Default voltages for AMD Thunderbird

[Technoid]

My motherboard are whacko and "overvolts" the CPU at default setting.

Now I would like to know what the official defalt voltage are for a T-B 1333Mhz?

I have it runing at :

In bios: 1,65

In MBM5 : 1,73

Seems to work fine...

NOTE:

I'm not overklocking at all!!

[Guru]

Most mobos set the default voltage to 1.8-1.85V for some strange reason!

[Technoid]

How much volt should it need then?

It works quite fine now and has passed the usual tests;

Win98SE boot without registerscrewup!

Both 3Dmark2000 &2001!

[The PIT]

I hate these things either the bios is lieing or the software is lieing. I thought the default voltage was 1.75v. As long as the cpu ain't cooking and doesn't bomb out theres no need to worry.

[Greebe]

the definitive source for your Q's...

http://www.amd.com/products/cpg/athlon/techdocs/pdf/23792.pdf

http://www.amd.com/products/cpg/athlon/techdocs/index.html

[Reverend Maynard]

Let me guess Technoid, Iwill KK266 mobo? I have the KK266-R and it does that exact same thing. Most of the time it runs at 1.75 volts if I set it to 1.65 in the bios. Every once in awhile it will boot at 1.73 when it is set to 1.65. I actually like this little problem as it allows me to get the voltage significantly higher without perfomring any hardware modifications. To answer your question, 1.75 is what your t-bird should be running.

[Technoid]

No it's actualy a MSI K7T Turbo - R!
It's a shithead monday example!

Its Fast as *** and stable as our swedish mountins but one of the heat sensors is dead the other one is static and it likes to cook the CPU!

My mothers computer has the model without Raid and it has working sensors and don't cook cpu's!

Acording to MBM5 my vcore is 1.75 now. (raised 0.025 in bios)

Thanks Greebe, Reverend Maynard & The PIT

[2Whyzzi]

I bet there is a mathematical answer to your question!

Circuts/Voltages in mobo are designed usually by calculations & percentages. MY rather cheesy explanation follows:

Power Supplies are supposed to be 5v. But the real world is never exact, and the BIOS might say the powersupply reads 5.08v. Close enough, right? Now, manufactures use voltage reguators to cut that lovely 5.08v down to size. But variences in metal quality/connection connuctivity, soldier, etc add/subtract to the power that finally gets to your CPU. So you can expect a
ratio of say +/-5% (I suggesting a guess here, I really don't know - if I felt like reading the IDs on the regulators I might be able to find out) on the conductivity on each voltage regulator. A typical AMD board goes through 6 or 8 voltage regulators (are they MOSFETs?) and if they tend to be on the upper +5% side, then you might get more than the requested voltage.

But then again I am NOT math wiz or a motherboard designer so what do I know...

Whyzzi

[Ali]

Well, going through that .pdf, looks like 1.75 is nominal. with 1.65 being min, and 1.85 being max (Table 6, part 7.7).

Never even thought about that. I have my new 1.4 set to default, and its on 1.85V, and sits at about 45C idle, and 61-62C full load.

This is with a G7DXR, which reads about 10C higher than other mobos though (so Ive been told). Ill have to try dropping the voltage a little, and see what happens.

Good to know the MURC is still usefull for something other that Matrox hardware issues.

Ali

[Greebe]

Ali, Table 7 part 7.8 (page 26) specs VCC_CORE maximum at Max+ 0.5V

ie 2.35vdc

[WayneHu]

It depends on the production series.

For example, some Athlon AXIA 0109 1G can operate with voltage as low as 1.5v. This is far below the default voltage. Mine can operate at 1.625v on KG7-RAID without any timing issue. But both BIOS and VIA HWM report 1.66v.

One more happy thing about this MB. With the new BIOS I received today, the 4 unbuffered dimm issue has been solved. Before that, I just could plug 2 unbuffered dimm without any stability issue.

In the spec of AMD761, it can only drive 2 unbuffered dimm. The reason why KG7-R can drive 4 is not magical. Abit just did the same thing with registered dimm. The only difference is the buffer is not on the module but the MB.

[WayneHu]

Just did the terrible thing suggested by my friend. Vcore = 1.4v (actual 1.43v). Prime95 and 3dmark 2001 run flawlessly. The temperature of the CPU never goes up much even the load is 100%. I should thank to the Copper heat-sink although I just put a 14 cfm 26 dB fan on it.

Actually, I never know my Athlon AXIA has such capaibility.
It seems the the 3rd line of AXIA beginning with "Y" has superior overclock and low-voltage capaibilities.

[Ali]

Greebe, I hadnt noticed that. I wonder if thats without clock though? I dont think I would like to run at 2.35V without a TEC or something.

Anyway, I tried to set my Vcore lower last night, from Default to 1.775 which is as low as my Mobo goes. Im sure the manual said 1.55v min or something. After doing that its still read as 1.85v by both MBM5 and the BIOS. I think somethings broken.

Ive read a bit about voltage problems with the gigabyte, but since mine was delayed for about 3 weeks due to a recall, I would expect the problem to be fixed. Bugger eh?

Ali

[Greebe]

You can run at those voltages as long as you do not exceed the max junction temp. Now whether that be via TEC/water cooling etc it doesn't matter... just as long as that specific spec isn't exceeded.

Of course now CPU's typically do not run well above their Max operational VCC_CORE voltage. It's like a bell curve go above or below a certain voltage and they won't work worth a hoot... try underclocking a cpu sometime... the same applies