intel D845GEBV2 with blown nippon chemi con

[trodas]

Re: intel D845GEBV2 with blown nippon chemi con

The most sure way will be, tough, use heatsink that blow the air on them, like Zalman 7xxxx series or there Thermalright series:



The fan on them blow the air right to the caps and mosfets, effectively cooling them and giving them much longer life. Rubycons MBZ won't hurt, tough. Maybe even increased capacity, if fit.

[gonzo0815]

Re: intel D845GEBV2 with blown nippon chemi con

Think this board is a little deamanding to the psu, as it is an 2 phase desing neding a lot of power. Then, the KZG aren`t long life types, e.g they are rated at 2000h @105°c. Derated acordingly to the guessed temperature of about 50 or 70°c (considering the closely placed mosfets and copper traces conducting heat to them) they probalby will work for about 8000h. Considering a weak psu will shorten life too. If esr has increased, more heat had be produced (considering the high ripple current in dual phase vrms this will be not nrglilbe). thus the ending is that all caps are blown.

As replacement i think some MBZ or other long life low esr will be fine.

[PeteS in CA]

Re: intel D845GEBV2 with blown nippon chemi con

I think MD hit it. KYs or HEs might be a little more tolerant of heat; you might check whether the cooling from the fans in the system are all working correctly. Your customer might need a little talk about keeping the space around the case clear, too.

[linuxguru]

Re: intel D845GEBV2 with blown nippon chemi con

I have the same board, though the manual is for the i845PESV (the alter-ego). The caps are different - instead of the Chemicon KZGs, there are 3 x 3300uF/6.3 Rubycon MBZ at the same locations. Instead of the Sanyo polymers, there are 7 x 560uF/4v Fujitsu SPAs (yellow) at the same locations. The green 1000uF/16v KZEs are the same. There's a Nichicon HD next to the DIMM slots. All small caps are 85c Nichicon VR or Rubycon YK (shame on Intel - they could have used 105c VZ or YXG instead, at negligible cost premiums).

I had some flakiness with the board, but that has been confirmed to be a PSU problem. These boards simply will not tolerate Deer/L&C crap.

[trodas]

Re: intel D845GEBV2 with blown nippon chemi con

kc8adu - looks like another Chemi-con caps killing board, or we take one star on the rating of these Chemi-cons down?

all the polymers were open too!

Don't belive it, care for a photo? Polymers should not and never blow out...??? WTF???


MD Willington - heat is definitively a factor. Mine Chemi-cons that blow on me on my DFI LP B was the hottest caps on the whole board. So, the lack of airflow defnitively can put a stresed enought cap life to less that year.
And since the ribs of the heatsink point in different directions that the caps are - they has next to none airflow. The heatsink will be hardly at 50°C, yet it sure deliver heat and all in all...

[MD Willington]

Re: intel D845GEBV2 with blown nippon chemi con

Ouch... not wanting to start anything here but is heat a factor? That sink is pretty close to the caps.

[kc8adu]

all the polymers were open too!