Twinx 1024 3200LL Strange SPD readouts.

Hi, I use CPU-Z to Read SPD. The problem is that just one of the modules reads the right Corsair info. The Other one reads as a Dallas Semiconductor chip! Why???? And in Aida 32 it won't read this chips SPD at all!!!! However, it's working fine and testing okay in memtest

I'd guess that that's NOT a Corsair mem stick! Sounds like a fradulent stick :(

[quote name='Wired']I'd guess that that's NOT a Corsair mem stick! Sounds like a fradulent stick :([/QUOTE] Here are the readouts CPU-Z version 1.22 Memory Modules Serial Presence Detect (SPD) Module #1 General Memory type DDR-SDRAM Manufacturer (ID) Corsair (7F7F9E0000000000) Size 512 MBytes Max bandwidth PC3200 (200 MHz) Part number CMX512-3200LL Attributes Number of banks 2 Data width 64 bits Correction None Registered no Buffered no Timings table Frequency (MHz) 200 CAS# 2.0 RAS# to CAS# delay 3 RAS# Precharge 2 TRAS# 6 Module #2 General Memory type DDR-SDRAM Manufacturer (ID) Dallas Semiconductor (7FA1FF0000000000) Size 512 MBytes Max bandwidth PC3200 (200 MHz) Part number CMX512-3200LL Attributes Number of banks 2 Data width 41775 bits Correction None Registered no Buffered no Timings table Frequency (MHz) 200 CAS# 2.0 RAS# to CAS# delay 3 RAS# Precharge 2 TRAS# 6

Actually, since everything else is correct, including the part number (CMX-), it might just be a bad SPD programming. I forget what's supposed to be on the PCB of the module, but I believe there's something. I'll alert RAM GUY to this thread. PLUR CK

[quote name='CandyKid']Actually, since everything else is correct, including the part number (CMX-), it might just be a bad SPD programming. I forget what's supposed to be on the PCB of the module, but I believe there's something. I'll alert RAM GUY to this thread. PLUR CK[/QUOTE] This seems strange to me. I am not sure if the other stick showed Corsair before. Can't remember. Is there any possibility that the SPD can be corrupted???? By the way, when using Nvidia system utility I just get "SPD DATA PARSE ERROR!" on the Dallas Stick. Aida Can't read it at all. And Sisoft is also blurry. Got this SPD with CPU-z. The stick tests okay with memtest and in prime. Strange. But one thing is for sure.... These Rev1.2 sticks will not overclock well in my FRN2-L. 213MHz bus is the max stable. And they will only run at cas 2.5. When I save CAS 2 or CAS 3 in bios it will default to Cas 2.5 no mather what. I guess soltek need a bios update. Have tested these sticks in other boards. And other sticks in this board. And it would save the Cas 2 setting

Technically, yes. The SPD can also have an error in it's original programming, they tend to crop up every now and again, it's a fact of life since nothing is perfect. PLUR CK P.S. I alerted RAM GUY to this thread, just be patient.

[quote name='CandyKid']Technically, yes. The SPD can also have an error in it's original programming, they tend to crop up every now and again, it's a fact of life since nothing is perfect. PLUR CK[/QUOTE] But is it possible it can be corrupted over time? After som use in the computer? because I'm not sure if it showed up correctly before.... Another error in the Second Sticks SPD: "Data width 41775 bits" It should be 64 bit.

Yeah, I saw that after my post... and you edited your post... had to go re-read. :p This does sound a bit wierd, but we should rule out a "fake" module first. Like I said, there may be a way to tell by looking at the PCB, but I can't remember it. RAM GUY will respond soon. BTW, just like any ROM, the SPD can get corrupted, but I'd think it'd be very rare. PLUR CK

Here is the SPD Readout from the the first stick. Which report as a Corsair. This is taken with Nvidia System utility. I noticed a strange thing here also. This is from the last info: "Module Revision Code: 0 (manufacture dependant) Module Manufacturing Date: Not reported Module Serial Number: 00000000 SPD Revision: 0.0" No serial? Strange? :) The Second module "Dallas", as I've said before just gave me an error with this program.... Memory Type: DDR SDRAM Error Correction: NONE Voltage Interface Level: SSTL (2.5V) Data Width: 64 bits Number of Physical Banks: 2 (double sided) Number of Logical Banks: 4 Number of Row Addresses: 13 13 Number of Col Addresses: 10 10 Bank Density: 256 268435456 MB Primary Ram Width: 8 1 bits Error Checking Ram Width: 13 13 bits Module Density: 268435712 MB Refresh Rate: 7812.500 us Self-Refresh: Yes Burst Lengths Supported: 2 4 8 CAS 2.0: Cycle Time (tCYC) = 5.000 ns fMax = 200.000 Mhz Access Time (tAC) = 0.600 ns fMax = 1666.667 MHz CS Supported: 0.0 clocks WE Supported: 1.0 clocks Min Delay, Bk/Bk Random Cols: 1 clocks Min Row Precharge Time (tRP): 10.000 ns Min Row Active delay (tRRD): 10.000 ns Min RAS to CAS delay (tRCD): 15.000 ns Min Actv to Precharge (tRAS): 30.000 ns Addr/Cmd Setup Time: 0.600 ns Addr/Cmd Hold Time: 0.600 ns Data Setup Time: 0.400 ns Data Hold Time: 0.400 ns Buffered Address & Control No Registered Address & Control No On-Card PLL: No FET Switch On-Card Enable: No FET Switch External Enable: No Differential Clock Input: Yes Includes Weak Driver: No Includes QFC Output: No Lower Vdd Tolerence: 0.2 V Upper Vdd Tolerence: 0.2 V Manufacturer's JEDEC ID Code: 00000000 009e7f7f Module Manufacturing Location: 1 (manufacture dependant) Module Part Number: CMX512-3200LL Module Revision Code: 0 (manufacture dependant) Module Manufacturing Date: Not reported Module Serial Number: 00000000 SPD Revision: 0.0

[quote name='RAM GUY']Not sure what might be the problem for sure! However, I would suggest we try and RMA and or replace the modules. In addition, when you submit the RMA please include a copy of your invoice along with the RMA Request Form. Please follow the link in my signature and provided it's not counterfeit we will be happy to replace them. Not to alarm you, just we have seen a few knock off’s! In addition, it does sound like the SPD is either just corrupt or has a bad or failing address that just happens to be where the Manufacturer code is stored. But let's convict it before we hang it.[/QUOTE] Hello, Thanks for the answer. They seem to be working okay in Memtest, Prime and such. So don't see the need to replace them. Is there any way I can find out if they are counterfeit? Anyway, If I'm gonna RMA them I will do it where I got them. Via my retailer. I live in Norway, and sending these modules to the USA and getting new sticks back will take forever! Lol!

Update! I'm running the Dallas stick alone in single mode now, and it gave me an Error in Prime 95 just after a few minutes! I'm gonna RMA it! Update 2: Tested with memtest 86 on SPD settings. Tried with 2.5 V and 2.7 Volts. Both gave me two errors in memtest at 200 MHz bus. I have tested this Stick ok before. So it was corrupted I guess. Have been in contact with my retailer, and they will replace it.

Hi, I'm back. It took ages for the reseller to test the RAM. The outcome: They sent back the same sticks! They had tested with memtest for 26 hours with several different Motherboards and had no errors. Since the last time I have bought myself a Abit AN7. Tested the RAM today and it passed 3 times in Memtest okay. But the SPD is still corrupt. I don't think they even checked that. As a side note this Corsair RAM won't even do 230MHz bus on my AN7. I'm using a Special beta bios which overclocks well. I bought some Cruicial PC3200 RAM several days ago. And had this RAM up to 235MHz bus at CAS 2.5, 3, 3, 11 timings :) And it's just half the price of the Twinx PC3200LL mem. These Rev 1.2 sticks are crap in my eyes. But I believe the Rev 1.1 sticks was good. This RAM are Way to expensive. Well, well.... See ya

[quote name='RAM GUY']If you want to over clock them I would try the same settings as you had set for the other memory! These are made to run at a aggressive setting at DDR400 and over clocking I would suggest Cass 2.5-3-3-7 and set the Dim Voltage to 2.8 Volts! In addition, you can tell if they are really our modules by looking on the PCB you should see Corsair printed on the circuit board on the edge of the module![/QUOTE] I have read that if you increase the Voltage past 3.1 V you will overshoot the built in VR on these Chips.

If you go past 2.9v you are VOIDING the warranty.

Is it possible?

Try to keep them at spec.

Will RMA them again. The Dallas Semiconductor Stick is flaky....