Fake Rubycons?

[momaka]

Re: Fake Rubycons?

Thanks for the advice I will keep that in mind for the future, because the size of these capacitors is rare.

The ones TC suggested will indeed work.

I have a number of 2.5V 820 uF Sanyo SEPC, Nichicon LF, Lelon OCRZ, and other similar polymers from Xbox 360 console motherboard and dead GPUs, so I use those instead. Same height as the 4V, 560 uF polymers, but higher capacity.

The MBZ are also very problematic with the heat, but if they are well refrigerated they will not present problems, I even have a 2001 MSi board with Rubycon MBZ in the VRM and they are perfectly fine.

Yes, MBZ and MCZ are fine, as long as they are cooled well *and* used. I find that a lot of my old-stock ultra-low ESR caps tend to discharge and even reverse-charge in storage rather quickly (less than two years) if not used. So if you have old equipment with those caps, it's a good idea to power it up once in a while and use it.

I have never seen the Rubycon MFZ series, the ones that I have seen are MBZ, surely they are custom MBZ for OEM.

Like I mentioned, MFZ is a custom-order series closely related to MCZ and *not* MBZ.
MCZ and MFZ are the lowest-ESR aqueous electrolytic caps you can get (or could get... they are no longer being made) from Rubycon. Only Nichicon HZ is slightly better and pretty much the "king" of ultra-low ESR caps. Like MCZ and MFZ, they also don't do well with heat, but generally seem to be a little more stable, so long as they have a manufacturing date code after year 2005 (so anything from 2006 and onwards is good... i.e. H06xx and newer.)

[kevin!]

Re: Fake Rubycons?

I've seen a few of those in the past. I replaced them with 560uF 4v polymers.... Low profile enough to clear heatsinks (same size as the originals). Never had one come back on me.

Thanks for the advice I will keep that in mind for the future, because the size of these capacitors is rare.

This motherboard is full of badcaps, all the capacitors have to be replaced. It has all the samyoung NXC capacitors in bad condition, as well as various teapo, and lelon .

MFZ is supposedly a very closely-related series to MCZ in terms of specs, if not equivalent.
They are really good caps, but don't do too well with heat, like most other ultra-low ESR aqueous electrolytics. The ones that failed there under the CPU heatsink don't surprise me at all. Despite being right in the flow of air from the heatsink, a lot of times they heat up a lot from the VRM. And the small case means they probably aren't rated for many hours endurance. So I'd say that's also a board designer mistake there (or was it? perhaps the manufacturer intended for them to fail after all.)

Either way, on socket 775 board like that with small caps under the heatsink, I always use polymers.

If I think the same, these MFC if they did not get to be in stress conditions, they would have had a longer useful life.
That if the board has worked 44,000h (according to the SMART of the HDD).
The board was recently cleaned, I suspect that this computer has been working with the heatsink covered for a long time, the computer came from a carpentry, and since there is a lot of dust in suspension, it must have been working with hardly any refrigeration and the capacitors have been damaged by heat.
I do not know the history of the computer, but the motherboard is clean and this indicates that the client had problems with the computer, and it was probably cleaned by a technician, but the computer was still not working properly due to badcaps.
The MBZ are also very problematic with the heat, but if they are well refrigerated they will not present problems, I even have a 2001 MSi board with Rubycon MBZ in the VRM and they are perfectly fine.

i thought the MF stood for Mother-Fucker given how small they are!

I'm starting to think the "F" in the series letters, stands for Foxconn.

Like the MFZs in post-version-1x XBoxes, IIRC.

I have never seen the Rubycon MFZ series, the ones that I have seen are MBZ, surely they are custom MBZ for OEM.

[Topcat]

Re: Fake Rubycons?

I think they are genuine. I recently replaced a faulty Foxconn motherboard with VRM capacitors leaking, and they are also rubycon, but not MBZ, they are MFZ series.

I've seen a few of those in the past. I replaced them with 560uF 4v polymers.... Low profile enough to clear heatsinks (same size as the originals). Never had one come back on me.

[eccerr0r]

Re: Fake Rubycons?

I'm still using the Foxconn board as a PVR with an E6700 and it's down to 3GB RAM. I wonder if getting a Qxxxx is worth it if I can't get 4GB to run in it anymore (Linux 64-bit rustc/gcc/llvm is probably the most cpu hungry app I run and they eat lots of RAM...may have to downgrade to 32-bit).

The Gigabyte board currently has a S775 preshott in it...

[momaka]

Re: Fake Rubycons?

I think they are genuine. I recently replaced a faulty Foxconn motherboard with VRM capacitors leaking, and they are also rubycon, but not MBZ, they are MFZ series.

MFZ is supposedly a very closely-related series to MCZ in terms of specs, if not equivalent.
They are really good caps, but don't do too well with heat, like most other ultra-low ESR aqueous electrolytics. The ones that failed there under the CPU heatsink don't surprise me at all. Despite being right in the flow of air from the heatsink, a lot of times they heat up a lot from the VRM. And the small case means they probably aren't rated for many hours endurance. So I'd say that's also a board designer mistake there (or was it? perhaps the manufacturer intended for them to fail after all.)

Either way, on socket 775 board like that with small caps under the heatsink, I always use polymers.

The Foxconn board has been the most troublesome board, but at least it seems always to be caps, caps, caps...

Well, with so many bad caps, I'm surprised it even started or tried to work. I usually don't even power on motherboards when I see them bad like that. Doesn't matter if you scope the outputs for ripple - bad caps are bad caps. Get rid of them and then it should work. (Unless of course the board uses an nVidia chipset - then all bets are off for any reliability.)

Should put the solid caps on the Foxconn, or at least get a whole set of solid caps for it... Just that it needs too many for a mostly worthless board (socket 775, microatx, DDR2, chipset graphics)

Eh, I wouldn't say s775 is worthless. With a low-end C2Q like the Q8200, Q8300, or Q9400, which can be had for pretty cheap, it will still make a great browser PC. 4 GB of RAM minimum if you intend to run a lot of tabs. Otherwise if less than 4 GB, Win7 32-bit will still get you good mileage even today. And with 8 GB of RAM and a decent older mid-range GPU like GTX 460 or 560 or Radeon HD6850, it will play most AAA titles from up to 2015 decently (like GTA V and Dirt Rally, for example.)

[eccerr0r]

Re: Fake Rubycons?

The Foxconn board has been the most troublesome board, but at least it seems always to be caps, caps, caps... unlike the other stricken board I have -- a Gigabyte -- and I still have no clue what's wrong with it (it has all solid electrolyte caps, at least it says so...)

Should put the solid caps on the Foxconn, or at least get a whole set of solid caps for it... Just that it needs too many for a mostly worthless board (socket 775, microatx, DDR2, chipset graphics)

[RJARRRPCGP]

Re: Fake Rubycons?

i thought the MF stood for Mother-Fucker given how small they are!

lol!

[stj]

Re: Fake Rubycons?

i thought the MF stood for Mother-Fucker given how small they are!

[RJARRRPCGP]

Re: Fake Rubycons?

I'm starting to think the "F" in the series letters, stands for Foxconn.

Like the MFZs in post-version-1x XBoxes, IIRC.

[kevin!]

Re: Fake Rubycons?

I think they are genuine. I recently replaced a faulty Foxconn motherboard with VRM capacitors leaking, and they are also rubycon, but not MBZ, they are MFZ series.

[eccerr0r]

Re: Fake Rubycons?

Actually it's a intel chipset graphics board, no dedicated GPU.

The board is a mishmash of different brands, these caps I don't know what it's connected to but they were close to the PLL. The chipset filter caps are a different set of caps that are also in bad shape but the waveform was not as bad as these. The CPU VRM caps have also burst, but probing with a scope they still look functional unlike these.

There are still a bunch that haven't burst, the only ones that have burst are the CPU (make a wild guess at brand... yep Nippon Chemi con KZG IIRC) and northbridge/chipset graphics chip VRM capacitors that burst (These...I believe were OST), and then these Rubycons. There are also some more caps on the CPU VRM but I don't recall what brand they were at the moment...

The OSTs were just simply dead just like these Rubycons. But the KZGs appear to still be doing their job despite popping. Probably not for long, but I wonder what I should do with this motherboard overall, I botched removing an OST and have jumper wires...

As for making the machine working again I decided to remove half of the memory - one pair of the 1G DIMMs - and the machine started working a lot better than before. Then I stuck in a pair of 512M DIMMs back, the machine appears to still be working and stable once more, unlike with just the Rubycon capacitor swap. Maybe it was just bad RAM even though the waveform at the capacitor is better than before?

Anyway the OSTs I had swapped earlier and those were very needed.

I really need to stock more caps but this is my own computer, not someone else's computer I need to fix...

[Topcat]

Re: Fake Rubycons?

Look genuine to me. They just wore out.

[Wester547]

Re: Fake Rubycons?

If the whole board is full of swollen capacitors it could have come out of a poorly ventilated and/or cooled case, or maybe certain FETs or coils on the board, neighboring said capacitors, dissipated too much heat over a prolonged period of time.

“The electrolyte ate through the foil layer”... well, I guess certain series whose foils are smaller (though also carefully etched to increased the surface area) and oxide layers are thinner so as to obtain larger capacitance and voltage values in smaller case sizes (miniaturization) may be more susceptible to this problem. But good series and brands should use ultra pure foil and proprietary oxidizers, depolarizers, and corrosion inhibitors in the electrolyte to eliminate a number of issues regarding long term storage, suppress any excessive hydrogen gas, and rein in the aggressive tendencies of the H2O based electrolyte. Rubycon’s main plant, in particular, borders upon a spring of incredible purity, which should go a long way to control the pH balance in the electrolyte. To my knowledge, long term storage for non-aqueous capacitors is only an issue in terms of foil deterioration rather than volatile electrolyte.

Of course, all those precautions don’t completely negate the possibility of failure even in the best case scenario (although they may significantly minimize it), and after enough time I guess even those aforementioned additives lose their effectiveness. But in this case I’d guess all the caps died from excessive heat.

[momaka]

Re: Fake Rubycons?

They look genuine to me as well.

Board may have ran very hot if you have a lot of them bulging. Either that or the board sat too long on the shelf and the electrolyte at through the foil layer (on that note, 6.3V and 10V electrolytics tend to be more susceptible to that than their 16V siblings.) Or as retiredcaps suggested, maybe they reached the end of their service life.

As for the motherboard still not working... nVidia chipset??

[stj]

Re: Fake Rubycons?

looking at the state of the print, i'd say they got pretty warm.

[retiredcaps]

Re: Fake Rubycons?

14 years = 122,640 hours.

Datasheet says

https://cdn.badcaps-static.com/pdfs/...a384835b7e.pdf

2000 hours for 105C.

Or

2000 hours at 105C
4000 hours at 95C
8000 hours at 85C
16,000 hours at 75C
32,000 hours at 65C
64,000 hours at 55C
128,000 hours at 45C

MBZ might have lasted its lifetime depending on temperature of the motherboard.

[eccerr0r]

Re: Fake Rubycons?

Ah okay, thanks. Assuming they are 105°C caps I'd think they could survive some heat. The stricken board has a lot of bad caps on them and I've been trying to find and replace the clearly bad caps and these stuck out. As replacing them didn't change the outcome (they were over 5Ω ESR), I'm at a loss what these capacitors were actually doing...

There are a bunch of other caps that have also burst but as I don't have spares on hand I had to make the decision to replace the ones who aren't filtering properly first (by 'scope).

[Per Hansson]

Re: Fake Rubycons?

They are probably genuine.
Rubycon MBZ does not like heat, well better said none of the high aqueous ultra low ESR models do.

[retiredcaps]

Re: Fake Rubycons?

They look real to me. Classic K vent. Correct bottom bung.

[eccerr0r]

I pulled these from my Foxconn board as I was seeing ridiculous ripple. After pulling them I find that their ESR was off the scale.

As I didn't have any spares I ended up jury rig soldered capacitors that were too big for the allotted space, unfortunately it didn't solve the problem despite their horrid ESR.

But this is most strange, are these fake? Surprised they went basically OC. Then again the area around these caps did get quite warm... plus probably 14 years old ...