The Corsair User Forums  

Go Back   The Corsair User Forums > Corsair Product Discussion > Memory

Reply
 
Thread Tools Search this Thread Rate Thread Display Modes
  #106  
Old 11-02-2004, 11:45 AM
Wired's Avatar
Wired Wired is offline
Administrator
Wired's PC Specs
 
Join Date: Aug 2000
Posts: 32,188
POST ID # = 152219
Wired Reputation: 32
Default

::looks at watch::

45 min early? Damn. Go back to sleep man! :)

Wait a minute... I don't WEAR a watch!!! HUH?!?!?!?!
__________________
No, I don't work for Corsair
...you need to flow like a river...
Forum Rules
Admin of this forum / webdesignforums.net / Petri's IT Forum / The Corsair Memory Steam Group
Founder/Creator/Admin of ZE SECRET PROJECT (Coming Soon)
Reply With Quote


  #107  
Old 11-02-2004, 03:27 PM
herman's Avatar
herman herman is offline
Registered User
 
Join Date: Dec 2003
Posts: 60
POST ID # = 152311
herman Reputation: 10
Default

Bump for tech guy on phone

k he found it now

Last edited by herman; 11-02-2004 at 03:37 PM.
Reply With Quote


  #108  
Old 11-02-2004, 04:11 PM
Wired's Avatar
Wired Wired is offline
Administrator
Wired's PC Specs
 
Join Date: Aug 2000
Posts: 32,188
POST ID # = 152329
Wired Reputation: 32
Default

Please update the thread with the results or anything you find out.
__________________
No, I don't work for Corsair
...you need to flow like a river...
Forum Rules
Admin of this forum / webdesignforums.net / Petri's IT Forum / The Corsair Memory Steam Group
Founder/Creator/Admin of ZE SECRET PROJECT (Coming Soon)
Reply With Quote


  #109  
Old 11-02-2004, 04:33 PM
herman's Avatar
herman herman is offline
Registered User
 
Join Date: Dec 2003
Posts: 60
POST ID # = 152343
herman Reputation: 10
Default

Nothing new. He said he would have customer service call me.

One thing he also said that the mem temp should be above "system" temp, where I think he means chipset.

He also said failing test five can be cause by chipset or cpu getting too hot. I kind of doubt that in my case, as even in prime95 running in-place fft power draw torture test for full run (16 mins.), CPU only gets to like 49, and chipset to 32. They get steady in about 6-7 mins. Prime95 blend test, that tests all the mem, gets them up to like 47 and 31 best I can remember.

I never did get a good answer about what max temp was, excpet for the more than system temp comment, or what temp they get to during QA testing. He assured me that, in my case at least, I did not get the same mem back during my last RMA.

We talked more, but most of it was asking what I tested, trying to find the thread, and waiting for customer service.


So I'm still in limbo.

Any of you RamGuys or Admins have a way to get answers from engineering, if you have that, it sure would be nice. Specifically, how high temp they should work, have they ever validated on Neo2, and stuff like that.

If someone says 60C or more, I'll probably just chill. No pun intended. :-)
Reply With Quote


  #110  
Old 11-02-2004, 04:59 PM
Wired's Avatar
Wired Wired is offline
Administrator
Wired's PC Specs
 
Join Date: Aug 2000
Posts: 32,188
POST ID # = 152348
Wired Reputation: 32
Default

Ram Guy is the only official Corsair employee here. I can poke him with a stick though :)
__________________
No, I don't work for Corsair
...you need to flow like a river...
Forum Rules
Admin of this forum / webdesignforums.net / Petri's IT Forum / The Corsair Memory Steam Group
Founder/Creator/Admin of ZE SECRET PROJECT (Coming Soon)
Reply With Quote


  #111  
Old 11-03-2004, 05:03 PM
herman's Avatar
herman herman is offline
Registered User
 
Join Date: Dec 2003
Posts: 60
POST ID # = 152519
herman Reputation: 10
Default

I have sent the new mem back now. I'll let you know what happens when I get the new mem.

I have a few more observations after thinking about and comparing all the various sticks I have had over the last two years.

Yes, more herman rambling, but please take consideration on this.

------

For the XL, the clips seem to cause the spreaders to start to "toe-out" from the bottom of the stick over time. This seems to be a function of the "frag tape" and the clip design, where the clip applies more pressure at the top. The ideal situation would look like this:

|..||..|
.|!||!|.
.|!||!|.
|..||..|
...||...

Where "|" is the speader or board/chip, the "!" is the frag tape bonding with the spreader and the chip, and "." is air. It looks like over time the frag tape starts to seperate at the bottom, and in some cases nearly the whole chip, depending on variaces in a given speader, board and final chip hight, where some sticks often end up with a bond that looks more like this:

|..||..|
.|!||!|.
.|.||.|.
|..||..|
...||...

And in some cases on some chips on some sticks like this, where only the edge of the chip is really touching:

|..||..|
.|"||"|.
.|.||.|.
|..||..|
...||...

It doesn't take much, especially with frag tape to cause a separation over time.

It looks like there are at least two different clip versions on the XLs, where on one version both sides of the clip are more circular at the top. I am assuming this is in an attempt to make the clip "miss" the protruding top of the spreader better and keep this from happening. The black XLs had this, and the PTs had the more streight design. In practice i'm not sure this works, as one of the black sticks, coincidentally the "bad" one, had a considerable amount of toe, and noticable gaps if you shine like one of those really bright led lights. There also seems variance in placement of the speader in that the spreader comes *very* close, if not touching the components at the top of the stick on the XLs. I know that once you stick the tape, you really don't want to ever restick it, so its not like you can start over.

I don't know if this is has anything to do with my problems. But I would think that it is in the reallm of possiblity that maybe a few of the chips could have over heated, and just stated flaking-out after that, as a gap can be considerably wose than no speader at all and make like a little chip oven. The chips themselves should internally get pretty hot almost immediately, which I would only see maybe minuites later as the whole speader gets hot. For chips with marginal contact, the temp could be a really good percentage higher than what I see.

-----

The spreader design you have has been around for a long time it seems. It may be time to give it an update to something more like, or better than, what I have seen from places liek OCZ and Ballistix, where ther is what looks like a more rigid steel clip that can apply pressure more evenly down the short vertical side of the spreader.

This becomes more crucial as these get hotter, and the XLs may be the first to really start having an issue in this situation.

Last edited by herman; 11-03-2004 at 06:06 PM.
Reply With Quote


  #112  
Old 11-03-2004, 05:24 PM
RAM GUY's Avatar
RAM GUY RAM GUY is offline
Corsair Product Guru
 
Join Date: Apr 2001
Location: Fremont, CA
Posts: 118,268
POST ID # = 152527
RAM GUY Reputation: 10
Default

Thanks for your comments, I will pass them on. However, we have looked at many different modules and heat spreaders and found that most of the rigid designs do not make contact with all the IC's just FYI. But we have talked about changing the H/S and I will let you know if that changes!
And please do let me know when you get the new modules!
__________________
Support accounts and tickets can be created at https://support.corsair.com.
Reply With Quote


  #113  
Old 11-03-2004, 05:29 PM
herman's Avatar
herman herman is offline
Registered User
 
Join Date: Dec 2003
Posts: 60
POST ID # = 152530
herman Reputation: 10
Default

Much thanks for all of your patience.
Reply With Quote


  #114  
Old 11-03-2004, 05:40 PM
RAM GUY's Avatar
RAM GUY RAM GUY is offline
Corsair Product Guru
 
Join Date: Apr 2001
Location: Fremont, CA
Posts: 118,268
POST ID # = 152533
RAM GUY Reputation: 10
Default

Not a problem, I am just sorry that you have had trouble with our products!
__________________
Support accounts and tickets can be created at https://support.corsair.com.
Reply With Quote


  #115  
Old 11-03-2004, 06:26 PM
herman's Avatar
herman herman is offline
Registered User
 
Join Date: Dec 2003
Posts: 60
POST ID # = 152549
herman Reputation: 10
Default

Just to be clear, the spreader half itself was staying relatively rigid, but both halves were starting to tilt in toward the top. It might have sounded like I was saying that the speader half itself was starting to warp.

Like || to /\ looking at the end.

Thanks.

Last edited by herman; 11-03-2004 at 06:28 PM.
Reply With Quote


  #116  
Old 11-08-2004, 05:20 AM
Walker Walker is offline
Registered User
 
Join Date: Oct 2004
Posts: 11
POST ID # = 153312
Walker Reputation: 10
Default

Well hopefully you guys have my memory and I have some new stuff on the way this week. I've been rock solid sense I pulled that ram from my PC. I hope the new stuff I get back lives upto its specs. I'm definitly giving my hard drives a workout with only a 256 in my pc I Cant wait to get my Gig back in I'm also wondering if you guys tested the stuff I sent in and can confirm my findings??
Reply With Quote


  #117  
Old 11-08-2004, 12:25 PM
RAM GUY's Avatar
RAM GUY RAM GUY is offline
Corsair Product Guru
 
Join Date: Apr 2001
Location: Fremont, CA
Posts: 118,268
POST ID # = 153375
RAM GUY Reputation: 10
Default

Walker,
All RMA'a are just replaced with new modules.. We do test the returned modules, but we do not get a Failure report unless we ask for it when the RMA is created. Unless you sent the modules directly to My ATT or one of our Tech support.
__________________
Support accounts and tickets can be created at https://support.corsair.com.
Reply With Quote


  #118  
Old 11-08-2004, 05:42 PM
herman's Avatar
herman herman is offline
Registered User
 
Join Date: Dec 2003
Posts: 60
POST ID # = 153501
herman Reputation: 10
Default

I got mem from my second RMA back. It works. No temp sensitivity or anything.

Memtest86+:
80 passes of test 5
5 passes of full 1-7

Second XL Corsair RMA passed with: defaults, 2225 1T @400, aggressive, SS off, vdimm 2.7, vcore 1.5, agp 1.5v, which was the first setup I tried when I got them.

I got XL v1.2 this time from the RMA, and the bad black Fry's sticks have gone back. Though before everyone starts asking for v1.2, I can definately say that the XL v1.1 is just fine as well, as at one point I had two v1.1 sticks from two different twinx sets that worked great even in dual together. Its just that I think I was very unlucky and possibly timing was bad with new A64s coming out and what Corsair had sent everyone. On the spreader thing for v1.2, ouch. Component placement and size is very similar to the BP (brainpower), and these definatly dont really fit the old spreader design very well.

Just for reference my RMA therad was here:
http://www.houseofhelp.com/forums/sh...2&page=1&pp=15


***********************************


Summary of my ordeal for anyone with similar issues.


Actually saw 3 different twinx 512x2 XL sets, where one stick, and only one stick per set, would fail on test 5 starting at around 45C-47C. I used nice Fluke brand calibrated multimeter for tesing. I also had very stable voltage under load across two different known good power supplies.

In the end all that mattered in my case was having two good sticks. This was no matter how many different settings I tried, over 4 XL pairs. All that seemed to matter for the three bad sticks was hitting ~46C or more in test 5. If one stick was bad it would fail in dual and individually after hitting that temp. To find bad stick individually you have to let it run a very long time, maybe even 100 passes, in test 5, for it to get hot enough. You may even need an Athlon 64 to pound it hard enough. Dual fails very quickly for MB where sticks in dual sit right next to each other, like 5-6 passes average, as the sticks get to 46C+ quickly because there is no real chance of air flow between the sticks. Lastly for those with newer good air flow cases, you might have to actually leave the case open to get higher temps for tesing individual. For others leaving it closed may get higher temps.


*** Setup ***

Corsair TwinX 1024 (2x 512) 3200 XL Non-Pro
AMD Athlon 64 3500+ 130nm Retail
MSI K8N Neo2 Platnum - BIOS 1.3
Reset to Safe Defaults

Stock Antec Sonata case with its stock TruePower 380s PSU
Zalman 7000A Cu, using BIOS 40C fan threshold, 3 deg. tolerance
BFG nVidia ti4600 - Original 4X AGP Version
WD Raptor 73GB, LG 52X DVD reader, Sony DVD R/W DL, Floppy

Memtest86+ 1.27
Bandwidth: 18113 L1, 4500 L2,
...2618 @ 2325 SS off, 2698 @ 2225 SS off,
...2619 @ 2325 SS on, 2699 @ 2225 SS on
Fluke 16 with bundled Fluke temp probe


*** tests ***

This is what I have tried for 2225. All fail test 5 unless otherwise noted.

PSU - Antec TruePower 380s, and Antec TruePower 430
BIOS - 1.37, 1.36, 1.3, 1.2
Vdimm - 2.8, 2.75, 2.7, auto
Vcore - 1.55, 1.525. 1.5, by startup
AGP - 1.60, 1.50
SS - on, off
aggressive - on, off
cmd rate - 1T, 2T
slots - 1&2, 3&4, 1, 3, 1&4 (to get sticks away from each other, no post)
sticks - O1&O2, P1&P2, B1&B2, O1, O2 (pass), P1, P2 (pass), B1, B2 (pass), P2&B2 (pass), R1&R2 (pass)
* O1 and O2 are the Original NewEgg sitcks, P1 and P2 are the Platinum sticks from the first corsair RMA, B1 and B2 are the Black shot in the dark Fry's sticks i bought after the first RMA failed, R1 and R2 are from the second corsair RMA I just recieved.

I have tried almost every combination, and definately tried what I'll call the "RamGuy combo": failsafe defaults, vdimm 2.8v, agp 1.6, ss on, 2225 1T, aggressive off (though I really think this should be supported if on).

Last edited by herman; 11-08-2004 at 07:02 PM.
Reply With Quote


  #119  
Old 11-12-2004, 05:12 AM
Walker Walker is offline
Registered User
 
Join Date: Oct 2004
Posts: 11
POST ID # = 154208
Walker Reputation: 10
Default

Well I recieved my ram yesterday, I ran memtest for a few hours with the settings at 400mhz@2-2-2-5. I'm glad to say not a single error and everything seems to be doing great now The performace is right on par with what I was expecting and I'm very pleased now. There has been quite a sharp learning curve jumping to the a64 for me but I'm very happy with my setup now. I do wish you guys had a little tighter screening process for memory thats advertised as top of the line, as no one wants to go thru buying a new computer and having nothing but trouble with it only to find out a part was not quite upto spec. When I shell out $300 dollars for ram and there is a big sticker on the top that says "Extreme Low Latancy 2-2-2-5" Thats what I expect, and nothing less will suffice. Ranting aside Corsair has always put out a good quality product, and I suspect that you guys will work out these issues. Thanks for the quick RMA and I hope I dont have to revisit these boards anytime soon with problems.
Reply With Quote


  #120  
Old 11-12-2004, 11:12 AM
Wired's Avatar
Wired Wired is offline
Administrator
Wired's PC Specs
 
Join Date: Aug 2000
Posts: 32,188
POST ID # = 154243
Wired Reputation: 32
Default

Walker, they have a very stringent testing process, however no testing process is foolproof. IIRC, the RMA rate for those sticks is something like 0.002 %.
__________________
No, I don't work for Corsair
...you need to flow like a river...
Forum Rules
Admin of this forum / webdesignforums.net / Petri's IT Forum / The Corsair Memory Steam Group
Founder/Creator/Admin of ZE SECRET PROJECT (Coming Soon)
Reply With Quote


Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT -4. The time now is 10:15 AM.


Powered by vBulletin® Version 3.8.7
Copyright ©2000 - 2019, vBulletin Solutions, Inc.