Unstable RAM on 680i

NachT

NachT

NachT

Associate
Joined
20 May 2006
Posts
1,068
I cant seem to get my RAM stable on the evga 680i.

My voltages are:

CPU - 1.38v
CPU FSB - 1.4v
nForce SPP - 1.25v
nForce MCP - 1.5v
HT nForce SPP <-> MCP - 1.2v
RAM - 2.2v

Are these voltages ok, do I need more volts on any of these?

My RAM is the Crucial 10th anniversary 4x 1GB @ 750mhz 5-5-5-12

I've slackened the timings this much just to try and get a pass on memtest, but still I'm getting errors... bit confused as to why they are still throwing up an error. Could the modules be failing? I have no spare DDR2 RAM to test in the board ( might be the motherboard? )

I've updated the BIOS to the latest one, still get an error in memtest at about 170% into the test.


What could I try? ( not sure if those SPP / MCP things have enough voltage or even if they're related to RAM.. )
 
Test each stick on its own.

Try upping the MCH a bit to boost it for running 4 sticks.
 
Cob said:
Test each stick on its own.

Try upping the MCH a bit to boost it for running 4 sticks.
Click to expand...

Do you mean MCP ? What is the MCP also, is it the south bridge ? And what would be a safe max for voltage on it?

I'll try what you've suggested, thanks :)
 
Yip MCP. It's the northbridge/Memory Controller Hub.

I don't know much about the nVidia voltages tbh, but up it by 0.025v-0.05v.
 
Cupra said:
4gb in 4x1gb on 680i can be dodgy get 2x2gb stick all will be fine
Click to expand...
can certainly agree with that. the memory i have in sig (4x1gb) wont work on my board at all. had to drop down to 2gb of pc5300 yet the strange thing is a friend tested my team stuff in his system and it worked fine:confused:
 
Last edited:
I've not had a problem up until recently with running the 4x 1GB sticks.

Before I left for work this morning I took 2 sticks out and set memtest away, using the same voltages as I originally posted. I got 2 errors at 250% through the test.

Going to try upping the voltages to the ones Reality mentioned and re-run the test.

If that fails, I'll either try some new RAM or take my RAM to a mates house and run the tests on his board.
 
Just re-ran memtest using 2 sticks in place, and upped all the voltages. About 20 mins into the test the PC just reset it's self. I take that as an error, so...

I guess 1 or more sticks are faulty, I'll run memtest on each stick by itself to find out what ones are faulty.
 
Little update:

As the PC reset when I upped the voltages to the ones Reality suggested, that could have been because of something over-heating right?

So I put the voltages back to the ones I was using originally and ran memtest with only 1 stick of RAM in place, 3000% into the test and not 1 error or anything.

So... I'm really confused now..

2 or more sticks and I get errors, 1 stick is fine. I have 4 sticks...
 
Update:

The RAM works fine now. Yeah... 6 hours 20 mins into memtest and not 1 error. It's a good thing, but don't know how its ok now. lol

Can I ask if anyone knows what one of these is the south bridge and north bridge?

nForce SPP - 1.25v
nForce MCP - 1.5v
HT nForce SPP <-> MCP - 1.2v

And if anyone would kindly explain what SPP and MCP stands for / what it is and does. And I just take it as the 'HT nForce SPP <-> MCP' is just the bridge between the two, like a bus?

I've searched but couldn't find explanations. I'm just one of those people who like to know what everything stands etc ^^

Thanks guys, btw I think when I put the voltages back down in the BIOS I left them a little bit higher than what they were. I guess that was the problem..

EDIT: Is it also possible for Memtest to start using the page file / virtual RAM at all? Or does it purely just test the physical RAM ?
 
You must log in or register to reply here.
Back
Top Bottom