Crucial Ballistix 2x1GB PC4000 kit gone bad (merged)

[jhyunl]

arrg, I bought these craping shiznit Crucial DDR500 on newegg 10/19/06 without knowing all these horror stories. Now, mine start showing the crazy window restart and random freeze in 3D games. More vdimm needed for stability, and start failing at memtest even under stock setting.

 

You know what? this time newegg put a $30 rebate on it for a month deadline, and my guess is RMAing won't be easy at all this time.

 

These sh%t sticks now fail at every single memtest86 at stock setting with no reasons. Its been just a week !! god damn it.

 

And, funny thing is I dont even own a DFI. I am an Abit nf8-v2 (754) and sempron palermo 2800 owner.

 

I am craping scared to go through all of these shiznit.....:mad:

[i_wheeldon]

haha, you really shouldnt have the TD sticks no more, does it say on the label/rapper TD or TD1?

 

If there TD's newwegg shouldnt be giving them out now. If there TD1's, god knows, maybe these ones dont like Abit boards.

[jhyunl]

I dont know what is the exact number yet. when i go home, I'll check it out.

 

These two sticks of memory modules on nf3 754 with OCed sempron palermo 2800+ are ridiculous. It doesn't have dual-channel issues, as 754 supports only single channel. I do never increased the vdimm over 2.8v, never ever, because my NF8-V2 only supports 2.8 as max vdimm.

 

Whats so funny is that windows memory diganotics report them to work 3-3-3-8, 1T, 2.6-2.8v working fine at 250FSB. But, not memtest. 3-4-4-8 1t works for memtest86, but FSBs any other than exact 250 will fail the test. When the sempron runs at stock so the memory is underclocked to 200, it fails memtest86 with no exceptions, I can't even play around FSB and HTT to get the maximum OC potential of sempron because memory fails whatever I do.

 

THE BIGGEST PROBLEM IS THAT THEY START SHOWING THE SYMPTOMS OF ERRORS IN MEMORY IN PRACTICAL APPLICATIONS, LIKE WINDOWS AND 3-D GAMING!!!! of course without failing at memtest86. This is very scary because the error will likely to be unknown whatever I would test.

 

Arrgg, this Crucial memory is the worst nightmare in my years of PC enthusiasm. And, I find my self-esteem dropping down because I fell for the cheap-arse Newegg's clearance marketing. They crapking lied and put a protection to save their a55 from their victims' RMAing by putting cheap 30bucks rebate. No refund or direct RMAing from Newegg if the rebate has been filed.

[i_wheeldon]

You've been done youth there, thats bad for newegg, Im in the UK and can send them right back to crucial on a 48 hour turn around, Pretty amazing.

 

Have you not contacted newegg though, i've only heard good things about them really, im sure they can sort it out for you. What your describing is the classic symptoms of the TD's

[jhyunl]

I haven't contacted newegg yet or crucial either. Because I am nervous but not confident enough to tell that they are really freaking bastards.

 

I was so anxious after reading about 75pages of discussions here, so I came back home from school early--Its friday... I find myself being impatient, but oh well, this must be done quick to avoid any long-term problems like RMAing 4-5 times unnecessarily.

 

The model batch number is CL11156X.4C, and of course these are z53 not L55. And, it is TD1 not TD. So, it is supposedly to have fixed those scary problems reported in this post. Then why is this happening?

 

Ok, I figured out that 754 memory controller on palermo may not be as efficient as 939, and the chipset on my crap nf3 764 board may not be as good as advanced models to cope with the large amount of 2G memory. So, I decided to test each stick individually to see if less number of modules can help to lessen the symptoms. There is also a higher chance of one module may have gone bad since last few days.

 

I put one of them into memtest86 before I head to school this moring. The setting as far as my board allows was 3-4-4-8,1T,2.6v.@DDR500. So, my sempron was working at 2.0G 250FSB 1:1 HTT multi 3 and a CPU multiplier of 8x-fixed. When I am back after 4 hours or so to check it out, it was OK.

 

Now, I set it to 240FSB, 3-4-4-8, 1T 2.6v to see if FSB other than 250 so underclocking can work when a single 1G memory is used. It passed the 1st sweep in memtest86 so far...... This never happened when both sticks were put together, btw.

 

I'll test overclocking the stick to about 260-270ish with 2.6v soon for an hour of memtest86 (although I know this isnt enough but time is running out). If these short tests are sucessful, I'll replace the current 1G with the other one to see if it performs likewise.

 

If they work fine individually and over- and underclocks well, I have no choice but to blame the combo of palermo and old design of cheap Abit nf3 board. And, I may have to keep these ballistix for now until they wear out its reliability someday. I can think of what I should then once if that happens. For now, I just hope each stick work separately.:sad:

[i_wheeldon]

Try them in different slots, move them around, your not using duel channel so its doesnt matter where you put them really.

 

But as you say, old 754 socket and old mobo are probably your problem

 

Make a sig and people will be able help you more

[jhyunl]

I haven't contacted newegg yet or crucial either. Because I am nervous but not confident enough to tell that they are really freaking bastards.

 

I was so anxious after reading about 75pages of discussions here, so I came back home from school early--Its friday... I find myself being impatient, but oh well, this must be done quick to avoid any long-term problems like RMAing 4-5 times unnecessarily.

 

The model batch number is CL11156X.4C, and of course these are z53 not L55. And, it is TD1 not TD. So, it is supposedly to have fixed those scary problems reported in this post. Then why is this happening?

 

Ok, I figured out that 754 memory controller on palermo may not be as efficient as 939, and the chipset on my crap nf3 764 board may not be as good as advanced models to cope with the large amount of 2G memory. So, I decided to test each stick individually to see if less number of modules can help to lessen the symptoms. There is also a higher chance of one module may have gone bad since last few days.

 

I put one of them into memtest86 before I head to school this moring. The setting as far as my board allows was 3-4-4-8,1T,2.6v.@DDR500. So, my sempron was working at 2.0G 250FSB 1:1 HTT multi 3 and a CPU multiplier of 8x-fixed. When I am back after 4 hours or so to check it out, it was OK.

 

Now, I set it to 240FSB, 3-4-4-8, 1T 2.6v to see if FSB other than 250 so underclocking can work when a single 1G memory is used. It passed the 1st sweep in memtest86 so far...... This never happened when both sticks were put together, btw.

 

I'll test overclocking the stick to about 260-270ish with 2.6v soon for an hour of memtest86 (although I know this isnt enough but time is running out). If these short tests are sucessful, I'll replace the current 1G with the other one to see if it performs likewise.

 

If they work fine individually and over- and underclocks well, I have no choice but to blame the combo of palermo and old design of cheap Abit nf3 board. And, I may have to keep these ballistix for now until they wear out its reliability someday. I can think of what I should then once if that happens. For now, I just hope each stick work separately.:sad:

[faheyd]

I haven't contacted newegg yet or crucial either. Because I am nervous but not confident enough to tell that they are really freaking bastards.

 

I was so anxious after reading about 75pages of discussions here, so I came back home from school early--Its friday... I find myself being impatient, but oh well, this must be done quick to avoid any long-term problems like RMAing 4-5 times unnecessarily.

 

The model batch number is CL11156X.4C, and of course these are z53 not L55. And, it is TD1 not TD. So, it is supposedly to have fixed those scary problems reported in this post. Then why is this happening?

 

Ok, I figured out that 754 memory controller on palermo may not be as efficient as 939, and the chipset on my crap nf3 764 board may not be as good as advanced models to cope with the large amount of 2G memory. So, I decided to test each stick individually to see if less number of modules can help to lessen the symptoms. There is also a higher chance of one module may have gone bad since last few days.

 

I put one of them into memtest86 before I head to school this moring. The setting as far as my board allows was 3-4-4-8,1T,2.6v.@DDR500. So, my sempron was working at 2.0G 250FSB 1:1 HTT multi 3 and a CPU multiplier of 8x-fixed. When I am back after 4 hours or so to check it out, it was OK.

 

Now, I set it to 240FSB, 3-4-4-8, 1T 2.6v to see if FSB other than 250 so underclocking can work when a single 1G memory is used. It passed the 1st sweep in memtest86 so far...... This never happened when both sticks were put together, btw.

 

I'll test overclocking the stick to about 260-270ish with 2.6v soon for an hour of memtest86 (although I know this isnt enough but time is running out). If these short tests are sucessful, I'll replace the current 1G with the other one to see if it performs likewise.

 

If they work fine individually and over- and underclocks well, I have no choice but to blame the combo of palermo and old design of cheap Abit nf3 board. And, I may have to keep these ballistix for now until they wear out its reliability someday. I can think of what I should then once if that happens. For now, I just hope each stick work separately.:sad:

 

I know you are in panic mode, but I bought mine AFTER reading 'most' of this thread. So, don't think you did something bad. I can only tell you mine are working fine, and I haven't even really tried to OC them yet. Good luck on your tests, you are doing the right thing by individually testing them.

[jhyunl]

The stick passed the first sweep of memtest86 on @DDR400, 2-2-2-5, 1T, 2.6v. This is getting interesting. Because I got tons of errors in the first sweep with two sticks under every combination of CAS 2 or 2.5 setting. At least so far, this stick appears working for good. wth?

[i_wheeldon]

It could be your modo or sempron processor letting the memory down, Especially the sempron, its a poor processor at best!!!

 

Your using top end DDR1 memory with below average MOBO & Processor, 2GB running 250Hz on that pair will most likly cause errors, Have you tested them at stock 200hz, basic??? See if that works?

[jhyunl]

I know. The only reason why I bought these ballistix kit was they popped up when I was looking around 2 x 512kit at Newegg. Those popular 2x512kit XMSes keep their price range around $120-150 if they are good ones. The ballistix 2G kit suddenly came out in a few weeks ago, and it was priced $189 after $30 mailin rebate. I was simply tempt to buy the 2G one by paying about 30-40 more bucks. Thought was a good deal then (and it must be if these two sticks of ballistix turn out to be OK). And, obviously, the ballistix would do a job anyway even if I throw one of those 2 sticks because they dont like to work together, and the performance won't be different from say from 2x512MB corsair XMSs DDR400 running with CAS 2 or 2.5. Believe me. Around then, my decision was the best because I did not hear anything about TD ballistix.

 

And, I did some research before buying this palermo core sempron expecting its 64bit memory controller (venice) should be able to deal with 2G sticks with no problem in single channel mode of course. I know lots of people suffer from the limitation of socket 754 in controlling more than 3 sticks of ddr ram, but I thought i was gonna be OK when they are 2x1G of double-sided. And, I saw many people report 2G of memory working OK on 754 nf3s. So, I was very certain that the ballistix kit must work as well if they are really quality memory (althogh I sensed something funky with the too good price....)

 

And, indeed when I first pop the sempron into the nf3 board, and OCed to 2.4, the performance was just a wonder ! Everything was very responsive, and totally different from my old P4 2.4c. But, I didnt realize something is not working properly.

 

Now I guess the memory controller on sempron may be the reason why these 2G sticks are not cooperating each other. And, memtest seems OK for now after a few sweeps of passes when plugged together.

 

Btw, each stick alone appears working fine upon these settings (not extensive test thought, just about 2 test passes for each).

 

200FSB-DDR400, CPU runs @2.4Ghz, 2-2-2-5, 1T, 2 passes

250FSB-DDR500, CPU runs @2.4Ghz, 3-4-4-8, 1T, 2 passes

267FSB-DDR5XX, CPU runs @2.51Ghz, 3-4-4-8, 1T, failed at first sweep

 

For the 200 and 250FSBs, vdimm 2.6 was just fine. For the 267FSB, vdimm increase to 2.65/2.70/2.80 did not help. They fail anyway. Over 2.8v, I dont know. My board is limited in adding vdimm.

 

So, it seems my batch does not like FSB over manufactures spec. But, if these truly work at least for 250FSB with no problems, I guess i may keep these two sticks.

 

OF course, they need to be tested during this weekend... And, I think there may be a high chance of me buying 939 venice or opteron and replace my sempron if the ballistix kit only works for more advance memory controller. Damn, what a waste of money.

 

Oh, the two sticks passed 3 sweeps of memtest86. I guess when it finished 4th sweep, I would go for overall stability test, and try some 3-d gamings and memory intensive applications. I guess it is likely to fail because it did last time after passing memtest.

[jhyunl]

I know. These ballistix kit just popped up when I was looking around 2 x 512kit at Newegg. Those popular 2x512 XMSes keep their price around $120-150 if they are good ones. And then, the ballistix 2G kit suddenly came out in a few weeks ago, and it was priced $189 after $30 mailin rebate. I was simply tempt to buy the 2G rather than 1G by paying about 30-40 more bucks. Thought was a good deal then (and it must be, if these two sticks of ballistix turn out to be OK). And, obviously, the ballistix would do a job anyway even if I throw one of those 2 sticks because they dont like to work together, and the performance won't be different from say from 2x512MB corsair XMSs DDR400 running with CAS 2 or 2.5. Believe me. Around then, my decision was the best because I did not hear anything about TD ballistix.

 

And, I did some research before buying this palermo core sempron expecting its 64bit memory controller (venice) should be able to deal with 2G sticks with no problem in single channel mode of course. I know lots of people suffer from the limitation of socket 754 in controlling more than 3 sticks of ddr ram, but I thought i was gonna be OK when they are 2x1G of double-sided. And, I saw many people report 2G of memory working OK on 754 nf3s. So, I was very certain that the ballistix kit must work as well if they are really quality memory (althogh I sensed something funky with the too good price....)

 

And, indeed when I first pop the sempron into the nf3 board, and OCed to 2.4, the performance was just a wonder ! Everything was very responsive, and totally different from my old P4 2.4c. But, I didnt realize something is not working properly.

 

Now I guess the memory controller on sempron may be the reason why these 2G sticks are not cooperating each other. And, memtest seems OK for now after a few sweeps of passes when plugged together.

 

Btw, each stick alone appears working fine upon these settings (not extensive test thought, just about 2 test passes for each).

 

200FSB-DDR400, CPU runs @2.4Ghz, 2-2-2-5, 1T, 2 passes

250FSB-DDR500, CPU runs @2.4Ghz, 3-4-4-8, 1T, 2 passes

267FSB-DDR5XX, CPU runs @2.51Ghz, 3-4-4-8, 1T, both sticks failed at first sweep

 

For the 200 and 250FSBs, vdimm 2.6 was just fine. For the 267FSB, vdimm increase to 2.65/2.70/2.80 did not help. They fail anyway. Over 2.8v, I dont know. My board is limited in adding vdimm.

 

So, it seems my batch does not like FSB over manufactures spec. But, if these truly work at least for 250FSB with no problems, I guess i may keep these two sticks.

 

I will visit this thread again to hear from you guys again soon.