Announcement

Collapse
No announcement yet.

Bad Rubycon caps in HP D530 SFF Motherboard

Collapse
X
Collapse
 
  • Page of 1
  • Filter
  • Time
  • Show
Clear All
new posts
Previous template Next
    #1

    Bad Rubycon caps in HP D530 SFF Motherboard

    Hi everyone. This is the first post from a newbie who found bad caps in some HP business PCs. BTW, love the great practical information on this site!

    I just bought a lot of 27 PCs, Hewlitt Packard D530 SFF, from an off-lease reseller. While imaging them I found 6 units have bad capacitors either leaking or domed on top, in all cases they are near the CPU.

    What's surprising is the failed caps are all Rubycon MCZ series, 6.3V/1800uf, which everyone here says are good. Who am I to disagree? But it makes me wonder if there is a design flaw in the motherboard or PSU that is causing premature cap failure? Can anyone recommend a replacement capacitor that would "beef up" the fault tolerance in case I'm right? The maximum size I could install is 8x25. Thanks!
    Tags: None
    #2
    Re: Bad Rubycon caps in HP D530 SFF Motherboard

    .
    Attached Files

    Comment

      #3
      Re: Bad Rubycon caps in HP D530 SFF Motherboard

      Thanks kc8adu. Didn't have a camera available.

      But now I do, so here are the pics...
      Attached Files

      Comment

        #4
        Re: Bad Rubycon caps in HP D530 SFF Motherboard

        is the PC well-vendilated with intake and exhaust fans within the chassis? if not, heat can easily built-up within, causing the internal temperature to rise. heat is the number one killer for capacitors and all other components in the PC, even if it is of the premium grade.
        Last edited by leekh; 06-09-2007, 10:18 PM.

        Comment

          #5
          Re: Bad Rubycon caps in HP D530 SFF Motherboard

          Originally posted by leekh
          is the PC well-vendilated with intake and exhaust fans within the chassis? if not, heat can easily built-up within, causing the internal temperature to rise. heat is the number one killer for capacitors and all other components in the PC, even if it is of the premium grade.
          I can't say for certain that heat is not an issue as we just started buying these models and don't have a long history with them. These small-form factor HP business PCs are not designed to be fiddled with by users - about the best you can hope for is all the fans keep working and the HP engineers did a proper job in the design. The D530s have one intake fan, a PSU fan, and a CPU fan that vents directly out the case.

          Since HP sells a lot of these you'd expect a lot of chatter on the 'net if they put out a lemon - the ultra-slim desktop designs were a poor design (too hot inside, as you suggested) and lots of people reported leaking caps, but this doesn't seem to be the case for the small-form factor designs.

          So a valid argument that I can't rule out, but I still have to decide whether to recap with the same Rubycons, or change to 10V caps, higher uf, etc. I know nothing about caps, hoping for some suggestions...

          Comment

            #6
            Re: Bad Rubycon caps in HP D530 SFF Motherboard

            Originally posted by DG in 54/40
            I can't say for certain that heat is not an issue as we just started buying these models and don't have a long history with them. These small-form factor HP business PCs are not designed to be fiddled with by users - about the best you can hope for is all the fans keep working and the HP engineers did a proper job in the design. The D530s have one intake fan, a PSU fan, and a CPU fan that vents directly out the case.

            Since HP sells a lot of these you'd expect a lot of chatter on the 'net if they put out a lemon - the ultra-slim desktop designs were a poor design (too hot inside, as you suggested) and lots of people reported leaking caps, but this doesn't seem to be the case for the small-form factor designs.

            So a valid argument that I can't rule out, but I still have to decide whether to recap with the same Rubycons, or change to 10V caps, higher uf, etc. I know nothing about caps, hoping for some suggestions...
            there are quite a number of cases whereby there are premium grade caps failure due to a number of reasons:
            1. overheating within the chassis. notice that only the VRM sections of your motherboard is affected. this area is always subjected to high heat due to the CPU and the mosfets.
            2. motherboard design flaw, that causes excessive ripples to the capacitors, causing to fail prematurely.
            3. faulty psu. a faulty psu can create excess ripples as well, take a big toll on the motherboard caps as well.
            4. failing neighbouring caps as well. as u have noticed, there are teapo capacitors around the VRM section as well. I suspect the teapos are the contributing factors as well as they are known to fail prematurely on motheboards due to the high surrounding temperatures. the rubies have to work doubly hard and becomes stressed, causing them to fail as well. teapos are not considered badcaps to me, its just that they are not suitable for motherboard applications. however, if they are used in premium grade power supplies, they are wonderful. they can even be on par with japanese caps, provided that the cooling is sufficient.

            To go for recap, the surrounding teapos have to go as well. to replace the dead rubies, rubycons mbz series are excellent replacements. for the teapos, panasonic fm series of similar rating are excellent replacements too.

            Comment

              #7
              Re: Bad Rubycon caps in HP D530 SFF Motherboard

              From my point of view, they would have better used some more Oscon Poylmer capacitors, as the initial design is probably made for them.
              Recently, all those vrms which are build with solid polymer caps in mind, are getting really hot, even the caps are uncomfortable to touch.
              So if you put an wet capacitor type in this kind of application, it is from my point of view a very typical design problem to save some bucks.

              Additionally the 8mm diameter caps are usually not a very good choice too, as they tend to have much smaller endurance as the bigger caps.
              And those small boxes are usually not very good cooled, due to noise complaints.
              The best solution for this problem, would be a recapp with some high grade Oscons and some high endurance capacitors for the Teapos (i would prefer some not that ultra low esr types like Panasonic FM, KZE or similar).

              Comment

                #8
                Re: Bad Rubycon caps in HP D530 SFF Motherboard

                I've seen a few Dell SX270's (not GX270) with vented rubycon caps. SX270's are mini-PC's, no ventilation, case packed very tight. cause was overheating. No cap is immune to heat.
                <--- Badcaps.net Founder

                Badcaps.net Services:

                Motherboard Repair Services
                ----------------------------------------------
                Badcaps.net Forum Members Folding Team
                http://folding.stanford.edu/
                Team : 49813
                Join in!!
                Team Stats

                Comment

                  #9
                  Re: Bad Rubycon caps in HP D530 SFF Motherboard

                  Thats the VRM output side, so PSU ripple is an unlikely suspect.. Do these SFF PC's by any chance have a Prescott core CPU in it?? Whether it's a P4 Prescott or Celeron Prescott doesn't make much difference.. I say 90% odds that the CPU in the socket is Prescott core, which are hard on well cooled VRM's, let alone SFF VRM's which are more likely to be overheated by a hot CPU.. I see the motherboard is a 4 phase design, which also says to me it's a Prescott core, as the older Northwood's didn't really "need" 4 phase..


                  My suggestion:

                  Recap the board as original, with MCZ's in the VRM output. Also unsolder the MCZ's that have not vented, and see how they look on an ESR meter.. Might check the date code on them and see if they're from the same batch.. If you've got Prescott CPU's in these SFF boxen, i would recommend pulling them and getting some P4 Northwoods drop in instead, as you could probably drop the heat in the VRM by a factor of at least 30% .

                  Comment

                    #10
                    Re: Bad Rubycon caps in HP D530 SFF Motherboard

                    Thanks everyone for your replies. I like to think myself somewhat handy with computers, but you guys are on a different level!

                    The CPU in the picture is a SL6PE Northwood core (sorry, gg1948). We could have Prescotts in other PCs as we bought a range from 2.4 to 2.8 Mhz, but I wonder if HP purposely avoided Prescotts?

                    What Teapos are you referring to, Leekh? I have 3 larger 16v/1200, and 10 6.3v/1000 scattered around, including 2 between the vented Rubycons and the northbridge. Or did you mean the solid caps in the VRM area with the vented Rubycons? I can't see a brand on the casing, but Teapo does make some 2.5v/820 polymer caps.

                    I'll take all your suggested recap options and look into what's available, but I'm probably going to be handicapped by the 8mm diameter, or the 1800uf for that matter, and end up with Rubycon again.

                    Comment

                      #11
                      Re: Bad Rubycon caps in HP D530 SFF Motherboard

                      Whoops, I have found out that the polymer caps in the VRM area are actually Oscons, something y'all probably knew. But replacing the Rubycons with more Oscons doesn't appear to be an option as I can't even get close to a comparable capacitance.

                      Comment

                        #12
                        Re: Bad Rubycon caps in HP D530 SFF Motherboard

                        That is not necessary, as usually the solid polymer caps can tolerate heat & high ripple current even with far lower capacitance than those MCZ.
                        The reason for the 1800uF is probably not that this amount of capacitance is required, rather than to obtain the high ripple current capability.

                        In short, anything in the rang between 820uF to 1500Uf would be ok if you use Oscons.
                        The only source for Oscons i know, is waifong.

                        Comment

                          #13
                          Re: Bad Rubycon caps in HP D530 SFF Motherboard

                          Originally posted by DG in 54/40
                          What Teapos are you referring to, Leekh? I have 3 larger 16v/1200, and 10 6.3v/1000 scattered around, including 2 between the vented Rubycons and the northbridge. Or did you mean the solid caps in the VRM area with the vented Rubycons? I can't see a brand on the casing, but Teapo does make some 2.5v/820 polymer caps.
                          the teapos are the green colored caps at the VRM input and VRM output surroundings. just take a closer look at the green colored ones.

                          Comment

                          Previous template Next
                          Working...
                          X