it goes to NC and the inverter takes over...the computers never even notice the switching.
@@gregorythomas333 Computer power supplies must stay online for 17 milliseconds before their output destabilizes, according to the ATX standard. I've never had a normal consumer UPS fail to successfully switch to battery power in that time. But over-engineering is fun, I suppose
@@gregorythomas333 Problem is most small relays are not really rated for mains switching on both NC and NC contacts, especially if there is a chance the 2 contacts will be out of phase, as there is a good chance the arc will be able to conduct enough to short the 2 contacts together in the arc.
Thank you for another great video. Thanks, especially, for the Snubber Network explanation
In our 5th grade classroom, a kid next to me said "I smell burning raisins!" moments before a florescent fixture above popped and smoked. Every time I smell burning raisins, I recall that kid's face.
We Merkins in 120 VAC land have it easier than the rest of the 240 VAC world. Things are not quite as catastrophic and 'fun' in the fireworks sense. We do have failures but with half the voltage, the results are not as spectacular!
My torture toy when I was growing up was an old TV transformer not much smaller than a microwave oven transformer. It had a HV winding that put out about 700 VAC - I didn't bother with the center tap. 😂😂 I tortured a lot of 1 watt carbon comp resistors with it.
Thanks, Clive for bringing back the memories of the smells.
Thanks for another great video Clive 🙂
Nice teardown Clive !
Good explanation of the use and reason for snubber networks. You'll find them across semiconductor switches too (mosfets, bipolar transistors, IGBTs) too with inductive loads to limit the rate of rise of voltage dV/dt from the flyback effect when the load is switched off.
Before they were as well understood it was not uncommon just to find a capacitor with no current limiting resistor across switch contacts. I had an amplifier like this with a C across the mains snitch directly and on an occasion I was doing some servicing, found that this capacitor had obviously been subjected to a large current that had blown it into 2 pieces and the foil inside had unwound like some Christmas decoration. It's now in my rouge's gallery along with other remarkable instances of total destruction like the KT88 power output valve that dissipated so much power that the anode melted and fractured its glass envelope due to thermal shock and an Altec compression driver diaphragm that was so overloaded by amplifier clipping that the aluminium dome ended up like confetti. Curiously the driver still made sound since the voice coil hadn't itself burnt out but it was fabulously distorted.
I'm curious if this PIR matched the power of the pump that ran that "water feature" and whether its inrush current ablated the relay contacts, since those things can be triggered repeatedly by birds and falling leaves etc..?
curious about a PIR on a water feature - haven't seen that myself yet. Normally they're on a timer, to get plenty of aeration. So this would only operate pump if someone approaches - or maybe it's a cunning anti-cat device!
That's also my thought. These small (and poor quality) relays can't stand the current of a motor driven pump and will arc themselves to death.
What an interesting experience. For a moment I thought I got smell-o-vision.
Turned out some dump truck next door somehow started burning up a belt or whatever it was, while I had the window open. Have to stop watching the video, have to get some coffee and snacks and watch them try to get the thing back running again. Molten belt seems to have fused quite nicely with a melted pulley and they don´t even get the thing started again.
The vise of knowledge should be paired with the hammer of curiosity for the really tough jobs.
Do you mean the X-Ray machine?
@@eDoc2020 No, the hammer of curiosity hasn't been seen for a while. I think it's due for a run.
The leakage current from that snubber circuit is more than enough to drive a solid state relay, but not a mechanical one. Leads to some fun problems when "upgrading" equipment.
There's nothing like a good catastrophic failure. The more spectacular, the better. "Sniff... Ah I Love The Smell Of Burnt Circuit Board In The Morning"
The "vice of knowledge" - if I heard that right. I love it.
Those pipe grips are the best nut crackers you can get.
I prefer vise grips.
Water feature--inductive load perhaps? That could be part of it. With resistive loads, these things usually last quite some time.
I love the smell of hot plastic in the morning...its smells like victory.
The cause was probably the water feature (likely pump, inductive load) overloading the relay. Such mains automation things usually have two figures: for resistive loads and for inductive.
On mention of it controlling a water feature my first thought is it's beig used to switch an inductive load.
Near the beginning the live connection in the choc block looks like it might have gotten rather hot, and the insulation on the live wire going into it looks as if the cable has gotten hot as well
Hurrah! The Vise of Knowledge. Perhaps the next one should be the Saw of Wisdom ;-)
Thanks Clive.... I've been filling up the comments sections of my favourite electronics channels with "can you explain snubber networks please" for ages.... :)
Love the vice opening technique....
I raided a couple of PIR modules from my dad's outdoor light assemblies when he bought self contained LEDs.
One has a mechanical relay, the other a solid-state relay capable of half brightness.
The first one, I'm going to "upgrade" it by having the line output push a relay or contactor.
I've got enough of those laying around that I've got options. Maybe that 4PDT, or that NEMA 0 3ph...
Just a thought, but how about cutting away the charred embers of the PCB, and see if the rest of the circuitry still works, or could be made to work? Purely for the fun of it. We know it's not going back together again, but pure interest/curiosity makes me want to know how much actually survived, since we already know what didn't.
The "Pipe Grips Of Knowledge" seen at its finest!
The first thing I would check for is a short or overload in the water feature. Excessive amperage could easily start the failure cascade you see on the board. Of course, an underrated relay would produce the same effect.
Always promising when Big Clive has to bring out the Clamp Of Persuasion!
My mental image of "The Clamp Of Persuasion" has changed for the worse since what may or may not be the start of World War III. As if the Scots have not had enough to put up with already!
Yeah, it is 'gripping'...!
.
.
.
Sorry...I couldn't resist.
I work with security systems and a lot of access control equipment comes with something similar to the encapsulated resistor and capacitor which you showed (or it's recommended that they are fitted). Most of our kit is 12 or 24 volts DC and those the snubber devices are fitted across the relay terminals and across positve and negati0ve near to the lock, which tends to be an inductive load like a solenoid bolt or a maglock (electromagnetic lock).
A company called RIFA makes gold coloured ones that famously dry out and go pop, with a delightful stink. I'm sure Big Clive has happened upon them.
The real problem with having such inferior products in the market is they create a demand for cheap goods, which creates pressure on otherwise safe product manufacturers to cut corners. On top of that is the fact that its impossible to tell whether something is safe or unsafe without doing an expert teardown. (Price is a good indicator but insufficient, because nothing stops a disreputable vendor from marking up the unsafe cheap products to the same level as the safe expensive ones.) We are going to have a number tragedies reminiscent of 19th and early 20th century West before this phenomen get fixed.
Which is why BigClive's channel is a service to humanity, if you think about it!
This is actually where the standarts should theoretically come in, but as you can see virtually no one certifies their devices for anything.
It is sad, i think it should be a more widespread concerne, but i doubt that it is gonna go anywhere without anyone pushing it, like with right-to-repair. If not for Lois Rossman i doubt it would go anywhere.
I try to look for certification marks (Intertek, UL, GS etc) when shopping for anything that plugs into the mains, but I don't really know how much that helps.
Edit: I know that CE means absolutely nothing in the real world. Same with RoHS.
TOS been a long time we last seen the vise of knowledge or X-ray, was beginning to miss them.
I think the clue is in the original covering letter "Water feature" to me indicates bl00dy big inductive pump, I suspect the relay contacts draw an arc when switching off creating so much heat the PCB turned into a resistor.
If I was Toni Peters I would be checking that the new PIR is rated correctly for the water feature Load.
I was confused by the water feature aspect. Home made cat scarer maybe? Or just wanting to impress the posty with a triumphant fountain but not use too much electricity?
We need more on the backstory for sure.
A very good Thursday morning to you sir
Oh, a Barbequed and crispy PCB and components, looovely, LOL. I wonder if it was down to inferior parts and should have been recalled.
Street light collector Mike Barford from streetlightonline gave me a Royce Thompson Monostar 1000 NEMA photocell that had water ingress, this had caused tracking across the Common and NC relay solder joints, in turn burning the PCB under the relay to a crisp and melting the relay, it also burnt a hole in the casing of the cell. I managed to get the thing working just by replacing the relay and bridging one trace, those things are bulletproof!
Thanks from Texas Clive
I've taken several of these sensors apart to modify them for low voltage operation, so I have a sense of how they are constructed. It would be helpful to know the operational history of this particular unit. Did it work properly when it was first installed or did it fail immediately? I suspect that it had been working OK in the beginning and failed sometime later due to water infiltration due to rain/fog or other sources. This enclosure is not hermetically sealed and there are multiple paths for water/moisture to enter. That could have initiated some arcing on the PCB under the relay and which lead to the meltdown. Also, switching the reactive load of the water feature pump motor (without proper snubbing) is also a likely contributor to this failure.
03:15 WOW, you could sell that close up photo as art.
And suddenly I vividly remember that foreman during my apprenticeship that would rip screwdrivers out of our hands if the size was like one size off for a screw… also he told us screw drivers aren't prying tools. We really learned a lot.
the funniest thing is that as you kept mentioning the smell... I could almost smell it myself. I know this smell so well. :D
Those relays don't like capacitive or inductive loads. That's why finding a PIR that allows more than 100 W LED or 5-8 light sources is hard to find. You will need a contactor to take up the load. I don't know if the astro clock I put in will survive since it controls a lot more light than I like. But the customer is always right.
quite informative. wonderful as always
I don't think these devices were ever designed to be used with an inductive load (water feature pump as mentioned). Have used many of these with a great deal of success but only with low powered lighting. Yes the tracking dose look a bit skimpy but I don't think that's it. Remember these things are built to budget and once in a while a defective component will show its head. Great video by the way 👏
I'm wondering whether the failure was caused by the diode across the relay coil. When the relay is de-energised, the diode clamps the back EMF and prevents the magnetic field from collapsing quickly. This causes the relay contacts to open slowly, which can result in lots of arcing and sparking, and degradation of the contacts, increasing their resistance. With time, this can cause overheating and even fire. It's so much better to protect the driver transistor with a zener diode between collector and emitter, with value chosen to clamp the back EMF below the Vce(max) of the transistor (rather than 0.6V with a regular diode). If you choose a high voltage transistor with suitable zener, the relay will last much much longer.
I'm always amazed at the amount of development or should I say guess work that goes into Chinese circuitry.
For a minute there I assumed you branched off into modern art appreciation.
2:15 that's usually my response when somebody asks me if I could open something for them. "Its not a matter of being able to open it but a matter of being able to close it again." 🤭
Nice forensics for that poor relay 👍🙂
This is almost exactly the same PIR module that i used to revive an old halogen motion detecting lamp - and due to the magic of having bought two, i have one open here. Mine has a date code of 2019-12-29 and is thusly slightly newer than the burnt up one here, but yes the non-existent spacing between the relay output and neutral is still here in this one. (Now feeling that if i ever use this spare module, it will be with some extra safety mods... like cut the neutral near output, use a seperate snubber, add a fuse..)
The relay model in mine claims to be a DAOHENG DHT78-A-24D-SS, rated (or "rated") 5A/250VAC, 7A/125VAC, 20A/14VDC. (Hopefully it is better built than that burnt up one lol).
Looking at the PCBs, the differences between the one in the video (2018) and mine (2019) are mostly PCB component options, mine has one extra hole for different length resistors for R1, and the removal of the through-hole option for C9 (that one is a bit odd, why remove unused pads? just to drill less holes? hm). Also my CDS light sensor has been mounted up way higher (next to the PIR, not down low) than the one in the video.
...due to the magic of having bought two..🤔😃
@@stevecann3394 I'll be the smart Alec and say I sense the technological connection.
@@rednammoc 😃😃
Of course that relay will switch 20 amps! (Needs to be mounted to an iceberg to dissipate the heat, mind you...)
its so fun investigating burned stuff like this. i feel like a CSIS agent.
I repaired a PIR light a month or so ago, the fuse wasn’t soldered in properly (inline) and the wire had fallen off!
A resistor of unknown value because it's cremated. Gold!
20 years ago one of those relays stuck closed in a bread machine, had the heating element on for some time and smoked out the apartment pretty bad. The bread machine never caught fire but it turned the bread into a small cube of carbon carbon. Since then little tiny 10 + amp budget (or now even grey market) relays always scare me when they switch any actual current (over 2 amps). If I could I'd like to rig up a good test setup with lots of amps and try a few.
So fun to see the pir sensor I used to use on the channel
Interesting internal construction. I'm using a bunch of eBay 12V PIRs to control external LV LED lighting. Ours use two PCBs, the IR module and the 12V PS and relay PCB. So far none of ours have let out the magic smoke but we have had a couple of dead ones over the years.
Head on the nail there.
That relay wasn't up to the task (under dimensioned). It started to arc, and the rest was just a slow chain reaction.
I've seen these very exact devices myself, and they've failed the very same way.
RIP PIR! :(
That looks like it didn't enjoy an inductive load. I don''t know about over there but here in Australia a lot of those water feature pumps are low voltage with a transformer to supply them. In any other application than a single lamp I use a real relay to switch the load.
There's a security light that pops and fizzes on the wall outside my unit.
I realize I'm a little late to the party, but you missed the charred bits on the output plug that you snipped off at the start. The red wire's plug end was charred, so I'm guessing this unit wasn't rated for the current that the attached water pump(s) were drawing.
Happy Thursday and I hope all is good with you.
Clive, you might find interesting a video I just put up on the Dollar store "$1.25 these days" LED lamps in the USA.
Did you consider the pump may have been more current than the relay is designed to carry. That would cause that and I've seen it before
very likely, whizz poor contact materials within the relay failed, overheating and melting the solder at the board, then the board starts charring and ZAP arc over. this happens on tons of appliances, (fridge, freezer, washer, dryer, dishwasher, oven, microwave, hvac boards, thermostats, garage door openers) yeah I could go on and on, but that's plenty to get the point across.
alternate ways, thunderstorms and lighting arced across leaving a carbon track. it got water inside and fizzled away.
👍
I don't trust those dubious relays. I recently bought a temperature controller and replaced the unbranded 250V 10A relay with an Omron. The relay footprint looks the same on your pcb here, the Omron G5LE series matches it.
I don't trust them for switching any significant load.
Excellent video thank you
The picture at 3:16 looks like a piece of modern art :D)
Is it possible the end user had the load wired between Load live and an Earth? Also you mentioned it was running a water feature, was it running a pump and as such an inductive load? Combining wrong wiring with an inductive load you could get huge spikes as the fields collapse across the windings of load or coil.
Could it have been a bad solder connection for the relay that caused all the heat? Looking at the photo the most damage seems to be where the connection is to the board.
It's possible.
My goodness! How did you manage to get that crystal clear image of that really in 3:07? I just wowed at it!
EDIT: *of that relay
Have you ever heard the story of the coconut headphones? It's about coping actions without first understanding the underlying principals.
That cap had a problem I've seen a bunch. You see, at the factory where electronic components are manufactured, they hermetically seal them with a little smoke inside. So, if they get a little hole in them, the smoke gets out and they don't work any more... A trainer at a new job many years ago told that tale to a customer who was an auto mechanic. The mechanic's high tension probe had quit (with a mysterious fan blade mark on the side of it). He'd connected the primary test leads directly to the high tension coil wire (should work in a pinch, right?). The 741 Op-amp that was straining to 'hear' the 100mV or so coming through the filter board received all 30KV that jumped over said filter right between the eyes. After that, when I was confronted with both primary and secondary wave traces missing, I'd look at the hall-effect secondary probe for high speed impacts and then tear down the scope to look at those 741's. I even saw one where all there was in its place were 8 pins standing there in testament to its existence and a tiny scorch mark... I suspect use on a C3I (computer controlled coil pack) at closer to 100KV... Poor little bastard never stood a chance. :) We often referred to this phenomenon as 'A Failure of the BDU to RTFB' (I was often tempted to put that on the repair invoice, but somehow refrained... :)
I have heard that you can buy the equipment to put the magic smoke back into components, but it's very expensive, and on a long lead time. Oh, and I also heard that the magic smoke is currently on back order due to a world-wide shortage of Unobtainium.
@@wiseoldfool Yeah, I've heard that. I myself have run up against this supply chain issue... :) Thanks Barry, didn't realize it was that damn Unobtanium being so dang hard to get... again. BTW, do you feel like Clive is an old friend after watching a few of his videos? One of the friendliest 'chaps' around, wouldn't you agree?
@@scottdebruyn7038 I sure do.
I bought the 12v variety of this PIR sensor and found it massively unreliable on my off grid 12v outdoor set up. Watching this has made me think otherwise on messing around with it further
Hi Clive, It's not quite as dramatic but have you ever tried Super glue debonder. It's dead cheap lasts forever and is instant Not in this case because it was destroyed before you touched it but if you wanted to save the item you're getting into, it might help. Brian Wharf
I'm supprised to see the ldr, I thought there was an issue with them or something with containing a heavy metal, so they don't produce them in products anymore, I guess this item is quite old, or do I just have bad info
Cadmium. They seem to be phasing them out over time.
soooooo it had a live-neutral spark gap. nice.