E6600 + EVGA 680i A1 + OCZ Reaper HPC PC2-8500

[Draetor24]

I'm looking for feedback with those who have similar setups as me with their overclocking results. I'm looking for a lot of details like voltages on everything, CPU and memory clocks, testing results, etc.

 

The problem I'm facing is this...

 

I was able to get my E6600 to a 1600FSB (3.6Ghz) @ 1.44750v. I was also able to get my OCZ memory to 1200FSB @ 2.3v. I could stress test 3DMark 06 and Orthos all day and pass the tests. I was playing games like Oblivion no problem for about 1 month. Suddenly, weird things started happening like blips with my mouse cursor and then hard freezes and lockups. I immediately did more Orthos testing and was receiving errors. I then underclocked my CPU and memory and kept testing with Orthos until I was stable again. Orthos ran stable for hours, but everytime I'd launch Oblivion, it would freeze or crash immediately.

 

I reset my BIOS to defaults and eliminated all O/C'ing to test and see if I gain stability again. My motherboard kept POST'ing C1 errors, which was a memory problem. I found out that (1) of my memory sticks was causing this error, and I was able to boot with the other stick fine. I RMA'd the RAM and got new RAM back. Booted up no problem, and ran memtest for about 6-7hrs before I started to O/C again. I was able to do the same CPU O/C and this time I set the RAM a little lower at 1120FSB to make sure I wasn't overclocking it too much this time. The memory I have is rated to run at 2.3v, so I run it at 2.3v.

 

Again, Orthos and memtest are passing no problems after many tests. About 2 weeks go by and I'm getting the same problems as before. Crashing and blips with the mouse, then Orthos giving me errors within 5 seconds of running. After many reboots and BIOS changes to underclock my CPU and memory again, I get the EXACT same thing with my memory. C1 LED errors on my board, and one of my sticks is bad again. I took both of my sticks to my friend's who has the exact same board and tested them, same results. I know for sure one of my memory sticks is gone bad again!

 

I've read reports stating that it's the motherboard's memory controller that ruins memory after it's already been damaged. I'd just like to get some feedback on users who use this board and/or this setup and what kind of results or hard times they are experiencing.

[jjsmillenia]

I feel your pain.

You are correct to say that the memory controller is ruining the ram.

My quick diagnostics for your problems is that quite possibly the board is not able to run that memory at the voltage that it needs. I believe that OCZ made it with too much voltage requirement. I have the same problem.

 

I had the same setup except with a c2d e6700. When i first recieved and assembled my pc I was able to run at some blazing speeds. I am a newbie when it comes to overclocking as this was my first experience. Though there was a lack of experience i did months and months of research. I was able to run stable @ 3.8 with 1.44 vcore and 1125 @ 555,15 @ 2.35. I had my northbridge v set to 1.4.

 

It ran fine for about 4 weeks. Then all of a sudden i started getting these IRQL_NOT EQUAL OR LESS, some stop errors, something like that. (sorry for my inaccuracies because i am at work right now). Now i am desparately trying to run stable. My pc will not accept and overclocks and i must underclock my memory to 800 just to load into windows.

 

I have run memtest and Orthos all day long. I get errors in the test 8 of the memtest. I get no errors for windows memory diagnostics.

I have formatted my installed and reinstalled windows hundreds of time (very little exaggerations).

 

I am upset because a 3500 custom built pc will not function to spec, and i feel like it was a waste.

I am going to do a test. My friend has some 800mhz 240pin memory from gskill. I am going to swap mine out and put his in. He is going to do the same. He has an msi 650i board . If he has the same problems i will rma this memory or ask for a lower set with less voltage.

 

 

 

 

 

 

 

 

 

I'm looking for feedback with those who have similar setups as me with their overclocking results. I'm looking for a lot of details like voltages on everything, CPU and memory clocks, testing results, etc.

 

The problem I'm facing is this...

 

I was able to get my E6600 to a 1600FSB (3.6Ghz) @ 1.44750v. I was also able to get my OCZ memory to 1200FSB @ 2.3v. I could stress test 3DMark 06 and Orthos all day and pass the tests. I was playing games like Oblivion no problem for about 1 month. Suddenly, weird things started happening like blips with my mouse cursor and then hard freezes and lockups. I immediately did more Orthos testing and was receiving errors. I then underclocked my CPU and memory and kept testing with Orthos until I was stable again. Orthos ran stable for hours, but everytime I'd launch Oblivion, it would freeze or crash immediately.

 

I reset my BIOS to defaults and eliminated all O/C'ing to test and see if I gain stability again. My motherboard kept POST'ing C1 errors, which was a memory problem. I found out that (1) of my memory sticks was causing this error, and I was able to boot with the other stick fine. I RMA'd the RAM and got new RAM back. Booted up no problem, and ran memtest for about 6-7hrs before I started to O/C again. I was able to do the same CPU O/C and this time I set the RAM a little lower at 1120FSB to make sure I wasn't overclocking it too much this time. The memory I have is rated to run at 2.3v, so I run it at 2.3v.

 

Again, Orthos and memtest are passing no problems after many tests. About 2 weeks go by and I'm getting the same problems as before. Crashing and blips with the mouse, then Orthos giving me errors within 5 seconds of running. After many reboots and BIOS changes to underclock my CPU and memory again, I get the EXACT same thing with my memory. C1 LED errors on my board, and one of my sticks is bad again. I took both of my sticks to my friend's who has the exact same board and tested them, same results. I know for sure one of my memory sticks is gone bad again!

 

I've read reports stating that it's the motherboard's memory controller that ruins memory after it's already been damaged. I'd just like to get some feedback on users who use this board and/or this setup and what kind of results or hard times they are experiencing.

[nismozcar]

Are you running active cooling or watercooling?

[road-runner]

I have heard reports a while back that the EVGA board was a RAM killer but people that had them where saying oh its fixed now with a bios update and RMAs for newer EVGA boards etc. Evidently the problem is still not fixed?

[kendellrt]

The A1 and the T1 versions fixed that and the Quad core issue, as well as added 1333 support. Sounds like a C1E error. That can be disabled in the BIOS.

[ccokeman]

The C1 post code on your diagnostic led is a memory initialization error. It does not see the memory. Jedec spec for DDR2 is 1.8 volts. Which OCZ ram do you have because most of their stuff is only good for 2.1 to 2.2 max vdimm. If you ram 2.3 on your memory without active cooling i would say you borked it seeing how you are getting the C1 error. Try dropping the volts on the memory and using 1 stick at a time to test it out. I had a stick that was fine at 2.1v but anything higher would put it out of commision. Test 1 at a time while its warm to find the bad stick of memory.

[kendellrt]

The C1 post code on your diagnostic led is a memory initialization error. It does not see the memory. Jedec spec for DDR2 is 1.8 volts. Which OCZ ram do you have because most of their stuff is only good for 2.1 to 2.2 max vdimm. If you ram 2.3 on your memory without active cooling i would say you borked it seeing how you are getting the C1 error. Try dropping the volts on the memory and using 1 stick at a time to test it out. I had a stick that was fine at 2.1v but anything higher would put it out of commision. Test 1 at a time while its warm to find the bad stick of memory.

Ok, yeah, what he said...

[Draetor24]

The OCZ Reaper series RAM is rated to run at 2.3v, according to OCZ themselves. I've run it at that from within the BIOS. I got my new EVGA board, and I'm removing all the crappy thermal glue they use and using my own Arctic Ceramique on the NB/SB and the mosfets protecting the memory controller.

 

I sent my RAM back to NCIX, and they said it's testing fine. They weren't testing it on the same board as mine though, and the 2 EVGA 680i A1 boards I tested the RAM on both had C1 errors. I'm going to try the RAM again once I get it and test it out. If I get C1 immediately I'm going to be pissed and send it back again to get different RAM all together.

 

Even though the RAM is rated to run at 2.3v, I should be able to run it cooler without underclocking the RAM, right? I won't mind running it cooler than 2.3v for the motherboard's sake, just so I don't hurt the poor old motherboard again.

 

By the way...this post here is exactly what I was experiencing with mine. Same hardware, and same EXACT situation.

 

http://evga.com/forums/tm.asp?m=69652&...mp;key=𑀔

Edited August 8, 2007 by Draetor24

[firky]

Quick fix for the C1 error is to spam the reset butten. also the evga is very fussy with ram. i ran my e6600 @ 3.2ghz - 3.4ghz for weeks on end then out of the bule one day it wouldn't boot c1 error could not get it to boot just wasnt happing cleaned the cmos twice use one stick of my ram tryed every slot still the same put every thing back how it was switched off and unpluged for an hour come back and it work just fine ..and never did it ever again!

[Draetor24]

There's definitely something wrong with the EVGA 680i A1 board, and if you read the thread in my link above, I think that one guy solved it. OCZ also recognizes that the 680i board has problems handling 1066 RAM at higher voltages, like 2.3v. It tends to overheat, even though the RAM can take it because it was intended to, the motherboard can not. With the correct thermal conductivity, you should be able to solve the problem according to that thread. I'm just waiting for my RAM to come back so I can test. I'll probably run my RAM at 2.1v instead of the 2.3v anyways, as I don't care so much about overclocking the RAM as I do the CPU. I just don't want another board to overheat and have to wait weeks again.

 

OCZ says to run the Reaper 1066 RAM at 2.1v instead of the rated 2.3v on the EVGA 680i board, according to one of the posts in that thread.

 

By the way, when I removed the gluey thermal paste that came stock on my new replacement board, it was done terribly. It was like filmy crust all around the edges of the chip, and not ON it. The mosfets protecting the memory controller chips had some crappy tape that didn't even touch all the chips. I removed it easily with one peel, and put Arctic Ceramique on each individual chip.

 

Oh...about the C1 error. It was giving me problems where if I reboot the computer, I kept getting different LED codes each time, like 5D, 7C, 9F, C1, etc. Sometimes after rebooting, I'd get a good boot and get the FF code, then after a couple more days, just C1 constantly. I could reset the CMOS with the jumper on board, power down, turn on, reset, no matter what I did, I kept getting C1.

Edited August 8, 2007 by Draetor24

[David_L6]

Did you just go straight to 2.3 volts on your RAM or did you work up to that setting by testing?

There's no need to start with high voltage if it isn't needed. Run just as much as needed and no more no matter what voltage is supposed to be safe.

 

I have a bunch of Crucial Ballistix but I'm not running 2.2 volts on any of it. It's in three different machines and I don't remember offhand what voltage two of those machines are running at but none of them have 2.2 volts on the RAM. The third one is at default (1.8 volts) because the motherboard doesn't have many options in the bios. It has an Intel DG965WH with an E6700 and 3GB of Crucial Ballistix PC2 6400. The other two are: D975XBX2 with a QX6700 @ 3.3GHz and 3GB of Crucial Ballistix Tracer PC2 8000 running a 2:3 divider, and a 680i SLI with a Q6600 @ 3.2GHz and 2GB of Crucial Ballistix PC2 8500.

[Draetor24]

I figured the RAM should be run at 2.3v minimum since OCZ rated it at that voltage. I didn't think I'd be able to get the rated 1066FSB on them if they were under voltaged. If I can run them at their intended 1066 with less than 2.3v, I'll be happy, but I figured it was completely safe and recommended to run them at 2.3v according to the manufacturer's settings.

 

My CPU seemed to be running great at 1600FSB (400x9) @ 3.6Ghz. I have it on water, and I think my voltage was 1.4875. Temps at stock were 19C idle, and 27C load. Temps with the O/C were 32C idle, and 56C load. Threshold is 80C, but I don't like anything over 60C, so I think it's good where it's at.

 

Pretty much is just the RAM/Mobo giving me trouble. If I can clock my CPU the same and leave the RAM running at it's intended 1066 without having to underclock it, I'll be happy. I'm going to run tests with Memtest and Orthos when RAM arrives and use a lower voltage @ 1066FSB. Any recommendations on starting voltage? I'm thinking 2.1v, since that's what I read from another thread stating there were problems running the 2.3v on 680i boards. Either case, Now that I cleaned up the board's crappy factory paste and applied my own AS5 and Ceramique to the memory controller, I think the heating issue will be a whole ton better.