Prometeia-shuts down on Windows starting, random reboots?

[Liquid3D]

Hello all, I really need help. Just got a new Prommie, and having problems.

First: each time system begins to start, or loads into WindowsXP desktop, the Prommie shut's down? The 8RDA alarm goes off while the Prommie goes through it's initial temp drop, until it reaches -45C, and then alarm ceases as system boots. But just as Windows begins, as HDD seems to run, Prommie shut's down? The CPU temp is reading 45C, yet the microfreezer is reading -45C. I checked the contact between micro-freezer head, and CPU-core (prior to tightening down micro-freezer) by ensuring thermal paste was compressed by microfreezer pressure (it was). If contact is problem, there are 3 washers per side now, I've only 2 more to add. I placed shim (per instructions) with cut-outs for protruding circuitry on TBreds ("SMP components"), HOWEVER; I'm wondering if the video anomaly I get even in BIOS at one time, is related to short being caused by large micro-freezer copper head making contact against these circuits (SMP?)? There's another shim which would cover these circuits, but I beleive it's simply intended for Palamino cores which have no protruding circuits around the core like TBreds.

Second: even if it were the case contact wasn't suffcient, and overheating was occurrng, wouldn't the BIOS, and MBM-5 read higher then 45C? That doesn't seem high enough to cause anomlies in BIOS when typing, random characters appeared? I thought perhaps 8RDA BIOS/MBM-5 couldn't read negative values, but MBM-5 can. I've checked and rechecked all the wiring ensuring the reset, and power switches, are connected to mobo from Prometeia, and then to case switches, and LED's. The tension when aligning micro-freezer is very tight, has anyone else experienced this? However directions state either sockets are compatable either vertical, or horizontal?

Third: I immediately shut-down system when I thought high temps were causing anomalies, pulled off micro-freezer and checked for condensation. There was none. So there seems to be enough contact on core from microfreezer to keep CPU at 45C, but even so the Prommie shut's down, when Windows starts, or after temps stabalize, system begins to boot, and HDD is accessed? I know it sounds weird, but that's what seems to happen, the systems begins to load Windows' the prommie slow's fans, and then shut's down? Obviously if temps are 45C to 50C contact isn't suffcient, but why the video anomalies in BIOS when typing? And would a copper surface making contact with circuits protruding up from sides of TBred core cuase a short? The anomalies seem either heat related, or perhaps Antec 480Watt isn't powerful enough? There's a Tyan tachyon on there, plus the voltage required by the heater in the rear cover of the Prometeia which is to prvent condensation. I didn't have this connected at first, not knowing the voltage, but did plug it in during half the process I'm explaining?

Please help my Prometeia brothers/sisters? Or anyone with knowledge of Prometeia? 

[Present]

YOOO,

hey doesn't look like anyone ever responded to your inquiry, so here is what i have found about prommies (and a couple questions of my own)...

i was having problems with the prommie restarting for no appearant reason (though it mysteriously was associated with cd-rom probing/access ;-). after some inspection, i determined it was power related, even though i was using an Antec TruPower 550w. seems the house wiring was very poor, and could not supply enough power for large draws. i moved the prommie to a CyberPower 1500AVR UPS (Uniteruptable Power Supply..battery backup) and it solved that problem.

the windows problem you are having though, may be unrelated. i am having the same problem i think. i have not yet found a solution.

my system "freezing" (sry lame pun), in both windows and linux. it seems to happen quicker if i don't run cpu intensive programs which makes it look like it may be temp related. it also seems much worse, if it is COLD in the room (cooler than 73f). i actually have 2 prommies, an M1 and an M2, and have been running quite stable on the M1 for about 6 months, until about 2 days ago, when i cleaned the air-filter i had installed on the front (netting cooler temps).

i'm trying to get the temps warmer in here right now while playing with the prommie settings to see if i can stablize it at a slightly warmer temp. it seems to run best around -30c. my system seems to become unstable @ -40c.

if anyone has any other possible suggestions they would be appreciated. i'll post what happens with warmer temps in a few minutes.

[Present]

my system spec's for anyone interested:

Prommie M1:
Antec 550w
Athlon 2500+ @ 2.5ghz
Radeon AIW 9000
Crucial PC3200 2@512mb ea
Rocket 404 (Hipt Raid controller)
Win2kPro

Prommie M2
Antec Tru 550w (all rails @ max, just trying max fan currently)
Athlon 2800+ @ ? (need to unlock multiplier 12.5.. more tweak time)
Radeon 7500
Corsair PC4000 2@512mb ea
Maxtor 80gb
LinMdk9.2/Win2kPro

[Present]

just purchased some an electrical compound which is a liquid, but turns to a rubbery insulating coating.

i hypothesize that some of the semiconductors around the core may be turning into a quasi-superconducting state. a light layer of the electrical compound might insulate enough to keep them slightly warmer than the core?

i doubt that the problem is from circuits on the board since the diode under the cpu usually is between -5 and 0, while the cooling head is approx -40.

edited: here's a pic with circles around the circuits...

http://www.pollination.us/images/athlon.jpg

could the close proximity of the evap head cause a change in conductivity?

i would really appreciate some observations/suggestions.

[Gizmo]

Originally Posted by Present

i hypothesize that some of the semiconductors around the core may be turning into a quasi-superconducting state. a light layer of the electrical compound might insulate enough to keep them slightly warmer than the core?

I suspect your problem stems from condensation. Although I have no direct experience with this myself, it stands to reason that if your head is hitting -40C, then there should be frost forming around the components on the top of the cpu unless something is being done to create an air-tight seal between the head and the CPU (not just the core, but the entire cpu). This is one of the few places where some cheap thermal dielectric grease like the stuff you can get at CrapShack might actually do some good. Anything that allows you to prevent condensation from forming will be good, because the condensate, while not a terrific conductor of electricity, can still cause problems.

Just my thoughts. Give it a weight equivalent to what it costs you.

[Present]

the evap head is sealed around the entire cpu, and sealed/insulated on the back of the mobo. i have had several other people suggest condensation tho, so i'm going to pull both systems apart tonite, dry them with hair dryer, put some grease on any components, update bios, then fire it up again. i'll post my results...

[Gizmo]

Originally Posted by Present

the evap head is sealed around the entire cpu, and sealed/insulated on the back of the mobo. i have had several other people suggest condensation tho, so i'm going to pull both systems apart tonite, dry them with hair dryer, put some grease on any components, update bios, then fire it up again. i'll post my results...

Note - that it doesn't necessarilly have to be condensation on the CPU itself. It COULD be on the mobo components in the area around the head that is uninsulated. If memory serves, the unit also has a small heater element that sits on the bottom side of the mobo under the CPU socket. This might not be functioning properly.

[Present]

woot!


AND THE WINNER IS!!!!

CONDENSATION!!

there was visible condensation on the evap head 5-10 sm droplets, and one droplet, pinhead size, on the bottom side of the cpu. there was also a whitish corrosion on 3 cpu pins.

after alcohol and a toothbrush to the backside, fill all contact pts with dielectric grease (that stuff rox!, i used it for all contacts on the mobo...)

we're up and running! 2500+ is running stable @ 2.5ghz

big thx to all who contributed.

one last question. what bridge do i change to get the upper multiplier range? both my 2500+ and my 2800+ will only work up to 12.5x.

i read somewhere that the last L5 bridge needs to be joined to enable 13-18.5, is this correct?

-or-

can i cut the last L3 bridge to add 8 to the multiplier?

-or-

i read the cutting bridge 3 on L12 has enabled some to hit higher fsb speeds on N7's but not A7N8X's? any comment on this? (i tried setting cpu jumper to 200 but it didn't help fsb speeds).

(ps the 2nd system had discoloration on the back of the mobo which was likely due to condenstaion as well. just waiting to know how to get the upper multipliers before i reassemble it)

[Present]

woot, WOOT, & woot!!!!

how simple could it be?

put a strand of VERY SMALL wire (hair size) in the cpu socket between pins AJ27 and AH26. now i have all the upper multi's working perfectly -AND- my mobo clocks stable to 215mhz. i'm checking the best FSB speed i can get right now, before i start bumping the multi, but i'm @ approx 2.9G right now.

many thx again to all who suggested and helped keep the cerebral matter active. also, a big thanks to John C who has a VERY informative site defining all the pins on the AMD and their purpose (well most of the important one ;-).

John's URL...
http://www.beachlink.com/candjac/index.htm

one last question, i noticed that the pin enumeration goes directly from AJ to AH and skips AI, anyone know the reason for this? simply curious.

many thx again,
potent