Spectacular solid state relay failure

I wired up a solid state relay to only turn on the hot water tank when electricity is cheapest. Using a 25 amp relay to switch about 22 amperes. I guess that was too close to it's rating. No issue for 4 months, then failed spectacularly.
My previous video of this set-up:
• Hot water heater time ...
"bigclivedotcom" has done a teardown and analysis of one of these:
• Teardown of an eBay 25...

[@jeremyturner2873]

It decided to become a plasma state relay.

[@AbdulKarim-fs5iw]

That cracked me up...

[@Petertronic]

A comment from the original video said: "The fakes do have a bad habit of catching fire."

[@davidchuang4927]

This is an application where a standard mechanical relay would be better. A SSW is more suited to situations where the load needs to be switched on/off rapidly or in short intervals, such as in the case of a PID controlled heater application. If the point was to maintain the tank temperature within a very narrow range at all times, then an SSR with a PID controller would be ideal. However, in this example, since the system is either on or off for long periods of time, a stout mechanical relay with any suitable coil voltage would have been more than adequate.

[@ThatEgghead]

David Chuang and some sort of safety device - fuse, PTC thermistor, etc.

[@TheMetalButcher]

Yeah they literally make water heater controllers for cheap.

[@user93237]

But why would a mechanical relay be better? You're only saying that SSW's strength lies in fast switching, which does not imply that it should not be used in slow switching applications.

[@Anonymouspock]

okorok Simple, very simple. It has a continuous coil current but no voltage drop problem. So they can switch large currents without heating up due to that current.

[@kamyarghofrani]

Mechanical relays have a much higher current rating/higher efficiency and are generally much less expensive.

[@First2ner]

at least you had a house to sell after this

[@MazeFrame]

TO-220 is not a lot of surface to sink away the heat. After one of those blew in a setup, I bolted the next one to an old Intel heatsink (one of those with copper core) and that fixed it.

[@thechumpsbeendumped.7797]

So glad you and your family weren’t hurt. Best wishes from Britain.

[@yamazakirichard767]

Thanks for posting your failures for us to learn from. This is something I have wanted to try myself but won't.

[@douglanglois456]

Kudos to you for publicizing the failure, and not just brushing it under the carpet. But I guess the "safety know-it-alls" have a point sometimes, eh?

[@yorgle]

I've found that spraying some isopropyl/rubbing alcohol on charred boards like that and a light scrub with a toothbrush makes it much easier to see what's happened after events like this. :}

[@ZenMuff1n]

wow. you're brave. I usually don't go over 2/3 of what the spec sheet says something like this can take.

[@Yonatan24]

...But how are you (me, inexperienced) supposed to know when not go over a rule like that? I bought a cheap Chinese high-speed 10A PWM dimmer for my homemade LED lamp, but it heated too much at 6-8 amps. so I glued two heat sinks to the tiny heat sink they supplied. Works perfectly.

[@bborkzilla]

I wouldn't trust those generic Chinese parts for anything.

[@ZenMuff1n]

It's half for super cheap stuff and cooling is a requirement. I know what dimmers you e talking about, and I burnt 2 of them making a stable system with remote controls. Nothing cheap chineese likes heat it seams. The rest(like trusted brands) it's 2/3.

[@TobyCowles]

If it is no name then I don't go over 1/3, if it is a made by a good known brand then I probably wouldn't go over 3/4 for something as high power as this. But all this stuff is just rule of thumb, all of the manufacturers have different quality standards and you never can be too carefull.

[@jameskelly1680]

Even if this was not imported from China, he is running 22A on a 25A device. NEC standards in the US say that you should never go above 80% in continuous use. This is 88%.

[@gardnmi]

Let's hope the folks that watched the first video and implemented your idea see this video too. Reading the comments in the original video someone mentioned that these knockoffs tend to catch fire....lol.

[@srmofoable]

I also warned of the dangers of those knock off Foteks. They're great for some experimenting but there is not a chance I would leave one hooked to mains unattended.

[@thebeststooge]

SrMofo Exactly. He was lucky this time but next time it may burn down his house. I do not trust their DC SSR for the same reason because, typical of the Chinese, they claim an Amp rating but you better halve, a third, or a fourth of their number to get a margin of safety. Spend the money for a real one (they are expensive but for a reason) as life is too precious to screw around with like this. edit: My other concern/rant was him using axle grease as a TIM. Seriously? That was adding on to a future failure and he, and I, had a few words about it. It worked didn't it? Was his main thing while mine was "It worked for now but it is not a TIM and not meant to be." He did two stupid things but luckily it did not cost a life/lives/property this time.

[@ClairdACDC]

Actually Matthias the more I think about this the worse it gets. You are a very knowledgable guy and a youtube star but this is just plain dangerous. In your build video for this you mentioned the ÿoutube safety experts" and had a sarcastic comment. I am and electrician from Australia and cringed at your original video. No earth on the mounting plate, crap wiring, limited protection and the wrong type of device for switching and holding in the current of a water heater. People think that Electricity is easy but seriously you could have burned your fucking house down to save a few bucks. Stick to the woodwork ( which you are excellent at ) and we qualified electricians will do our stuff............... ( p.s you probably wont read this anyway )

[@knucklecorn]

I agree 100%.

[@vedranlatin1386]

Other then this one probably being a fake, what have you got against SSR switching a heater?

[@vlogerhood]

Well for one thing, is it rated for continuous duty at that amperage?

[@ClairdACDC]

The amount of misguided switching capacities with SSR's are endless. They say "50a" but after they blow up you realize that the 50 amps was an AC7 rating ( light domestic loads ) Contactors on the other hand are tried and proven to last up to 1,000,000 operations and if he had used a 3 pole contactor that costs about $15 and wired all 3 poles in series there would never have been a problem.

[@davidwootton683]

bigclivedotcom took one apart on his KZbin channel. Worth watching! He had the same type rated at 10 amps and no more.

[@SuperGrover]

I've never had luck with the solid state relays in those situations. Good old school oversized mechanical contactors for the win...

[@RadioactiveOwl1]

So glad you're still doing videos despite your tendinitis :)

[@Havreflan]

The safety-know-it-alls saw this coming. You got lucky.

[@dieselscience]

Yet for some odd reason NONE of them will call it what it is - cheap Chinese crap.

[@CPD-KD6-3.7]

I actually tried to tho something like this a while ago. The main difference between ours is that I used a proper contactor and an extra circuit breaker. Great job though, not ignoring it and posting it!

[@marcsimonsen1578]

Loved the teardown and forensics! Looking forward to the direction you are taking us!

[@treynamy]

Thank you for this Matthias. I did a similar installation on two water heaters in remote locations using contactors like you'd find on an air conditioning unit. Less susceptible to heat issues I think. I paid about $10 each on Amazon, and they had 24v, 110v and 240v activation inputs for the contactor. But, this video makes me want to turn them on while at room temp and monitor the temps at all the connection points. Thanks!

[@matthiasrandomstuff2221]

I thought about using a contactor, but the contactor itself also uses a lot of power, whether the heater is heating or not. So that really cuts down on efficiency.

[@treynamy]

Gotcha. Yep, efficiency is a different goal. Mine is a remote property, occasional use (10-15 times a year), so reliability/ease of use was my priority.

[@jesperdahl1486]

Contactors use only a few Watts, and ONLY when on, why do you think they consume power when off ? If only your ancestors had not fought to keep tyranny, you could have been part of a educated country, and not a Socialist hell hole, hell bent to win the race to the bottom.

[@treynamy]

Wish I could delete replies, but alas, only he can. His goal for this project was to conserve energy. And his point is that even if the Water Heater has reached temp, and is therefore "off", the contactor will still consume energy holding the circuit closed to provide power to the heater if it needs it. Granted his relay thingy performs the same function, but at a much lower power consumption rate. So, the main lesson here is, don't be a tool.

[@sssiod]

I would have installed something like that in a Hoffman type enclosure. An enclosure is not really made to keep people out but rather contain the electronics. In a failure such as this there is not much risk burning things down around it from inside a metal box.

[@JeffDM]

You're supposed to use a heat sink at those current levels, a thin plate of metal doesn't count. The one usually sold for that looks like an inverted tree. Also, Fotek SSRs are known to be fraudulently relabelled at a higher amp number to fetch a higher price. UL has a web page to help you spot the authentic ones.

[@stefantrethan]

I'm just glad that deathtrap is gone and nobody got hurt, you were lucky the fire didn't spread. If you look at the size of the heatsinks that come optional with these relays to achieve their full rating (well, not the chinese ones but proper ones) you know why this was bound to fail.

[@_cb336]

Yeah, right on, brother. Deathtrap were never the same after that first album.

[@roldac]

Good job on diagnosing this potential fire hazzard Fire Marshall Wandel. I set up my solid state relay with a cooling fan but will monitor it closer for potential failure.

[@gordonlawton8815]

Best practice is to use an SSR for fast switching applications and always to put a mechanical contactor or relay in series. This serves to isolate the power if the SSR fails as they fail closed. You should not use an SSR by itself regardless of the quality. So the safety issue is the result of poor design not just poor quality.

[@PaulScott_]

Thanks for sharing the good and the bad!

[@i_might_be_lying]

I've disassembled bunch of those solid state relays from different ebay sellers and found that 100% of triacs are grossly underpowered (12 - 16A, while relay is rated for 25A) and some of the relays had clearance issues, meaning that if some surge condition would happen - the relay will disintegrate, just like yours did. My advice - don't use those cheap SSRs from ebay, buy them from Digikey, Mouser, etc. (yes, they will be much more expensive).

[@Unaboner]

I had the exact same experience. Took apart a cheap eBay 25A rated SSR and found a 12A triac

[@pedrobravo6972]

Hi, this is not a product failure. It is a problem of your installation. The problem that happened to you was not to place a heat sink, as the specification says. If the circulating current is greater than 20 amps, you must include a fan.

[@JamesBiggar]

I've had trouble with this brand before too (yes folks, with a proper heat sink).

[@danmorphis7759]

James Biggar the ones you find on eBay/AliExpress are actually counterfeit. Real ones cost about $40/ea. The UL even has a warning on them.

[@ScottPankhurst]

Fotec are OK. it's the fake ones that are the problem.

[@richb313]

I have used these for many years in Industrial Applications with little problems. We mainly use them to energize motor Contactors and the like when using Programmable Logic Control Applications.

[@jmsaltzman]

If it says "Fotec" or "Foteh" it's probably not legit :)

[@Dirk3672_StupidYT]

FauxTech - legit, eh?

[@theroboticscodedepot7736]

WOW, I had the impression relays were very safe, but lesson learned! May be a good idea to enclose in a steel box with holes for air flow like a power supply in a computer. I had one fail on me while I was working in my office. It poured smoke out but was never close to catching anything on fire.

[@arthurmorgan8966]

3D printing community uses these relays for mains heated large printer beds, almost always attached to thermal fuses for thermal runaways.

[@36736fps]

Best part of your channel is showing and analyzing when things go wrong. Hope your insurance agent is not a subscriber.

[@SurajGrewal]

This is why my Dad used to DIY stuff like these in the 90s. He had an electric stove power reducer. It worked for 15 years, then the 2 triac that were in it, went bad and it allowed full power to the stove. A simple replacement fixed it

[@accountshanahan]

Wow your a brave man using that in your house! Lucky you didn't burn it down. For the loads involved with a water heater I would have that relay operating the coil on a contractor and let the contractor do the heavy switching on the heater circuit. (Don't know if a contractor is a thing in the US, in the UK its basically an industrial type relay with failure protection and a much bigger spark gap so they can be rated to very large loads). Love the channel and good luck with the house move

[@kwad3d10]

What would you suggest to switch on a 350w power supply? I could be left on for months at a time

[@patpalermo7629]

Bigclive did a teardown on one of these. turns out the triac was only rated for 12A. That's why I bought the 40A version - hopefully it will handle at least 15a

[@girder123]

Just buy a quality product from a reputable distributor, add some safety margin to your application and you'll be ok - no need to risk your life and property with cheap ebay crap

[@Umski]

From experience, triacs dissipate a lot of heat - approx 1W per Amp - I have a BTA26 triac (self made SSR) which switches anything from 1W - 2900W (0-13A@240V) into a water heater but it is fully variable, not just on or off. I noticed that when it was constantly running at full power (solar PV diversion) it started to do weird stuff as ambient temperatures crept up - measured it with a IR temperature gun and it was hitting over 100C on the heatsink - far too close to the junction failure temperature. Re-calculating the size of heatsink, I went from 3.1W/C to a 2.7W/C (twice the height) and it's been fine since (I even considered forced cooling with a fan at one point!). For your example, I don't see the point of using 'complex' electronics for something which is either on or off for a long period of time (minutes/hours) - there are too many points of potential failure - personally I'd stick with a relay/contactor arrangement for purely on/off switching. Glad to see it was only the casing that melted itself however and not anything else nearby :o

[@roypage5659]

I really appreciate your electrical knowledge. I wonder if you would consider some sort of electrical class. When you start talking about electrical stuff I really get lost and would love to understand more what you are talking about. You have a gift in explaining stuff.

[@davidwootton683]

bigclivedotcom took one apart on his KZbin channel, it is worth watching.

[@JohnBurgessMusic]

I would never use those cheap ebay electrical components in any high power or safety critical application. Doesn't even matter what the datasheet says for that TRIAC and it's current handling, there's no reasonable guarantee that it's a genuine part, or it could be a factory reject. A certified mains rated mechanical relay from a reputable supplier might cost a lot, but it's cheap when you consider that it will be switching high power loads, in your house, and largely unsupervised - probably while you're sleeping. Scary stuff if it goes wrong.

[@antalz]

Anything that I plug into the mains or runs unattended has to come from a reputable european shop for me. My projects involving mains power all run on Mean Well power supplies. That small risk of burning your entire house down isn't worth it.

[@aaroncameron1494]

I am going to go ahead and get this out... "Yo this channel is lit "

[@petermuller6923]

Wow, that cheep Chinese SSR lasted for 4 months at 88 % load running for longer periods of time dayly? Not bad. These Chinese ratings are very optimistic and i would not ever put more than 30-50 % load on these 4 $ SSRs. Sometimes, Chinese manufacturers label 12 A ones as 24 A SSRs. You should also never use a high quality SSR on nearly full load in continuous use. For 22 A, i would always use something in the 40 A range.

[@petermuller6923]

Wow, thank you for confirming my own experiences and what i have seen at BigClive's KZbin channel. I had not thought, that you even get a 4A Triac in a "60A" SSR. You absolutely never know, what's inside unless you take it apart and destroy it. That is unbelievable... A fireproof box is a must-have for these ones. A really good idea. Never switch loads with an SSR, when the load running forever may burn down your house or destroy something valueable. Almost everytime, an SSR fails, it will get continuously conductive and the load will stay on forever. Lights are not very critical, but heaters or motors in some applications.

[@briandyck4637]

Whenever I have to use a relay for something, I usually use a magnetic type relay with a rating at least double the rating of the load that it’s going to be switching.

[@Elaphae]

Very glad that you did not have a house fire. 🔥 Interesting tear down too. (Ever watch Diodes Gone Wild?)

[@BrokenLifeCycle]

I'd use a solenoid relay. That way, it doesn't draw much power (if any) when off. If you do choose to use that option, make sure it's a fully enclosed one or at least in a very clean housing. I have experiences with exposed lead relays, and those tend to get stuck closed when the contacts get dirty. The downside here is that these give off a loud click when it turns on or off.

[@richb313]

Actually more likely is that because we are in summer months the odds of a Brown Out increase due to increased loading on the grid because of Air Conditioners. The Triacs used in those SSR are quite robust and can easily absorb brief surges but if the voltage drops all bets are off. A safer way to have controlled it would have been to use the SSR to control a contactor with a 120 Volt coil.

[@zinckensteel]

I saw the aftermath of this exact same failure in a piece of lab equipment - in the same semi-clean room as a mass spectrometer used for C-14 analysis.. ..those were some very stressed out biologists.

[@paulmeynell8866]

Some times these just fail! We replace all the 25A with 75A now much more reliable

[@dasstackenblochen9250]

Well it is a 3 $ SSR after all... also SSRs *need* surge protection and pre-fusing when connected to mains. That might have prevented this spectacular failure, or limited its impact. 20 A on a 25 A SSR is also really pushing it - no way you can get the required thermal resistances with just that small heat sink.

[@g.z6161]

the datasheet says that you need a heat sink after 5A and a fan at

[@Rachotilko]

Much more likely explanation: a lightning surge. According to the datasheet, the SSR is rated to withstand 600VAC. However, in case of current/voltage surge, the voltage will inevitably rise much higher in case no protection was in place (GDT, MOV, TPSD).

[@TheRainHarvester]

Was that relay sold as a water heater relay? (Just curious if ones sold as a hot water heater relay are known to have a problem)

[@s0nnyburnett]

This is the point where being cheap can become very expensive and may result in the loss of something irreplaceable. Good god man.

[@gondala]

well, I think in this example it is showed how important to make automation system. I mean, it should have used the current limiter to make sure the current will not over 20 Amp. It is also important to add temperature sensor near the SSR to detect if the temperature is higher than it should be. If those two things happen, then the system will turn off the heater automatically and an alarm is turned on (before the SSR is failed or burned).

[@MrFaaaaaaaaaaaaaaaaa]

Did you use wires in screw terminals? I have seen screw terminals burn up because they were used with wires instead of proper ferrules on the end of the wires

[@STONEDay]

Here in Canada if a non-CSA approved electrical device like this burns your house down and the insurance company is able to prove that is the case then you're not likely getting any cheque from the insurance company!

[@coldfinger459sub0]

The transistor inside that unit is actually only rated at 50% of whatever is printed on the label on the faceplate this has been highly publicized and documented fight test performed by other people and at that if you really want to be safe you would only run something at 50% of that 50% which would make it 25% and It have to be heat sink not mounted on a small piece of tin they literally need a large aluminum and he seems to dissipate the heat. It’s burnt up because it was applied incorrectly for the wrong application with the wrong rating without a heat sink

[@kadmow]

Was that a genuine rated 25A SSR, or a cheapy off ebay (underrated) ??

[@g.z6161]

that was a fake you can see it says made in Taiwan instead of Taiwan made....

[@jmsaltzman]

There are many, many counterfeit Fotek SSR's for sale online :( To say the issue is Chinese SSR's is not accurate: Fotek is a Chinese company and makes great relays.

[@ianjefferson9518]

I belive Fotek is Taiwanese. Indeed these fake controllers work but only at about 1/2 or 1/4 their rated current. One issue I encountered with 40A fake foteks is that they fail in the on state so regardless of signal they pass current. Same application as Mathias's - immersion heater. I have used these quite successfully for up to 1500W ( 10-12A) heater circuits but at 4800W (20-22A) they fail on. I have a rainbird relay purchased from a slightly more reputable supplier and it works fine at these loads.

[@jmsaltzman]

Ah yes, Taiwanese, my mistake. Seems like de-rating the SSR's by more than half is a good idea! A colleague recently ordered a few for cheap, and they came with "Foteh" on the SSR's labels, in non-branded packaging :|

[@OldCurmudgeon3DP]

@@ianjefferson9518 I have Inkbird brand 40A on my project. 4500W @ 240V so about 1/2 of the claimed rating. Heatsink that came with it is the size of your fist. The enclosure has forced air circulation. I've checked after 10min operation and the hottest spot was about 100F. I used cpu thermal paste before bolting on the heatsink too. So far the only runaway was due to a failed MCU. Left the mechanical thermal safety cutout in the mains just in case; to prevent a ka-boom.

[@michaelconnors9906]

I'm glad every body is ok.

[@Thepracticalengineer]

ooops! glad it was only the relay that caught fire

[@nicolek4076]

Someone is channelling Big Clive.

[@spokehedz]

I'm just glad everyone is safe and nothing of 'value' was exploded!

[@DavidGillooly]

What was the part number on the Triac?

[@Greebstreebling]

I had one destroy itself, mine was a DC version switching 12V, switched by a microcontroller. It melted badly, so I made my own using two MOSFETs as a low side switch.

[@IslandHermit]

It looks like the power transistor is using the metal base as a heatsink and only relying on the springiness of the leads to ensure contact between the two. I can't see that being terribly effective at heat dissipation.

[@schsch2390]

I think a close look at when he opened the unused SSR, the triac was fused and riveted to the base plate. I agree with the speculation that the SSR was cheaply made for its rating.

[@SharkyMoto]

use a contractor next time, that thing is made to switch heavy loads. we use this to switch on the heaters in the whole house and its working since about 40 years now. you have to change it every 5-10 years because it fails and doesnt switch anymore but at least it wont burn down. the thing gets mounted inside the fusebox too so its extra security

[@uoit01]

I have the same type of set up on my electric hot water heater (3000watt) that I use in the in-floor heating system of my shop. It's been in use for 4 years. Only runs the heater at night. I used the same relay but 40amp as I had read that they were prone to overheating when used at their top rating ;) mine is also on a large aluminum heat sink. It gets warm but not hot. Electromechanical relay is actually a better option for this use but I opted to use the triac as it was a lot cheaper. I would just upgrade your hardware and resume saving money.

[@beyonddeath123]

Had these fail too, only use name brand SSR's now and havent had an issue since. Crydom or Omron are ones I've had good success with, but 20-30$ cad for one.

[@AtlasReburdened]

Electronic component ratings are the MAXIMUM rating before failure. You wouldn't drive an axle rated to shear at 2000Nm at 1800Nm cont. load that has a startup surge like a cold heating element does. My general rule for ratings with relays is if you want assurance that it will get the job done you de-rate a component by 25%, if you want it to last a few years you de-rate your component by 50%, if you want it to last for the foreseeable future you de-rate by 66%. Put another way; Fast job: ComponenRating = SurgeCurrent x 1.25 A few years: ComponenRating = SurgeCurrent x 2 Foreseeable future: ComponenRating = SurgeCurrent x ~3 And you should calculate SurgeCurrent by measuring the resistance of the element while cold and dividing your mains voltage by that number in Ohms instead of using a clamp meter to measure it in operation. Solid states don't need contact ware considerations so for your application I would slap a 60A version of that same SSRelay on and call it pretty good.

[@AtlasReburdened]

Oh, and mounted on a thin piece of galvanized steel isn't the best idea ever, but if you properly de-rate your components you could probably get away with it.

[@tommays56]

We use SS relays all the time and one that size actually needs and epic fined heat sink with good air flow to run remotely close to its rating

[@iglesiasleonardo1742]

How much power does your heater draw? The triac probably burned because there wasn't enough cooling to support your load

[@hugobose2254]

So in short Matthias, Dont forget to switch things off when your going away lol great talk through

[@jacilynns6330]

The single most popular fail of any electronic device built outside of north america is "automated" soldering. There is way to many cold solder joints that add resistance overheating circuits. In north america these are checked for on any device made. usually fixed before it is sold to consumer. Sadly many of these great businesses are being put out of work because of the volume coming out of china. Note india/japan/tiawian/korea made components are becoming much the same for lack of quality control.

[@deserteagles948]

A question I've had for a while... How do you amplify wattage?

[@OutOfNameIdeas2]

Amplify wattage???????? You get a bigger power supply?

[@a2akustik]

Wow 22A on the china knock off an sheet of aluminium as heat sink with bad air convection, thanks for showing this.

[@andrewbieger5004]

Maybe I am missing something about what you were trying to accomplish. Why go with this electronic switch when an old school mechanical clock type HWH timer works fine? My timer (all 220volt) has been running for 25 years, switching a 4500Watt element, with a one hour "heat" in early morning and a 2 hour "heat" starting at 6PM. The pins lock down on the clock wheel, and I have infinite adjustment for on and off cycles. The internal contacts on these mechanical units, just like those on a pump clock or pool pump relay are rated for plenty of amps, and you can't beat the cost, based upon its life cycle. I am glad the electronic unit didn't burn your place down. Old school works just fine, IMHO. Good luck with the move.

[@Fuogor]

What's this? A bit of BigClive on Matthias Wandel channel? Just by the title I was expecting it to be a Clive video.

[@lucasdJAdam]

I don't know if it's remotely possible, but can you rig the one you have left to burn. Would be a nice vid.

[@movax20h]

That is why you avoid cheap SSR from ebay, China and Futek. The main problem with Fotek (Taiwan) is that there is actually a lot of fake / counterfeit ssrs around from China. Recommeneded: Crydom, Omron, Ametek, Opto22, Weidmuller, Sharp, Altech, Carlo Gavazzi. There might be more, like Panasonic, but who knows. And buy via trusted distriubutor, like digikey for example. Seriously, anything that can go into flames when you are not around should be of highest quality. But honestly I think you failed to properly secure cables, or the screw terminal connectors corroded or are of poor material, and they started heating up. Also, it is likely mechanical relay would be better here. Unless you need switching off and on every few seconds, go with relay.

[@printxii]

What were your thinking!.

[@mauriziofigini]

bigclive?

[@marcoschwanenberger3127]

I don't think even Clive can analyse anyhting of this mess.

[@mauriziofigini]

exactly, he would try to blow up the other one straight away!

[@matambale]

In the video "Teardown of an eBay 25A Solid State Relay. (SSR)" (at five minutes into the video) Clive shows that the triac used is actually only rated for 12A.

[@TheGlen1985]

I was thinking the same thing

[@rickpalechuk4411]

The things you do to put content on the Internet 😄 You must freak the wife out on a regular basis 😉 Cheers

[@aksting]

These type of solid state relays require a good heatsink unless your using < 50% of power depending upon ambient temperature and % ON time. The steel sheet metal is a poor heat sink. You should have used a 50 amp SSR with an appropriate Aluminum heatsink with thermal compound/pad. Better yet, use a contactor.

[@ro_yo_mi]

I'll bet that ghost stunk up the whole house.

[@allluckyseven]

It looks like you have unintentionally created vanta black.

[@chimaeria6887]

I'm pretty sure BigClive tested one of those knock-off Fotek SSR before and they actually are rated for about 12A, not the 25A they claim. If you had bought a real one, i'm pretty sure nothing would have happened.

[@RippoZer0]

The triac in these relays is very underrated compared to the claim made by the seller. I bought a 40 amp one and tore it down to find a 16 amp triac. This means that it should never be run at more than 16 amps. I think the 25 amp ones came with a 12 amp triac. The electronics of these relays is OK but the build quality is shifty. I think it's worth buying a genuine name brand if you need any kind of lifespan or fire safety. You can get by on these cheap ones if you buy one with a suitable triac.

[@kalmdwn7711]

The real Fotek has a stylized "O" in their logo.

[@yamamatalk7002]

Thats why hvac unit use contactor or relay not ssr but as we saw with ai and another technolgy advancement who knows ssr gonna be good at handling big load or even controlling heavy load too

[@iamGomer]

I had hope for an AvE type quote "Release the schmooo"

[@gregfeneis609]

I think it's more likely that not-quite-tight-enough screws plus heat cycling (SSR mounted on warm WH) lead to the screw terminals coming loose. The poorer the connection gets, the more extreme the heat cycle goes. When the connection gets poor enough to start arcing, the solder will melt and things will burn. Cheap SSRs will have crappy screws that you can't get tight enough before the screw driver cams out, to properly clamp on bare wire. Best to do proper crimp terminals and fasten the screws to them.

[@tseckwr3783]

One possibility is that the mechanical connection (screw terminal) where you hooked in the wiring was not tight enough causing a slight resistance -- coupled with high current translates to heat. The terminal heats up, the solder liquefies, liquid solder drips on the grounded heat sink bridging the terminal to ground providing a current path. Arcs and sparks result. Additionally, the circuit board carbonizes. More heat, more flames. Too bad You did not have had one of your raspberry pi cameras trained on this. It was likely quite a show. Glad no real damage resulted. No people or pets were harmed....

[@nukularpictures]

I really dont understand why you didnt fit a thermal fuse and a normal fuse along with this ssr. I mean of course the choice of that really was poor but it would have been much safer if you would have something limiting the power if it fails. I would recommend of using a high quality ssr in series with a fuse and a thermal fuse. Also a propper heatsink would be required. If you would want to do it neatly you should also put it all in a nice fireproof enclosure and then it should work for ages without having that problem or risk ever again. For me saving those few dollars in the installation wouldnt be worth the risk of my house or family. And if lets say 100 dollar for a proper installation is too much then I'd say it is not worth it anyways. Not for the risk it creates.

[@firepower9966]

I use these in a pid controlled coffee machine, they fail occasionally (1 or 2 years use) open circuit. but so cheap so just replace now buy multiples so have spare, Last one seems to be lasting long time.

[@kittyfanatic1980]

You put WAYYYYYYYYY too small a heatsink for that SSR. For a 30a load (what most water heaters pull) you need active cooling or a radiator system.

[@andrzejp8782]

Got exactly the same with mine that looked identical. I got rated for 40A and used for EV charger drawing 15A and after 2h my fire alarm took off ;)

[@mr_spud_nugget3723]

looks like a job for bigclivedotcom :P

[@galacticovergrow]

He would have to find one at the pound store first :P

[@sixtofive]

Glad there was no serious damage

[@wtimber]

Didn't expect that, if I ever make one I will include a fuse and an ice cube relay.

[@alfredneumann4692]

Nearly 5 KW switch power...... i never would use a solid state relay for such a hammer. And this little thing made 50W heatpower? Never. Perhaps with a good heatsink and a fan.

[@mildyproductive9726]

I think probably more like 40W. But that is a proper roasting for any component in an enclosed box with inadequate heatsink. I think 2W will make a bare TO-220 boil water.

[@UndernetSystems]

Rip in peace my dear triac

[@Yonatan24]

RIP, RIP in peace, or did you mean rest in pieces?

[@SuppaflyZSM]

Do electricity prices fluctuate so much that it's worth allowing the stored water to cool and have to be heated up more at one time vs maintaining a constant temp throughout the day?

[@mckenziekeith7434]

It never, ever saves money to keep water hot instead of letting it cool down in between uses. Every minute that the tank is full of hot water, it is losing heat to the environment. The hotter the water, the more heat it loses. The cooler the water, the less heat it loses. The amount of heat added to the tank is exactly equal to the heat lost over time. Of course, there is no point in running out of hot water, though. And nobody wants to wait for the whole tank to heat up just to take a shower.