The Hall of Shame - Badcaps Photo Montage

[xnRS]

I hope you are accepting soldered out capacitors

Sometime ago I got some non-working ATX PSUs for free... And as I leave in not rich country, these were cheap faulty PSU with many bulged badcaps. I already resoldered most of capacitors and got little 'box of shame' with died crapacitors. Some of them are even from Motherboards, so here we go:

First comes my most favourite one - FHY MKX 470uf 200V, it's so cool, that it even made new vent for himself as main one isn't working as it should. Saw many bulged caps from this brand(not by my self, but they are badcaps 100%).Soldered this one out from Modecom FEEL 500W.



Second one is with interesting name - Forever, sadly this one didn't lasted forever. Soldered it out from some crapy old ATX PSU.



Next one is Sapcon, soldered 5 of those capacitors from a Raidmax KY-600ATX.



Another 'con' badcaps - Canicon, Soldered this one out from no-name crapy old ATX PSU.



Next is SD, I got 4 of them soldered out from some old crapy ATX PSU.



And another 'con' crapacitor - Ricon. Soldered out from some old crapy ATX PSU.



The capacitor I see on many old Microlabs is BH. I didn't saw them bulged except for this one, I didn't measured their capacitance as I don't have capacitance on ESR tester, maybe will test them if get one on my hands.



p.s on photo of capacitor you can't see leters BH as they are covered into some thing that maufacturer puts on capacitors. Here is the other BH capacitors soldered on some died microlab.



Asia X - popular one, some manufacturers put them on new PSU(Like Aerocool on their VX series), have 4 them bulged and soldered out.



CapXon - same as Asia X, but I think they became better now, have some old 2 bulged caps.



Next is 3 interesting ones

Nichicon copy - Nicon, soldered out from some old crapy ATX PSU, they put caps with those name on factory, what a shame...



Fake Rubycon MBZ ! 1000uf 6.3V didn't found any info about those in datasheets, found them soldered onto MSI P965 NEO(or NEO-F), not from factory, as old NCC KZG were bulged(yes-yes KZG caps at their best, also there will be some series of NCC in the list).



And the final one is - NCC TMV 680uf 4V, they have no vent on top, so they bulge on bottom, got 3 of them from some old ASUS motherboard.



p.s sorry if image size is too big, just wanted to show my little 'box of shame'
Also i didn't mentioned 3 other caps - OST from old Powermans or In Wins, Teapo, and of course, NCC KZG, i got many of them bulged...

[RukyCon]

I can't see your photos, which leads me to guess they timed out while you were writing the post (since i can see they're all Badcaps URLs, so you didn't use 3rd party image hosts). I've had the exact same thing happen, and usually, you can still see the photos, but no one else can.

[xnRS]

Originally Posted by RukyCon

I can't see your photos, which leads me to guess they timed out while you were writing the post (since i can see they're all Badcaps URLs, so you didn't use 3rd party image hosts). I've had the exact same thing happen, and usually, you can still see the photos, but no one else can.

Crap... I will try to fix it

[RukyCon]

Yeah, it's a pain when this happens, i think it might be a bug. Now that first cap must have shorted out and burned a hole through the can, I wonder what it looked like to see it fail, probably pretty exciting if i were to guess.

Based off the name, the sapcon capacitor must be particularly good at sapping the life out of whatever it's installed in, granted, every other capacitor you've shown would also be good at doing just that.

[xnRS]

Originally Posted by RukyCon

Yeah, it's a pain when this happens, i think it might be a bug. Now that first cap must have shorted out and burned a hole through the can, I wonder what it looked like to see it fail, probably pretty exciting if i were to guess.

Based off the name, the sapcon capacitor must be particularly good at sapping the life out of whatever it's installed in, granted, every other capacitor you've shown would also be good at doing just that.

Well, first cap for some reason didn't exploded a vent, and had rust or some electrolyte residue under the sleeve... I think vent didn't made it job well because of factory defect... Sapcon is did well in sapping life of poor Raidmax KY(S)-600(yea 600, 18 amps on 12v, 600, 100% sure)ATX... But forever is best name out of all capacitors here, it becomes even better after this capacitor didn't lasted 'forever'...

[RukyCon]

Illinois Capacitor 47uf 25v.
https://www.badcaps.net/forum/attach...1&d=1622627220
The one above tested at 213pf, but I also pulled two more that had failed, the second one tested at 173pf, and the last one was open. These were likely made around 1983, and were likely exposed to a lot of heat during their lives.

Originally Posted by xnRS

Well, first cap for some reason didn't exploded a vent, and had rust or some electrolyte residue under the sleeve... I think vent didn't made it job well because of factory defect...

I've seen the vent on a bunch of cheap capacitors fail, likely due to it not being stamped deep enough into the tops of the capacitors, leaving very shallow vent marking that don't break when too much pressure builds up in the capacitor.

Originally Posted by xnRS

...But forever is best name out of all capacitors here, it becomes even better after this capacitor didn't lasted 'forever'...

I think I've seen the forever brand here before, and it goes to show that no matter what you name your crappy caps, they will always be crap.

[xnRS]

Originally Posted by RukyCon

I've seen the vent on a bunch of cheap capacitors fail, likely due to it not being stamped deep enough into the tops of the capacitors, leaving very shallow vent marking that don't break when too much pressure builds up in the capacitor.

That's what I meant tbh

[momaka]

Originally Posted by xnRS

First comes my most favourite one - FHY MKX 470uf 200V, it's so cool, that it even made new vent for himself as main one isn't working as it should. Saw many bulged caps from this brand(not by my self, but they are badcaps 100%).Soldered this one out from Modecom FEEL 500W.

https://www.badcaps.net/forum/attach...1&d=1622553381

Wow, a failed *primary* cap. FHY must be absolutely terrible then, because even other crap cap brands usually rarely have issues with the high-voltage (200V) caps failing on the input of a PSU. Perhaps it's also possible the PSU saw a power surge? (Probably not though, and these GHY are just that bad. )

Originally Posted by xnRS

Second one is with interesting name - Forever, sadly this one didn't lasted forever. Soldered it out from some crapy old ATX PSU.

https://www.badcaps.net/forum/attach...1&d=1622553381

I think what they meant with the "Forever" brand is Forever Crap.
I've encountered this brand only with small caps in various cheap appliances (a toaster or mini-oven, IIRC.) Never checked if they were bad, as I didn't have a way to back then.

Originally Posted by xnRS

Next one is Sapcon, soldered 5 of those capacitors from a Raidmax KY-600ATX.

https://www.badcaps.net/forum/attach...1&d=1622553446

Another 'con' badcaps - Canicon, Soldered this one out from no-name crapy old ATX PSU.

https://www.badcaps.net/forum/attach...1&d=1622553381

I suppose if it's anything ending in -con and not a Japanese brand... then it probably is a con.
The Can-I-Con is a known old "classic" bad cap (and probably what started the -con jokes.)

Originally Posted by xnRS

CapXon - same as Asia X, but I think they became better now, have some old 2 bulged caps.

CapXon are downright terrible too... though on very rare occasions (in low-stress GP applications), they sometimes (very rarely) have lasted 10+ years. I do consider them a notch better than the "no-name" brands, like most of the caps with a -con in the name listed above... but not that much better.
I still see CapXon failed too often to consider them trustworthy.

Originally Posted by xnRS

Fake Rubycon MBZ ! 1000uf 6.3V didn't found any info about those in datasheets, found them soldered onto MSI P965 NEO(or NEO-F), not from factory, as old NCC KZG were bulged(yes-yes KZG caps at their best, also there will be some series of NCC in the list).

https://www.badcaps.net/forum/attach...1&d=1622553381

That actually looks like a genuine MBZ. Just curious, what made you conclude it is fake/counterfeit? The vent and the text seem reasonable with other Rubycons, at least from the picture provided. Perhaps the bung?

Originally Posted by xnRS

And the final one is - NCC TMV 680uf 4V, they have no vent on top, so they bulge on bottom, got 3 of them from some old ASUS motherboard.

https://www.badcaps.net/forum/attach...1&d=1622553381

Yeah, I think these are similarly unreliable as KZG and KZJ, especially on CPU VRM. I got a few bad ones, also from the CPU VRM of an ASUS (P5GC-MX) motherboard. Like yours, they have no vent so they bulged from the bottom. I'll see if I can upload a picture later.

[momaka]

OK, here is the (single) failed UCC TMV cap I have:





Note: the cap is not original to that EVGA GeForce 7600 GT video card. I removed it, along with the other TMV caps seen in the background, from this ASUS P5GC-MX motherboard while doing a preemptive recap. They tested good at the time of the removal, so I set them aside for re-use on non-important hardware. In the case of the above 7600 GT video card – its RAM was damaged from the original Sacon FZ caps that failed. I replaced the RAM and caps, which managed to get the card to display video again. But it still showed some artifacts, so I pulled my good caps for use on other hardware and put these 4v 680 uF UCC TMV capacitors in there as a filler. They did test good prior to being installed on the 7600 GT. But over the years with the video card mostly sitting on the shelf (played with it maybe once or twice in the last 5 years), I guess these TMV caps are starting to fail now. Note that the TMV caps in the background of the above image still look OK. I haven’t pulled them out to check them, but I imagine they might be due for the same failure in the future. In any case, I’m not too worried, as that card is more of a “toy” / waste of time at this point. If I ever repair it, it will certainly be top nominee for the Worthless/Pointless Repairs thread.

But yeah… United Chemicon TMV series apparently aren’t the most stable of caps either. And like xnRS noted, they don’t have vents, so they will bulge from the bottom instead (and possibly leak!) Good thing I removed them from that P5GC-MX.

[Alex@]

Hi,

My old Audio Dynamics amplifier (ca2000e) started to take time to power on, sometimes 10 seconds sometimes 10 minutes. The filter caps appear to be the problem, 2 of them are starting to bulge on the top, not to easy to see in the photos. They all check out ok on a peek esr meter so I guess that these type of meters are of limited use in testing capacitors.

Anyway they will be replaced.

[momaka]

Well… looks like I might have to rethink/re-check (and possibly re-do) some of my personal equipment recaps that [re]used 6.3V 2700 uF Rubycon MFZ series from Xbox 360 motherboards. Here’s why:







The first picture above is from a Biostar NF4U motherboard that I recapped around late 2016 / early 2017 (or possibly early-2019.) That motherboard never saw much use after the recap until I tried to build a PC with it recently. It seems that within a few months after using the board, one of the replacement MFZ caps failed. Date code on the bulged cap is T0541. The other ones around it with the black marker dot on the vent have a date code of T0532 and don’t appear to be bad so far. The one in the center of the group also has T0541 date code and has slightly higher capacitance (around or under 3000 uF), but still within 20% spec. Not sure if that’s indicative of anything, but I pulled it, just in case.

The second picture is of a scrap Xbox 360 motherboard I kept in my parts bin (inside the house, so not exposed to temperature or moisture extremes.) The single failed MFZ on there has a date code of T0602, just like the others next to it. Strangely enough, that was the only cap that was failed. The others next to it measured normal capacitance and ESR, if I remember correctly.

But those aren’t the only failed ones. I also have a failed MFZ with date code of T0544 that I installed as a replacement for a failed UCC KZG 6.3V 1500 uF on a Leadtek GeForce 8600 GT video card. I only put occasional use on the card while I was testing new coolers on it, so it didn’t see too much use or heat, but also didn’t sit on the shelf for too long (2 years max.) Another two failed MFZ came from an ECS P4ibms motherboard that I recapped preemptively a long time ago. Like the other hardware above, it was used very infrequently for CPU cooler tests, due to there being some other issues with the motherboard. On this one, the two failed MFZ have a date code of T0609. And finally, there is another failed T0541 MFZ in my spare caps bin. I don’t think I actually [re]used this one at all and it just bulged while in storage all these years (though I did reform many of the caps a few years apart.)

So all in all, that’s 6 total failed Rubycon MFZ 6.3V, 2700 uF caps pulled from various Xbox 360 motherboards. Again, date codes for these are T0541, T0544, T0602, and T0609.

In addition to that, I also have a few MFZ with other date codes (namely T0541, T0603, T0609, and T0632) that are starting to show slightly high capacitance (around 3000 uF or less), but still well within 20%... and one T0544 cap that measures close to out-of-spec on capacitance (above 3000 uF), but still not over 20% yet. None of these are bulged and appear to reform OK. However, after sitting disconnected from their voltage source while still charged, many of them fall down to about 2.5-2.7V rather quickly. So I think they definitely might be suffering from a degraded oxide layer.
*update 09/2021*
The T0544 with close to out-of-spec capacitance has failed (bulged and leaked) sitting on the shelf after I came back from visiting family this summer. Likewise, one of the caps with the T0603 date code mentioned above has also failed the same way.
/update

I guess it’s best to keep an eye on these. They seem to not like sitting in storage without use. More to the point, I regularly see some of them measure up to -0.3V (yes, negative 0.3V) after sitting in storage for a very long time without use, regardless if I left them with a full charge between their plates or not. On that end, I also notice there are a few sellers on eBay (and probably other places online too) that have genuine-looking Rubycon MFZ. Most likely, they are leftovers from old Xbox 360 motherboards or who knows what other production line and thus probably quite old at this point. From what I know, MFZ series was never directly available for sale to “the public” from Rubycon and is a custom series closely related to the MCZ series. So like the MCZ series, they probably aren’t very tolerant of heat… and moreover appear to be slightly unstable when left unused on the shelf for a long time, especially with advanced age. In my case, I reformed most of my stock a few years apart (and before each re-use as well), so I think the failures just come down to their electrolyte being slightly unstable, as was the case with quite a few other ultra-low ESR caps of the era. After all, Nichicon had their own problems with the HM, HN, and HZ series between 2001 and 2005. And United Chemicon KZG seems to have had problems for even longer, especially for a few particular sizes, like the 6.3V 820 uF and 3300 uF, among others. Sanyo WF didn’t seem to show any problems at first, but I find many are now starting to bulge and go high capacitance (scroll back a few pages.) Perhaps the only one that appears to have skirted around these issues with lowest ultra-low ESR caps is Panasonic with their FL series… but they weren’t used as much (IME), so hard to tell if the failure rate is the same or not.

And it’s probably worth mentioning again, that my Rubycon MFZ stock all came from scrap Xbox 360 motherboards, some of which have been subjected to BGA rework and/or multiple reflows by the repair shop I got them from – i.e. exposure to high heat. In fact, I suspect the two failed T0609’s might be a result of that. But in any case, I figured I have enough of these failed now that it may be worth noting it here in this thread. Overall, though, they’ve done *OK*. I think back when I began scrapping Xbox 360 motherboards, I started with about 50 to 60 of these Rubycon MFZ caps in my stock. Some of them are still installed on various PC hardware and they still appear to be doing their job. I currently have only about 20 left in stock. So 8 failed caps out of 30+ used isn’t super-terrible, all things above considered and their age. But again, I suggest for people to keep an eye on these and/or replace preemptively in any critical equipment. Seems like they are nearing EOL now that they are close to 15 years old or more.

[PeteS in CA]

Ruh-roh Raggy .....

[momaka]

While still on the discussion of ultra-low ESR series of caps, here is one more to beware of: Nichicon HZ series with date codes that start with “H05” – i.e. HZ made in 2005. In particular, the only problem I’ve seen was with the 6.3V, 2200 uF caps in 10x20 mm size. I don’t have any other Nichicon HZ caps with H05 date codes and different ratings, so I can’t say how widespread the issue is. But in the case of my 6.3V, 2200 uF H05xx stock, at least half of the caps have failed or are on their way to fail at this point.

Although it may seem normal for these caps to be bad now, given that they are over 15 years old, I started seeing the first failures of H05xx HZ back around 2013-2014. So I think the issue of unstable electrolyte is certainly there. Of course, I will note that these too, like the Rubycon MFZ 6.3V 2700 uF caps mentioned above, came from the same donor Xbox 360 motherboards I scrapped (in fact, the 6.3V 2200 uF HZ and 6.3V 2700 uF MFZ were often interchangeably used, probably to due having very similar ESR and ripple current spec.) With that said, I don’t know if this has played any part in the issue here or not. I suspect not, because the same HZ caps that have a date codes starting with “H06” have been absolutely solid for me – not a single failure so far. In terms of count, I think I used/pulled about an equal amount of H06 and H05 HZ caps – about 20-30 from each. And I’ve used each more or less in equal amounts, too – sometimes often mixing H05 with H06 caps on the same piece of equipment and in the same circuit (VRMs of a few Biostar motherboards, among other hardware.) The result was that many of the H05 HZ failed, but the H06 didn’t.

In terms of failure mode, the H05 HZ caps seem to behave very similarly to the failing Rubycon MFZ mentioned above. First, the capacitance starts to increase approximately +20% to +30%. At this point, the caps are still functional and still appear to have good ESR (strong spark when short-circuiting a charged cap – a method I use to crudely test if ESR is actually low and not just my ESR meter getting fooled.) But then anywhere from 3 months to a year (or sometimes more), the cap will eventually go higher in capacitance (I’ve seen up to 100% increase), then bulge and leak, regardless of how much it was used. In contrast, the Rubycon MFZ caps seem to go a lot quicker once their capacitance has gone high.

Regardless of these small differences, the results are the same in the end: 6.3V 2200 uF HZ with H05xx date codes eventually have failed for me. I now have only a handful left in my stock that I re-use for testing purposes and on non-special equipment only. In fact, the only H05 HZ caps that haven’t gone bad (yet) in my stock or on equipment appear to have specific date codes of H0550 and H0551. I counted about 13 of these total (5 in stock and 8 installed on various GPUs and motherboards.) Meanwhile, the following specific date codes (and how many) are of my HZ caps that have gone bad:
H0532 (3x)
H0533 (4x)
H0544 (5x)

I may have a few more bad and “good” ones (possibly still in equipment that I forgot to count.) But overall, the total so far is about at least 12 failed and 13 still reading/looking OK.

Here are a few examples of the failed ones:






^ The first card above is an XFX 6800 XT video card that I recapped back in 2014. The card didn’t see much use (other than occasional testing) up until about 2016. Around early 2016, I put it into a PC and used it for maybe 6 months very regularly. After that, I pulled out the video card and stored it. It’s seen very infrequent use since then. Around mid 2019, I found the single 6.3V 2200 uF Nichicon HZ on the card to be bulged and leaking, just sitting in storage (actually, on my wall as decoration.)

The story of the XFX 8600 GT (2nd picture) is more or less similar: recapped in early 2016, used infrequently, installed in PC around early 2018 and found cap bulging on the card just a few months afterwards.

3rd picture is of an eVGA GeForce 6200 (6600LE) video card, also with similar use/non-use pattern... and again, same failure.

Lastly, I did a mixed polymer-electrolytic recap on a Biostar NFS327-A7 motherboard in 2018 (no picture here.) For 3 of the CPU VRM output caps, I used 3x HZ 6.3V 2200 uF – 2x with H05 date code and 1x with H06 date code. In 2020, I pulled the board out to put in a temporary built to test it out more thoroughly. This is when I found the 2x H05 HZs bulging. Needless to say, I changed these with H06 HZs, and it has been OK.

In conclusion, it looks like Nichicon HZ caps – at least those rated for 6.3V, 2200 uF, and date code starting with “H05” – should NOT be trusted, especially the earlier batches. I guess we will see if those H0550 and H0551 HZ’s are any good as time goes by, since I plan to leave them in the equipment they’re in for the time being.

But while on the topic of Nichicon and date codes, I should mention that I also have a ton of Nichicon HN series with H05 date codes - specifically caps in 16V, 1500 uF, and same 10x20 mm can size from the same scrapped Xbox 360 motherboards. With these 16V, 1500 uF HN caps, I have not seen a single failure, and I have used them on all kinds of hot and abusive circuits. I also have 16V, 1500 uF HM series that came stock with a few MSI motherboards, and those are also OK. So indeed whatever goof-up Nichicon may have had with their ultra-low ESR series (HM, HN, and HZ) seems to have been fixed right around 2005, because many are still OK, save for the HZ 6.3V 2200 uF mentioned above.

Anyone else have any different (or similar) experience?

[MahdiA]

Hi
Luckly, when the appearance show the bad cap
And
Misery, when the appearance is good but disen't work.

[Behemot]

Originally Posted by momaka

Well… looks like I might have to rethink/re-check (and possibly re-do) some of my personal equipment recaps that [re]used 6.3V 2700 uF Rubycon MFZ series from Xbox 360 motherboards. Here’s why:

Got a report from some guy that some of those MFZs I've sent him bulged in storage. Not sure what he did with it, I've checked my two boxes of them caps, everything looks just fine.

[goodpsusearch]

I have seen Panasonic ultra low esr caps bulge in storage. I think it was Panasonic FJ, not FL that was mentioned above. I think that it is wise to avoid using ultra low esr caps when possible, especially now that most of those series of caps have been out of production, so the stock available is getting old.

btw I missed you guys

[xnRS]

So, conclussion is:
1.Don't use old Ultra-Low ESR Caps
2.Don't reuse old Ultra-Low ESR(And Normal) Caps
3.Use Polymers instead

[dmill89]

In a case of "when good caps go bad", my Brother HL5250DN was having issues "coming back" after a loss of power and I found this Chemi-Con KMG 47uF 35v cap on the standby rail of the PSU with a .14Kohm ESR , I guess after 16 years of 24/7 use (this printer has been plugged in continuously since I bought it back in 2005) it finally "bit the dust". After replacing the cap (I fully recapped the secondary side of the PSU though this was the only cap that tested wildly out of spec), the printer is working fine again.

[momaka]

^16 years is pretty good for a GP 105C cap. Had the manufacturer gone with an actual series meant for PSUs (like UCC KY, among others), more than likely that printer would have still kept going. Good to hear that you got it working, though. I wonder how many other Brother printers would get discarded because of this. I suppose I better keep my eyes open then. Not that I need a printer (already have a Brother HL1440 that's still working OK-ish), but always good to keep things out of landfills, especially if I can get them for free or really cheap and the fix is just a few small caps.

Originally Posted by xnRS

So, conclussion is:
1.Don't use old Ultra-Low ESR Caps
2.Don't reuse old Ultra-Low ESR(And Normal) Caps
3.Use Polymers instead

Yeah, more or less... and if you can.
Some older motherboards do really need the higher capacity caps. But that's not to say ultra-low ESR caps are still needed. The next best low ESR grade should do fine too - stuff like Panasonic FR/FM/FS, Rubycon ZLH/ZLQ, UCC KZH/KZN, and Nichicon HW. Moreover, these can be mixed with polyerms when there are multiple spots on the motherboard connected in parallel - this will give close-to-original capacitance *and* good low ESR from the polymers.

[Uranium-235]

Originally Posted by momaka

^ The first card above is an XFX 6800 XT video card that I recapped back in 2014. The card didn’t see much use (other than occasional testing) up until about 2016. Around early 2016, I put it into a PC and used it for maybe 6 months very regularly. After that, I pulled out the video card and stored it. It’s seen very infrequent use since then. Around mid 2019, I found the single 6.3V 2200 uF Nichicon HZ on the card to be bulged and leaking, just sitting in storage (actually, on my wall as decoration.)

Those XFX 7000/8000 cards almost ALWAYS had bad caps. I had next door neighbor over a decade ago with three of them