ET: 9nda3j & TCCD 1T problem

[rE-vErb]

if ET comes by this thread, please inform us 9nda3j users about whether a BIOS fix would be possible for this issue, and if it is in the works.

i have searched around the web and found out that there is a problem running TCCD modules at 1T(dual channel DS) on 9nda3j MBs where the exact same setup(CPU and memory) on several other MBs would run fine on 1T.

although this could be specualtion, but it also seems that the 9nda3+ and 9nda3j are not that identical as they seem to be(i mean not just the extended options), they don't seem to share the same BIOS(or do they? they seem to be different files, though dated the same).

i have seen users with no problems running DS tccd modules at 1T with brain power PCBs on 9nda3+ but none on 9nda3j.

i have switched CPUs, memory sticks and BIOS settings, including the BIOS itself to an older version, and possibly almost every method to get this going.

i had found an exact same situation posted on a russian review site, the link is as follows :
http://www.overclockers.ru/lab/18375.shtml

this problem could be specific to tccd with brain power PCBs or even more specific to just the Corsair TwinX 4400C25(but as i know of, if the IC chips and the PCB are the same, its just a matter of SPD timings set by the manufacturer, and it should be easy to have no problems with them with a new BIOS)

edit: forgot to mention; it's probably a venice core specific issue

[XeroHouR]

I'm using an epox 9nda3j with pdp patriot xblk. It works 1T fine.

[rE-vErb]

XeroHouR:
is that at the settings on your sig?
333 1T 1.5-2-2-9 1:1? with 2x512MB memory?
did you do a vdimm mod?
the pdp patriot xblk has the same IC chips(tccd)and the same PCB(Brain power) as my corsair 4400C25, if that is the case... then maybe its a SPD recognition problem on the bios level?

[RTarget]

I haven't noticed any problems with TCCD on my 9NDA3J either. OCZ Plat. Rev. 2 @ 260x10 2.5-4-4-10, 2.8 VDIMM.

[rE-vErb]

RTarget:
is that with a venice? or winnie?

[XeroHouR]

mine is venice, its 2x256 though, and its not 1:1 its divided, i forget how, but I'm not running that setup anymore right now. no vdimm

[rE-vErb]

Originally Posted by XeroHouR

mine is venice, its 2x256 though, and its not 1:1 its divided, i forget how, but I'm not running that setup anymore right now. no vdimm

2x256 would be 2 SS modules in dual channel...
the problem is that the enhanced memory controller on the rev.E cores should be able to handle 2 DS modules in dual channel at 1T, or 4 SS modules in dual channel at 1T, whereas winnies do not guarantee(but good setups can) that.

thanks for being more specific it always helps to narrow down problems

[azool]

Hi,

With some peoples we organised a groupe order of 9NPA+ Ultra and we have the same problem.

Someone can't go over 280Mhz with 1T timing with his TCCD wich run on DFI @ 300 1T at same timings with no problems. (Memtest errors on EPoX and not on DFI)

I have Gskill LA sticks and no way to go further 275Mhz 1T even with CAS 3.

Any possible bios issue ?

[rE-vErb]

azool:
in my case, i can't even do 1T at everthing else auto 1:1 @ 200Mhz(default). i hope i can even dream about overclocking at 1T.

[Judderman]

I have the same issue with my setup and my Corsair memory, I cant get past 220 HTT (1:1), this is my 2nd board that I have had no joy with at 1t.

I previously owned a Abit KV8-Pro and couldnt get past 220 at 1t but hoped that a change of board, chipset and CPU might help, but sadly didnt.

My only hope now seems to be a modded bios.

[Rocketman6800]

I have my 2 X 512 MB Geil Ultra running at 2-3-3-6 1T (~200Mhz) dual channel mode. No problems at all. My board revison is 2.1.

[rE-vErb]

Originally Posted by Rocketman6800

I have my 2 X 512 MB Geil Ultra running at 2-3-3-6 1T (~200Mhz) dual channel mode. No problems at all. My board revison is 2.1.

is that on a venice?...

[Rocketman6800]

It's a Winchester.

[rE-vErb]

i'm getting pretty sure that this is a rather simple problem relating to the 9nda3j bios not coping well with the new revision E memory controllers... any thoughts on this ET?
i have yet to see anyone with a venice or san diego core on a 9nda3j who uses 1T setting(at even default 200Mhz FSB) with their double sided tccd chips on a brainpower PCB, dual channel.

i've given up much hope of a mod bios because through some research, i found out that there aren't much information known(or hacked) about the nforce3 chipset compared to nf2 or nf4, and the only modded bioses(on the msi k8n neo2 pe) are easy mods which only enable the hidden settings. i could find no information about switching between ram tables or any releated to ROMSIP on nf3.

though it could only be a mere problem specific to such components, many enthusiasts are choosing tccd with brainpower PCB memories and many are going to upgrade to the new cores. i hope an official bios fix is coming up...

[Hysteria]

i hate 9nda3j...i just bought'it and i can't run 1:1 on 250 Mhz with my winnie no matter what HTT multi or voltage i use.
the modules r 2x 512 corsair DDR 500 8-4-4-3 1T

on Abit AV8 i could run @ 260 1:1 8-4-4-3 1T 2.8v with no problems at all, and on this board i cannot run even at 220 1:1 without a BSOD.

tried every possible setting and no go.

what's wrong with this board? is there a fix for this?

on corsair page i saw that my memory modules are compatible with 9NDA3+.

aany help would be appreciated.

my 9nda3j is rev 2.1 and it's a crappy board, i think i'll throw it over the window tonite if i'm not gonna manage to get it to work @ 250 1:1.


sorry for my crappy english

[EPoX Tech]

I also am sorry for your crappy English, or better put, your poor use of English vocabulary and grammar and a need to use abrupt or dirty language in a disturbing manner.

[MONKEYMAN]

Amen.

I have never had an Epox board with any major problems and the one I did have that had a minor problem was Rma'd with ease, I dont know why people have to come into a forum to hurl around their own abusive opinion, if you had discussed the topic further on the forum you would have probably come to a solution or a reason for your troubles.

[rE-vErb]

Originally Posted by MONKEYMAN

Amen.

I have never had an Epox board with any major problems and the one I did have that had a minor problem was Rma'd with ease, I dont know why people have to come into a forum to hurl around their own abusive opinion, if you had discussed the topic further on the forum you would have probably come to a solution or a reason for your troubles.

i neither had any major problems, EPoX was and is my favorite choice since the days back to the VIA MVP3 chipset when i used K6-2s.

and of course, this is not a "major" problem either, because at 2t settings everything works fine. But i have found out that this minor problem could be related simple bios misimplementation, and non other board has the same issues, including the fact that with a winchester it wasn't a problem, and also providing a link to a review site which came upon the exact same problem - which led be to believe that this is not only my problem. That is why i am hereby formally requesting help to ET.

also i am sorry to mention this, but for a few who have misunderstood this thread as a problem with overclockability: it is not, and please don't whine about your own problems with overclocking, EPoX has no responsibility to help you out in such matters.

I will clearly restate the main issue here:
it occurs on a rev.E CPU with 9nda3j M/B using 2 DS tccd memory chips on brainpower PCBs. 1T command rate does not work, and rev.E CPUs are supposed to work 1T with 4 SS modules or 2 DS modules in dual channel(i am not talking about OC'ing, but on default 200Mhz SPD settings). and there was no such problem with rev.CG winchesters.

[Hysteria]

Originally Posted by EPoX Tech

I also am sorry for your crappy English, or better put, your poor use of English vocabulary and grammar and a need to use abrupt or dirty language in a disturbing manner.

sorry for that, but i'm angry cause i've just spent half of my monthly fee on a mobo that doesn't do the job properly. I'm sorry if you do not understand, but living in eastern europe it's very tough.

i've never had any problems with Epox untill now and i've had Kha+,8RDA,8RDA3+ and 4PDA5+.

after 5 hours of tests i still cannot increase memory frequency much over 200, highest setting i can manage is 9*292 HTT / mem @ 193 Mhz

[EPoX Tech]

But the memory controller isn't on the mainboard - so why be angry with us ?