8RDA+ HD corruption @200Mhz FSB?

Vampiregabe · Aug 4, 2003

I have the 1.1 revision. Pushed the Vdd to 1.82. Pushed my 2100+ to 1.80 vcore with a 10 multi. Water cooled. Running Corsair XMS 3200@ 100% of FSB =400DDR dual channel.

Boots just fine. In Windoz I get a number of strange errors (an error has prevented windows from validating license, even though it is registered) and the vga mode is screwed, like 600x800 even though it is normally 1024x1280.

The HD is a WD80G 8M cache.

The memory timmings are 6,2,2,2.5.

Am I doing something wrong? Where is the limiting factor?

I can, however run SiSandra @ 180X13 which clocks out better than a 3200XP.

Just not stable for very long, like 1hr or so.

zabomb4163 · Aug 4, 2003

8rda has pci lock. hd corruption from too high of bus is impossible

Vampiregabe · Aug 4, 2003

Well thanks for your thoughts zabomb4163, if it is not HD corruption, then what is preventing the booting into the OS?

I am well aware that the 8rda+ has a pci lock, which is WHY I am trying to find clues as to why this is happening?

Thanks anyway.

glock19owner · Aug 4, 2003

you will get a HD corruption if your ram timings are too high...and 99% of the times you can get around that issue without having to reformat if you lower the ram timings...or up the memory voltage...

Vampiregabe · Aug 4, 2003

Thanks Glock,

OK. Since I would like to keep the 6,2,2,2.5 raising the vdimm will have to be the solution. So raise the vdimm? To what 2.7? 2.8?

What is a safe vdimm for the XMS series ram w/o active air flow?

glock19owner · Aug 4, 2003

since you have heat spreaders on XMS's...you can run up to 2.9...now if upping the voltage doesn't correct the problem...then it will be memory timings...

Vampiregabe · Aug 4, 2003

OK. I will try that. Thanks for the info. I could also try and actively cool those ram sinks.

Thanks.

glock19owner · Aug 4, 2003

Vampiregabe said:
OK. I will try that. Thanks for the info. I could also try and actively cool those ram sinks.

Thanks.

On your video card...or on the mosfets?

sticking a side mounted blow hole fan...2 actaully...one right in front of the CPU and one right in front of the AGP slot will help...the one above will help cool the CPU and mosfets and help a little on the ram also...the one below will help cool the video card and any PCI cards you might be using...

Vampiregabe · Aug 4, 2003

I already have mosfet sinks w/o active cooling (looking for a small quiet fan).

Video card (Gforce 4200ti) water cooled with ram sinks. Since I am not that big into games I don't feel the card needs active cooling on the sinks.

Also begining to look for a small silent fan for the memory heat spreaders.

I already have active sb cooling, and the nb is water cooled.

At full load at 180x13 I was reading cpu 47C and mb 42C with radiator fans on lowest setting.

Going to raise the ram vdimm when I get home.

glock19owner · Aug 4, 2003

You should be good to go then on cooling...but I would add a few intake and exhaust fans to lower the case temps though...they don't have to be real loud or powerfull...a few 32 to 40cfm fans would work...would see around a 5c or more case temp drop...if not intakes...then at least exhaust...to help port out the hot air from the WC until...those are pretty high MB temps...

squeakygeek · Aug 4, 2003

Speakinging of ram and voltage, do you think it's safe to continuously run my Buffalotech DDR400 at 2.9v? (it gets 230mhz) I have no heat spreaders or anything.

bad4u · Aug 4, 2003

For CAS=2.5, Trp=2 and Trcd=2, try setting your Tras to 9.

That should stop any HD problems AND increase your memory bandwidth, which you can easily verify by running Sandra memory bandwidth bench mark with various Tras settings.

Mushkin recommends DDRAM Tras timing be set to CAS+Trp+Trcd+2 clock cycles. Count CAS=2.5 as 3 clock cycles.

Hope this helps!

glock19owner · Aug 4, 2003

squeakygeek said:
Speakinging of ram and voltage, do you think it's safe to continuously run my Buffalotech DDR400 at 2.9v? (it gets 230mhz) I have no heat spreaders or anything.

If you have good airflow in your case...if not and they get hot...like too hot too touch...then turn the voltage down...

calvin · Aug 6, 2003

8rda has pci lock. hd corruption from too high of bus is impossible

Relative to boards with no lock the tendency is noticeably reduced. Still a possibility, however, as pushing the mem too hard can produce data errors that can in turn lead to hdd corruption, glock19 correct in this regard. 2.9v on ddr400 mem is not excessive, and will aid stability at the higher settings.

Vampiregabe · Aug 6, 2003

Ok. Ran it up to 200FSBx10 with 1.84 on the nb. Memory settings at 2,2,2,9 with 2.9Vdimm. 1.825 Vcore.

Got into the OS and hell broke loose. Something like 3-4 errors, and screen res @ 800x600.

Shut it down.

Increased the Vcore to 1.850 Vcore. No change.

Dropped the Mem to 2.5,3,3,9. No go.

I was, however able to run 180X13 for almost two hours like this.

Strange. This board hates the 200FSB. The mosfets are now active cooled which is keeping the temps at 38-39 on the mb and 47 cpu. This is @ 2340Mhz and full load. 42cpu and 36-38 idle. Ambient temp is around 80F.

Anymore ideas??
Maybe get the 8RDA3+?

bad4u · Aug 6, 2003

Suggest you resolve whether it's the board or memory that's the problem.

To test board FSB limits, set multiplier to 5, and set 'Memory Speed' to run asynchronously at 60% of FSB. These settings will not stress either the CPU or memory, and will allow you to push the FSB to the limit. Use Prime95 for stability test.

To test memory limits, set memory to tight timings. Set 'Memory Speed' to run asynchronously at 120%. This will allow you to push memory frequency without stressing either CPU or board. Begin testing at 166MHz FSB (199MHz memory speed). Use Memtest86 Test#5 to speed up testing, since that's usually the most-failed test. To use Memtest Test#5 only, use key sequence C255EnterEnter.

I thought my problem was my 8RDA+, but it ran Prime95-stable at 240MHz, tested at above.

My limiter proved to be my Corsair XMS PC3200C2, which runs Memtest-stable at 2/3/3/9 timings at 207MHz FSB in synchronous mode, but requires relaxed memory timings above 207 FSB in synchronous mode.

Hope this helps!

DaveSauce · Aug 7, 2003

I would add active cooling to the northbridge, or even change the NB heatsink to an orb or something. I'm not up on the most recent 8RDA+ boards, but the older NF2 chipsets do not officially support 200mhz FSB. I'm running at 195x11 right now, and I'm doing fine. But once I get into 200+, I start to have stability problems.....just a thought.

shadowdr · Aug 15, 2003

All that voltage really isn't necessary.I run my 2100 at 1.5,which gives 1.68 in actual reading.dim set to default and run 7-3-3-2.be sure and set agp to 66 mhz.I have a raid card which is much more sensetive to hd corruptin at higher speeds and all the tweaking i have done has been to lower the voltages.excessive voltage leads to excessive heat.Any higher voltage on my 2100 leads to corruption of the on die memory,not my system memory.It still runs 200x10 at these settings on my 1.0.

8RDA+ HD corruption @200Mhz FSB?

Vampiregabe

Member

zabomb4163

Member

Vampiregabe

Member

glock19owner

Member

Vampiregabe

Member

glock19owner

Member

Vampiregabe

Member

glock19owner

Member

Vampiregabe

Member

glock19owner

Member

squeakygeek

Member

bad4u

Registered

glock19owner

Member

calvin

Member

Vampiregabe

Member

bad4u

Registered

DaveSauce

Member

shadowdr

Senior Member

Similar threads