cannot push HTT over 235-340 : chipset, BIOS or CPU ?

[hagar]

Hi all,

 

First, the sys specs :

 

SLI-DR, 510-2 bios

3700+ SD watercooled

2*512 MB Adata BH-5 "vintage"

6800 ultra stock cooling

74 GB Raptor on NF4 sata1

120 GB + 80 GB EIDE drives

LG4163B DVDR

OCZ 520W PS, all 4 power inputs connected

XP SP2

 

It runs rock stable at stock, the memory does 1.5-2-2-5 @ 2.6V dual. To avoid any memory or CPU issue, configued a 9X multi and 4:3 divider, and a 2*LDT multi. Then moving HTT upwards using clockgen, I get an instant reboot somewhere between 235 and 240. Needless to say that the CPU and memory run lower freqs than stock when this happens. Same thing running the CPU at stock multiplier, when first looking for the CPU limit. So it seems to be a motherboard issue (heat, bios or hardware). Kind of disappointed when I see 500+ HTT on this board. I hope it's somehow bios related, or there is a mess with this sandiego memory controller. An idea anyone ?

[Sharp]

Hello,

 

What voltages are you using for the Vcore/chipset/LDT?

How hot is the PWM?

 

Did you plug all the power connectors into the board? (4 in total)

[hagar]

Hi Shark,

 

As I'm running everything underclocked (except for the HTT obviously) everything is at stock voltage : memory 2.6, LDT 1.2, Chip 1.5. Changing any of these doesn't change anything, tried many combinations.

 

PWM is quite hot : reading 44°C (MBM5 + DFI settings) because there is no airflow as the CPU is watercooled (idle at 26°C, 23°C ambiant). The chip is idle at 53°C because of the hot airflow from the 6800 ultra I guess.

[Sharp]

Hello,

 

Max out the LDT and chipset voltage, this should rule out any voltage problems. (quite safe)

 

Add some extra cooling incase the temps are causing a problem.

 

And maybe start fooling around with the CPU Vcore voltage.

 

When changing the HTT frequency/FSB you must increase the Vcore.