ASrock Conroe E4400 Overclocking problems Part 2

Have finally got round to investigating the problem.

Repalced ram with Crucial DDR2 667. It will now POST to 256 MHz with
PCIE Aysnc at 114MHz. NB There is a PCI bus lock so that keeps it at
33.3MHz anyway.

DRAM timings are from SPD (makes no diff if not). These are at 255.8
MHz and a 1:1 ratio:

3,3,3,9,12

What I think is killing the overclock potential is the Vcore - I can
only use Auto or High usefully (haven't tried Low). Thishas resulted
in a Vcore pretty much the same as stock at 200MHz; 1.2v

Any suggestions to unlock the potental of the E4400?

"Cornelius" wrote in message
...
Have finally got round to investigating the problem.

Repalced ram with Crucial DDR2 667. It will now POST to 256 MHz with
PCIE Aysnc at 114MHz. NB There is a PCI bus lock so that keeps it at
33.3MHz anyway.

DRAM timings are from SPD (makes no diff if not). These are at 255.8
MHz and a 1:1 ratio:

3,3,3,9,12

What I think is killing the overclock potential is the Vcore - I can
only use Auto or High usefully (haven't tried Low). Thishas resulted
in a Vcore pretty much the same as stock at 200MHz; 1.2v

Any suggestions to unlock the potental of the E4400?

Without being able to raise the vcore higher than 1.2V, I doubt you will go
much further. I had an E6600 that would take over 1.35V and get to 3.6Ghz
very solidly with liquid cooling. That is a 1.2Ghz increase from 2.4Ghz
stock. I believe Phil Weldon got his E4400 up to 2.7-2.8Ghz on an NV 680i
chipset MB with decent air cooling. Being able to raise the vcore is very
important in getting a decent overclock.


Ed

On Thu, 4 Sep 2008 07:55:54 -0500, "Ed Medlin" ed@ edmedlin.com
wrote:


"Cornelius" wrote in message
.. .
Have finally got round to investigating the problem.

Repalced ram with Crucial DDR2 667. It will now POST to 256 MHz with
PCIE Aysnc at 114MHz. NB There is a PCI bus lock so that keeps it at
33.3MHz anyway.

DRAM timings are from SPD (makes no diff if not). These are at 255.8
MHz and a 1:1 ratio:

3,3,3,9,12

What I think is killing the overclock potential is the Vcore - I can
only use Auto or High usefully (haven't tried Low). Thishas resulted
in a Vcore pretty much the same as stock at 200MHz; 1.2v

Any suggestions to unlock the potental of the E4400?

Without being able to raise the vcore higher than 1.2V, I doubt you will go
much further. I had an E6600 that would take over 1.35V and get to 3.6Ghz
very solidly with liquid cooling. That is a 1.2Ghz increase from 2.4Ghz
stock. I believe Phil Weldon got his E4400 up to 2.7-2.8Ghz on an NV 680i
chipset MB with decent air cooling. Being able to raise the vcore is very
important in getting a decent overclock.


Ed

Any hacks to allow Vcore adjustment then?

Thanks

"Cornelius" wrote in message
...
On Thu, 4 Sep 2008 07:55:54 -0500, "Ed Medlin" ed@ edmedlin.com
wrote:


"Cornelius" wrote in message
. ..
Have finally got round to investigating the problem.

Repalced ram with Crucial DDR2 667. It will now POST to 256 MHz with
PCIE Aysnc at 114MHz. NB There is a PCI bus lock so that keeps it at
33.3MHz anyway.

DRAM timings are from SPD (makes no diff if not). These are at 255.8
MHz and a 1:1 ratio:

3,3,3,9,12

What I think is killing the overclock potential is the Vcore - I can
only use Auto or High usefully (haven't tried Low). Thishas resulted
in a Vcore pretty much the same as stock at 200MHz; 1.2v

Any suggestions to unlock the potental of the E4400?

Without being able to raise the vcore higher than 1.2V, I doubt you will
go
much further. I had an E6600 that would take over 1.35V and get to 3.6Ghz
very solidly with liquid cooling. That is a 1.2Ghz increase from 2.4Ghz
stock. I believe Phil Weldon got his E4400 up to 2.7-2.8Ghz on an NV 680i
chipset MB with decent air cooling. Being able to raise the vcore is very
important in getting a decent overclock.


Ed

Any hacks to allow Vcore adjustment then?

Thanks

Unless the MB allows it, not to my knowledge. Most mid/top line motherboards
are more overclocking friendly and have more adjustments than a lot of the
more "budget" boards. I would hold on to what you have and wait until Intel
releases the Nehalem processors which will use a different socket and
promise to be very good overclockers in a range of prices. I generally use
Asus motherboards because they have been reliable for me and have all the
adjustments I need for overclocking if I choose to do so.

Ed

On Thu, 04 Sep 2008 21:39:15 +0100, Cornelius
wrote:

On Thu, 4 Sep 2008 07:55:54 -0500, "Ed Medlin" ed@ edmedlin.com
wrote:


"Cornelius" wrote in message
. ..
Have finally got round to investigating the problem.

Repalced ram with Crucial DDR2 667. It will now POST to 256 MHz with
PCIE Aysnc at 114MHz. NB There is a PCI bus lock so that keeps it at
33.3MHz anyway.

DRAM timings are from SPD (makes no diff if not). These are at 255.8
MHz and a 1:1 ratio:

3,3,3,9,12

What I think is killing the overclock potential is the Vcore - I can
only use Auto or High usefully (haven't tried Low). Thishas resulted
in a Vcore pretty much the same as stock at 200MHz; 1.2v

Any suggestions to unlock the potental of the E4400?

Without being able to raise the vcore higher than 1.2V, I doubt you will go
much further. I had an E6600 that would take over 1.35V and get to 3.6Ghz
very solidly with liquid cooling. That is a 1.2Ghz increase from 2.4Ghz
stock. I believe Phil Weldon got his E4400 up to 2.7-2.8Ghz on an NV 680i
chipset MB with decent air cooling. Being able to raise the vcore is very
important in getting a decent overclock.


Ed

Any hacks to allow Vcore adjustment then?

Thanks
Ok. Have now got it to run at 266 with Vcore stuck at 1.20v:
CPUz reports:
Core speed = 2661.4MHz
Rated FSB = 1064.5MHz
DRAM Freq = 266.1MHz (i.e. 1:1)
CL= 4
tRCD = 4
tRP = 4
tRAS = 12
tRC = 16

Had to set DRAM timings manually as SPD kept them at 3,3,3,9,12 which
froze XP a little while in.

To get it to boot need to set the PCIE bus upwards eeach time so now
runing a PCIE of 119MHz. Anythin less fails. Will be trying to
increment cpu upwards towards 280MHz...

Ok. Have now got it to run at 266 with Vcore stuck at 1.20v:
CPUz reports:
Core speed = 2661.4MHz
Rated FSB = 1064.5MHz
DRAM Freq = 266.1MHz (i.e. 1:1)
CL= 4
tRCD = 4
tRP = 4
tRAS = 12
tRC = 16

Had to set DRAM timings manually as SPD kept them at 3,3,3,9,12 which
froze XP a little while in.

To get it to boot need to set the PCIE bus upwards eeach time so now
runing a PCIE of 119MHz. Anythin less fails. Will be trying to
increment cpu upwards towards 280MHz...

It is best to relax the memory timings and speed to something very
conservative while overclocking the cpu. Once you have your cpu where you
want it then work on memory speeds. That way you don't have to worry about
what is failing you, your memory or cpu speeds. I like to keep things as
simple as possible............:-)



Ed

Ok, now have it running at 269 MHz and a PCIE bus at 120 MHz (not used
anyway). It runs stable like this but any attempt to reduce the PCIE
bus even to 119 MHz results in a boot failure. Wierd.

Both Dhrys and Whets are improving linearly.

Oh, and the Vcore is still running between 1.200 to 1.21 volts. DRAM
timings are as before. Won't need to change these until (and unless) I
hit 333 MHz clock speed...


Ed Medlin wrote:
Ok. Have now got it to run at 266 with Vcore stuck at 1.20v:
CPUz reports:
Core speed = 2661.4MHz
Rated FSB = 1064.5MHz
DRAM Freq = 266.1MHz (i.e. 1:1)
CL= 4
tRCD = 4
tRP = 4
tRAS = 12
tRC = 16

Had to set DRAM timings manually as SPD kept them at 3,3,3,9,12 which
froze XP a little while in.

To get it to boot need to set the PCIE bus upwards eeach time so now
runing a PCIE of 119MHz. Anythin less fails. Will be trying to
increment cpu upwards towards 280MHz...

It is best to relax the memory timings and speed to something very
conservative while overclocking the cpu. Once you have your cpu where you
want it then work on memory speeds. That way you don't have to worry about
what is failing you, your memory or cpu speeds. I like to keep things as
simple as possible............:-)



Ed

"GAK" wrote in message
...
Ok, now have it running at 269 MHz and a PCIE bus at 120 MHz (not used
anyway). It runs stable like this but any attempt to reduce the PCIE
bus even to 119 MHz results in a boot failure. Wierd.

Both Dhrys and Whets are improving linearly.

Oh, and the Vcore is still running between 1.200 to 1.21 volts. DRAM
timings are as before. Won't need to change these until (and unless) I
hit 333 MHz clock speed...


There is no need to mess with the PCIE bus anyway. It is not going to help
with the OC. I don't understand why it causes a boot failure unless it is
somehow tied into the SATA controller or clock generator somehow on your MB.
I have never seen a MB that has overclocking capabilities like yours that
doesn't have any voltage adjustments. It seems that you are a slight Vcore
rise from getting a pretty decent overclock. Slight increases in Vcore
sometimes don't raise temps at all, and if they do it is usually minimal.


Ed

Ed Medlin wrote:
"GAK" wrote in message
...
Ok, now have it running at 269 MHz and a PCIE bus at 120 MHz (not used
anyway). It runs stable like this but any attempt to reduce the PCIE
bus even to 119 MHz results in a boot failure. Wierd.

Both Dhrys and Whets are improving linearly.

Oh, and the Vcore is still running between 1.200 to 1.21 volts. DRAM
timings are as before. Won't need to change these until (and unless) I
hit 333 MHz clock speed...


There is no need to mess with the PCIE bus anyway. It is not going to help
with the OC. I don't understand why it causes a boot failure unless it is
somehow tied into the SATA controller or clock generator somehow on your MB.
I have never seen a MB that has overclocking capabilities like yours that
doesn't have any voltage adjustments. It seems that you are a slight Vcore
rise from getting a pretty decent overclock. Slight increases in Vcore
sometimes don't raise temps at all, and if they do it is usually minimal.


Ed

This could be the original thread here. Asrock Conroe 1333-D667.

http://groups.google.ca/group/alt.co...1d99c68f3ecbcb

I wonder if Speedstep is disabled ?

Paul

'Ed Medlin' wrote:
There is no need to mess with the PCIE bus anyway. It is not going to help
with the OC. I don't understand why it causes a boot failure unless it is
somehow tied into the SATA controller or clock generator somehow on your
MB. I have never seen a MB that has overclocking capabilities like yours
that doesn't have any voltage adjustments. It seems that you are a slight
Vcore rise from getting a pretty decent overclock. Slight increases in
Vcore sometimes don't raise temps at all, and if they do it is usually
minimal.
_____

My E4300 that you mentions runs happily at 150%, 2.7 GHz with the CPU core
voltage set at 1.250 volts, BELOW the stock speed encoded in the CPU of
1.325 volts. Of course that is on an overclocking friendly EVGA 680i
motherboard that has the slightly ridiculous ability to increment CPU core
voltage in 0.005 volt steps. The same CPU takes 1.375 volts to run at 3.16
GHz. Even though there are dozens of voltage and speed parameters that can
be set in the motherboard, the only necessary one for a stable 150%
overclock was increasing the FrontSide Bus speed and setting the Memory
Clock to suit the DDR2 used (DDR2-800 or DDR2-1200 in my system.)

Phil Weldon


"Ed Medlin" ed@ edmedlin.com wrote in message
...

"GAK" wrote in message
...
Ok, now have it running at 269 MHz and a PCIE bus at 120 MHz (not used
anyway). It runs stable like this but any attempt to reduce the PCIE
bus even to 119 MHz results in a boot failure. Wierd.

Both Dhrys and Whets are improving linearly.

Oh, and the Vcore is still running between 1.200 to 1.21 volts. DRAM
timings are as before. Won't need to change these until (and unless) I
hit 333 MHz clock speed...


There is no need to mess with the PCIE bus anyway. It is not going to help
with the OC. I don't understand why it causes a boot failure unless it is
somehow tied into the SATA controller or clock generator somehow on your
MB. I have never seen a MB that has overclocking capabilities like yours
that doesn't have any voltage adjustments. It seems that you are a slight
Vcore rise from getting a pretty decent overclock. Slight increases in
Vcore sometimes don't raise temps at all, and if they do it is usually
minimal.


Ed