UPDATE: quite a few eagle eyed viewers spotted that R14 was missing. R14 is one of the balance resistors across the supercaps. That falling off could certainly have eventually triggered the event we saw here. I cannot find the resistor, but upon inspection of the pads it was clear that the resistor was soldered in there at one point, but the joint appears like it cracked so to speak. So it could have been a poor solder joint that ultimately lead to the failure. Indeed, the unit likely wouldn't have survived long at all if R14 was missing from production. The resistor could have fallen out on handling and transport to the lab. I have spoken to the designer and the balance resistors were chosen to prevent imbalance and hence overvoltage in the supercaps, and there have been no other reported failures in over 70k units installed. He also concluded that the unit wouldn't have survivded this long if missing from production. In any case the latest design they are sending does not have series supercaps like this. I'll try and do a teardown when I get it.
The problem is that its a mystery and a lot of people can't bear unsolved mysteries.. Maybe something caused R14 to heat up and that's why it fell off.
@@Lucien86 From inspection of the joint it looks like it was a dry joint to begin with. How it actually fell off I can only presume durign the transportation from the panel to the lab. I did kinda toss it around in the car.
30 odd years ago a colleague got a zap from a board across his hand - no damage. All such incidents had to be reported. Some time later he was asked about the incident with some health and safety guys. He showed them what happened with another board. Of course the caps got him again - much swearing.
Reminds me of a Darwin I read about. Some guy got got electrocuted while working on a scaffold (I don't remember the entire incident), and the insurance company didn't believe what happened, sent somebody to investigate, he recreated the incident exactly and electrocuted himself.
If dave was in Germany and would have been working company hours during this incident, an ambulance would have been called and he would need to stay 24h at a hospital 😀
@@AnoNym-zi5ty Wow. Compare this to America where I took a 1000W plasma generator across the heart (instant burns on both hands). I got to drive myself to the hospital and they said I was fine to leave after 4 hours.
The last time I got shocked was because I used a 1000V insulation tester to check some 5kV cable for suspected damage while it was still on a reel (it turned out to be fine). What I had not considered was that the conductor and foil shield together essentially made the cable a big capacitor, and that charging it up to 1000V and discharging it across my thumb would feel less than pleasant. 0/10 would not recommend.
@@Okurka. It's a known failure when you put supercaps in series. There were a bunch of HP Raid controllers that had similar issues where they used supercaps in series as the backup battery to keep the cache alive if the power fails, and they failed in a similar manner. When the supercaps are new, they share the voltage equally across them, but as they age, they don't match anymore, and eventually the voltage across one of them is beyond the ratings, and then it pops.
@@gorak9000 I can't believe a decent engineer wouldn't have predicted this [that series supercaps would require balancing circuitry]... then again I've seen series supercaps paired direct from the manufacturer, and sold at "reputable" wholesalers, and I've seen "high end" equipment with parallel mosfets, which obviously failed, so apparently decent engineers no longer exist? Or maybe they're too pressured to compete, in so-called "productivity," with others' lack-of-care? [edit: unrelated aside follows] "Sorry to hear the device is dead ;)" "The device is dead"?! They're lucky they [it] didn't burn down his house! [editted, after the rant that follows and apology at the end, in hopes of better clarity of my intent, should anyone come along 'fresh' hereafter. I did get a bit extreme in the "burning down the house" part. Hey, I'm human. But, as I recall , nearly all my comments hereafter are related to the concept of unbalanced series-caps, parallel mosfets etc., *not* this particular device's implementation of series-caps, except where explicitely stated]
@@gorak9000 Totally agree. Chamberlain / Liftmaster has the same problem with their garage door control panels. Will give you free replacements, but same flaw in the replacement with the two caps in series. Much worse failure mode - will open your garage door in the middle of the night.
I've had a few bites from 120Vac (usually a bad ground from a guitar amplifier) but the ones that caused more injury were DC - at Tektronix (Wilsonville), someone 30 meters away blew off their finger on a 5V, 120A power supply... got their wedding ring across it. - 12V DC (motorcycle gel-cell battery) fused a metal watchband to a lady's wrist when I was working at the place that made the first AED. - ME? I got 5kV in one hand and out the other troubleshooting an early-production AED which had a badly mis-wired High Voltage section... someone in production got lost q/all the white-only 6kV silicone wire and somehow failed to connect up the dead-man relay properly so the 50ohm dump resistor was never across the Defib Cap... not even the 1000:1 voltage divider was properly in-circuit, so when I checked the cap looked like it was drained. I got within about half an inch and a BEAUTFIL Bright Blue Spark leapt from the cap and met my fingertip... then up through my arm, across my back (the scapulae met in the middle) and down through the other hand, jumped the anti-static mat (breaks down at 600 volts) and back to the unit chassis, which was siting on the mat. Stupid me, I forgot about the mat's breakdown voltage and thought I was "Safe" with my hand almost 2 feet from the unit. Yer not the only Dumb one, Dave. I think my T-shirt should say : Safe Not By Design... By Luck
The gunk under the resistor is from the glue used in the reflow. That is a double-sided loadout of SMT parts, they would be all reflowed at the same time, the ones on the bottom have a dot of glue to hold them on during the reflow process. When the resistor heated up to such a high temperature, it burned/melted that glue under it.
And people on the internet complain about north american breaker boxes. They're a lot tidier than that mess. That's what happens when you're on the bottom of the world - spent most of your time just trying to stay on the surface of the planet - no time to worry about tidy breaker boxes.
As a German i thought to myself "What the Fck are they installing over there" and got flashbacks to a breaker box i replaced at my grandfathers house (installed in 1934) which was evenly messy but there was also about 5x more stuff in it like 20 breakers as well as a retrofitted RCD. But i would have never thought that in an Industrial nation installations on plywood could have been legal in the 80ies.
@@minespeed2009 I don’t think the board is wood, it looks like some sort of phenolic resin type material. Somewhere between plastic and the cheaper kind of PCB.
I nuzzled a charged disposable camera flash cap once. I immediately took the battery out and put myself on "you don't get to play with this anymore" time-out.
Ahh this reminds me of the memories of taking Kodak cameras apart at 10 years old and being blasted by the 400v 100uF capacitor! I never trusted a circuit board without discharging every cap for years!
.... Complacency is our constant companion, can't believe the othher day I was using a bare jewelers screwdriver to bend and close the female end of a (IEC C7 ) mains cable. Just luck I did the neutral first before realizing it was live. (FFS !).
When I was 11 years old my dad was repairing a big CRT and he just had left it open with no one around. I walked in and touched the back of the tube and it felt like my arm got kicked in the balls by Tom Dempsey. I was too socked to even cry or anything. I just said nothing and went back to my room and play video games. When I asked my dad about it later he said the tube itself is a capacitor and it was probably charged around 10kV.
@@felixar90 .... Yeah, I was so glad to see the end of CRT's & Video recorders. I Made a living from repairing them, don't miss KV's shooting up my arm.
When detecting the 2 supercaps in series I immediately got concerned about them being balanced. @gorak9000 mentioned this too: supercaps don't like to be overcharged to a higher voltage and when placing supercaps in series there should always be some protection to prevent that from happening. Balance resistors are great for small caps (we used to design high voltage capacitor banks that way) but for those large supercaps balancing resistors are not enough. At least some zener diodes or even better one of those active balancing chips especially designed for supercaps should have been used. But writing this I do remember that, back then, I was advices to just place balancing resistors over my supercaps by a component supplier.
In HP servers of generation 7 and 8 there are supercaps installed for the RAID controller. They saved a cent by not populating the capacitor module balancing resistors. Now we replace blown caps all day. And it's always one cap bulged because of overvoltage and one ok. (2 caps in series)
Solar system analytics fail... meh, I'll watch later. Dave gets zapped-INSTAWATCH! 😂 I predict 6:11 will be "Most Replayed" once the watch graph appears.
Dump it mate! I grew up on 220V 50 Hz. Got zapped endless times. Then went to England and got zapped on 240V Hey the extra 20 make you taking extra care. Then I had 117V 60Hz, that was just a near pleasant cribble. Next stop: Western Australia in 1982: They had 254V mains!! Never even wanted to get zapped.
You probably never worked with CRT TV's... 25kV and the picture tube acts as a perfect capacitor. It is a good advice to discharge the voltage properly before you try to remove the HV plug. (called the pacifier)
That second SuperCap might have had problems also as you can see the shrink-wrap cracked on it as well (18:33). Although there is a chance it was damaged by it's neighbor.
If they are in series and one has failed it is likely to have severely overstressed it neighbour. Both should be replaced, preferably two from the same batch.
The crispy-fried resistor sort of makes sense since it looks to be part of the snubber network on the primary side. I would hazard to say that D8 that looks kinda cooked as well might be a TVS (Edit: It is. A 200V Zener. I'm sure it got toasty piping hot!). With it there absorbing some of the excess energy the supply probably sort of worked, but once it cracked there was nothing to stop the flyback voltage from sailing away, probably from having the output loaded down or some other issue. I'd guess the caps started to go short-circuit, or possibly it was no longer in regulation and running in some form of quasi-open loop and overvolted everything.
Must have been a 'zap Dave day', as I got a 'tingle' changing the back light led strips in my LG TV, even though it had been powered down for over 8 hrs!.. Darned led driver psu caps... Cheers from a miffed Dave in the UK..😟
I have a suggestion Dave. You should video a break down of the new one checking PCB and quality of solder joints on the new unit, if you do replace it. Could help with future diagnostics.
A decade ago I was living in a little Victorian era cottage built in the 1860's, so because of its age the power was installed many years after it was built, with all the conduit installed on the walls and ceilings. It also still had the original fuse box complete with ceramic fuse holders and the obligatory packets of fuse wire so you could repair a fuse when it melted. None of that smart meter rubbish!
The first house I ever lived in and grew up in had exactly one fuse, and the only power outlet was part of the lighting circuit. As we began to need more outlets, we eventually had that old outlet removed and replaced with a modern outlet with modern wiring. (Ended up with three fuses: One for lighting and two for power)
@@melkiorwiseman5234 In our house, we still have everything on a single 10A breaker. We have many outlets and lights, but everything is on that one breaker... And aluminum wires, most outlets have no ground.
I never understood how you don't get shocked in your videos... I just assumed it was because you're an amazing EE but its good to see you're also human! haha
As an EE who only works with 24V/3.3V systems professionally, I'm suuuuper paranoid when I take mains stuff apart in my free time when stuff breaks... Rubber gloves and every primary cap is checked twice.
New one will fit in the same box, just will have a nice big DIN rail box to hold all the breakers and other DIN rail mounted units, with the meters now all being on the top row in a line. Plenty of space then for extension. When you replace the heater, insulate all the hot water pipes both outside and inside the roof, so as to minimise the heat loss from the long runs.
I’ve lost count as to how many supercaps I’ve changed out at work on the bench. They are notorious for leaking and destroying boards, components and tracks.
It's because they put them in series with each other with no protection or voltage balancing. When they're new, they're well matched, so share the voltage equally, but as they age, they don't match anymore, and the voltage creeps up on one until it's above the rating, and then it pops.
@@gorak9000 I like you. Hopefully some aspiring engineers learn a thing or two from your time. ... OP: As for doing repairs, if you see them in series, prb best to replace both, regardless if one's OK, because a new one will put that much more strain on the old one.
Yep, I've had a few shocks like that myself. Each and every time it's from powering up a piece of gear on the bench after repairing it, and stupidly handling the board with my bare hands, moving it to get a better angle for probing. One time I could even feel the 50Hz on the secondary of the transformer, and I had enough time to think "that's probably not good" before I started to put it down, evidently coming into contact with the primary in the process, which caused the muscles in my hand to twitch and throw the board across the room. The last time my arm flew off and smacked into the side of my desk. That one really hurt. I couldn't even close my hand to pull out the power plug.
Last time I did it was when computers still used AT power supplies... I took mine out to clean the dust out, and when I put it back in it didn't power up.... I figured I probably had the switch wired wrong, and started unplugging the terminals... Live. I ended up with a pair of burn marks on my finger which scarred (and one is still visible still 25 years later), and a bleeding hole on the back of my hand where I smacked it into the corner of a CD drive. I've been a combination of careful and lucky ever since.
I saw the caps as you split the case and thought, “hope he checks for charge on the caps before he proceeds”. I usually check with a meter but the thumb works good too! 😅. I learned the same lesson 30 years ago. Love your videos!
The isolation between AC mains and secondary circuit does not look safe on the power supply board. The 3G antenna is accessible so there should be double/reinforced insulation on that power supply. This kind of equipment should be designed taking over voltage category III requirements into consideration.
@13:24 Looks like one of the supercaps died. It's a little bowed, and there's the tell-tail signs of schmoo leaking out of the bottom. That would explain the "slow" overload that cracked a resistor and melted the diode. That gunk under the resister is very likely the glue used during manufacture (wave soldering / IR cooking the underside.) I'm really surprised there aren't fuses all over that thing, being mains connected and all. (Yes, I've seen resistors crack. Even the big 1960's era pinned ones -- inside an arcade console powersupply. Moisture gets into them and then heat/cold cycles take their toll.)
Is this a common way of building electric panels in Australia? I'm just wondering, bc most of the devices in there are DIN-Rail compatible, mounted in a tiny box, with a tiny rail, mounted to a panel. The big meters look like for ordinary 3-point fixtures. Of course every country has its standards, but as a German electrician, this looks a bit odd to me. We use mostly standardized cabinets with DIN-Rails, 3 point fixtures and wiring conduits. This way the insides are more accessible and serviceable. Though, interesting to see how it's made in other countries.
Once got a 1000VDC zap from a PMT power supply by somehow touching my probe with a metal braided shield touching the back of my hand - man I had 3 fingers numb for a few months doc even sent me to the neurologist where they did an EMG with a sharpie and marked where the damage was (ulnar nerve I think) - thank goodness it healed on its own - high DC voltage is no joke m8 glad you're OK
.... I was thrown accross the room as a child investigating my brothers 'reel to reel'. The drywall had my body shape imprinted (Like a Warner Bros. cartoon).
Since it's a two-sided load, one side of the board will have all the SMT components glued down so they stay on until it gets cycled through the reflow oven. That may be the source of the bunk under some, but not all, of the destroyed/lost components.
Forty years ago I was fault finding a Yankee made telecommunications impulsive noise test-set that had been converted from its original 120V to 240V for operation here in Australia. The bloody thing weighed a tonne, the Yanks had obviously built the case on it out of leftover Sherman Tank steel panels from WW-2. Anyway I wanted to turn it over on the repair bench to measure some voltage readings on the rear circuit board and so I grabbed one of the case front metal handles with my left hand and put my right hand under it to lift what I thought was just the metal chassis but as soon as I lifted it off the bench I copped it with full-bore 240AC surging through my entire body. I lost all control of my muscles and was left standing there with 240V rippling through my chest, and shaking violently and uncontrollably with this damn thing held in a vice like death grip that I couldn’t release and the only thing that saved me was that after a period of time that seemed like it would never end it literally tore itself out of my hands because of its sheer weight and fell crashing down upon the bench, whereupon, I fell crashing backwards and just by luck happened to fall into the seat of the chair positioned behind me. That was, and still is, the closest thing that I had, and have, ever experienced to me leaving this mortal coil, on that day. To close this story, when I later investigated what had caused me to receive this life threatening shock, I found that whoever had previously done the conversion on it from the original Yankee 120V wiring, to the 240V Australian supply, had left the majority of the original Yankee Mains wiring in place , which included the on/off switch on the front panel breaking the Neutral, instead of the Active, and the ‘un-switched’ Active wire then continuing around to the rear of the unit and being terminated upon a ‘bare’ completely un-encased, 3AG fuse holder that was positioned right in the middle of the low voltage 12 volts secondary power supply, and it was that bare, and live, 3AG fuse holder, that the fingers of my right hand had touched when I lifted it off the bench. Unbelievable stupidity, by the Yanks, in the first place, for such a dangerous design and build, on the damn thing, and by the unknown idiot, who subsequently did the 240V Australian conversion on it, and finally by me, for relying upon the front panel mounted power switch, on it, to turn it ‘Off’ instead of turning it Off at the bench power point, before I tried to lift it to turn it over. It taught me a very valuable lesson that day, however, to NEVER, ASSUME, that ANYTHING, Mains Powered, is wired correctly, before you work upon it…!! 🙄
Some caps you just cannot discharge, especially there are caps on the other side of a transformer. Discharge a cap on one side, the caps on the other side charge up. This happened in a module in a fire control black box in an attack aircraft while I was in the Navy. We sent back the module repaired, bare as with the other type of modules with the document in its own heavy duty vinyl weather bag. Well, it bit a guy in supply and henceforth that module and others like it had to be in a separate heavy duty document bag or other container. Kudos for the SA guy answering this vlog!
I love how, in spite of limited first hand electronics experience, just UA-cam has by now burned in my mind: see a big cap on something recently disconnected, bridge it before getting your hands anywhere near
As a young TV tech trainee, I picked up a colour TV picture tube. When the HT ultor got near my chest, wack, instant discharge! Ouch, that hurt! I managed not to drop the tube and never did that again... lesson learnt!
Ah - the memories of copping a boot of a capacitor... I had the pleasure of fixing a Fluke 5215A Precision Amplifier with a burnt out section of board in the 2kV line... That was a long repair job.
3:12 You and Citroen have something in common. Citroen uses one gray color wire on a specific car, and you use red wiring on that panel! DO NOT let a dog solar electrician work on that panel. Dogs can not see red, just gray! =P
@@MattyEngland Most modern wiring insulation is soy-based. You can smell the soy beans. This is why rodents love to chew up wiring in cars. Are you related to Mickey Mouse? =P
@@kensclark Just because dogs don't appreciate the entire spectrum of color that humans do, that does not mean they are unable to perceive different colors. They just may not see the “true” color of an object. For example, the color red appears dark brownish-gray or black to a dog.
@@alansmith4734 Uh... That is exactly what I just replied... They see colors but not like we do. I replied in reference to you saying that all dogs see gray.
We seem to take bleeder resistors for granted these days. It took a similar experience with a 1000W PC PSU to finally buy a cap discharge tool since it nearly melted my screwdriver and scared the s... out of me!
Years ago I was running a small bare tube HeNe laser and even though I was well aware that the tube acts like a cap andholds a charge for several minutes after being turned off I managed to brush it with my right pinky while moving another HeNe tube and oh boy did that hurt like hell. I don't know how much voltage was still on the tube but I know it operates on 1250VDC and had enough on it to make my finger go numb for a good 20 minutes. Of course the one time I didn't put gloves on was the time I managed to touch a charged tube. That's one mistake I hope to never make again.
My guess for the rotated resistor is that it was hanging vertically in your panel and gravity pulled it down after it cracked and desoldered. (haven't finished the video, maybe that conclusion is drawn later)
When I was 15 I got thrown by a 400V DC pull up cap for the CRT screen I was playing in. Cap the size of beer can. My hand was sore and shaking (like the rest of my body) for a few hours. Think I was thrown from the DC shock, as I did have my other hand on part of the chassis. At 15 I didn't know that alone was a no, no. Never mind that the sucker on the back of the tube could carry +400V for long after the CRT was off. I got "bit" by a laptop power brick last year. Unplugged it and grabbed the mains plug. ZAP! Obviously those scrubber caps are getting higher value for efficiency!
10F at 2.7V equates to around 36J. So definitely quite a jolt. The diode and resister must have reached at least 220°C in order to melt the solder. With the parts of the resistor being light enough for capillary action to keep them on the board.
That split rotated R24 reminds me of an incidents at work back in the day when IC sockets were still widely used, we took a lightning strike which came down into the racks. Burnt ICs were literally physically ejected from there sockets.
Don't forget the leaking electrolyte may have bridged connections on that socket/board edge connector and this caused the resistor to overheat (sinking current somewhere it shouldn't?).
Thats why I used a 50-100 ohm resistor to charge the supercap, but provided a diode bypass for discharge. This way, even if the cap fails short, you only heat up the resistor way within its power handling capabilities.
One morning I came to work and the internet was down.Investigating the issue I found the router was off. A cap had blown in the wallwart PSU. This had vented through the case by breaking the plastic earth pin off the plug (UK double insulated type) and ejecting the PSU partway out of the socket.
Well I guess many people can relate to that capacitor shock. Reminds me of the time one of our suppliers send me one of their new and "improved" replacement models for verification, but they forgot to add the discharge resistor and I found out the hard way when unplugging it and got shocked on the dangling plug. Since then I'm paranoid and short every plug on a metal surface after pulling it. I guess in your case it isn't really needed since usually you wouldn't have a plug on that thing, so depending on your standards it may not be needed, but in my case it clearly violated some basic electrical codes/standards here that required the voltage to drop to safe levels in something like 5 seconds.
R24 most likely rotated during a EOS type failure. We make a design EOS/TLP pulsers and its common to see components physically eject themselves from PCBs.
Speaking of electrical zaps - years ago I was mowing the paddock with an old Victa which had the fuel tap just above the spark plug, and the only way to turn the thing off was to close the fuel and let it run out. I woke up about 10 feet behind the mower on my ass - the mower was stopped so it either ran out while I was out, or I stole all the spark....... probably 10 years old at the time.
When I read "Dave gets zapped", I was anticipating it. When you started talking about the capacitors, I thought, that would be the moment. And in the next second - ZAP!!! Just could not predict it more precisely ... 😁 I know it hurts! Once I got it across my thumb and index finger while repairing an extension cord. It was like a heavy hammer hitting my hand. As they say, unplug before open ... 😃
I remember when I had my first zap from a capacitor😂 I was about 12 or 13 years old and I was taking a TV apart to see how it worked 😆 the TV had been turned off for weeks, maybe months! I was quite shocked 😂
The diode may have been the initial failure, with the leakage causing AC to appear across the super caps which they would not have liked.The excessive current draw then cooked the primary side resistor.
I remember the time I learned about caps. I tore apart a disposable camera, the flash cap got me good. I was and am still surprised how long it held its charge. Love it
I've seen a resistor fail like this, but only once before... A customer's vehicle was struck on the little aerial on the roof of the vhicle, and the control unit for the telematics had failed inside the MCU itself. It seems as the though the resistor slowly heats up until the soldr melts, however the resistor continues to heat up slowly until it fails. When it seperates, the component breaks in half due to the heat and the solder is still molten, so the surface tension of the solder on the 2 halves of the resistor pull it in stange ways...
What are the Manufacturer’s Environmental Specs? I’d bet somewhere it says 25C max ambient temperature… If you measured your exterior electrical cabinet’s temperature at 12 noon, then say at 2pm then at 5pm what temperatures to you find? I’d bet you will find temp readings >30C. I recall problems in Jakarta with 32C often in the shade! Your solar monitor might need to be located inside the building and/or some exterior solar shading is needed over your electrical panel…
@@yadabub It's a known failure when you put supercaps in series for higher voltage ratings. There were a bunch of HP Raid controllers that had similar issues where they used supercaps in series as the backup battery to keep the cache alive if the power fails, and they failed in a similar manner. When the supercaps are new, they share the voltage equally across them, but as they age, they don't match anymore, and eventually the voltage across one of them is beyond the ratings, and then it pops.
Correct me if I'm wrong, but the scoring in the metal is a fail-safe to prevent too much pressure building up; but that doesn't mean that the bottom 'rubber gasket' can't give way earlier. As a matter of fact, leaking caps usually leak from the bottom. It's just to prevent it from exploding should the rubber gasket not give way.
@@Rob_III Yeah, the intentional weak points in the top of the can are a pressure release, but the rubber seals can and also do fail. That's why when you're intentionally blowing up electrolytic caps for fun or profit (by hooking them up with the polarity backwards and or above their rated voltage), the little caps that don't have the "vent" in the top of the can give better explosions than larger caps that have the vents.
Charged capacitors can be beasts... I once had a power supply disassembled on a Friday and continued working next Monday and then it bit me in my finger! The dam cap was still charged.
it seems like power on/off cycles (I dont know if there was any) constantly overloaded the diode because of those beefy hsr super caps. thats my guess. I felt the zap while watching. Hope thumb is recovered. love your videos. take kare.
It got hot enough to melt the connections, then cracked, then gravity took over. The gunk underneath is what's left of the glue. When they do low end boards like this they'll glue the parts on the bottom then use a mass soldering technique like wave soldering.
That failed capacitor looks like it has small bulge on top, not very noticeable but it is. Also tracks from second capacitor to card connector look broken and lifted a bit. About that resistor my guess is, that it got so hot that it almost desoldered itself and violently cracked in half, thus rotating like that, with solder solidifying right after that.
As you were holding the board I knew you'd get zapped. It was pure luck you weren't stung earlier. lol. I've only seen resisters like that once. It was caused, or at least what the engineer who came to fix it said, by lightning. It had struck one of the phone line relay boxes and had caused numerous, he said almost 30, wifi routers, to fail in the immediate area. He didn't have that many spares in his van so would come back the nest day. He left the damage one with me and told me to throw it away. But I wanted to see inside. numerous SM components had come off and I saw one like yours all melted and twisted.
Supercap failure due to over voltage due to poor series voltage sharing or due to long term AC currents causing heating and electrolite loss over time? SC's have pretty high internal resistances so perhaps the long term ripple was too much for it?
Hi Dave, I haven't seen a flipped torn component yet, but I have experienced a self desoldering triac in a TO220 case in our customer's rig. That's why I'm not writing, you already know why you got hit by that capacitor, I dare say I figured it out, because you forgot to put the jazzyk at the right angle and that caused a disturbance of the field of concentration on the distribution of components in the board 😁 All that remains is to say, that they have it nicely designed. Nice day 🙂 Tom
UPDATE: quite a few eagle eyed viewers spotted that R14 was missing. R14 is one of the balance resistors across the supercaps. That falling off could certainly have eventually triggered the event we saw here.
I cannot find the resistor, but upon inspection of the pads it was clear that the resistor was soldered in there at one point, but the joint appears like it cracked so to speak. So it could have been a poor solder joint that ultimately lead to the failure. Indeed, the unit likely wouldn't have survived long at all if R14 was missing from production. The resistor could have fallen out on handling and transport to the lab.
I have spoken to the designer and the balance resistors were chosen to prevent imbalance and hence overvoltage in the supercaps, and there have been no other reported failures in over 70k units installed. He also concluded that the unit wouldn't have survivded this long if missing from production. In any case the latest design they are sending does not have series supercaps like this. I'll try and do a teardown when I get it.
the super caps popped probably over charged
The problem is that its a mystery and a lot of people can't bear unsolved mysteries.. Maybe something caused R14 to heat up and that's why it fell off.
When your product fails and you get a free failure analysis from EEVBlog
@@Lucien86 From inspection of the joint it looks like it was a dry joint to begin with. How it actually fell off I can only presume durign the transportation from the panel to the lab. I did kinda toss it around in the car.
Although the missing resistor is connected with another supercapacitor._
6:11 Used the BigClive method of "I'll just go ahead and discharge this cap with my finger to be safe..."
Or perhaps going the electroboom route.
30 odd years ago a colleague got a zap from a board across his hand - no damage. All such incidents had to be reported. Some time later he was asked about the incident with some health and safety guys. He showed them what happened with another board. Of course the caps got him again - much swearing.
Helluva recreation, eh?
Reminds me of a Darwin I read about. Some guy got got electrocuted while working on a scaffold (I don't remember the entire incident), and the insurance company didn't believe what happened, sent somebody to investigate, he recreated the incident exactly and electrocuted himself.
If dave was in Germany and would have been working company hours during this incident, an ambulance would have been called and he would need to stay 24h at a hospital 😀
@@AnoNym-zi5ty Wow. Compare this to America where I took a 1000W plasma generator across the heart (instant burns on both hands). I got to drive myself to the hospital and they said I was fine to leave after 4 hours.
@@AnoNym-zi5ty if Dave was working on the clock he would have went to hospital for an ECG
The last time I got shocked was because I used a 1000V insulation tester to check some 5kV cable for suspected damage while it was still on a reel (it turned out to be fine). What I had not considered was that the conductor and foil shield together essentially made the cable a big capacitor, and that charging it up to 1000V and discharging it across my thumb would feel less than pleasant. 0/10 would not recommend.
@Ralph Reilly no, I rate the experience 0/10, and I would not recommend it. That's not a double negative.
As an electronics tinkerer, if you don't get zapped every so often, are you even doing anything? Thanks Dave for keeping it real
Oh no! Sorry to hear the device is dead. We'd love to help you out and will be in touch!🌞
@@Okurka. It's a known failure when you put supercaps in series. There were a bunch of HP Raid controllers that had similar issues where they used supercaps in series as the backup battery to keep the cache alive if the power fails, and they failed in a similar manner. When the supercaps are new, they share the voltage equally across them, but as they age, they don't match anymore, and eventually the voltage across one of them is beyond the ratings, and then it pops.
@@gorak9000 hgzgazezva der r
@@gorak9000 der rf er
@@gorak9000 I can't believe a decent engineer wouldn't have predicted this [that series supercaps would require balancing circuitry]... then again I've seen series supercaps paired direct from the manufacturer, and sold at "reputable" wholesalers, and I've seen "high end" equipment with parallel mosfets, which obviously failed, so apparently decent engineers no longer exist? Or maybe they're too pressured to compete, in so-called "productivity," with others' lack-of-care?
[edit: unrelated aside follows]
"Sorry to hear the device is dead ;)" "The device is dead"?! They're lucky they [it] didn't burn down his house!
[editted, after the rant that follows and apology at the end, in hopes of better clarity of my intent, should anyone come along 'fresh' hereafter. I did get a bit extreme in the "burning down the house" part. Hey, I'm human. But, as I recall , nearly all my comments hereafter are related to the concept of unbalanced series-caps, parallel mosfets etc., *not* this particular device's implementation of series-caps, except where explicitely stated]
@@gorak9000 Totally agree. Chamberlain / Liftmaster has the same problem with their garage door control panels. Will give you free replacements, but same flaw in the replacement with the two caps in series. Much worse failure mode - will open your garage door in the middle of the night.
I've had a few bites from 120Vac (usually a bad ground from a guitar amplifier) but the ones that caused more injury were DC
- at Tektronix (Wilsonville), someone 30 meters away blew off their finger on a 5V, 120A power supply... got their wedding ring across it.
- 12V DC (motorcycle gel-cell battery) fused a metal watchband to a lady's wrist when I was working at the place that made the first AED.
- ME? I got 5kV in one hand and out the other troubleshooting an early-production AED which had a badly mis-wired High Voltage section... someone in production got lost q/all the white-only 6kV silicone wire and somehow failed to connect up the dead-man relay properly so the 50ohm dump resistor was never across the Defib Cap... not even the 1000:1 voltage divider was properly in-circuit, so when I checked the cap looked like it was drained. I got within about half an inch and a BEAUTFIL Bright Blue Spark leapt from the cap and met my fingertip... then up through my arm, across my back (the scapulae met in the middle) and down through the other hand, jumped the anti-static mat (breaks down at 600 volts) and back to the unit chassis, which was siting on the mat.
Stupid me, I forgot about the mat's breakdown voltage and thought I was "Safe" with my hand almost 2 feet from the unit.
Yer not the only Dumb one, Dave.
I think my T-shirt should say :
Safe
Not By Design...
By Luck
Dave has an ElectroBOOM moment!
My thinking exactly
Thirded! ^^)
The gunk under the resistor is from the glue used in the reflow. That is a double-sided loadout of SMT parts, they would be all reflowed at the same time, the ones on the bottom have a dot of glue to hold them on during the reflow process. When the resistor heated up to such a high temperature, it burned/melted that glue under it.
..... Excellent detective work Mate.
Mixed with electrotyle for extra taste
@@EEVblog MMMMM yummy too!
@@EEVblog 🤔 that mixing fixation glue and electrolyte would create a new kind of explosive mixture tearing the component base 😶........ 😂
Your fuse box isn't old, it's just goofy as fuck to anyone who's from a normal country with normal things.
And people on the internet complain about north american breaker boxes. They're a lot tidier than that mess. That's what happens when you're on the bottom of the world - spent most of your time just trying to stay on the surface of the planet - no time to worry about tidy breaker boxes.
@@gorak9000 central European fuse boxes would like a word.
The whole fully non standardized “bolt stuff into a board and wire it up” style was used here before wwii. And even that wasn’t this free for all.
As a German i thought to myself "What the Fck are they installing over there" and got flashbacks to a breaker box i replaced at my grandfathers house (installed in 1934) which was evenly messy but there was also about 5x more stuff in it like 20 breakers as well as a retrofitted RCD. But i would have never thought that in an Industrial nation installations on plywood could have been legal in the 80ies.
@@minespeed2009 I don’t think the board is wood, it looks like some sort of phenolic resin type material. Somewhere between plastic and the cheaper kind of PCB.
I love how the electrical panel is just perf-board and random hookup wires…
im now starting to think that my completely passive reactions to being shocked might speak more towards my carelessness than my experience
220 or 120? Makes a difference 😉
I nuzzled a charged disposable camera flash cap once. I immediately took the battery out and put myself on "you don't get to play with this anymore" time-out.
Ahh this reminds me of the memories of taking Kodak cameras apart at 10 years old and being blasted by the 400v 100uF capacitor! I never trusted a circuit board without discharging every cap for years!
.... Complacency is our constant companion, can't believe the othher day I was using a bare jewelers screwdriver
to bend and close the female end of a (IEC C7 ) mains cable. Just luck I did the neutral first before realizing it was live. (FFS !).
I did that too! After the first time I got bitten, I definitely made a habit of discharging those flash caps.
When I was 11 years old my dad was repairing a big CRT and he just had left it open with no one around.
I walked in and touched the back of the tube and it felt like my arm got kicked in the balls by Tom Dempsey.
I was too socked to even cry or anything. I just said nothing and went back to my room and play video games.
When I asked my dad about it later he said the tube itself is a capacitor and it was probably charged around 10kV.
I've had this experience, my finger got 2 red points where i touched the leads for more than one month
@@felixar90 .... Yeah, I was so glad to see the end of CRT's & Video recorders. I Made a living from repairing them, don't miss KV's shooting up my arm.
When detecting the 2 supercaps in series I immediately got concerned about them being balanced.
@gorak9000 mentioned this too: supercaps don't like to be overcharged to a higher voltage and when placing supercaps in series there should always be some protection to prevent that from happening.
Balance resistors are great for small caps (we used to design high voltage capacitor banks that way) but for those large supercaps balancing resistors are not enough. At least some zener diodes or even better one of those active balancing chips especially designed for supercaps should have been used. But writing this I do remember that, back then, I was advices to just place balancing resistors over my supercaps by a component supplier.
In HP servers of generation 7 and 8 there are supercaps installed for the RAID controller. They saved a cent by not populating the capacitor module balancing resistors. Now we replace blown caps all day. And it's always one cap bulged because of overvoltage and one ok. (2 caps in series)
They are 2.2k balance resistors and were carefully chosen. See the other comment thread. Mine is the only known failure in over 70k devices installed.
Solar system analytics fail... meh, I'll watch later.
Dave gets zapped-INSTAWATCH! 😂
I predict 6:11 will be "Most Replayed" once the watch graph appears.
Dump it mate! I grew up on 220V 50 Hz. Got zapped endless times. Then went to England and got zapped on 240V Hey the extra 20 make you taking extra care. Then I had 117V 60Hz, that was just a near pleasant cribble. Next stop: Western Australia in 1982: They had 254V mains!! Never even wanted to get zapped.
You probably never worked with CRT TV's... 25kV and the picture tube acts as a perfect capacitor. It is a good advice to discharge the voltage properly before you try to remove the HV plug. (called the pacifier)
@@PlaywithJunk Yep, 1 zap with 25kV and the tube went zsssssss.....
@@LawpickingLocksmith Normally the tube will be fine but your finger hurts... 😁
You must admit, that diode held on for dear life before it actually failed.
That second SuperCap might have had problems also as you can see the shrink-wrap cracked on it as well (18:33).
Although there is a chance it was damaged by it's neighbor.
If they are in series and one has failed it is likely to have severely overstressed it neighbour. Both should be replaced, preferably two from the same batch.
so super they weren't anymore and I'd say they are the reason for failure. maybe compounded by his local climate, as compared to others.
The crispy-fried resistor sort of makes sense since it looks to be part of the snubber network on the primary side. I would hazard to say that D8 that looks kinda cooked as well might be a TVS (Edit: It is. A 200V Zener. I'm sure it got toasty piping hot!). With it there absorbing some of the excess energy the supply probably sort of worked, but once it cracked there was nothing to stop the flyback voltage from sailing away, probably from having the output loaded down or some other issue. I'd guess the caps started to go short-circuit, or possibly it was no longer in regulation and running in some form of quasi-open loop and overvolted everything.
Must have been a 'zap Dave day', as I got a 'tingle' changing the back light led strips in my LG TV, even though it had been powered down for over 8 hrs!.. Darned led driver psu caps... Cheers from a miffed Dave in the UK..😟
Supercap in Series is like charging battery in parallel the Voltage get unbalanced or overvoltage on one side over the specifications
Right... but *series*
@@ericwazhung right let me change that, I'm not that good at writing English comment
@@satria4195 But same with batteries... series... But, yes, parallel can have its own set of problems.
I have a suggestion Dave. You should video a break down of the new one checking PCB and quality of solder joints on the new unit, if you do replace it. Could help with future diagnostics.
Thanks for the videos, you are extremely smart. The vapors of the electrolyte condense above them when they got hot, causing the upper gross deposit.
A decade ago I was living in a little Victorian era cottage built in the 1860's, so because of its age the power was installed many years after it was built, with all the conduit installed on the walls and ceilings. It also still had the original fuse box complete with ceramic fuse holders and the obligatory packets of fuse wire so you could repair a fuse when it melted. None of that smart meter rubbish!
The first house I ever lived in and grew up in had exactly one fuse, and the only power outlet was part of the lighting circuit. As we began to need more outlets, we eventually had that old outlet removed and replaced with a modern outlet with modern wiring. (Ended up with three fuses: One for lighting and two for power)
@@melkiorwiseman5234 In our house, we still have everything on a single 10A breaker. We have many outlets and lights, but everything is on that one breaker... And aluminum wires, most outlets have no ground.
I never understood how you don't get shocked in your videos... I just assumed it was because you're an amazing EE but its good to see you're also human! haha
As an EE who only works with 24V/3.3V systems professionally, I'm suuuuper paranoid when I take mains stuff apart in my free time when stuff breaks... Rubber gloves and every primary cap is checked twice.
@@AnoNym-zi5ty Yep, I'm super paranoid about anything over 24V DC
If I ever met an EE that claimed he had never been zapped by a cap or live mains, I would assume they were either a liar, or still an intern.
@@nickwallette6201 look at Mr Gatekeeper right here, deciding who's EE and who's not.
I'm relatively sure Davis is human, but is also good at editing out the "Ow, FUCK!" moments when necessary...
New one will fit in the same box, just will have a nice big DIN rail box to hold all the breakers and other DIN rail mounted units, with the meters now all being on the top row in a line. Plenty of space then for extension.
When you replace the heater, insulate all the hot water pipes both outside and inside the roof, so as to minimise the heat loss from the long runs.
I’ve lost count as to how many supercaps I’ve changed out at work on the bench. They are notorious for leaking and destroying boards, components and tracks.
Soldered-in batteries are worse I think...
It's because they put them in series with each other with no protection or voltage balancing. When they're new, they're well matched, so share the voltage equally, but as they age, they don't match anymore, and the voltage creeps up on one until it's above the rating, and then it pops.
@@gorak9000 I like you. Hopefully some aspiring engineers learn a thing or two from your time.
...
OP: As for doing repairs, if you see them in series, prb best to replace both, regardless if one's OK, because a new one will put that much more strain on the old one.
@@gorak9000 These morons don't even put a string of resistors in series, each resistor being in parallel with a respective cap?
Yep, I've had a few shocks like that myself. Each and every time it's from powering up a piece of gear on the bench after repairing it, and stupidly handling the board with my bare hands, moving it to get a better angle for probing. One time I could even feel the 50Hz on the secondary of the transformer, and I had enough time to think "that's probably not good" before I started to put it down, evidently coming into contact with the primary in the process, which caused the muscles in my hand to twitch and throw the board across the room.
The last time my arm flew off and smacked into the side of my desk. That one really hurt. I couldn't even close my hand to pull out the power plug.
Last time I did it was when computers still used AT power supplies... I took mine out to clean the dust out, and when I put it back in it didn't power up.... I figured I probably had the switch wired wrong, and started unplugging the terminals... Live.
I ended up with a pair of burn marks on my finger which scarred (and one is still visible still 25 years later), and a bleeding hole on the back of my hand where I smacked it into the corner of a CD drive.
I've been a combination of careful and lucky ever since.
" Zappy zap! I just got zapped! " Needs to be on a T-shirt.
isnt the gunk just the glue which holds the smd parts onto the board before the board goes through the soldering machine?
Mixed with electrolyte for extra gooyness.
i can imagine what that zap felt like, i got zapped when i touched a charged flash capacitor in an old disposable camera!
I saw the caps as you split the case and thought, “hope he checks for charge on the caps before he proceeds”. I usually check with a meter but the thumb works good too! 😅. I learned the same lesson 30 years ago. Love your videos!
The isolation between AC mains and secondary circuit does not look safe on the power supply board. The 3G antenna is accessible so there should be double/reinforced insulation on that power supply. This kind of equipment should be designed taking over voltage category III requirements into consideration.
Perhaps the antenna jack is isolated from the main board. Power measurement is certainly easier if the measurement circuitry isn't isolated.
Opening the case halves is easy by inserting flat steel pieces in the rectangular holes on one side and releasing the plastic latches.
@13:24 Looks like one of the supercaps died. It's a little bowed, and there's the tell-tail signs of schmoo leaking out of the bottom. That would explain the "slow" overload that cracked a resistor and melted the diode. That gunk under the resister is very likely the glue used during manufacture (wave soldering / IR cooking the underside.)
I'm really surprised there aren't fuses all over that thing, being mains connected and all.
(Yes, I've seen resistors crack. Even the big 1960's era pinned ones -- inside an arcade console powersupply. Moisture gets into them and then heat/cold cycles take their toll.)
The input side is fused.
"Matsushita, they're actually good capacitors" - said Dave's thumb after testing them
ElectroBoom approves of the pain inflicted in this video.
Can you check if there is a voltage balancer for the super caps - just had recently the same issue.
Is this a common way of building electric panels in Australia? I'm just wondering, bc most of the devices in there are DIN-Rail compatible, mounted in a tiny box, with a tiny rail, mounted to a panel. The big meters look like for ordinary 3-point fixtures.
Of course every country has its standards, but as a German electrician, this looks a bit odd to me.
We use mostly standardized cabinets with DIN-Rails, 3 point fixtures and wiring conduits. This way the insides are more accessible and serviceable.
Though, interesting to see how it's made in other countries.
Surface tension in the solder pool will rotate components, I expect thats what happened to the half resistor.
Once got a 1000VDC zap from a PMT power supply by somehow touching my probe with a metal braided shield touching the back of my hand - man I had 3 fingers numb for a few months doc even sent me to the neurologist where they did an EMG with a sharpie and marked where the damage was (ulnar nerve I think) - thank goodness it healed on its own - high DC voltage is no joke m8 glad you're OK
.... I was thrown accross the room as a child investigating my brothers 'reel to reel'.
The drywall had my body shape imprinted (Like a Warner Bros. cartoon).
@@robbieaussievic I was thrown across the room as a kid from a live chassis tv I was fixing.
@@EEVblog That's where the term fly-back transformer comes from!
@@EEVblog Yeeah the same here from a cheap ATX power supply. Bloody live heatshinks! Should have known better....
Since it's a two-sided load, one side of the board will have all the SMT components glued down so they stay on until it gets cycled through the reflow oven. That may be the source of the bunk under some, but not all, of the destroyed/lost components.
Forty years ago I was fault finding a Yankee made telecommunications impulsive noise test-set that had been converted from its original 120V to 240V for operation here in Australia. The bloody thing weighed a tonne, the Yanks had obviously built the case on it out of leftover Sherman Tank steel panels from WW-2. Anyway I wanted to turn it over on the repair bench to measure some voltage readings on the rear circuit board and so I grabbed one of the case front metal handles with my left hand and put my right hand under it to lift what I thought was just the metal chassis but as soon as I lifted it off the bench I copped it with full-bore 240AC surging through my entire body. I lost all control of my muscles and was left standing there with 240V rippling through my chest, and shaking violently and uncontrollably with this damn thing held in a vice like death grip that I couldn’t release and the only thing that saved me was that after a period of time that seemed like it would never end it literally tore itself out of my hands because of its sheer weight and fell crashing down upon the bench, whereupon, I fell crashing backwards and just by luck happened to fall into the seat of the chair positioned behind me. That was, and still is, the closest thing that I had, and have, ever experienced to me leaving this mortal coil, on that day. To close this story, when I later investigated what had caused me to receive this life threatening shock, I found that whoever had previously done the conversion on it from the original Yankee 120V wiring, to the 240V Australian supply, had left the majority of the original Yankee Mains wiring in place , which included the on/off switch on the front panel breaking the Neutral, instead of the Active, and the ‘un-switched’ Active wire then continuing around to the rear of the unit and being terminated upon a ‘bare’ completely un-encased, 3AG fuse holder that was positioned right in the middle of the low voltage 12 volts secondary power supply, and it was that bare, and live, 3AG fuse holder, that the fingers of my right hand had touched when I lifted it off the bench. Unbelievable stupidity, by the Yanks, in the first place, for such a dangerous design and build, on the damn thing, and by the unknown idiot, who subsequently did the 240V Australian conversion on it, and finally by me, for relying upon the front panel mounted power switch, on it, to turn it ‘Off’ instead of turning it Off at the bench power point, before I tried to lift it to turn it over. It taught me a very valuable lesson that day, however, to NEVER, ASSUME, that ANYTHING, Mains Powered, is wired correctly, before you work upon it…!! 🙄
Some caps you just cannot discharge, especially there are caps on the other side of a transformer. Discharge a cap on one side, the caps on the other side charge up. This happened in a module in a fire control black box in an attack aircraft while I was in the Navy. We sent back the module repaired, bare as with the other type of modules with the document in its own heavy duty vinyl weather bag. Well, it bit a guy in supply and henceforth that module and others like it had to be in a separate heavy duty document bag or other container. Kudos for the SA guy answering this vlog!
I love how, in spite of limited first hand electronics experience, just UA-cam has by now burned in my mind: see a big cap on something recently disconnected, bridge it before getting your hands anywhere near
Dave trying to do a Mehdi by touching that cap before checking it has some bite left in it.
I am soooo relieved seeing the backside of your fusebox - I can now claim my fusebox looks slightly better organized than Dave’s!
As a young TV tech trainee, I picked up a colour TV picture tube. When the HT ultor got near my chest, wack, instant discharge! Ouch, that hurt! I managed not to drop the tube and never did that again... lesson learnt!
oh.. nothing like the good old mains-rectified 300vdc, sure works better than coffee to wake you up in the morning
Ah - the memories of copping a boot of a capacitor...
I had the pleasure of fixing a Fluke 5215A Precision Amplifier with a burnt out section of board in the 2kV line... That was a long repair job.
3:12 You and Citroen have something in common. Citroen uses one gray color wire on a specific car, and you use red wiring on that panel!
DO NOT let a dog solar electrician work on that panel. Dogs can not see red, just gray! =P
Discrimination, I can smell red
@@MattyEngland Most modern wiring insulation is soy-based. You can smell the soy beans. This is why rodents love to chew up wiring in cars. Are you related to Mickey Mouse? =P
Dogs have color vision. They do not see in gray.
@@kensclark Just because dogs don't appreciate the entire spectrum of color that humans do, that does not mean they are unable to perceive different colors. They just may not see the “true” color of an object. For example, the color red appears dark brownish-gray or black to a dog.
@@alansmith4734 Uh... That is exactly what I just replied... They see colors but not like we do. I replied in reference to you saying that all dogs see gray.
We seem to take bleeder resistors for granted these days. It took a similar experience with a 1000W PC PSU to finally buy a cap discharge tool since it nearly melted my screwdriver and scared the s... out of me!
6:12 Electroboom is proud of you!
Years ago I was running a small bare tube HeNe laser and even though I was well aware that the tube acts like a cap andholds a charge for several minutes after being turned off I managed to brush it with my right pinky while moving another HeNe tube and oh boy did that hurt like hell. I don't know how much voltage was still on the tube but I know it operates on 1250VDC and had enough on it to make my finger go numb for a good 20 minutes. Of course the one time I didn't put gloves on was the time I managed to touch a charged tube. That's one mistake I hope to never make again.
OMG.
I was totally okay with this "fuse board," it was just, "alright, so it's old, whatever."
THEN IT OPENED.
My guess for the rotated resistor is that it was hanging vertically in your panel and gravity pulled it down after it cracked and desoldered. (haven't finished the video, maybe that conclusion is drawn later)
I remember one morning walking in to work doing a small DVD warranty job, and the cap got me. No coffee needed that morning Dave.
When I was 15 I got thrown by a 400V DC pull up cap for the CRT screen I was playing in. Cap the size of beer can. My hand was sore and shaking (like the rest of my body) for a few hours. Think I was thrown from the DC shock, as I did have my other hand on part of the chassis. At 15 I didn't know that alone was a no, no. Never mind that the sucker on the back of the tube could carry +400V for long after the CRT was off.
I got "bit" by a laptop power brick last year. Unplugged it and grabbed the mains plug. ZAP! Obviously those scrubber caps are getting higher value for efficiency!
OK to leave those CTs open-circuit with the system live? I always thought that was a no-no, or do the CTs have burden resistors built in?
I believe they are just normal current output type with no voltage conversion resistor. They are disconnected.
10F at 2.7V equates to around 36J. So definitely quite a jolt.
The diode and resister must have reached at least 220°C in order to melt the solder. With the parts of the resistor being light enough for capillary action to keep them on the board.
I brushed my thumb over a charged capacitor from a old disposable camera that I took apart years ago and it put a hole in my thumb, lol.
That split rotated R24 reminds me of an incidents at work back in the day when IC sockets were still widely used, we took a lightning strike which came down into the racks. Burnt ICs were literally physically ejected from there sockets.
so, that's where those 5 or 10 watts were going that you couldn't find in the power audit video
Ah yes, the old magic smoke! Just like toothpaste, once it gets out it never goes back in!
Don't forget the leaking electrolyte may have bridged connections on that socket/board edge connector and this caused the resistor to overheat (sinking current somewhere it shouldn't?).
Thats why I used a 50-100 ohm resistor to charge the supercap, but provided a diode bypass for discharge. This way, even if the cap fails short, you only heat up the resistor way within its power handling capabilities.
LOLz, happens to us all. Normaly you don't get zapped, but when something is fulty you never quite know due to the fult condition.
Great analysis! Thanks for showing, including your little zap.
One morning I came to work and the internet was down.Investigating the issue I found the router was off. A cap had blown in the wallwart PSU. This had vented through the case by breaking the plastic earth pin off the plug (UK double insulated type) and ejecting the PSU partway out of the socket.
6:00 - 6:24 lol, that's what she said! :) Happens to the best of us mate. Hope there were no lingering burn marks
Well I guess many people can relate to that capacitor shock. Reminds me of the time one of our suppliers send me one of their new and "improved" replacement models for verification, but they forgot to add the discharge resistor and I found out the hard way when unplugging it and got shocked on the dangling plug. Since then I'm paranoid and short every plug on a metal surface after pulling it. I guess in your case it isn't really needed since usually you wouldn't have a plug on that thing, so depending on your standards it may not be needed, but in my case it clearly violated some basic electrical codes/standards here that required the voltage to drop to safe levels in something like 5 seconds.
R24 most likely rotated during a EOS type failure. We make a design EOS/TLP pulsers and its common to see components physically eject themselves from PCBs.
Interesting video and great to find the point of failure - also great to send the vid to the maker for an excellent feedback tool. Cheers mate.
That's cool - Point-to-point wiring behind a pegboard with DIN mounts everywhere. Never seen that before.
Speaking of electrical zaps - years ago I was mowing the paddock with an old Victa which had the fuel tap just above the spark plug, and the only way to turn the thing off was to close the fuel and let it run out. I woke up about 10 feet behind the mower on my ass - the mower was stopped so it either ran out while I was out, or I stole all the spark....... probably 10 years old at the time.
Way too crowded in there. Overheated for sure.
That's not at all what happened, but I won't bother explaining it to you - you clearly think you know more than everyone else
*shifty eyes* appears to notice a lack of RCD (after noticing the same during a power company directed fault test at my place)
When I read "Dave gets zapped", I was anticipating it. When you started talking about the capacitors, I thought, that would be the moment. And in the next second - ZAP!!! Just could not predict it more precisely ... 😁
I know it hurts! Once I got it across my thumb and index finger while repairing an extension cord. It was like a heavy hammer hitting my hand. As they say, unplug before open ... 😃
It's a spoiler, but I gotta get the clicks.
I read Dave gets zapped,and the video opens with you at the box. I thought "oh no!" Then you open the device and I saw those caps, then I knew!
I remember when I had my first zap from a capacitor😂 I was about 12 or 13 years old and I was taking a TV apart to see how it worked 😆 the TV had been turned off for weeks, maybe months! I was quite shocked 😂
Clive will love this part around the 6 minute mark hahaha
He constantly checks caps by sticking his finger across them. Just ironic that it should bite EEVblog by accident.
Biggest zap I've done is 120v ac. One time barely noticeable, another time I was white for the afternoon. Glad to see you'll make a full recovery.
Oh yhea, and then there was the leaky spark plug wires from an HEI (high energy ignition) system. Memorable.
@@bertblankenstein3738 Don't get that across both your hands...
The diode may have been the initial failure, with the leakage causing AC to appear across the super caps which they would not have liked.The excessive current draw then cooked the primary side resistor.
Are the three voltage dividers too close to each other for the phase-to-phase potential?
I remember the time I learned about caps. I tore apart a disposable camera, the flash cap got me good. I was and am still surprised how long it held its charge. Love it
If we can give a minute of silence for a resistor 🤔 we need more for someones thumb! 👍👍👍👍👍
I've seen a resistor fail like this, but only once before... A customer's vehicle was struck on the little aerial on the roof of the vhicle, and the control unit for the telematics had failed inside the MCU itself. It seems as the though the resistor slowly heats up until the soldr melts, however the resistor continues to heat up slowly until it fails. When it seperates, the component breaks in half due to the heat and the solder is still molten, so the surface tension of the solder on the 2 halves of the resistor pull it in stange ways...
What are the Manufacturer’s Environmental Specs? I’d bet somewhere it says 25C max ambient temperature… If you measured your exterior electrical cabinet’s temperature at 12 noon, then say at 2pm then at 5pm what temperatures to you find? I’d bet you will find temp readings >30C. I recall problems in Jakarta with 32C often in the shade! Your solar monitor might need to be located inside the building and/or some exterior solar shading is needed over your electrical panel…
That's kind of weird that the supercap didn't fail out the top. The metal is scored there to send the explosion up and away from the PCB.
Perhaps the can was damaged during manufacturing.
@@yadabub It's a known failure when you put supercaps in series for higher voltage ratings. There were a bunch of HP Raid controllers that had similar issues where they used supercaps in series as the backup battery to keep the cache alive if the power fails, and they failed in a similar manner. When the supercaps are new, they share the voltage equally across them, but as they age, they don't match anymore, and eventually the voltage across one of them is beyond the ratings, and then it pops.
Correct me if I'm wrong, but the scoring in the metal is a fail-safe to prevent too much pressure building up; but that doesn't mean that the bottom 'rubber gasket' can't give way earlier. As a matter of fact, leaking caps usually leak from the bottom. It's just to prevent it from exploding should the rubber gasket not give way.
@@Rob_III Yeah, the intentional weak points in the top of the can are a pressure release, but the rubber seals can and also do fail. That's why when you're intentionally blowing up electrolytic caps for fun or profit (by hooking them up with the polarity backwards and or above their rated voltage), the little caps that don't have the "vent" in the top of the can give better explosions than larger caps that have the vents.
@@gorak9000 This is the same as charging without a balancer, for example, 18650 - a huge mistake
If you didn't already know, Matsushita is the parent company of Panasonic and Technics.
Charged capacitors can be beasts... I once had a power supply disassembled on a Friday and continued working next Monday and then it bit me in my finger! The dam cap was still charged.
it seems like power on/off cycles (I dont know if there was any) constantly overloaded the diode because of those beefy hsr super caps. thats my guess. I felt the zap while watching. Hope thumb is recovered. love your videos. take kare.
It looks more like the leaked electrolyte shorted the output of the power supply and caused an overload, which is why the output diode was overheated.
It got hot enough to melt the connections, then cracked, then gravity took over. The gunk underneath is what's left of the glue. When they do low end boards like this they'll glue the parts on the bottom then use a mass soldering technique like wave soldering.
That failed capacitor looks like it has small bulge on top, not very noticeable but it is. Also tracks from second capacitor to card connector look broken and lifted a bit.
About that resistor my guess is, that it got so hot that it almost desoldered itself and violently cracked in half, thus rotating like that, with solder solidifying right after that.
As you were holding the board I knew you'd get zapped. It was pure luck you weren't stung earlier. lol.
I've only seen resisters like that once. It was caused, or at least what the engineer who came to fix it said, by lightning. It had struck one of the phone line relay boxes and had caused numerous, he said almost 30, wifi routers, to fail in the immediate area. He didn't have that many spares in his van so would come back the nest day. He left the damage one with me and told me to throw it away. But I wanted to see inside. numerous SM components had come off and I saw one like yours all melted and twisted.
Supercap failure due to over voltage due to poor series voltage sharing or due to long term AC currents causing heating and electrolite loss over time? SC's have pretty high internal resistances so perhaps the long term ripple was too much for it?
I got a cap zap from an Amstrad satellite receiver box in the 90's. It had been on the shelf at least a week before I opened it.
Hi Dave, I haven't seen a flipped torn component yet, but I have experienced a self desoldering triac in a TO220 case in our customer's rig. That's why I'm not writing, you already know why you got hit by that capacitor, I dare say I figured it out, because you forgot to put the jazzyk at the right angle and that caused a disturbance of the field of concentration on the distribution of components in the board 😁 All that remains is to say, that they have it nicely designed.
Nice day 🙂 Tom
Who HASN'T been zapped from a board?
Let's just say it happened way more often than I am comfortable admitting.