It also fried the front USB ports (or at least the chip that controls them along with the 1394 port), but putting in a USB 2.0 card has rectified that for now.
I don't mean to bump again, but I do have another question that has been encircling my mind. ^^;
Are capacitors by the CPU supposed to get hot when the CPU is under full load? In hot enough weather, too hot to touch? In my case, I mean the 105C (Nichicon HD and they're gray) capacitors left of the CPU. One, in particular, gets noticeably hot even when the CPU is only at 49C for an hour. At 53C it gets almost too hot touch, which leads me to be of the belief that it would be too hot to touch if the temperature were any higher for a prolonged span. I know capacitors release heat, but are 105C capacitors next to the CPU (one, in this case, in particular, is hotter than the others) supposed to be hot? Could it be either A) that the PSU in there does not have 80% efficiency (it's a 300W Hipro so I doubt it has that) so it's giving off more ripple or B) that the CPU fan expels heat to the left, where the vents are, and that the 105C capacitors that are getting hot are left of the CPU? I know those capacitors are rated at 105C but I have a drudging time believing that a capacitor could last when it gets too hot to touch in hot weather (when the CPU is at 100% full load). One 105C capacitor gets hotter than the CPU heatsink itself does (and hotter than the air the PSU fan is pushing out, along with the CPU fan), so....
As well as the Chemicon PS capacitors.... but I noticed that the top of the capacitors are too hot to touch, period (only when the CPU is at 100% load; when at 0-5% load, they are hardly warm).... what does that mean?
Assuming the heat is not radiated from nearby components or conducted from nearby components through copper traces ...
The terms "low impedance" and "low ESR" are used here more or less interchangeably,but they are not identical. The impedance of a cap is the vector sum of the ESR, the ESL reactance and the capacitive reactance (the latter two partly cancel each other, with the net reactance, inductive or capacitive, being frequency dependent). ESR = Equivalent Series Resistance; ESL = Equivalent Series inductance (L). The ESR dissipates power when ripple current flows through it, the formula being (I)(I)(ESR). So if you have 15 milliohms ESR and 3A of ripple current, that little 10mm diameter capacitor is dissipating about .135W. And the only good heat dissipation path is through the top (for a lytic with a plastic sleeve).
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.
****************************
The threshold for "too hot to touch" is somewhere around 55C-60C for most people. 60C is not bad for 105C cap. But touch is not a good test for whether the part is too hot to have a long useful life.
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.
****************************
60C would be okay in cold weather?? ^^; For a 105C cap... because that's how it felt at the top in weather that was 60F (at 100% load on the CPU, which was at around 50C, for about an hour). It only concerns me that it felt as such in cold weather. But if touch isn't a good test of heat, that means that touch is not a reflection of the temperature of the part (necessarily)?
I hate to bump this thread but I figure this thread deserves an update.
There are bad capacitors on that motherboard.... three HMs with a date code of Week 20, 2002 sit by the CPU.... it's funny since I always heard people say the D845PEBT2 was not affected. Maybe the earlier ones weren't since that particular PC was bought in January of 2003. Still too bad. I know it can be recapped but I dunno if something that old is worth the effort, especially considering the board has already suffered slight damage from the failure of the last two PSUs. Kinda feel silly though that I squandered people's time asking so many questions about a board whose problem was right under my nose all along. ^^; Hard to believe it lasted at least over 9,000 hours but more likely 11,000+.... probably because it's not a thin case like those of the Dell Optiplexes.
[In general] Jap companies are king at QA. . CH & TW companies are king at pinching pennies.
The companies which outsource and rebadge units from Taiwan and China could also be pinching pennies.
My first choice in quality Japanese electrolytics is Nippon Chemi-Con, which has been in business since 1931... the quality of electronics is dependent on the quality of the electrolytics.
Comment