Announcement

Collapse
No announcement yet.

ASUS NCT-D first bad caps after 10 Years !!!

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

    #21
    Re: ASUS NCT-D first bad caps after 10 Years !!!

    I'm puzzled you mean leave ML alone, or don't use polymers at all?

    Comment


      #22
      Re: ASUS NCT-D first bad caps after 10 Years !!!

      Just for the 100uF capacitors , use electrolytic capacitors for those. Use electrolytic or polymer capacitors for the others, as you wish.

      Comment


        #23
        Re: ASUS NCT-D first bad caps after 10 Years !!!

        Mariushm's talking points are well-explained and spot-on. But there's just one thing I'd like to clarify about the endurance/load life/lifetime ratings for solid conductive polymer capacitors. Unlike most low impedance / ESR wet capacitors manufactured by the big four, polymers are NOT tested for however many (in this case, 2,000) hours at 105ºC at the maximum rated ripple current, just the maximum rated voltage with a 1000Ω resistor in series so as to ensure that the peak voltage does not exceed the rated voltage. If you were to test them with the maximum ripple current as well, that would reduce the lifetime to 1/10th of the rated lifetime, as such a high ripple current would cause a 20ºC internal core temperature rise (or a core temperature of 125ºC). For an example, if you use Nichicon's lifetime calculator and apply the maximum rated ripple current to their polymers which are rated for 2,000 hours, you will find that the lifetime drops to only 200 hours for the aforementioned reasons:

        http://www.nichicon.co.jp/english/pr...01.html?sr=PLF

        The 131,400 hours (15 year) limit of both the Nichicon and Illinois calculators stems from the fact that due to the eventual deterioration of the rubber bung, no manufacturer guarantees a shelf life beyond 15 years for electrolytic capacitors (Nichicon, Illinois, Nippon Chemi-con, Elna, NIC Components, and even Taicon, Man Yue [Samxon], Teapo, and Su'scon have all admitted to this more or less). That includes polymers too. I would suspect that epoxy resin-coated rubber bungs would allow the capacitors to last a bit longer (20 years or 175,200 hours). Whilst liquid in and of itself can't pass through rubber bungs, water vapor, humidity, or very small molecues certainly can, and any seal that isn't hermetic (hermetic seals would be glass-metal, glass, metal-metal, ceramic, etc), or rather polymeric (rubber, plastic, any epoxy) will allow the liquid electrolyte to dry out over time (moisture ingress will cause a steady ESR rise in polymers as well, since it will eventually evaporate to steam).

        Epoxy-coated rubber can still fracture and no epoxy can completely bond to the tinned lead wires. Capacitors with high H2O content are bound to dry up sooner because they are more prone to evaporation than non-aqueous electrolytics (barring quaternary ammonium salt electrolytes), although any solvent will eventually permeate and pass through their rubber bungs, it just happens at a slower rate.

        I guess what I'm trying to say is, capacitors that aren't hermetically sealed are going to have a limited shelf-life no matter what. So the idea of fully recapping the board, even if only one KZG went bad, is a prudent one.

        Comment


          #24
          Re: ASUS NCT-D first bad caps after 10 Years !!!

          Also, sorry to double post, but as an addendm and a correction, I'm fairly certain the Lelon capacitors are in fact LTEC and not Lelon (the TH series, 100uF 16V). But since LTEC series that tiny are known to dry out very quickly, replacing them is also a very good idea (again though, with electrolytics and not polymers as mariushm suggested).

          Comment


            #25
            Re: ASUS NCT-D first bad caps after 10 Years !!!

            Thank you very much for taking time to explain about polymer caps.
            I replaced almost all of the caps, but I used regular electrolytic Rubycon caps. PC works OK.

            I had tested the desoldered caps and new Rubycon ones, every one read 0.00Ohms
            That is the reason I left some unreplaced. But I do have spares just in case.

            The video rendering that I do uses image sequences so no acceleration there of any kind, no GFX or CPU. Why more cores would be better, I do not have money right now to replace the PC.

            Good luck to you my friends

            Comment

            Working...
            X