Jump to content
Sign in to follow this  
jhbodle

Crucial Ballistix 2x1GB PC4000 kit gone bad (merged)

Recommended Posts

Me too!

 

The plot thickens. BIOS 201 seems to give slightly more errors than 218 but the memtest screen I have attached shows a strange phenomenon. Note how the screen shows that memtest is on test number 4. Now here's the weird part: the red line showing errors in test 2 is still incrementing upwards even thorough tests 3 and 4. This didn't happen on BIOS 218 (and this is the first time its happened on 201 too).

 

Maybe it's because I had to put a copy of Memtest 1.51 on a floppy and boot from that with the 201 BIOS as 218 had it built in. Either way, I don't have a clue how I can get memtest and my DFI board to like my Ballistix RAM.

 

Man that screen capture is not showing the settings I am using. I never run down that low so mine stays set as I showed earlier; ALL the time. These boards seem to like it looser than what the memory specs show and if it is not loose the errors come, but the looser settings give me room to GO on UP and get past any give-away from loose. It is a balancing act that if not undestood and followed to the letter almost with very minor vaiables then all bets are really off. Just how I see it. And then as always with putering>YMMV.

 

RGone...

Share this post


Link to post
Share on other sites

I was just trying lots of different settings trying to experiment. I have attached a screenshot I have just run using the exact settings you suggested in this thread. Errors in tests 3,4,5,6 and 7. Let me know what you think.

Share this post


Link to post
Share on other sites
I'm going to flash bios 201 now, I think is the only one having not tried. As your settings with 218 bios it "works" for me, but gives errors in test 4, 6 and also sometimes in test 3. What I noticed is even when running 250x9, sometimes the board reboots after IRQ checking, maybe 1 of 10 times. With 250x10 always reboots.

 

I hope a new bios is Ballistix friendly

 

YEAH! We (ballistix users) have to join forces and hope for the best ;)

 

Thanks all

Share this post


Link to post
Share on other sites
I was just trying lots of different settings trying to experiment. I have attached a screenshot I have just run using the exact settings you suggested in this thread. Errors in tests 3,4,5,6 and 7. Let me know what you think.

JH, man I have no idea unless you really plainly have some bad memory. You have no "idea" how badly beatup this Crucial PC3200 stuff I have is. I beat the hale out of it on the NF3 after I swapped it to AG for my PC4000 stuff.

 

I have run 3.3 volts on it and run it tight and loose and just beat the Pee out of it and yet it runs better apparently than yours. That is strange to me all around as the only reason I was fooling with the stuff up that high was to test it for you guys. I am not overly found of the stuff and like my TCCD better, but to see yours just wither on the vine makes me wonder.

 

Maybe put 3.2Volts on the memory instead of the 3.1V and see if it helps.

 

Dang I cannot remember what the symptoms are for that hitting on every test nearly with an error. I have seen it before on the Nforce boards and it is a setting that needs to be changed. See each and every board acts differently. And that is not really 'accurate'> each memory controller in the cpu itself acts somewhat differently and then couple that with a board it is plugged into and things just get hard to fathom.

 

After returning from DFI, CA> I have done nothing but beat up this Ultra-D RevA02 board I have and right now have a 939 sempron in and got the Corsair XMS Expert in that many have trouble with and loosened up like I normally run>>well h*ll I am running 285FSB on memory that hardly wanted to get past 275FSB before. But I am running some "really" off the wall timings that everyone leaves alone because they follow the rules or guidelines laid out. Hehehehee. Not me man, I am off in never never land it pays dividends in my board getting off its' hiney and hauling b utt.

 

I get thru playing with the TCCD, I may revisit the Ballisitix stuff in a different board and processor and see if I get an 'entirely' different setup and if so will let you guys know. But right now I am having way too much fun running the Pee out of this once not so fast memory that is flying now. Installed winDoors at 285x9 last night for a lark just to see how it would all go as the install of winDoors has to be the hardest thing to do 'well' and not have errors. But these new off the wall settings. Well they rok for me.

 

RGone...

Share this post


Link to post
Share on other sites

I have given up on my 2 x 512 of PC4000 Ballistix CL2.5 (BL6464Z505). I had the incrementing upwards errors on my first 2 sticks and the second pair just kept failing MemTest 86+ if left running for any length of time.

 

I have swapped them at Overclockers UK for the second time with no problems and have now got 1GB (2 x 512MB) OCZ PC4000 Dual Channel Gold Series EL-DDR CAS2.5 Revision 2.

 

I will try and see what I can get out of these.

 

Cheers, ProSpec

 

P.S. good luck to anyone stiil trying to get their Ballistix to run ok, maybe it only needs a Bios revision to get them up and away!

Share this post


Link to post
Share on other sites
Dang I cannot remember what the symptoms are for that hitting on every test nearly with an error. I have seen it before on the Nforce boards and it is a setting that needs to be changed. See each and every board acts differently. And that is not really 'accurate'> each memory controller in the cpu itself acts somewhat differently and then couple that with a board it is plugged into and things just get hard to fathom.

RGone...

 

Thanks for trying RGone. Could you (or Angry Games) let the DFI BIOS programmers know about this issue (apparently others seem to be having it too) or are you certain that the BIOS is not the problem? I only ask because the only thing I have changed is the Asus motherboard, which means:

 

-The RAM is OK

-The CPU is OK

-The nForce4 chipset can handle Ballistix OK

 

 

This can only leave the DFI BIOS, right? Feel free to correct me if i'm wrong...

Share this post


Link to post
Share on other sites

After a LOT of testing and configuration, I think i've found a way to get Crucial Ballistix RAM to work with DFI boards without issue. Until a BIOS is released that makes Crucial Ballistix RAM more compatible, this is the best option. The results speak for themselves:

 

2x Crucial Ballistix 512MB [email protected] 2-2-2-6, 2.7v on DFI Lanparty UT nF4 SLI-D

 

MemTest stable and Windows Memory Diagnostic stable!

 

PICT0037.JPG

 

Use the exact settings below and let me know if they work for you.

 

- Flash to BIOS 201 and clear CMOS settings

- Install Crucial Ballistix in YELLOW MEMORY SLOTS

- Plug power cables into BOTH the Molex and Floppy power connectors on the motherboard

 

Power up the machine and enter the BIOS. Set your CPU to run at stock speeds and stock HTT speeds, then enter the settings below. Anything I don't specify, leave at default or tune at will.

 

* STANDARD CMOS FEATURES:

- Halt On: No Errors

 

* ADVANCED CHIPSET FEATURES:

Everything DISABLED except:

- SSE/SSE2 Instructions: Enabled

MAC PHY IEEE Test: BypassSET

 

* GENIE BIOS SETTING:

- FSB Bus Frequency: 200

- LDT/FSB Frequency Ratio: 4.0

- LDT Bus Transfer Width: down 16 up 16

- CPU/FSB Frequency Ratio: [your CPU's default multiplier]

- PCI eXpress Frequency: 101MHz [possibly 105MHz for ATI cards]

- K8 Cool n Quiet Support: Auto

- Cool n Quiet MAX FID: Auto

- CPU VID Startup Value: 1.425v [for 90nm CPUs only - experiment with 130nm]

 

- CPU VID Control: 1.425v [as above]

- CPU VID Special Control: Above VID * 110%

- LDT Voltage Control: 1.40v

- Chip Set Voltage Control: 1.80v

- DRAM Voltage Control: 2.70v :D [higher causes errors on my setup]

 

* DRAM CONFIGURATION:

- DRAM Frequency Set: 200 (DRAM/FSB:1/01)

- Command Per Clock: Enable

- CAS Latency Control: 2.0

- RAS to CAS Delay: 2

- Min RAS active time: 6

- Row Precharge time: 2

- Row Cycle Time: 7

- Row Refresh Cyc Time: 14

- Row to Row Delay: 5

- Write Recovery Time: 3

- Write to Read Delay: 2

- Read to Write Delay: 5

- Refresh Period: 3120

- Write CAS Latency: 5

- DRAM Bank Interleave: Disabled

 

- DQS Skew Control: Increase Skew

- DQS Skew Value: 0

- DQS Skew Control: Level 6

- DRAM Data Drive Strength: Level 4

- Max Async Latency: 7.0

- Read Preamble Time: 5.0

- IdleCycle Limit: 256 Clocks

- Dynamic Counter: Disable

- R/W Queue Bypass: 16x

- Bypass Max: 7x

- 32 Byte Granularity: Disable (8 Bursts)

 

 

Let me know if it works for you! I will keep updating this thread with more aggressive timings as I find them - watch this space!

Share this post


Link to post
Share on other sites
Also I have a question for you, how do you run 2x Leadtek's 6800GT in SLI? The Heatsinks is massive and I suppose you modded them or replaced with another heatsink.

 

Sorry mate, only just saw this! I used the original heatsinks - all I had to do to make them fit is remove the black plastic shroud from each card. It's a tight fit, but it works!

 

By the way guys, check out this thread I just made for a possible solution to the Ballistix problems:

 

http://www.dfi-street.com/forum/showthread.php?t=6719

Share this post


Link to post
Share on other sites
That's right. Just one problem though - his settings don't work for me, or some other people either. He's also running looser timings.

I think you should experiment with lower multis and more fsb. The Ballistix I'm running now 244 mhz 1:1 2.5-2-2-7 with 2.7v, multi @ 9, htt 4x, vcore 1.275v. With more voltage and same timmings should do 250 / 255 easy. That is stable with all benchs for days on. I'm using bios 218, it is the same mhz that you're running but more performance.

 

When using multi 10x I have to run the mem async 180/200 divider, 166/200 doesn't work :confused: , for example 260x10, mem @ 234mhz 2.5-2-2-6, 2.7v. OK the Ballistix doesn't run perfect atm but hopefully it is quite usable.

Share this post


Link to post
Share on other sites
I think you should experiment with lower multis and more fsb. The Ballistix I'm running now 244 mhz 1:1 2.5-2-2-7 with 2.7v, multi @ 9, htt 4x, vcore 1.275v. With more voltage and same timmings should do 250 / 255 easy. That is stable with all benchs for days on. I'm using bios 218, it is the same mhz that you're running but more performance.

 

When using multi 10x I have to run the mem async 180/200 divider, 166/200 doesn't work :confused: , for example 260x10, mem @ 234mhz 2.5-2-2-6, 2.7v. OK the Ballistix doesn't run perfect atm but hopefully it is quite usable.

 

Hmm, you could be right. Actually, now I've finally got the memory stable at stock speeds, I can start to overclock it and see how high it will go. My aim is 233MHZ 2.5-2-2-7 cos I got that easily on my old Asus board.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
Sign in to follow this  

×