Memory CAS problem

  • Thread starter Thread starter Paolo
  • Start date Start date
P

Paolo

Hi!
I have a fujitsu-siemens amilo 1424, with 2x256MB pc2700 ram.
I cannot set CAS by bios, since it's automatically detected, and from
memtest I get CAS: 2.5/3/3/10.
I got from ebay 2x512MB PC2700 CL2.5 ram, but when I use it, memtest
indicates CAS 2/3/3/10, and the test fails.
I performed the test mixing old and new banks (therefore with
512+256MB) and everything is fine (CAS 2.5/3/3/10).
Why is my motherboard detecting the wrong CL? is my new memory faulty?
is there a way to force it to 2.5 anyway, even if my bios doesn't
allow me to change it?
I have seen the software wpcrset, but I am looking for a more
definitive solution.
Thank you for your help,
Paolo
 
Paolo said:
Hi!
I have a fujitsu-siemens amilo 1424, with 2x256MB pc2700 ram.
I cannot set CAS by bios, since it's automatically detected, and from
memtest I get CAS: 2.5/3/3/10.
I got from ebay 2x512MB PC2700 CL2.5 ram, but when I use it, memtest
indicates CAS 2/3/3/10, and the test fails.
I performed the test mixing old and new banks (therefore with
512+256MB) and everything is fine (CAS 2.5/3/3/10).
Why is my motherboard detecting the wrong CL? is my new memory faulty?
is there a way to force it to 2.5 anyway, even if my bios doesn't
allow me to change it?
I have seen the software wpcrset, but I am looking for a more
definitive solution.
Thank you for your help,
Paolo
Click to expand...

CPUZ (cpuid.com) has a tab on the main window, that says "SPD".
It will display the contents of the serial presence detect
EEPROM on the DIMMs, in an easy to read format.

CPUZ also has the "About" tab. In there, is a function called
"Registers Dump". That button makes a text file. In the text
file, you'll see a table of numbers. This is the raw SPD table
in hexidecimal. This is what my RAM looks like. There are eight
rows of numbers, of which the first four are the most important.

Dump Module #1
0 1 2 3 4 5 6 7 8 9 A B C D E F
00 80 08 07 0D 0A 02 40 00 04 50 60 00 82 08 00 01
10 0E 04 04 01 02 20 C0 00 00 00 00 28 28 28 28 40
20 70 70 40 40 00 00 00 00 00 37 46 30 28 50 00 00
30 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 85 ...

To decode the contents of an SPD table, you can use this
document. This is the JEDEC standard for DDR SPD encoding.

http://web.archive.org/web/20030417070529/http://www.jedec.org/download/search/4_01_02_04R11A.PDF

CAS latency is a function of clock speed. The slower the clock speed,
the lower the CAS can go. What is your current memory clock speed ?
(Maybe CPUZ can help with that too.)

Paul
 
Thank you Paul,
I used cpuz, and everything seems to me ok with the old memory:

....
Memory type DDR
Manufacturer (ID) Samsung (CE00000000000000)
Size 256 MBytes
Max bandwidth PC2700 (166 MHz)
....

Frequency (MHz) 133 166
CAS# 2.0 2.5
RAS# to CAS# delay 3 3
RAS# Precharge 3 3
TRAS 6 7

BUT...with the new one I get

....
Memory type DDR
Manufacturer (ID) (0000000000000000)
Size 512 MBytes
Max bandwidth PC2700 (166 MHz)
....
Frequency (MHz) 133 166 166
CAS# 2.0 2.5 3.0
RAS# to CAS# delay 2 3 3
RAS# Precharge 2 3 3
TRAS 6 7 7
From the chipset section, I have the following
Click to expand...
Memory Frequency 166.7 MHz (3:5)
CAS# 2.5
RAS# to CAS# 3
RAS# Precharge 3
Cycle Time (tRAS) 7

I tried to used both the new 512MB banks, and (before windows crashed)
I got
Memory Frequency 166.7 MHz (3:5)
CAS# 2
RAS# to CAS# 3
RAS# Precharge 3
Cycle Time (tRAS) 7


Is it normal to have 2 sets of parameters for 166MHz for the new
banks?
where is my chipset getting 2/3/3/7 which is not stored in those
banks?
I can say I'm more confused than before :)
Can anyone understand what is happening?

Paolo


ps. here are the dumps
Dump Module #1 (256MB, working)
0 1 2 3 4 5 6 7 8 9 A B C D E F
00 80 08 07 0D 09 02 40 00 04 60 70 00 82 10 00 01
10 0E 04 0C 01 02 20 C0 75 70 00 00 48 30 48 2A 20
20 80 80 45 45 00 00 00 00 00 3C 48 30 2D 55 00 00
30 00 00 00 00 00 00 00 00 00 00 00 00 00 00 10 0E
40 CE 00 00 00 00 00 00 00 01 4D 34 20 37 30 4C 33
50 32 32 34 46 54 30 2D 43 42 33 20 30 46 05 06 08
60 05 A9 E9 00 57 43 44 56 33 44 53 00 00 00 00 00
70 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00


Dump Module #2 (512MB)
0 1 2 3 4 5 6 7 8 9 A B C D E F
00 80 08 07 0D 0A 02 40 00 04 60 70 00 82 10 00 01
10 0E 04 1C 01 02 20 C1 60 70 75 70 3C 28 3C 28 40
20 60 60 40 40 00 00 00 00 00 37 41 28 28 50 00 01
30 00 00 00 00 00 00 00 00 00 00 00 00 00 00 10 87
40 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
50 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
60 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
70 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
 
Paolo said:
Thank you Paul,
I used cpuz, and everything seems to me ok with the old memory:

...
Memory type DDR
Manufacturer (ID) Samsung (CE00000000000000)
Size 256 MBytes
Max bandwidth PC2700 (166 MHz)
...

Frequency (MHz) 133 166
CAS# 2.0 2.5
RAS# to CAS# delay 3 3
RAS# Precharge 3 3
TRAS 6 7

BUT...with the new one I get

...
Memory type DDR
Manufacturer (ID) (0000000000000000)
Size 512 MBytes
Max bandwidth PC2700 (166 MHz)
...
Frequency (MHz) 133 166 166
CAS# 2.0 2.5 3.0
RAS# to CAS# delay 2 3 3
RAS# Precharge 2 3 3
TRAS 6 7 7

Memory Frequency 166.7 MHz (3:5)
CAS# 2.5
RAS# to CAS# 3
RAS# Precharge 3
Cycle Time (tRAS) 7

I tried to used both the new 512MB banks, and (before windows crashed)
I got
Memory Frequency 166.7 MHz (3:5)
CAS# 2
RAS# to CAS# 3
RAS# Precharge 3
Cycle Time (tRAS) 7


Is it normal to have 2 sets of parameters for 166MHz for the new
banks?
where is my chipset getting 2/3/3/7 which is not stored in those
banks?
I can say I'm more confused than before :)
Can anyone understand what is happening?

Paolo


ps. here are the dumps
Dump Module #1 (256MB, working)
0 1 2 3 4 5 6 7 8 9 A B C D E F
00 80 08 07 0D 09 02 40 00 04 60 70 00 82 10 00 01
10 0E 04 0C 01 02 20 C0 75 70 00 00 48 30 48 2A 20
20 80 80 45 45 00 00 00 00 00 3C 48 30 2D 55 00 00
30 00 00 00 00 00 00 00 00 00 00 00 00 00 00 10 0E
40 CE 00 00 00 00 00 00 00 01 4D 34 20 37 30 4C 33
50 32 32 34 46 54 30 2D 43 42 33 20 30 46 05 06 08
60 05 A9 E9 00 57 43 44 56 33 44 53 00 00 00 00 00
70 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00


Dump Module #2 (512MB)
0 1 2 3 4 5 6 7 8 9 A B C D E F
00 80 08 07 0D 0A 02 40 00 04 60 70 00 82 10 00 01
10 0E 04 1C 01 02 20 C1 60 70 75 70 3C 28 3C 28 40
20 60 60 40 40 00 00 00 00 00 37 41 28 28 50 00 01
30 00 00 00 00 00 00 00 00 00 00 00 00 00 00 10 87
40 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
50 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
60 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
70 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
Click to expand...

The SPD for the 512MB DIMM is not in very good shape.
There really shouldn't be two entries for 166MHz. The
manufacturing information section is also zeroed out,
which isn't really that good either.

The BIOS doesn't appear to handle the badly designed SPD
at all well. So the BIOS is also at fault here. It should
be using CAS3, if anything, when the 512MB stick is present.

If the computer does not have an option to set the speed
manually, then using the four DIMMs, to trick the BIOS
into slowing the CAS down a bit, is about all you can
do. Or, return the 512MB memory, and get some branded
memory to take its place. Crucial, Kingston, Samsung etc.

Paul
 
Hello Paul,
unfortunately I have just 2 slots, so I will ask the seller for a
refund or a replacement.
I thank you very much for the useful information you provided, and for
answering so fast to my questions!
Paolo
 
Back
Top