Congrats, looks like you were right - software does it.

YMBOC,

You are a true PIONEER. Unfortunately you have totally lost me on this one.

Anyway you could "dummy it down" a little so us mere mortals can figure out what exactly it is you did?

I think I understand soldering the socket and adding the new bios chip but where did you get the chip? Did it involve any programing (flashing)?

I know just enough about this stuff (from my XBOX mod days) but after that I am lost. Sounds like you were recompiling a kernel or something????

More info PLEASE!

THX

Thanks for the kind words... but no more of this 'mere mortal' stuff. I'm just a guy who's having a run of good luck... I honestly didn't think I could pull it off when WHS was BSODing.

So it did involve (basic) bios modding. The only reason I got an IC socket soldered is because I expected to screw up at some point and I didn't want to 'brick' my board. I had a spare bios chip which I knew was compatible by cross-referencing it from the IOSS Bios Savior selection table (google it).

The spare BIOS chip (not shown) came from an asus nforce2 board which I also sacrificed in trying to practice surface-mount (SMD/SMT) soldering... which I quickly learned I was not cut out for / didn't have the right tools (ie: hot air rework station or just plain soldering finesse). FYI the original bios chip is currently not installed in my MSS.

Basically using a program called CBROM (v1.82 -- available from forums where bios modding is discussed) I replaced two components of the MSS aka "SJD4" bios with those extracted from donor bios files.

Specifically...
I extracted "AGESACPU.ROM" from a foxconn bios... which is what got the CPU recognized but on its own but still didn't resolve the issues with WHS / WinPE / WHS Install (ie: hangs & BSOD's)... Then I 'extracted' "ACPITBL.BIN" from a shuttle bios... which is what finally got windows to boot. That this foreign acpi table did the trick is just dumb luck, but I'll take it. There may still be side effects of doing this that I haven't encountered yet (I'm thinking the system may fail activation upon restore if an OEM key required for WHS activation is kept in the ACPI table).

I replaced those two components of the "SJD4" bios by first 'releasing' them and then adding them in with cbrom. Finally, booting off of flash media, I used awdflash to flash the bios (using the switches spec'd in 'hot flash' how-to's for good measure).

Aside: Any mention of kernels was related to linux's ability to override a system's ACPI table... its required on many laptops if you want to get all the extra buttons (on laptops) working with linux... its how I could have gotten the x2 chip working (in linux) despite windows refusing to touch it.

Totally COOL

I read it once...I will read it again ....I will "Google" AND someday I may actually attempt this.

YES you are definitely a step above (actually many steps) us mere mortals and your modesty is quite refreshing.

SO what temps does the X2 processor run at?

Sad to say I'm having issues... I've re-applied paste 3 times 3 different ways (and I'm anal about doing paste properly)... screws on the HS bottom out too early and not enough pressure is exerted to spread the paste as well as I'd like... (so rethink doing the AS5 recommended method for x2's -- audi's as5 mod method seemed to get decent & thin coverage when applied in an X but still nothing like what I'm able to achieve on my q6600).

Bios temperature reads much higher than the WHS console... I've been reading about issues with the temperature probes in the brisbane cores... and since its the bios that decides when its time to turn off the computer .... its turned off a few times in my testing (and not just under stress testing).... top it off the rear fans aren't speeding up as the chip gets hotter (I was able to override with speedfan though).

Silly thing is chip is supposed to be only 10Watt hotter than the stock sempron (wth!). And since the bios menus are limited I can't undervolt / underclock the chip. Maybe a be-2300 would run cool enough to avoid the issues I'm having but the be-2400 is all that was available locally.

Aside: I was able to perform a 'factory initialization' / restore without any 'activation' issues.

Followup: Take a look at this update to an abit bios and this [url=http://www.hardforum.com/showthread.php?t=1170002[/url]thread[/url]

Anychance when/if you feel you have the bios running properly you might post a copy of it?

*If* I get past this latest brisbane overheating/overvoltage dilemma I will likely at least post very detailed instructions on how to reproduce the bios I'm using... possibly a batch file to automate the process, a 'diff' file or maybe an md5 file so people can be sure they've got the same file I'm using... I'm a bit wary to post the actual bios -- again if I even get it working, things are looking bleak (again) -- because that IP actually belongs to the original developers.

That said those bios modding forums post their bios' with impunity and apparently without out issue so maybe I should just roll the dice and post it.

I wish there was someone helping me on this stuff... doing it all yourself with precious little guidance is getting a bit old... and to that end: Anyone know which bios module decides how much voltage to supply the processor?

And perhaps the final nail in the coffin for the BE-2400... Bios reports that normal voltages (1.18V) are being supplied to the CPU. I tried to find a testpoint to confirm with my multimeter but didn't find the one I wanted.

maybe a BE-2300 or a 2100 (if someone manages to export one from china) would be cool enough but the BE-2400 is just too hot. Frankly I think its absolute BS that AMD gave this processor a 45watt TDP.

If someone has also purchased an X2 AM2 and wants to try it in their MSS, PM me for a bios... your best bet is with a 45Watt 90nm chip if it exists (I haven't looked)

So this is it for my now-reduced-to-nifty-looking-paperweight-be2400.... (I have no other AM2 systems)... the stock sempron will have to go back in for me... enough money wasted.

hopefully solaris build 89 will have something exciting in store for me... so I can 'perk up' a bit again (zfs root & boot is scheduled for that release and if I'm lucky they'll have snuck in raidz root & boot as well).

Ymboc, the BE-2400 is a good 45 watt processor if you get the right voltages to the right pins. You're on the right track that the bios controls the voltages to the pins. Look what a difference the correct bios implementation made with the abit motherboard running a BE-2400, going from overheating quickly to 30-35 degrees with just the bios change. Now they also used a fan with the heat sink for that motherboard so it indeed may be too hot for the server. The lowest wattage 90mn dual processor I have seen is 65 watts which would probably be too much for a heat sink. A low profile fan might do the job. You are doing a great job so I would hate to see you quit now.

Bios reports 1.18 Volts being supplied to the processor under the monitoring screen... I have little reason to doubt it except that I'd like it to be wrong.

I've checked a few of the bigger caps & mosfets I figured might be carrying CPU voltage but didn't find what I was looking for.

IMO these 45watt processors are the same as the 65watt processors but have been speed/voltage binned. Some of the 45watt chips will be even lower wattage than 45 and some of them will be very close to 45 or above...

I do believe that I've run out of dumb luck.

So... I Re-Checked the unit with a multimeter -- somehow a test point I thought I already tested yielded what I wanted. Test points include either end of The "R82"-things(inductor?) as well as one end of the "NEe / eSB"-things (rectifier diode?).

So I can confirm 1.2 Volts are being supplied to the CPU and put this to bed.

Even if it's not a voltage thing and maybe just a temperature sensor problem... it is said Abit took a long time to release their bios update that addressed the Brisbane overheating / temperature sensor issue. Abit would have had all the resources they needed to fix it as well as contacts within AMD if they needed support in doing so... I on the other hand... well, you know.

Again, if anyone does pick up a BE-whatever or X2-whatever to test... PM me for a bios that will let you boot into Windows and/or properly identify the CPU.

I can certainly understand that! Maybe a rapidshare link will mysteriously show up someday which I'm sure will have nothing to do with you

Nah not rapidshare... personal correspondence... I want to know if & when someone's found an X2 chip that doesn't overwhelm the unit.

Someone *else* roll the dice? please? They're only $54 bucks at newegg... shame they wont ship 'up here'... maybe if you pick one up you can get a second one for me too? any takers? I'll reimburse...

Ymboc are in the us? do know the spec and info for 2100 x2 AMD processor?do you have use a bios mode to use that processor?