Troubles with OCZ Platinum Rev.2 TCCD

[lemonlime]

Hey Guys,

My trusty old dual channel PC3200 kit of OCZ Platinum Rev.2 seems to be on its way out. Its been running flawlessly in my second PC for quite some time. Yesterday I noticed some instability in Windows and sure enough, it failed Orthos after only several minutes with no overclock at all. (This PC is running totally stock). I followed up with Memtest86+ and looped test #5. I got only one error after 45 minutes, but clearly something is wrong.

I increased vDIMM from 2.7 to 2.8V and loosened secondary timings (like trtw etc) but I'm still seeing occasional MemTest86+ errors. The memory is running at 2-2-2-5 1T @ 200MHz defaults. This RAM did 300+ MHz @ 2.5-4-3 1T without issue when it was my main machine :)

I threw in a 512MB stick of value ram and it ran flawlessly overnight, so I think I've ruled out the rest of the PC. Here are the specs:

Athlon 64 X2 4200+ Manchester E4 revision
DFI LanParty NF4 Ultra-D
2x 512MB OCZ Platinum Rev.2 PC3200 (2-2-2-5 1T @ 2.8V)
PCP&C 510 Express (tested fine)

What do you think? RMA time?

[TheLostSwede]

Hey Lemon,

Run each stick separately at 2.8 and 2-2-2-5 timings, alphatimings to default/failsafe settings. Run a full memtest on each stick and let me know in what test the error occurs.

[lemonlime]

Hey Lemon,

Run each stick separately at 2.8 and 2-2-2-5 timings, alphatimings to default/failsafe settings. Run a full memtest on each stick and let me know in what test the error occurs.

Will-do :)

[lemonlime]

Call me crazy but I can't get these sticks to show any signs of instability individually. Both sticks ran memtest for hours without any errors on their own. :confused:

I cleared the CMOS and will install them in dual channel again to see what happens.. odd.

Thanks Mackanz

[lemonlime]

Ok, finally had a chance to troubleshoot some more. Each stick tests just fine individually, but I still get some odd instability with two sticks.

I've cleared the CMOS and reverted a lot of the secondary timings to 'Auto'. I'll post back with an update to see what becomes of that. I'm thinking it is a quirk with the board.

Changing 'DRAM Response Time' from 'Normal' to 'Fast' seems to have improved stability but I have not tested long enough to be sure yet.

[TheLostSwede]

Hey lemon,

After a long leave of sickness i amfinally home from the hospital again. This have unfortunatly meant that a few of OCZ customers have had a useless customer service from me. How long have you gone in your testing? Dram response time is essential on some bioses, but impossible to tell exactly on which bios.

[primesuspect]

Welcome back, marcus! :eek2:

[TheLostSwede]

Thank you sir :)
How nice of you Brian!

[RyderOCZ]

and he didn't even tell me ;D

[TheLostSwede]

;D

Your fault boss, you wheren't online when i got home :P

TTY later Eric.

[lemonlime]

No worries at all! Glad to hear you are feeling better :)

DRAM Response time appears to have corrected the issue. I do remember (quite some time ago) that this was a very sensitive option with when I was tweaking TCCD.

All is well now. It passed MemTest and Orthos for 8+ hours at 220MHz (I bumped it up to be extra sure of the stability). I sold the mobo/cpu/ram last week, so no more testing will be done :)