back to article ASUS recalls motherboards that flame out thanks to backwards capacitors

A few weeks ago a curious thing started happening to gamers whose PCs are powered by the ASUS Z690 Hero motherboard – their systems started to catch fire. Reports of smoldering boards started surfacing on Reddit and various forums. ASUS issued a recall on the product, or rather what they call a “replacement program.” With …

  1. Little Mouse
    Coat

    Shocking.

    1. Chris G

      A board that gives the capacity to smell the burning in a hot game.

    2. Bitsminer Silver badge

      A reversal of fortune.

    3. Tom 7

      I'm fucking gob-smacked! Does no-one do ERC these days? And do they not have some connection between the CCT and the thing that puts the components into the motherboard?

      1. Rattus
        Boffin

        In their defence

        Given the numbers involved this looks like 1 reel of capacitors was the wrong way around (in the tape - not the reel being loaded the wrong way because tape reels are asymmetric)

        That is highly unlikely to happen with just one MPN - most likely a different MPN being used due to the wonderful supply problems we are having - and this brand / MPN happens to come the other way round.

        You would hope that the technician loading the reel might have noticed, but that is a big ask to spot a stripe being the other end of a part, especially when you are probably loading a fast loader - not the pick and place machine directly.

        Next, AOI should have spotted the capacitors were the wrong way round - and a good line will have AOI cameras after pick and place and after the oven. However these are probably only configured to 'look' for missing or miss-aligned parts NOT to check that the part is the correct way around.

        Then we get to test and inspect.

        (I will state that I have not gone digging into this story, having only read this article so here I can only offer an opinion)

        Motherboards are a high enough value part that they are probably all tested...

        They are likely put onto an automated test station (a test fixture, possibly a bed of nails) that will power the board and take measurements to check if the unit is functional (and probably to program any programmable devices at the same time).

        Given that the fault is a 'memory capacitor' - I think they mean a capacitor in the memory power supply (or perhaps they mean a bulk capacitor, which will almost certainly be in the input or output of one of the switch mode power supplies) - in any-case it is common practice to have several capacitors in parallel to build up bulk capacitance in order to achieve a longer operating life because capacitors (especially wet Aluminium Electrolytic) degrade over time. Thus a missing (or a reverse polarity) capacitor is likely to have little or no effect on initial operation of the system. Failure of a reversed capacitor is unlikeley to happen in the first few hours - so again this is would probably not be spotted...

        Had this been the case on boards destined for engineering test (as opposed to end of line) then I would expect the fault to be spotted, but I believe it is quite understandable that this fault would get through production. That said now that the horse has bolted you can bet that this particular door will be closed.

        /Rattus

        1. martinusher Silver badge

          Re: In their defence

          Good analysis. I don't think the capacitor was aluminum because they're too bulky, I've only seen tantalum or high density ceramic used. I'd guess it was a tantalum because it won't fail immediately. It must have been fairly high value because it was downstream from a transistor acting as a high current series regulator.

          There is a design fault, though. Given the likely effect of a faulty capacitor -- the board burning up --

          the designer should have borne this in mind and either put some kind of current limit on the regulator or a sacrificial zero ohm resistor ("fuse") that would burn up when a fault happened. They didn't and I'd guess the reason for this is that the board is cost reduced to the max -- it might be a 'high value product' but its definitely not industrial grade design.

          1. lnLog

            Re: In their defence

            Those are polymer caps, tantalum are too expensive / bulky for that usage (voltage/capacity combination).

          2. ChrisC Silver badge

            Re: In their defence

            "There is a design fault, though."

            Bit of a stretch to suggest that - this wasn't a faulty capacitor where the failure mode would typically (unless you were insane enough to use tantalums) be relatively benign and could be taken into account at the design stage, this was a *mis--fitted* capacitor, which really isn't something a designer should ever need to consider for a part that's permanently fitted at the factory. As someone who does earn a living designing industrial-spec systems, dealing with this sort of fault doesn't even get raised as one of the wilder "what if" scenarios during the FMEA meetings.

            And I suspect it's quite likely the upstream power supply would have had some form of current limiting/fusing - even a basic entry level motherboard is unlikely to be designed with the same level of penny-pinching zeal as something where the possibility of saving even a fraction of a penny might be taken as a matter of routine - but if the failure mode of an incorrectly fitted part is that it acts as a low resistance load to that power rail, then there's a reasonable chance that it'll reach a steady state whereby it's sinking not quite enough current to trip the limiter/fuse (and bear in mind that some of the power rails on a modern motherboard will need to have their limiters/fuses set quite high to cope with the intended demand on the rail), but more than enough to cause it to get hot, at least up to the point where it fails entirely.

          3. Anonymous Coward
            Anonymous Coward

            Re: In their defence

            If it was ceramic I believe those are usually nonpolar, i.e. it doesn't matter the orientation of the capacitor.

            This appears to be a tantalum or polymer capacitor, both of which are polarized and will fail if installed backwards.

            I'm guessing polymer because I have seen the effect of a tantalum put in backwards, and it is an instant fireball which a zero ohm resistor wouldn't break before the tantalum ignites. Apparently what catches fire is the MOSFET from the surge of current from the capacitor shorting out.

        2. Tom 7

          Re: In their defence

          Its a long time since I worked on these things (80..90s) but you COULDNT put things in the wrong way round! And if a supplier provided them in packaging the wrong way round they were toast.

          1. the spectacularly refined chap

            Re: In their defence

            "Couldn't" in particular situations only. This kind of issue is fairly common at the assembly phase, generally they will be picked up in QA but not everything gets caught.

            Reels can't be loaded onto the machine the wrong way round but sure the reel can be wound in the opposite orientation to expected. That doesn't make it "wrong" if it is the orientation specified. As others have stated current supply chain disruption may mean alternative options have had to be sought out.

            That's assuming they are reel packed to begin with. If they were ammo packed the burden of getting it right is on the operator.

            Then there are setup issues. I doubt a dedicated line exists for these boards, there simply won't be the demand for them to justify it. It'll be a general board assembly plant, the order will come through to spit out a few thousand of this particular board. Someone will look at the order and set up the machine as appropriate. Or possibly as here, screw up and write the lot off.

        3. ChrisC Silver badge

          Re: In their defence

          Indeed. Been there, done that etc... IIRC for us it was a "drop in" replacement optoisolator which came reeled in the opposite orientation to the one it was replacing. Might have been a drop in replacement in terms of schematic and PCB design, but most assuredly NOT a drop in replacement as far as the SMT line was concerned. At least in that case, the incorrect orientation meant the boards couldn't get past production test, so customers were none the wiser and we only had to rework about 100 boards from that initial build run.

          I also have a vague memory of occasionally seeing components available with both standard and inverted orientation reels, presumably as a way of avoiding problems like the above, so that provided you ordered the right part number then you genuinely would get a drop in replacement for whatever other version of the component you were already using...

      2. cyberdemon Silver badge
        Devil

        Does no-one do ERC these days?

        Never mind ERC (electrical rules check is a schematic analysis that would not have caught a capacitor fitted backwards by the pick n place machine) but does nobody apply power to the boards before they leave the factory? Aren't they supposed to have a self-test in the BIOS that would at least say "no CPU or RAM installed, but i'm otherwise OK" ?

        1. Unicornpiss

          Re: Does no-one do ERC these days?

          Since the boards were happily working in users' systems... um, until they weren't anymore, applying power at the factory (which no doubt is done as a basic test before they leave), wouldn't have caught it. It really would take some sharp-eyed QA inspector to notice it.

          1. cyberdemon Silver badge
            Paris Hilton

            Re: Does no-one do ERC these days?

            Errr, when an electrolytic capacitor is connected backwards, it usually explodes pretty much immediately after applying power..

            Nowhere in TFA do I see it mentioned that they were "happily working in users' systems" but I haven't checked Reddit, obviously.

            Although it DOES say "may cause debug error code 53, no post, or motherboard components damage" - all of which can be detected by testing at least to POST, at the factory.

        2. Tom 7

          Re: Does no-one do ERC these days?

          ERC should be part of an integrated system so that the pick and place was driven by the cct diagram. If the capacitor was the wrong way round in the cct then the ERC would have picked it up. There whole engineering paradigm then was - people fuck up, that's really expensive so lets make sure there is no way for them to fuck up. There is a little custom PCB place I visited not far from here and anything with a run of more than 50 boards is all automated because its simply not worth letting humans get it wrong.

          Pissing off a customer costs more than the couple of hours of computer time involved in dotting the T's Crossing the i's and pointing out to an extremely well trained but human cct designer that you do not connect that bit of a capacitor to a net that connect to that bit of a fet.

    4. This post has been deleted by its author

  2. Mishak Silver badge

    Reminds me of a place I used to work at

    They used to get part-time staff in to work nights when they needed extra capacity.

    They paid them peanuts, gave the first group no training, and were surprised when all of the first batch of board failed when they hit the test cell the next day.

    Turns out it would have been a good idea to teach them that diodes and some capacitors needed to be put in the right way round.

    1. Red Ted
      Mushroom

      Re: Reminds me of a place I used to work at

      A variation on this was a batch of through hole tantalum capacitors that there labelled with the + ambiguously close to the other pin.

      So when you powered the board up for test, there would be a bang and the capacitor body would vaporise. This left just the two legs poking out of the PCB that were surprisingly difficult to find!

      Note quite this bad, but the first few times it did make you jump! -->

      1. Boothy

        Re: Reminds me of a place I used to work at

        Reminds me of a company I worked at in the mid 90s, one of the items we built, for some custom gear, was some dual rail PSUs that provided +24v and -24v lines (plus a centre 0v line).

        The company was stingy, so always bulk bought the cheapest components, including the caps.

        Some batches would be quite poor quality, regularly popping under load test, despite being supposedly within spec.

        We didn't have a separate area for testing, just a bench in the same room where everyone worked, and people got fed up of jumping in their seats when they went pop! Quite loud sometimes!

        So we ended up building a box out of some spare lumber we had, we then lined the internal sides and top with foam for sound proofing (but not the bottom, not a good idea to have a hot running PSU under full load sat on flammable foam! We had an air gapped sheet of aluminium in the bottom to sit the PSUs on).

        Was much better after that!

        We eventually built a rig just to test the caps, to save on re-soldering later!

    2. Tom 7

      Re: Reminds me of a place I used to work at

      They let people put the components onto the PCB? What is this the 70s?

      1. This post has been deleted by its author

      2. Robert 22

        Re: Reminds me of a place I used to work at

        There was probably a machine involved, but if the parts were loaded incorrectly ....

        1. Tom 7

          Re: Reminds me of a place I used to work at

          The packaging these things come in mean you CANT put them in the wrong way. Or if you can then you NEVER use that supplier again,

    3. Mr. V. Meldrew
      Coat

      Re: Reminds me of a place I used to work at

      "They used to get part-time staff in to work nights when they needed extra capacity."

      Surely they needed "extra capacitors"? :-)

      Jacket icon for obvious reasons......

  3. Dwarf

    OS reactivation

    Not being a windows user any more, do they still lock the OS licence to the motherboard ?

    If so, then who pays for the OS reactivation after the failed motherboard has been replaced

    1. Binraider Silver badge

      Re: OS reactivation

      El Reg readers typically being wise enough to avoid being early adopters I'm not sure how many we'll see here to report back, but yes, I would love to hear how this plays out.

      New laptop board? It's a new laptop according to MS activation. Get a new license.

      1. Anonymous Coward
        Anonymous Coward

        Re: New laptop board? It's a new laptop according to MS activation. Get a new license.

        Much as I am averse to MS and their products, I think that changing such a core component is almost always likely to end up looking like a new laptop, and so I would expect any bog-standard MS activation attempt to indeed come to that conclusion.

        And at best you would have to expect to jump through a few hoops to convince them otherwise (which is not to say I don't think the hoops should exist -- there should be a way, but I would still expect it to be more than a little annoying)

        1. Steve Kerr

          Re: New laptop board? It's a new laptop according to MS activation. Get a new license.

          It's more likely these boards were part of self builds or small company builds so they would not have OEM licenses for windows.

          I originally purchased a Win7 license when it was released (not OEM tied).

          I've been thru this when I had an i7 die and the cost of a replacement was laughable, it was more cost effective to build a new machine and transplant some bits that I could (decided to do a pimped out box, with RGB lighting and stuff - never again will I do this after the 12-14 hours it took over 4 days with huge amounts of internal wiring!)

          So, new case, motherboard (switched to AMD), RAM - only things that remained were the original SSDs and the graphics card.

          I booted it and rather expected it to blue screen and die and have to reinstall windows, instead it installed all the correct drivers, didn't complain and asked to relicense which took less than 5 minutes to sort with the license key.

          So as much as I hate to say it, but kudos to M$ for making a seemless migration with the same install into what was pretty much a totally different PC with a different CPU and drivers and not had any problems since in the 2 1/2 years since it was built.

          1. Boothy

            Re: New laptop board? It's a new laptop according to MS activation. Get a new license.

            Similar here.

            I switched from an old i7 to an AMD Ryzen a couple of years back, I just moved everything across into the new build, basically a new case, Motherboard, CPU and memory, but same 2.5" SSD boot drive etc. Everything just worked after a bit of updating, even re-authenticating itself after a while.

            My system does a full image backup of the boot drive (now an M.2 NVME SSD) to a NAS, I periodically do a recovery test to the old i7 system, which now has an old HDD in it. On boot, a bit of driver updating, and it works again!

          2. Robert 22

            Re: New laptop board? It's a new laptop according to MS activation. Get a new license.

            OEM vendors, such as Dell, will usually install a version of Windows that is licensed for the motherboard. If you rep[lace the motherboard with anything different, the OS is unlikely to activate.

            If you have a transferable version of Windows, it is often helpful to link your license to your Microsoft account. Note that the OS key that appears in the registry is a generic key that can't be used for activation.

            1. Rob Daglish

              Re: New laptop board? It's a new laptop according to MS activation. Get a new license.

              If I remember correctly (it's been a few years since I did the tests) at least HP and Lenovo (possibly Dell) expect that the engineer that replaces the board will correctly licence it when they set it up - although it's a horrible process to go through and requires *exactly* the right version of utility pen from said company, who seemingly won't release it to anyone under any circumstances, which was too much like hard work for many of my colleagues...

    2. Pascal Monett Silver badge

      Re: OS reactivation

      I don't think a new license will be required. Borkzilla is capable of understanding that you change your hardware. I think that you have five activations possible before they lock down activation, but even then, you can call them and explain the situation.

      Of course, we're talking honest people here. If you activate your license on four different computers and use them concurrently or within a given time frame, don't be surprised if your key gets rescinded.

      In this case, the situation is pretty clear : yes, you had activated on a certain configuration, but that config is gone and you're activating on a new one.

      Honest people should not have any trouble.

      1. Anonymous Coward
        Anonymous Coward

        Re: OS reactivation

        Even on W7 and possibly XP - an MS Office licence was tied to the motherboard too. Since then IIRC it has become the MS norm.

      2. Rodderstoo

        Re: OS reactivation

        This honest person had trouble.

        Attempting to reactivate Windows by phone. The recorded message was of a woman with a strong southern US accent who spoke too fast and who, after I had input the long string of characters that she read out, wanted me to hit the (unintelligible) key. I played the recording back many times and never managed to determine what key was required. I had to give up and install Linux instead..

    3. Roger 11

      Re: OS reactivation

      Windows 10 does. I learned it the hard way.

  4. Anonymous Coward
    Anonymous Coward

    Never cross the streams!

    Or capacitors it seems.

  5. Anonymous Coward
    Anonymous Coward

    “fastest we’ve seen a brand go from there’s a problem to we’re handling it, that we’ve ever seen.”

    Fire damage related lawsuits wait for no one. Not even if they involve personal injury.

    1. Eclectic Man Silver badge

      Liability

      Yes indeed. I suspect that the prospect of burning down someone's house and the attendant claims for personal injury, and other potential damage concentrated their minds wonderfully. After all, if it burst into flames when in 'normal use' then it is, in UK law "not of merchantable quality", and there is no exemption clause in any contract that will get you out of liability for that.

    2. John Gamble

      Sadly, there were manufacturers that were dragging their feet despite the Now With Extra Flames feature, which is why JayzTwoCents could make that sadly unironic statement.

      2021 was a good year for exploding power supplies and flaming cases.

  6. Nugry Horace
    Mushroom

    At least it wasn't one of the big ones... https://assets.publishing.service.gov.uk/media/5acb313ded915d5a90e44be4/R052018_180320_Guildford.pdf

    1. jackalek

      This is actually a very good read! Thanks for that.

  7. TheManCalledStan
    FAIL

    QC fail

    I suspect overly tight QC being overridden.

    Given the different brands of capacitors being used, the original QC camera specs would go red everytime an alternative capacitor was used.

    So instead of recoding the QC cameras to look for the stripe only (instead of all the features shown on the cap), they switched it off...

    1. Snowy Silver badge
      Flame

      Re: QC fail

      Yes a very big QC fail makes me not want to buy any stuff made by them.

      Seeing the board was made using a PCB assembly machine does make me wonder if the magazine holding them was put in backwards or they where made by someone different and where oriented backwards to the normal supplier. Which would make it a buying error too.

    2. Bartholomew

      Re: QC fail

      Electrolytic capacitors are usually larger through hole parts and are generally hand soldered onto the board after the pick and place machine and the boards have gone through the ovens. Basically any non-SMD (Surface Mount Device) parts end up typically being hand soldered. So the guess would be the "new" staff were trained how to solder on parts, but not that some parts are polarised and need to be soldered on one way around only (plus to plus and minus to minus to match the part up with what is shown on the silkscreen layer of the PCB (Printed Circuit Board)).

      1. the spectacularly refined chap

        Re: QC fail

        I can't remember the last time I've seen anything hand soldered on a motherboard. Pick and place is the order of the day for through hole stuff too, there is a reason resistors and other axial components generally come on tape and reel.

        The only place I have recently seen hand soldering in a generic computer would be inside power supplies. Even there a lot is clearly wave soldered, frequently it is only the connection to the mains connector that is done by hand.

      2. This post has been deleted by its author

    3. Tom 7

      Re: QC fail

      I suspect overly tight QC being overridden? Obviously NOT overly tight QC.

  8. Version 1.0 Silver badge
    Happy

    It was fun when I saw that happen

    As a kid I went to the Electronics Club in Rugby and I was in the club one evening when a capacitor blow up, it was just like Christmas with little bits of silver floating all over the place ... luckily the EL34 valves were not damaged so the user got the audio amp up and running again quite quickly. The guys running the club were teaching us Morse code, they were excellent at it because they worked at typing Morse letter by letter (not word by word) to submarines.

    1. Paul Johnston

      Re: It was fun when I saw that happen

      Not sure how you type Morse word by word but anyone who can do Morse is better than I ever was, despite doing O-Level Seamanship and have tried to get a "Full Amateur Radio License". Passed the electronics exam but never the Morse for the full fat version :-(

      1. Tom 7

        Re: It was fun when I saw that happen

        A bit like emoticons - frequently used words get their own code or abbreviations. Speeds things up enormously. Best not used when trying to get a nuke underwater though!

        1. Version 1.0 Silver badge

          Re: It was fun when I saw that happen

          Transmitting to a submarine is quite effective when the radio transmission frequency is about 14kHz.

    2. David 132 Silver badge
      Boffin

      Re: It was fun when I saw that happen

      I recall two capacitor-related incidents from my time at school, oooh several decades ago in the late 80s.

      The first was when our young, hip, down-with-da-kids Physics teacher (Mr Fisher, if these increasingly senile memory cells don’t lie) was teaching us from the front of the room while idly fiddling with a rather large electrolytic that he had in his hands. Eventually the inevitable occurred and he touched the two fly-leads together, which led to a large spark, singe’d fingers, and a classroom of boys all learning the word “fuck!”, which of course none of us innocents had ever heard before. Nope.

      And the second was a year or so later, when we were actually learning about capacitors. One of my chums seated on the back row of benches behind me obviously thought it would be fun to connect every electrolytic he could get his hands on together in series, then charge the two ends of the chain from the lab 12v bench supply.

      Suddenly there was a loud “BANG” and we all turned round to see Bloggs Minor sitting there with an innocent expression as wisps of capacitor material drifted slowly down and settled on him like dandruff. The teacher for that class, who was infinitely more world-weary, just rolled his eyes and carried on with the lesson.

      Ah, happy days. A big “hi” to anyone reading this who thinks “gosh, I saw that, I think I was at that school…”

      1. Mike 137 Silver badge

        Re: It was fun when I saw that happen

        Yes, capacitors can be fun. In my younger days I built a bank of a dozen 30V 100A linear PSUs on a research project, each with one Farad of electrolytic capacitors for smoothing. Some of the post docs were being patronising about the safety precautions I designed in - one stating "that's just an electric kettle" - so I switched one of the PSUs off and shorted the capacitor bank with a pocket screwdriver, the last 3mm or so of which just evaporated. Point proven.

      2. drgeoff

        Re: It was fun when I saw that happen

        Difficult to do anything exciting with 12 Volts and a lot of electrolytics in _series_.

        1. David 132 Silver badge

          Re: It was fun when I saw that happen

          Dunno if you’re trying to call me a liar, but that’s what happened. I can only relate what I saw…

  9. Paul Johnston
    Flame

    Neat!

    “In our ongoing investigation, we have preliminarily identified a potential reversed memory capacitor issue in the production process from one of the production lines that may cause debug error code 53, no post, or motherboard components damage,” said ASUS in an announcement.

    I'm going to start using "debug error code 53" for when it all goes pear shaped.

    Obvious icon usage!

    1. Ken Moorhouse Silver badge

      Re: "debug error code 53"

      It is always DNS, even when it has nothing at all to do with the internet.

  10. Arthur the cat Silver badge
    WTF?

    But, but …

    Years of watching SF taught me that reversing the polarity was a good thing!

    1. David 132 Silver badge

      Re: But, but …

      It is. Do you see any mention in the Asus story coverage of an ongoing Klingon invasion? Nope? There you go then. Reverse your polarities with confidence. Make it so.

    2. Antony Shepherd

      Re: But, but …

      You only reverse the polarity of the neutron flow!

  11. Jwetvent

    Abit tale of fail

    Surprised no one mentioned, years ago, if memory serves,in the late 80s or early 90s Abit had a capacitor problem as well. Failing motherboards sans the smoke & fire. It was eventually traced to a single individual stealing the aluminum electrolytic capacitor recipe from a company he worked for and giving it to his new employer. Turns out somehow the recipe wasn't quite up to snuff & months in, high priced gaming boards by the droves were going belly up. I may be a little foggy on the details so feel free to correct me.

    1. Anonymous Coward
      Anonymous Coward

      Re: Abit tale of fail

      I had several Abit motherboard builds that died from their capacitor problem. Bulging can tops was the symptom for me. After that I used Asus instead.

    2. Anonymous Coward
      Anonymous Coward

      Re: Abit tale of fail

      a.k.a "Dell disease".

      Those caps ended up all over the place.

      1. Anonymous Coward
        Happy

        Re: Abit tale of fail

        The 'capacitor plague' of the late 90s/early noughties primarily affected Abit, IBM, Dell, Apple, HP, and Intel.

        It was the capacitor brands that were to blame, and those included such 'well-known' names as Tayeh, Choyo, or Chhsi. It was difficult to pin down (or do anything to if you did) any of the manufacturers responsible.

        Abit was the only company to admit that it was faulty Taiwanese capacitors that were to blame, but they didn't identify the brand they used. This was significant since they effectively admitted liability.

        I was on tech support at the time for a computer retailer in the UK and we took lots of calls about this, as well as contacting customers who hadn't had issues to replace their motherboards free of charge once we became aware of the issue. At the time, we saw it as 'a faulty batch of capacitors', but it became a bigger problem when it began to affect different vendors - many of whom denied any problem at the time. If the machine had Abit in it, job done. It was harder going with the others though - though I stress, at the time.

        A lot more is known or guessed now at what happened, and how. But when you're in the middle of it, things are different.

        1. Binraider Silver badge

          Re: Abit tale of fail

          I used to like Abit motherboards. I guess this is why the brand died! (Though I had umpteen other machines of that era die to capacitor fail).

          Not least of which a Crapple G5 Quad with unobtainium replacement PSUs.

          1. Anonymous Coward
            Anonymous Coward

            Re: Abit tale of fail

            Abit got into hot water in Taiwan when it was found they were operating through a handful of companies all registered at the same address, with a capital value of just a few dollars. Then it was suspected the board was embezzling funds.

            The following year the capacitor thing happened. From what I can gather, the lawsuits from that - and the fact they had long warranties and readily replaced goods compared to their competitors - sent them under.

        2. MarthaFarqhar

          Re: Abit tale of fail

          You didn't work for a company that sounds like Piglen, did you?

          I kind of remember the V69 towers with Pentium III contenders had this issue, we used to set ourselves Pit-Stop style challenges who could strip a board out and have the machine reassembled, and powered up in the quickest time. I can't remember the exact time, but when they sent a few with a tech rather than just the boards themselves, he was rather impressed how quick the turnaround was from us.

          1. Anonymous Coward
            Happy

            Re: Abit tale of fail

            No, it was Dixons/PC World.

            When I started, and it was operated by Dixons, there was an entrance 'exam' and you pretty much had to be able to build a PC to get through it (as well as know how the components worked).

            When I left, after (and while) it was still outsourced to Capita, a lot of the people they took on didn't know what or where the 'ON' button was.

        3. Rob Daglish

          Re: Abit tale of fail

          Our local water company had a fleet of Lenovo M52/55/57 ish machines which were all very long in the tooth, and the caps on those would frequently swell, causing various issues. It got to the point where you could look at the board and know if it was giving disk errors or graphic errors by which clump of caps had swelled. Spent about three years doing motherboard swaps on those - old board out and off to central workshops for new caps, refurbed board from central workshops back in, job sorted. Eventually they were replaced with a fleet of HP Elitedesk machines, and we went from actually doing some work to recording serial numbers in and out as they moved to a whole machine swap out with repairs done at their head office.

  12. FlavioStanchina

    "may cause debug error code 53"

    Was expecting debug error code 451.

    1. X5-332960073452
      Mushroom

      Re: "may cause debug error code 53"

      Only if you are from the 'left side', the 'right side' would be 233

      Burning pile of >>>>>>>

  13. Anonymous Coward
    Anonymous Coward

    Part of my learning curve was that small tantalum capacitors were polarised - with what appeared to be a non-obvious marking. Bought a Motorola 6800 evaluation kit - which had a bare circuit board and ICs only. Soldered the many tantalum capacitors on to it in an aesthetic fashion. Luckily the first one was the right way round - thus so were the rest with that alignment.

    The final two had a 90 degree orientation to the rest. Murphy's eye was off the ball - and those two were put in the correct way too.

  14. Jo Bloggs
    Stop

    Normal for Asus

    I have seen so many manufacturing defects from Asus in the past that I already don't ever buy anything made by them. This new feather in their cap comes as no surprise to me. It's a shame as technically some of their stuff is very good, it just too often doesn't work at all or fails too soon.

    1. First Light

      Re: Normal for Asus

      I'm sorry to hear that as I just got an Asus Tuf Gaming laptop (not affected by this recall). I got it for the MIL STD 810H rating as I will need it where I am going (Tropics) and the glue behind the screens of the last two laptops I had melted from the heat and humidity, destroying them. Oh well, fingers crossed!

  15. normal1

    It's fairly easy to tell if your board is effected?

    Just look for the glow and flickering?

    1. Tom 7

      Re: It's fairly easy to tell if your board is effected?

      Gaming motherboards - no need for a case!

  16. Anonymous Coward
    Anonymous Coward

    Reverse polarity is rapid death but electrolytic caps always need care

    I sometimes use integrated stepper motor drivers that come with the control PCB attached to the back of the motor in a metal case and they started failing. I opened some up and looked inside to find the smoking remains of a 50V capacitor on the 48V circuit which due to the sealed box always runs hot. I am amazed that any of them survived normal use. Pushing the machine along the axis by hand I could easily generate a back EMF of >60V.

    I complained to the manufacturer but they said that the 50V capacitor was fine to use at 48V and don't worry about the heat.

    Moral of the story is pick electrolytics with 50% over voltage rating and go for 5000h at 105C as the absolute minimum if they will run warm.

  17. ecofeco Silver badge
    Coat

    So?

    They are incapacitated?

  18. Conundrum1885

    Capacitor

    I also had the infamous glowing tantalum fail, this time on an HP dv9000 aka "Put a T7300 in a laptop with completely inadequate cooling and it will probably work".

    This one emitted smoke as well, luckily got to it in time and took it out.

    Laptop ran fine after that!

  19. Anna Logg

    Memory capacitor gone TITSUP?

    Is that a mammary capacitor then?

  20. David Gosnell

    School memory

    I remember in an A-level physics lesson the teacher preparing a demonstration of blowing up a capacitor, outside the lab window. Suddenly and unexpectedly there was a flash, bang and large shower of sparks from across the lab. One of the students had jammed a large electrolytic straight into a 240V AC socket. He said he thought there was a 50:50 chance it would be the right way round...

  21. Unicornpiss
    Flame

    Had an "Elite Group" mobo do this to me

    It started making a sizzling sound and let out the magic smoke. Fortunately I was home and near the PC, though it probably would have been safe enough in a metal case.

    In this circumstance nothing was installed backwards, it was just the poor quality components that the manufacturer specd for the board.

  22. TeeCee Gold badge
    Headmaster

    "Netectives"

    FREEZE! NEOLOGISM POLICE!

    UP AGAINST THE WALL, HANDS ON YOUR HEAD AND DO NOT FUCKING MOVE YOU BASTARD!

  23. pdxalc

    Cap Backwards

    Just measured the voltage on the suspect capacitor. It is 5.0 V on the non-polarity side and the cap polarity band is toward the CPU like others that have led to failure. Ugh. So far the MB is ok and the cap and mosfets are not hot.

    Got a notice from ASUS notice through Newegg. Chatted with an ASUS rep who wants me to send the MB in for repair. I told him that was not acceptable as I cannot afford to wait through the long repair process. I need this computer for client engineering studies. He is bumping up to higher level and I am supposed to hear back next week.

    The rep sent me an RMA form which has a link to purchase a shipping label from ASUS. It is an affront to expect the buyer to pay to return this expensive MB to rectify a manufacturing defect caused by failure of ASUS quality control.

POST COMMENT House rules

Not a member of The Register? Create a new account here.

  • Enter your comment

  • Add an icon

Anonymous cowards cannot choose their icon

Other stories you might like