Jump to content
Corsair Community

TWIN2X2048-6400C5 issues with Asus P5B Deluxe in Dualmode


fubu

Recommended Posts

Hi there,

 

I am using TWIN2X2048-6400C5 in dual-mode on an Asus P5B Deluxe Wifi/Ap (Bios 804).

I experience random reboots in both XP and Vista.

Memtest doesn't find any errors.

In my tests the system was stable only using one stick (I tested both sticks seperately). Setting the timings to values posted in earlier threads here didn't help.

 

Any ideas?

Link to comment
Share on other sites

Have you set the Vdimm to 2.1v?

 

Is this the specifications of your PSU?

 

http://i41.photobucket.com/albums/e269/Ropey123/GPS-400AA-101A_1.jpg

 

Are you overclocking your CPU or Video Card?

 

Your 12V rails are 14A and 15A respectively. A bit low I would think.

Link to comment
Share on other sites

If you are not gaining errors in Memtest with single stick and dual channel, then the issue is elsewhere. What are you doing when the issues arise? Are they specific to a certain event/program/game?

 

I would look to the mainboard and PSU as possible issues. Can you remove the DRAM and other components singly and perform a testbed analysis with different components? Video Card, Hard Drive, PSU, etc?

 

Are you running on a stock cooler?

 

Download Core Temp and post your load and Idle temperatures.

 

http://www.thecoolest.zerobrains.com/CoreTemp/CoreTemp.zip

 

Your PSU is minimal for such a system. You have a 14A and 15A pull on the +12V lines. Most of the power on your system is being drawn on the +12V1 line, so it makes sense to figure out what is being drawn from this line. The likely suspect is the high-powered 7800GT VGA card, which consumes a significant amount of power. The 3.3v runs the DRAM and the +5v runs the optical and hard drives. I would not worry about them although they do draw power. I would test with a multi-meter to see if your CPU is drawing power from both +12V lines. This would have poor effects with your 7800GT VGA card.

 

7800GT

Idle Pull of the 7800GT = 133 Watts

Load Pull of the 7800GT = 236 Watts

 

http://www.anandtech.com/video/showdoc.aspx?i=2496&p=10

 

E6600 Power Consumption Prime95

 

Idle Pull of the E6600 Stock = 65 Watts

Load Pull of the E6600 Stock = 151 Watts

 

http://www.behardware.com/articles/623-9/intel-core-2-duo-test.html

 

The trend is pretty clear.

Link to comment
Share on other sites

The errors occur in CS:Source and sometimes during video encoding.

I have tried to remove any component but hdd, ram and graphics-card.

Even with this minimal configuration blue screens occur.

Also I tried to connect the PCI-Express Power on the Geforce with the PCI-Express-power-cable and with two hdd-power-connectors - no difference.

The Cooler is "in a box". I will try a Geforce 6600LE instead of the 7800GT. I have unfortunately no other PSU to test with.

I will download Coretemp and post the results later today.

The Load Pulls of the CPU and GPU are impressive, perhaps I will go and buy a new one after work.

Is this PSU good enough? http://www.be-quiet.net/be-quiet.net/index.php?StoryID=15 (BeQuiet! BQT E5-550W)

 

Thanks for helping!

Link to comment
Share on other sites

First, have you tested Memtest on your DRAM in dual channel mode? I would test with the 6600GT as that will lighten the graphics load and show if the error is coming from that direction. You could be issuing errors due to your CPU thermal output so I would like to see the Core Temp results on both Idle and load.

 

Download and run Orthos:

 

http://sp2004.fre3.com/beta/orthos_exe_20060420.cab

 

Post your idle Core Temps (with Core Temp) and your Load (Orthos running for 1 hour) Core Temps please.

 

Regarding the PSU:

 

Spezifikation: +3,3V: 30A, +5V: 30A, +12V1: 18A, +12V2: 18A, +12V1: 18A, +12V2: 18A ,-12V: 0,5A, +5VSB: 3A

 

They seem to have errored on the specifications as they are showing 2 X 12V1 and 2 X 12V2. This gives you 36A on the two rails. It should be enough, but still, you would do well to test with the 6600GT.

 

Test with the 6600GT and also Memtest with both Dimms in Dual channel mode.

Link to comment
Share on other sites

Yes, I have tested the system with memtest in dual-channel mode. No Errors. I will try gaming with the 6600 and run the orthos-test. Will post the results this evening.

 

here are the specifications for the PSU:

http://i127.photobucket.com/albums/p139/fubu1/bq.gif

I think there are 4 rails a 12V... this would give 72A?

Link to comment
Share on other sites

Hi,

 

I tested following stuff:

VDimm set to 2.1 - BSOD right after startup..

Orthos-Test: one error after 1 hour:

1:FATAL ERROR: Rounding was 0.5, expected less than 0.4

1:Hardware failure detected, consult stress.txt file.

But there is no stress.txt...

CoreTemp said the temperatures of both cores during the orthos-test were about 66°C, but Asus Probe said in this time, the temperature was 49°C. Which programm is right?

Tonight I set VDimm to 1.8V and let CS running (were usually most BSOD occur)... NO CRASH!

What do you mean? Could it be possible that the mainboard or psu delivers wrong voltages to the ram?

When I am back home I will try to test with another PSU and/or set the precharge delay to 15...

Link to comment
Share on other sites

CoreTemp said the temperatures of both cores during the orthos-test were about 66°C, but Asus Probe said in this time, the temperature was 49°C. Which programm is right?

 

I would follow Core Temp as it is the correct reading. If you are on a stock cooler then Orthos Load of 66C is about right. I would personally look to good aftermarket air cooling but I like my parts running cool.

 

Tonight I set VDimm to 1.8V and let CS running (were usually most BSOD occur)... NO CRASH!

What do you mean? Could it be possible that the mainboard or psu delivers wrong voltages to the ram?

When I am back home I will try to test with another PSU and/or set the precharge delay to 15...

 

It could well be that your board is overvolting the DRAM. Quite possible. Set the precharge as RamGuy states. He knows his stuff. Test with Orthos and the Vdimm at 1.8v. Any errors? It's good you are a tweaker and try things. You will very likely find stability on your system. :)

Link to comment
Share on other sites

Interesting - just found this thread:

http://www.houseofhelp.com/forums/showthread.php?t=54227&highlight=1.8V

I called my wife at home - memtest is now running for 8 hours without any error with Vdimm=1.8V.

Seems really that the board is giving too high voltage to the rams. Is this a defect with the board or the psu? I will check that with another psu tonight...

If it's not the psu, perhaps it is the defect with this board-revision :confused: ?

Tomorrow I will post my result.

Link to comment
Share on other sites

Hi,

 

memtest and CS:S were running for hours without any errors with Vdimm=1.8.

When I changed any memory-timings, the system becomes unstable. Setting ‘Write to Precharge delay’ from 10 to 15 didn't made any difference. Only with setting by SPD (it gave me 5-5-5-18) it was stable. But there were still errors in the orthos-test (blend-test) after one hour or so.

Unfortunately I wasn't able to test with another PSU yesterday.

But my wife (love her so much for this (but not only for this :biggrin: )) is going to change the motherboard today and buy a new PSU (be quiet! straight power 600W).

If the retailer changes the mainboard today for a new one without sending it back to ASUS I will post my new results tomorrow. Otherwise we will have to wait :(:

Link to comment
Share on other sites

Yesterday I tested with a new psu (bequiet 600w) and a changed mobo.

Results are the same as before. Only with Vdimm=1.8 and ram settings determined by spd the system seems to be stable so far. Setting a higher Vdimm the system becomes unstable and with Vdimm=2.1 it even won't boot into Windows.

I think I will go to my retailer and replace the memory-sticks.

Link to comment
Share on other sites

Okay, here we go.

Yesterday the new memory-sticks arrived and - what a surprise - the system is stable now. Tested 3 hours with orthos blend-test, played games, let memtest run - no errors or crashes anymore.

Here the summary:

Old ram displayed no errors in memtest but crashes in games and failure in orthos test after an hour. Both sticks individually ran fine, but in dualchannel-mode the errors occur. The only halfway stable setting was Vdimm=1.8V and timings by SPD(5-5-5-18).

 

Something that I noticed about the new memory-sticks: although it is the same revision like the old one (5.1) the sticks look different. The outer border of the black cooling unit is matt on the new sticks, the old ones had shiny borders...

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...