Can't adjust CPU Voltage?

hwmgr

Distinguished
Jan 19, 2012
53
0
18,630
I have a Zebronics ZX-945-15 motherboard, with a Pentium 4 HT 2.8GHZ.

Currently i have been running @ 1.8GHZ, coz the system just restarts after windows loading screen @ 2.8ghz.
I thought it to be heating problem, and brought & installed a new cooler & hs, but still i can't clock it to 2.8ghz.

I think that i have been having high CPU voltage, causing the heating..

Here's what CPU-Z shows:
7156909713_6041ab447a.jpg

Normally, voltage would fluctuate anywhere b/w 1.65-1.80.
Is this an OK'ish range, or on the higher side???
I have checked my bios, it does'nt support any voltage regulation options.

Is there any other method to lower the CPU voltage, hardware or software?
 
Ah, so you are basically enforcing Speedstep.

I would guess the CPU is fried. Increased resistance due to aging. Either that or the voltage regulation on the board has gone.... either way it's probably not worthwhile given the gains to be had from even the cheapest modern system.
 
From your description it seems like the issue is not too much voltage, but too much resistance in the CPU. Any lowering of the voltage, if possible, would just cause it to become unstable.

Have you inspected the board for leaking/bulging capacitors? I think this is less likely but worth checking out.
 
G

Guest

Guest
No VID? That's strange. Without the VID value I can't recommend a pin mod.

"VID is the recommended voltage the processor should be running at at the current power state, this is a predetermined value, programmed by the CPU manufacturer. When the processor settings, especially the VCore is kept at default settings, Core Temp will provide a good approximation of the real VCore, but if you change the setting using the BIOS or by other means, Core Temp's VID readings should be ignored."

Transistor degradation is possible but I'd say the board is simply not fully compatible with that processor. Additionally I can't locate any product websites with potential BIOS updates. Seeing that this is an older low-end board I'd guess there aren't any updates.
At this point I'd give up on this CPU/motherboard combination if I were you.
 
G

Guest

Guest
As a last resort you could try "RightMark CPU Clock Utility". It helped me modify CPU multiplier and voltage on a board without SpeedStep support.
 
G

Guest

Guest
I need the VID for the max. power state. That's not necessarily 1.4V. For my pin modded E5700 (VID range 0.8500V-1.3625V) the max. VID was 1.2875V. Without this value the result of the pin mod could have been far too high/low (0.5V-1.6V according to the datasheet).
 

hwmgr

Distinguished
Jan 19, 2012
53
0
18,630


Here's Dump for Rightmark clock:
RightMark CPU Clock Utility CPU info dump
-----------------------------------------

System CPU #0 (physical/1st core)

CPUID EAX EBX ECX EDX
00000000 00000005 756E6547 6C65746E 49656E69
00000001 00000F41 00020800 0000441D BFEBFBFF
00000002 605B5001 00000000 00000000 007C7040
00000003 00000000 00000000 00000000 00000000
00000004 00004121 01C0003F 0000001F 00000000
00000005 00000040 00000040 00000000 00000000
80000000 80000008 00000000 00000000 00000000
80000001 00000000 00000000 00000000 00000000
80000002 20202020 20202020 20202020 6E492020
80000003 286C6574 50202952 69746E65 52286D75
80000004 20342029 20555043 30382E32 007A4847
80000005 00000000 00000000 00000000 00000000
80000006 00000000 00000000 04006040 00000000
80000007 00000000 00000000 00000000 00000000
80000008 00002024 00000000 00000000 00000000

MSR EDX EAX
00000017 000A0000 00000000
0000002A 00000000 00000000
0000002C 00000000 0E12000E
0000002D FFFFFFFF FFFFFFFF
00000198 00000E2D 00000E2D
00000199 00000000 00000E2D
0000019A 00000000 00000002
0000019B 00000000 00000000
0000019C 00000000 00000000
0000019D 00000000 00000E2A
000001A0 00000000 20840089
000001A1 00000000 00000000

System CPU #1 (logical/2nd core)

CPUID EAX EBX ECX EDX
00000000 00000005 756E6547 6C65746E 49656E69
00000001 00000F41 01020800 0000441D BFEBFBFF
00000002 605B5001 00000000 00000000 007C7040
00000003 00000000 00000000 00000000 00000000
00000004 00004121 01C0003F 0000001F 00000000
00000005 00000040 00000040 00000000 00000000
80000000 80000008 00000000 00000000 00000000
80000001 00000000 00000000 00000000 00000000
80000002 20202020 20202020 20202020 6E492020
80000003 286C6574 50202952 69746E65 52286D75
80000004 20342029 20555043 30382E32 007A4847
80000005 00000000 00000000 00000000 00000000
80000006 00000000 00000000 04006040 00000000
80000007 00000000 00000000 00000000 00000000
80000008 00002024 00000000 00000000 00000000

MSR EDX EAX
00000017 000A0000 00000000
0000002A 00000000 00000000
0000002C 00000000 0E12000E
0000002D FFFFFFFF FFFFFFFF
00000198 00000E2D 00000E2D
00000199 00000000 00000E2D
0000019A 00000000 00000002
0000019B 00000000 00000000
0000019C 00000000 00000000
0000019D 00000000 00000E2A
000001A0 00000000 20840089
000001A1 00000000 00000000


Here's what utility says:
7162536797_e46596508f_z.jpg

Can't find any of the settings for CPU multiplier and voltage...
 
See your actual CPU voltages already exceed VID by quite a bit, even while your processor is running at a reduced frequency. Either your board is trying to compensate for increased resistance, or it's damaged in some way.

I also have to think that that much voltage will have generated excess heat. I think the CPU can TAKE more voltage than that... maybe well over 2.0, but it would need good cooling.
 
G

Guest

Guest
Did you monitor your vCore in CPU-Z while RMClock is running?



The following pin mod could potentially lower you voltage but since your hardware is already running out of spec, indicating an incompatibility or hardware failure it is RISKY! No guarantees!!

Consider solutions like a new motherboard, CPU or PC.

Using a strand of short (!) wire bent into a "U"-shape, bridge the following pins:
socket478mod.jpg


 

hwmgr

Distinguished
Jan 19, 2012
53
0
18,630


yes, it's the same vcore reading as it was without RMClock running:
7162969979_089b716e15_b.jpg
and can you further explain the 'risks'??
that if it can fry the processor or mobo??
 
G

Guest

Guest
Yeah. Maybe. I don't really know.
What I'm trying to say is that the the board/CPU combination is behaving abnormally and the pin mod may not cause the desired effect.