Announcement

Collapse
No announcement yet.

Rubycon MCZ and Chemicon KZG mass failures

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

    Rubycon MCZ and Chemicon KZG mass failures

    Hello everyone,

    I was doing some further research into the premature failures of Rubycon's MCZ series and Chemicon's KZG series.

    After working with a handful of the same motherboards which used varying caps from both brands and different series, there was always an 80% chance that these KZG or MCZ capacitors failed prematurely-- whereas other quality low-ESR series didn't.
    I also had a handful of other computers which were unrelated to my original five that contained blown MCZ and KZG.

    For example. if we had a bunch of 2200uF MCZ or KZGs on one board, and then 2200uF MBZ/ZLH on another, the MCZ/KZG were *always* catastrophically blown up. Generally the KZG will just open up and dry up, whereas the MCZ will spill their guts.

    Now, supposedly these caps (MCZ/KZG) offer low ESR-- but if they fail prematurely; this potentially means they aren't even *deliverying* that low-ESR. Even upon release from the factory due to their inherent instability. Similar to how chinese G-Luxons or TEAPOs are *not* delivering any of those ratings as stated on their sleeves.

    Having discovered Rubycon's rather interesting RX30 series, I decided to replace all 1500uF MCZ/KZG based capacitors with RX30 (which also offer more voltage tolerance).


    Here you can see I replaced the MCZs with RX30. Some of the rather small desoldered MCZ caps are near the top of the board for comparison. I think one problem these low-ESR capacitors have is that... they are just too small.

    After replacing MCZ & KZG with RX30s, the computers booted up fine and were tested for extended periods of time.

    In further studies, I discovered RX30 has rather impressive abilities to suppress noise in audio circuits. After recapping the audio capacitors, my line in mixer, and various sound cards-- the white noise and any line hum is rectified.


    So that's another use for them. Really quite flexible.

    Anywho, I hope that may be useful to anyone who is looking to replace the MCZ or KZG that have popped/making their system unreliable. I'd also recommend Rubycon ZLH or Panasonic FL in place of extra sensitive ESR applications.

    #2
    Re: Rubycon MCZ and Chemicon KZG mass failures

    A thread on another forum is rather dated but it has interesting commentary on Rubycon YXG/ZA/ZL/YXF performing poorly in audio/PS decoupling/bypassing applications and Panasonic/Matshushita FC performing well:

    Recently I used some Rubycon YXF caps (470uf/16V) for local decoupling of the output stage of a DAC and compared to Panasonic FC 220uf/25V the sound was rather harsh and muddy. Rubycon YXG was the same story. Same experience but only much worse with Rubycon ZA. Rubycon ZL seems to have the same character but to a lesser degree.

    IMO I cannot say that Rubycons are really bad for audio stuff, might have been some ringing going in there with the bypassing caps (tried both EPCOS/SIEMENS ceramic 0.1uf/X7R or WIMA MKS 02/0.01uf). Had the very same problems with the Rubycons when used for the input stage PS decoupling in an amplifier. The sound was nice, but it did not have that open&firm character that Panasonic FC gave me. I find Panasonic FC a much better choice for PS decoupling. Maybe the Rubycons have too low ESR for my particular applications? Don't know, that's my personal experience with those.
    This may not be the same thing as you make mention to but as you also made of note, it isn't only the Taiwanese/Chinese capacitors that are ill favored for audio. The RX30 series may perform well because they are rated up to 130*C. It may also have to do with how the performance of especially low ESR electrolytics falls off faster (at very low and higher temperatures) than less aqueous capacitors.

    FYI, I don't think Teapo, OST, or G-Luxon lie about the capacitance or voltage rating, or even ESR rating (though obviously the ripple current rating can't be true, at least not for long, because they fail much quicker than their Japanese equivalents). No Taiwanese capacitor really does well on a motherboard so they probably just fail much quicker there, hence the shoddy performance. I don't know that much about audio so I'm just making hopefully good guesses about the subject. I also wonder if the bad Rubycon MCZs you're seeing are fakes... and I also wonder if the impurities in Taiwanese/Chinese mined aluminum foils and cans are responsible for the problems you're hearing with them in audio (as well as Taiwanese/Chinese electrolyte).

    Also, I have a pair of Altec Lansing 2.1 speakers whose power supply is littered with Ltec, CapXon, and Jamicon, and I think it has the best bass I've ever heard. Don't know how they managed that. :P
    Last edited by Wester547; 07-25-2013, 08:14 PM.

    Comment


      #3
      Re: Rubycon MCZ and Chemicon KZG mass failures

      those are not rated for anywhere near the ripple current that mcz or kzg are.
      so not good for mobo use.might be great for audio use though.

      Comment


        #4
        Re: Rubycon MCZ and Chemicon KZG mass failures

        Originally posted by kc8adu View Post
        those are not rated for anywhere near the ripple current that mcz or kzg are.
        so not good for mobo use.might be great for audio use though.
        That's my whole point though-- MCZ and KZG are *not* actually performing at their specified ratings on paper. So you have systemboards operating under these caps with rather high ESR and low microfarad ratings: due to premature failures and poor electrolyte. So... the RX30s are actually an upgrade to KZG and MCZ.
        Not to mention, they'll last longer (i.e. not fail silently or violently explode) and will be able to handle higher voltage.

        ZLH and/or FL would be better in a low-ESR circuit I agree; but MCZ and KZG can be safely replaced with RX30 because they are not actually low-ESR capacitors. On paper, yes-- but not in reality.

        Comment


          #5
          Re: Rubycon MCZ and Chemicon KZG mass failures

          ^ You mean their ESR rating even brand new don't conform to that of the datasheet judging by the ESR meter? Or are you just talking about failed ones?
          Last edited by Wester547; 07-30-2013, 05:50 PM.

          Comment


            #6
            Re: Rubycon MCZ and Chemicon KZG mass failures

            Originally posted by Wester547 View Post
            ^ You mean their ESR rating even brand new doesn't conform to that of the datasheet judging by the ESR meter? Or are you just talking about failed ones?
            I have a hunch that new ones also don't conform to the datasheets due to the electrolyte being inherently unstable. Unfortunately I don't own any new ones to perform this test on.

            Although it would be *very* interesting to test the ESR ratings of [fresh] MCZ and KZG over a period of a year each month running under normal operational conditions.
            --> particularly conditions that cause them to eventually pop-- but none of the other brands (ZLH, FL, MBZ, WG, etc).

            Comment


              #7
              Re: Rubycon MCZ and Chemicon KZG mass failures

              Yes there have been reports of later dated MBZ/MCZ caps failing prematurely (They're out of production now). I guess it's safe to stick with Nichicon HM/HN/HZ.
              "We have offered them (the Arabs) a sensible way for so many years. But no, they wanted to fight. Fine! We gave them technology, the latest, the kind even Vietnam didn't have. They had double superiority in tanks and aircraft, triple in artillery, and in air defense and anti-tank weapons they had absolute supremacy. And what? Once again they were beaten. Once again they scrammed [sic]. Once again they screamed for us to come save them. Sadat woke me up in the middle of the night twice over the phone, 'Save me!' He demanded to send Soviet troops, and immediately! No! We are not going to fight for them."

              -Leonid Brezhnev (On the Yom Kippur War)

              Comment


                #8
                Re: Rubycon MCZ and Chemicon KZG mass failures

                I still wouldn't even consider using RX30s on PC hardware. If you want it to handle the heat better, then polymod.
                I love putting bad caps and flat batteries in fire and watching them explode!!

                No wonder it doesn't work! You installed the jumper wires backwards

                Main PC: Core i7 3770K 3.5GHz, Gigabyte GA-Z77M-D3H-MVP, 8GB Kingston HyperX DDR3 1600, 240GB Intel 335 Series SSD, 750GB WD HDD, Sony Optiarc DVD RW, Palit nVidia GTX660 Ti, CoolerMaster N200 Case, Delta DPS-600MB 600W PSU, Hauppauge TV Tuner, Windows 7 Home Premium

                Office PC: HP ProLiant ML150 G3, 2x Xeon E5335 2GHz, 4GB DDR2 RAM, 120GB Intel 530 SSD, 2x 250GB HDD, 2x 450GB 15K SAS HDD in RAID 1, 1x 2TB HDD, nVidia 8400GS, Delta DPS-650BB 650W PSU, Windows 7 Pro

                Comment


                  #9
                  Re: Rubycon MCZ and Chemicon KZG mass failures

                  Originally posted by theokretes View Post

                  So that's another use for them. Really quite flexible.

                  Anywho, I hope that may be useful to anyone who is looking to replace the MCZ or KZG that have popped/making their system unreliable. I'd also recommend Rubycon ZLH or Panasonic FL in place of extra sensitive ESR applications.
                  what audio card is that?

                  Comment


                    #10
                    Re: Rubycon MCZ and Chemicon KZG mass failures

                    Originally posted by logic7 View Post
                    what audio card is that?
                    Roland SCC-1: Vintage ISA sound card. Sought-after as well. Now rare.
                    Recovering a BEFSR41 v1 and v2 router from solid red DIAG Light
                    I have two v2s and one v1.

                    I am still looking at these boards nearly every day.

                    What I'm doing: Planning an upgrade of my mining setup from Block Erupters to Red Furys. Though, if the Block Erupters don't sell, I will keep using them for a while.

                    Comment


                      #11
                      Re: Rubycon MCZ and Chemicon KZG mass failures

                      Originally posted by Ami Sapphire View Post
                      Roland SCC-1: Vintage ISA sound card. Sought-after as well. Now rare.
                      Ah, I wanted one of those for years. Ended up with a SB16 and Yamaha DB50XG daughtercard. I loved that combo.

                      Comment


                        #12
                        Re: Rubycon MCZ and Chemicon KZG mass failures

                        Originally posted by mockingbird View Post
                        Yes there have been reports of later dated MBZ/MCZ caps failing prematurely (They're out of production now). I guess it's safe to stick with Nichicon HM/HN/HZ.
                        Interestingly enough I never had any MBZ fail on me. It was *always* the KZG and MCZ that were dead. I'd take MBZ over MCZ any day of the week.
                        Due to the fact Nichicon's HM(M) series has a terrible history, I stay away from it in the same manner as MCZ/KZG. There are better caps than HM anyways.

                        Originally posted by c_hegge View Post
                        I still wouldn't even consider using RX30s on PC hardware. If you want it to handle the heat better, then polymod.
                        Unfortunately polymers cannot be used to replace huge capacitors at 1500uF and beyond (it would be like trying to recap your PSU with polymers)-- the whole circuit would have to be reworked to use smaller uF values in series to make up for the huge 1500uF.

                        Also, the RX30s are better for the computers than the MCZ and KZG that were originally there. I've tested my computers for weeks now and they are running great.
                        Please note that the CPU filtering already uses polymers (like any proper systemboards). The RX30s in these higher values are near at the back of the boards. I also over-rated the uF on capacitors in the audio circuit too. This is perfect if you want better bass response and to filter out any unwanted noise.

                        Originally posted by logic7 View Post
                        Ah, I wanted one of those for years. Ended up with a SB16 and Yamaha DB50XG daughtercard. I loved that combo.
                        I actually also recapped my SB16 as well, sounds five times better:

                        Comment


                          #13
                          Re: Rubycon MCZ and Chemicon KZG mass failures

                          Originally posted by theokretes View Post
                          Interestingly enough I never had any MBZ fail on me. It was *always* the KZG and MCZ that were dead. I'd take MBZ over MCZ any day of the week.
                          Due to the fact Nichicon's HM(M) series has a terrible history, I stay away from it in the same manner as MCZ/KZG. There are better caps than HM anyways.
                          Only 2001-2004 datecodes of HM/HN are potentially bad. I say "potentially" because Nichicon made it sound as though there were only a few batches at that time. I believe a few members here have documented some HM/HN capacitors with 2001-2004 datecodes lasting several years of 24x7 use, which further affirms that belief. That being said, there's no way to make the visual distinction, so I recommend recapping any HM/HN capacitors with 2001-2004 datecodes upon sight. 2005 datecodes and after should be fine. Those series of capacitors are somewhat sensitive to heat with relation to the amount of water in their electrolyte (which expands considerably with heat), but they are fairly reliable.

                          MBZ is harder to say - we have seen too many failed here but they don't seem that awful. Hard to say why though - could be fakes, could be problematic batches, etc.

                          Those Sound Blaster cards often use 85*C Wincap, Jamicon, or Nippon Chemi-con capacitors, or even Nichicon, but of 85*C ratings and usually general purpose too... not surprised a recap helped.
                          Last edited by Wester547; 08-02-2013, 11:55 PM.

                          Comment


                            #14
                            Re: Rubycon MCZ and Chemicon KZG mass failures

                            Originally posted by Wester547 View Post
                            Only 2001-2004 datecodes of HM/HN are potentially bad. I say "potentially" because Nichicon made it sound as though there were only a few batches at that time. I believe a few members here have documented some HM/HN capacitors with 2001-2004 datecodes lasting several years of 24x7 use, which further affirms that belief. That being said, there's no way to make the visual distinction, so I recommend recapping any HM/HN capacitors with 2001-2004 datecodes upon sight. 2005 datecodes and after should be fine. Those series of capacitors are somewhat sensitive to heat with relation to the amount of water in their electrolyte (which expands considerably with heat), but they are fairly reliable.

                            MBZ is harder to say - we have seen too many failed here but they don't seem that awful. Hard to say why though - could be fakes, could be problematic batches, etc.

                            Those Sound Blaster cards often use 85*C Wincap, Jamicon, or Nippon Chemi-con capacitors, or even Nichicon, but of 85*C ratings and usually general purpose too... not surprised a recap helped.
                            The smaller canned MBZs could have issues-- equivalent uF ZLH caps are twice the size... obviously Rubycon realized you couldn't get away with making the cans too small and changed that. I also replace MBZ (they're old capacitors), but I just haven't seen a SINGLE ONE failed or explode like an MCZ/KZG-- and this is comparing loads of the same boards with varying caps on over 20 systems. That's not a coincidence. None of these capacitors were counterfeit either

                            I have a board with lots of HM(M)s which are running normally-- however, I don't recommend them to anyone because... there are *so* many better alternatives. Why buy plain boring nichicons when you could get panasonic or more exotic ELNA RJFs?
                            The only series from nichicon which is particularly useful is VX because axials are important if you have vintage electronics. I guess radials can be bodged in, but I like doing stuff properly.

                            The soundblasters that I have *were* populated with chinese wincaps and jamicons. They were all bloated and useless. Pathetic. Sewage water has more electrolyte than those things.

                            Comment


                              #15
                              Re: Rubycon MCZ and Chemicon KZG mass failures

                              A bit of personal background ... I've worked in P/Ss since 1980, except for some 4 years at a motor control company (which is pretty similar in many respects). Ultra-low impedance caps such as Rubycon MBX/MCZ, Nichicon HM/HN or UCC KZG/KZJ weren't what I worked with - 2000 hours rated life (at max rated ripple current and temperature) wasn't appropriate for products expected to last 10 or 20 years (the motor controllers). But Nichicon HE, Panasonic FM, Rubycon ZL and UCC KZE are of that electrolyte type (not quite as low impedance) and are rated for 5000-10000 hours life. So I compared a part in Rubycon's ZL and RX30 series, 3300uF, 10V (the lowest voltage rating available in RX30):

                              ZL: 12.5mm x 25mm; 18 m-ohm impedance; 2.77A ripple current; 5000 hours;

                              RX30: 16mm x 25mm; 34 m-ohm impedance; 1.90A ripple current; 4000 hours.

                              Here is the corresponding data for Nichicon's PW (7000hours) and PA (5000 hours) series parts:

                              PW: 12.5mm x 25mm; 30 m-ohm impedance; 1.95A ripple current;

                              PA: 12.5mm x 25mm; 22 m-ohm impedance; 2.28A ripple current

                              Other than being rated for 130C rather than 105C, Rubycon's RX30 is more similar to Nichicon's low impedance PW series than Rubycon's almost ultra-low impedance ZL series (water-based electrolyte) or Nichicon's almost ultra-low impedance PA series (non-water-based electrolyte).

                              All that said, the RX30s might do very well for a long time in your mobo if the mobo design has enough margin in ripple tolerance, in its regulator compensation, and in the ripple current to which the caps are subjected (or if the user isn't stressing that area of the mobo to the max). But if is one of the biggest words in the English language.
                              PeteS in CA

                              Power Supplies should be boring: No loud noises, no bright flashes, and no bad smells.
                              ****************************
                              To kill personal responsibility, initiative or success, punish it by taxing it. To encourage irresponsibility, improvidence, dependence and failure, reward it by subsidizing it.
                              ****************************

                              Comment


                                #16
                                Re: Rubycon MCZ and Chemicon KZG mass failures

                                Originally posted by PeteS in CA View Post
                                A bit of personal background ... I've worked in P/Ss since 1980, except for some 4 years at a motor control company (which is pretty similar in many respects). Ultra-low impedance caps such as Rubycon MBX/MCZ, Nichicon HM/HN or UCC KZG/KZJ weren't what I worked with - 2000 hours rated life (at max rated ripple current and temperature) wasn't appropriate for products expected to last 10 or 20 years (the motor controllers). But Nichicon HE, Panasonic FM, Rubycon ZL and UCC KZE are of that electrolyte type (not quite as low impedance) and are rated for 5000-10000 hours life. So I compared a part in Rubycon's ZL and RX30 series, 3300uF, 10V (the lowest voltage rating available in RX30):
                                Unless I'm mistaken or unless no one is privy to this by way of the datasheets, I don't think the rated lifetimes are wear out mechanisms (those aren't published and I believe that the capacitors don't actually fail in the stress tests on the datasheets, just that it is stated that they will last at least that long at 105*C and at that ripple current, at 100KHz). Of course, I have seen Nichicon state that "HE" is "longer life" than "HD" on the datasheet, so I could be wrong on that. Nichicon HD and Rubycon YXG (both of aqueous electrolyte) are rated up to 5,000 hours and 6,000 hours respectively (KZH being rated up to 6,000 hours) but I recollect them both doing poorly in your stress tests, along with KZH (though YXG has higher ESR and lower ripple ratings).

                                However, RX30 may just be suited best for audio all together (not necessarily every other application), which theokretes is using.
                                Last edited by Wester547; 08-04-2013, 03:00 PM.

                                Comment


                                  #17
                                  Re: Rubycon MCZ and Chemicon KZG mass failures

                                  I've pulled and re-used 100's of caps from Xbox 360s, and almost never seen any bad ones except on systems which were intentionally overheated or reflowed on too high of a temperature (thus popping the caps). The Xbox 360s usually have a mix of Nichicon HN and HZ, Rubycon MCZ* and MFZ**, Panasonic FL*, Chemicon KZJ*, and Sanyo WG*.
                                  * - indicates only 16V 1500uF caps
                                  ** - indicates only 6.3V 2700uF caps or 6.3V 820uF caps
                                  All the ones that looked good also tested good on both ESR and capacitance meter (yes, even the KZJs). The ones that were slightly bulged (due to reflow overheating) would test good most of the time and their vent on top could be pressed back into place without feeling too much pressure from the inside of the can. This is good evidence IMO that water-based electrolyte expands a lot more with heat.

                                  Comment


                                    #18
                                    Re: Rubycon MCZ and Chemicon KZG mass failures

                                    Originally posted by momaka View Post
                                    I've pulled and re-used 100's of caps from Xbox 360s, and almost never seen any bad ones except on systems which were intentionally overheated or reflowed on too high of a temperature (thus popping the caps). The Xbox 360s usually have a mix of Nichicon HN and HZ, Rubycon MCZ* and MFZ**, Panasonic FL*, Chemicon KZJ*, and Sanyo WG*.
                                    * - indicates only 16V 1500uF caps
                                    ** - indicates only 6.3V 2700uF caps or 6.3V 820uF caps
                                    All the ones that looked good also tested good on both ESR and capacitance meter (yes, even the KZJs). The ones that were slightly bulged (due to reflow overheating) would test good most of the time and their vent on top could be pressed back into place without feeling too much pressure from the inside of the can. This is good evidence IMO that water-based electrolyte expands a lot more with heat.
                                    Chemicon *KZJ* is perfectly fine.
                                    Chemicon *KZG* is a known afflicted series.

                                    Some are less inclined to believe MCZ has serious problems, but it does. Heat isn't the cause of their premature deaths because.. on the same boards with MBZ or FL-- they were not bulged or damaged in any way-- whereas the same boards with KZG and MCZ failed *every time*. There may be the odd cap out which didn't bulge. And this is out of over 20 boards.

                                    Once a capacitor starts to bulge like that, it could also indicate it being dried up inside. The best way to tell is to pop them open and take a peek.

                                    A lot of MCZs I squished open (the ones that weren't bone dry) had some very dark (almost black) rubyjuice. Whereas the MBZs had a much lighter colour.

                                    Comment


                                      #19
                                      Re: Rubycon MCZ and Chemicon KZG mass failures

                                      Originally posted by theokretes View Post
                                      Chemicon *KZJ* is perfectly fine.
                                      Chemicon *KZG* is a known afflicted series.
                                      KZJ isn't a commonly seen series, but on this forum you will find notable KZJ failures and it is expressly stated in the KZJ datasheets that they use the same unstable electrolyte KZG uses so I would peg KZJ bad, along with KZV (have also seen premature failures) and TMZ/TMV (a custom order from Chemi-con).

                                      Some are less inclined to believe MCZ has serious problems, but it does. Heat isn't the cause of their premature deaths because.. on the same boards with MBZ or FL-- they were not bulged or damaged in any way-- whereas the same boards with KZG and MCZ failed *every time*. There may be the odd cap out which didn't bulge. And this is out of over 20 boards.
                                      Depends on how hot the MCZ were running and what application they were in. I wouldn't be surprised if it was a heat related failure. Heat does expedite the wear out mechanism of lytics.

                                      Also, I meant that there seemed to be only a few bad (overfilled?) batches of HM/HN in 2001-2004. I think the capacitors do well in Xbox 360s because they are in close proximity to powerful airflow, either a single 92mm fan or dual 70mm fans in the rather older models, from Delta, Nidec, or Sunon.
                                      Last edited by Wester547; 08-05-2013, 02:02 AM.

                                      Comment


                                        #20
                                        Re: Rubycon MCZ and Chemicon KZG mass failures

                                        Heat is almost never your friend in electronics (except vacuum tubes, of course; aside from exotica, vacuum tubes are alive and well in output stages of microwave power amplifiers). I reflexively cringe when I see how pocketed P/S O/P caps often are, and I think such compromised cooling is even less healthy for caps with water-based electrolytes.

                                        In my cap testing (some 5 years ago, now!) I found the water-based electrolytes seemed to expand more than did conventional electrolytes. Among series with water-based electrolytes, some parts did better in my torture test than others (e.g. HE did better than HD, ZL did better than YXG, etc.). I don't have any explanation to offer, just the empirical observation.
                                        PeteS in CA

                                        Power Supplies should be boring: No loud noises, no bright flashes, and no bad smells.
                                        ****************************
                                        To kill personal responsibility, initiative or success, punish it by taxing it. To encourage irresponsibility, improvidence, dependence and failure, reward it by subsidizing it.
                                        ****************************

                                        Comment

                                        Working...
                                        X