Archived from groups: alt.comp.periphs.mainboard.asus (
More info?)
Paul wrote:
> In article <nospam-0307041625370001@192.168.1.177>, nospam@needed.com
> (Paul) wrote:
>
>
>>In article <nospam-0307041343430001@192.168.1.177>, nospam@needed.com
>>(Paul) wrote:
>>
>>
>>>In article <_ozFc.67553$G%.24737@tornado.fastwebnet.it>, "Gimbo"
>>><nospam@nospam.xx> wrote:
>>>
>>>
>>>>I tested these CPUs on a P4PE-XTE. (The BIOS is the 1006 version):
>>>>
>>>>Celeron 2400 400fsb
>>>>Celeron D 2660/256/533
>>>>Pentium 4 2400/1024/533
>>>>Pentium 4 2800/1024/800
>>>>
>>>>The only CPU that works is the Celeron 2400. With the others,
>>>>the mainboard does not power up.
>>>>
>>>>It could be a BIOS problem?
>>>>
>>>>Thanks!
>>>>Gimbo
>>>
>>>Yes, it is a BIOS problem.
>>>
>>>I examined a couple of BIOS files for P4PE family in the last
>>>couple of days and found:
>>>
>>>P4PE1007 BIOS supports family codes:
>>>
>>>0F0A, 0F12, 0F13, 0F21, 0F23, 0F24, 0F25, 0F27, 0F29
>>>
>>>These codes indicate that P4C Northwood 512KB cache is supported,
>>>the older Celerons and P4 processors are also supported.
>>>
>>>The CeleronD probably draws too much current to be safe on that
>>>board. Typical numbers are Vcore ~ 1.0V and current 73 amps. You
>>>can look up the exact figure at processorfinder.intel.com .
>>>
>>>The other two processors are Prescott, and it doesn't appear the
>>>BIOS is ready to handle them. The very fastest Prescotts draw
>>>too much power to be safe. My _guess_ is about 60amps from Vcore
>>>is a reasonably safe value.
>>>
>>>Always consult this web page before buying processors:
>>>http://www.asus.com.tw/support/cpusupport/cpusupport.aspx
>>>
>>>HTH,
>>> Paul
>>
>>Wow, I just took my own advice, and checked
>>
>>http://www.asus.com.tw/support/cpusupport/cpusupport.aspx
>>
>>There are some CeleronD processors listed in there for
>>your board. But you need a revision 2.0 motherboard. I thought
>>all these P4PE variant boards were the same, but I see Asus
>>is still providing new BIOS for your board, as there is a
>>1007.001 beta BIOS for the P4PE-X/TE.
>>
>>The P4PE-X/TE has a two phase Vcore circuit, so that
>>part hasn't changed.
>>
>>The P4PE-X/TE 1007.001 beta bios supports family codes:
>>0F04, 0F12, 0F13, 0F14, 0F21, 0F23, 0F27, 0F25, 0F29, 0F30
>>
>>Now, the Asus cpusupport page says 2.8A/533 90nm
>>processor is supported. On the Intel processorfinder
>>page, one of the 2.8A processors has family code
>>0F33 and the other if 0F34. Power consumption is
>>radically different between them. One runs at 1.25-1.55
>>volts and the other at 1.0 volts.
>>
>>The CeleronD SL7C7 2.8A is listed as family code 0F33.
>>So, again, I don't see support in the BIOS.
>>
>>So, it looks like Asus claims you need a revision 2.0
>>motherboard to run the Prescott processors, and yet
>>the BIOS support isn't correct for them.
>>
>>Maybe the power supply in your computer wasn't strong
>>enough to run them ? That is another possible reason
>>they didn't run.
>>
>>Contact Asus and ask them to get their story straight.
>>All these new processors from Intel give me a headache.
>>
>>Perhaps the revision 2.0 motherboard has a voltage
>>regulator that can supply the lower voltages needed ?
>>I don't even know what voltage regulator chip is being
>>used on those boards. I cannot get any device on this
>>page to line up with the partial part number I see
>>in the picture of the motherboard in the manual.
>>The RT9602 is a dual MOSFET driver, so this is a
>>two chip Vcore power circuit, but I cannot see the
>>whole number on the other chip RT92xx.
>>
>>http://www.richtek.com.tw/ch/product.asp?level1=2
>>
>> Paul
>
>
> OK, I looked up some specs from the Intel site. The first
> spec applies to Socket 478 boards, and the second to
> Socket 775 (included for completeness).
>
>
ftp://download.intel.com/design/Pentium4/guides/25288503.pdf (S478)
>
ftp://download.intel.com/design/Pentium4/guides/30235601.pdf (S775)
>
> S478 motherboards can be designed for four different types of
> processors.
>
> Intel® Celeron® Processor 478_VR_CONFIG_C
> 0.13 Micron 478-Pin Package
>
> Intel® Pentium® 4 Processor 478_VR_CONFIG_C
> Intel® Pentium® 4 Processor 478_VR_CONFIG_D
> 512-KB L2 Cache, 0.13 Micron, 3.40 GHz
>
> Intel® Pentium® 4 Processor P4EE 478_VR_CONFIG_D
>
> Intel® Pentium® 4 Processor on 90 nm 478_VR_CONFIG_A
> 478_VR_CONFIG_B
>
> Configuration Iccmax VR_TDC Dynamic_Icc RLL TOB Maximum VID
> 478_VR_CONFIG_A 91A 80A 70A 1.24 ±19 mV 1.4 V
> 478_VR_CONFIG_B 78A 68A 55A 1.3 ±25 mV 1.4 V
> 478_VR_CONFIG_C 70A 63A 50A 1.5 ±25 mV 1.6 V
> 478_VR_CONFIG_D 91A 80A 70A 1.5 ±19 mV 1.6 V
>
> The CeleronD datasheet is here:
>
ftp://download.intel.com/design/celeron/datashts/30235301.pdf
>
> The CeleronD has a load line curve that is different than all
> of the above, but the closest match I can find is CONFIG_C.
> The CeleronD doesn't use dynamic VID, a VRD 10.1 feature, so
> most P4 motherboards designed for CONFIG_C should be able to
> work with the CeleronD. For the 2.8GHz/533/256KB 90nm Celeron:
>
> VID 1.25-1.4v
> I_max 73 amps
> Power 73 watts (power drawn under non-max conditions, called TDP
> thermal design power, or "how big to make the
> heatsink")
>
> Intel, of course, cannot list the current at TDP, so you divide
> 73W by 1.25 or 1.4V, to come up with an estimate. 58.4 amps or
> 52 amps would be an estimate. The CONFIG_C lists the current
> at TDC as 63A, enough capacity to handle 58 or 52 amps.
>
> So, the only error I see from this exercise, is the Intel
> processorfinder entry that lists the CeleronD as running at
> one volt. This spec sheet says 1.25 to 1.4 volts, and that is
> a more believable number.
>
> Some Prescott S478 info is here:
>
ftp://download.intel.com/design/Pentium4/datashts/30056102.pdf
>
> Prescotts are divided into two groups, low power ones and
> high power ones. A 2.8A/533 or a 2.8E/800 are both classed as
> CONFIG_B types, and run with a VID of 1.25 to 1.4 as well
> (makes sense, as the CeleronD is related to them). The absolute
> max current is 78A, and the Thermal design power TDP is listed as
> 89W. Dividing 89W by 1.25 or 1.4 volts, gives 71.2A or 63.6A.
> Blame the extra power on the larger cache.
>
> Whether a low end Prescott will run or not, depends on whether the
> motherboard was designed for both CONFIG_B and CONFIG_C. Since
> Asus never lists design parameters for the Vcore regulator, the
> only way to guess, is to analyse the circuit itself, by reading
> the datasheet for the specific Richtek regulator being used. It is
> possible that the spec for CONFIG_B didn't even exist when the
> P4PE was designed.
>
> I find it hard to believe a two phase regulator can produce 70 amps
> of current, but that is what the specs call for.
>
> I expect all the problems are BIOS related.
>
> HTH,
> Paul
Hey paul you're doing all the work
I wish ASUS would say something :/
Eventually someone will buy a Celeron D and post on this forum if they
got it working hopefully
Maybe I should buy it and return it if it doesn't work.
Facebook
Twitter
Instagram