mynameisyoung - yep, the oficial bios allow you only 3.3 or 3.2Vdimm, I can't check precisely becuase my page just gived up, however following the link I posted in previous post you learn how to set up to 3.8Vdimm voltage
I don't see much pls having get troubles with therir DFI mobos, but I coule be, of course, mistaken. But there is nothing like stress-free overclocking
Overclocking is allways about getting close to the borders of possibility and somewhat even over the edge...
Yes, mine Epox 8RDA+ died after the capacitors close to my mosfets bumped too much (mosfets heatsink temp 132 degrees, capacitors temp 120 degrees) and I added replacement capacitors from the back of the mobo and one of them exploded. That is last time I saw my 8RDA+ mobo working
It happen at modest 200x12.5 with 1.850Vcore. From this time I got VERY suspicious about mobos, that having capacitors close to mosfets and even - hell what a stupid idea - placed the mosfets between the capacitors
And if you can check photos, Abits are designed like this. Therefore I fear that they end up such badly, as my Epox...
DFI using a bit different design witch I hope will survive my fanless case. For more infos/photos about my Epox feel free visit this galery and have funn:
http://ax2.wz.cz/galerie.php?p=wc&c=7&d=1&v=v2
Maxamus - I got two Abit mobos. One ST6R. It never allowed me to go over 120Mhz FSB and any other divider settings that 1:3:3 locked up the mobo for good. Tell me about Abit and top brand again, please
Second Abit mobo I have is KD7-G. Not bad one, same mosfets design as DFI and handled 166x15 with 1.850Vcore (that put my Epox after 20min to it's end) for more that week w/o troubles
Folding 24/7 ... So I say it depends. But don't say that Abit is paing attention to details and stuff like that. Driver for the gigabite network can't be found anywhere on Abit page for the KD7-G and you have to search over net for yourself. My conclusion - Abit is overrated.
apocalypse80 - well, its likely that the 3.8Vdimm trick require more rencent bios and perhaps even more recent revision of the mobos? I don't know. Todays revisions working well with the latest bioses and that is what matter IMHO. Yup, for BH-5 rams this is a jackpot, but presonally I think that the dude is rather insane
You know how long it took you to convince me to run my Mushkins at 2.91 instead of 2.77, right?
So you perhaps got the idea, what I think about 3.8V
The good news is, that even w/o cooling the mosfets the mobo can make it. This giving my the opinion, that even with my fanless enviroment the mobo is capable of doing like 225x12 with 1.9Vcore and 3.0Vdimm stable for long time
And my point it, that there is virtually no Abit mobo that can get at 1T to 260MHz, not even over
I can't fight the accusing of the DFI users from their blind loyality to DFI
Well, they are, erm, different
There are even some things that are pretty bad from DFI, like the fact that their latest official bios are almost half year old (!) and w/o the latest users likely to experience cold boot issues with mobile Bartons and so on... I quess we wait and see - you know, Im really perefectionist and I will not overlook any shortcoming of the mobo
Maxamus - well, goog question. Even I have electronics degree, I'm still not sure WHY my Epox died... Or what happens. The irony of life is, that few days before it happen I chatted with friends and it finds out, that even I stressing my components hard with very high temps, bad airflow and fanless cases, I come out as the best one from my overclocking pals, because - at this time - I have ZERO died componets into my overclocking path
Now I got one dead - or almost dead, It seems that the mobo want kick in, showing FF on the diagnostic led as okay, ready to go, but there is no voltage past the mosfets (only 0.3 - 0.5V instad of 3.3, 5 and Vcore/Vdimm) and I can't replace them, because I can't desolder them, because there is a heatsink on them what I glued by ArcticSilver epoxy and I can't remove it
So I don't know.
From what you suggest, I should say - memory pushed to its limits? Not a problem for mobo, just overheating the NB, if it's cooled enought, not a problem. Sooner you get very bad errors that you fry the NB
Pushing CPU to it's limits? Dude, get a good water or phase cooling, cool the mosfets too, and you are happy
Too much voltages? Well, yep, using 3.6V on 3.3V rail cause the mosfets overheat even more, but allow you to get more stability on very high Vcores and such. I tried 2.20Vcore and it worked. Maybe there is where the death of the Epox start, because about from this time the mosfets started overheating very much... My quess is, that the mosfets start overheat because the capacitors capacity after the 2.20Vcore get lover because they got too much overheated and therefore the mosfets starting overheating even more, and this process iniciated the death end circle... Maybe. It's one of my ideas how it happend...
So far, the mobo is not possible to destroy, unless you use hammer - IF you pay attention to the temperatures and cool the overheating components - and sometimes coolina pasively is not enought.
The AIRFLOW is need
I recently tried to cool-down the NB on my replacement MSI KT6V mobo and ended up with at least 3-4x bigger NB heatsink
(
http://ax2.wz.cz/show.php?p=wc&id=94&c=7&d=1&v=v2 ) but since my airflow is next to zero, It did not helped much, in fact, I got about 1 degrees down at best...
Lesson lerned - I reorganized some cables, and even with only passive airflow my temps droped 9 degrees for NB and 4-6 degrees for CPU. Still, top FSB of this MSI crap is about 174Mhz. When I used the mobo w/o case on my testing desk, It folded stable at 185Mhz. What this telling you is obvious, is not it?
It's about the cooling.
Watercool the mosfets, phase cool the CPU, install 4 hi-flow fans to your case and you can get any overclock you ever want w/o worry about the mobo
Of course, unless you selected Abit mobo with mosfets between capacitors, that can't be cooled
BTW, there is one of my very strange recent toughts. Don't cool with heatsink something, what is not overheating DENGEROUSLY. Why? Because for example, what happen, if I cool my HDD? Let's imagene installing a ZM-2HC2 silent HDD heatpipe cooler on my HDD. My HDD temp probably drop from 57-63 degrees about 10 or even more degrees down, but where the heat go? I tell you. To the case and to heat-up the rest of the components!
So, I better did not cool it, because the temp are well with the operational limits, because when I cool it, the heatpipes draw the heat to the air inside of my case more effectively. Effectively heating up my case and other components. Im going to say that with more heatsinks and bad airflow you sometimes can do more worse that good, and I, for example, never install a heatsink on the SB - just because the temperature hitting 60 degrees? To the hell with SB
It can handle it.
So far, I starting to get convinced that too much BIG heatsinks actually killed my Epox mobo into my fanless case.
Sounds weird? Come on, I whould like to heat better theory or explaination that how heatsinks "support airflow" instead of restricting it, and how for example "cooling things down" did not increase temps into my case...