Big Shuttle problem, please help

015734

Solid State Member
Messages
13
I have a Shuttle ST20G5. The problem started last night when I started to play a game of Supreme Commander Forged Alliance which I have done many times before. When the game loaded I got a BSOD Stop error. When I tried to restart, the computer would not even boot to bios. I had to power down using the button and unplug PSU then reboot and only then could I get back into windows. When it said Windows has recovered from a serious error... it said it was a driver problem. Thinking video card, I updated my drivers but the problem now appears whenever I reach or even get close to 100% cpu.

The shuttle came with a 2.2ghz 3500+ and a 6800gt graphics card by xfx. Not happy with the performance, I updated the card to an EVGA 7950gt about 3 weeks ago and the processor to a 4800x2 about a week after that. It ran smooth for 2 weeks and SUP COM FA ran great. However, now in bios the 12volt + rail only measures 11.73 volts and some of the other voltages including the CPU are also low. Could this be a PSU problem? Video Card? CPU? RAM? I know a 4800 and 7950gt are a bit for the stock PSU, however, other peopel have used them fine before. Or... could this be a software problem? I have all the latest updates for windows including optional ones, however, I can no longer even check for updates, as the computer will shut off
 
as you said, im guessing its a psu problem, 12v should be 12.10, 12.2 something in there, it should never go below, just a couple tenths of a volt above, also when the cpu gets to 100% its obviously taking more power, and if your power supply is going then it cant provide enough power to the processor thus the bsod, so i would advise getting a new psu, a good 600watt psu should be fine with a couple 20A rails
 
Thanks for the quick response. I should have mentioned that the stock supply is a 240watt Shuttle and the upgraded one is a 300watt. I would love a 600, however, the PSU will physically not fit in the SFF case :(

Aany way that this could be related to RAM. Right now I have two 1gb sticks from Nspire. It would be easy enough to testso... what do you think?

Edit: I tried the computer without my 7950gt and I still ran into a BSOD almost immediatly... which elimanates the Video card and also makes me doubt the power supply is the culprit...
 
hmm, this is VERY weird... how did the voltages look w/o the video card in?

it could possibly be the RAM, do you have another computer you could test the ram in? or a buddies pc you could pop it in to test? i would suggest getting a ram testing program, but if your pc is bsod'ing thats obviously not possible right now... just see if you can pop it in another pc and give it a try

what do your temps look like processor wise? judging by the original graphics card that came in that pc, im guessing its at least a year old?? might try blowing compressed air through the cpu heatsink, also does the fan spin ok?
 
Thanks for staying with me. Right now I am typing this on the machine, as it is fine as long as I don't put it under load. I should have checked the voltages, however, I didn't think to get a reading in bios ... and now I have the card back in. One thing which makes me wonder whether or not it is the power supply is that if it crashed with and without the card, that means that the GPU is fine, and that means that it isn't the 3.3 or 5volt rail giving me problems. I have a DMM right next to me and was wondering if there was any way to measure the voltage on the rail right now (pull out DVD rw power and meaure that?)

The CPU temps are fine, especially for a SFF case. I just installed it about two weeks ago, used Arctic Clean and then Silver 5. The temps idle in the uper 30's, low 40's which is fine, as the Silver 5 can take up to 200 hours to break in and there is not much airflow in a SFF case as small as the st20g5... That is why I have to put faith in the ICE system used in most shuttles

Oh one more thing, the PSU is warm to the touch right now, bout 90-95 degrees surface but not extreamly hot (feels good to cold hands ;))
 
Alright, an update. I have narrowed the problem down to RAM slot number one, as both of my 1gb chips work fine in slot 2, however, if I try either in slot 1, the machines gets a BSOD when I try to do something intensive.

Is there any way to fix this or what even caused it?

If it comes to replacing the motherboard, is it worth it? I would run on only a gig, however, games like Supreme Commander and a few others need more, not to mention Google Earth, etc.

Where could I replace the motherboard, as it is not a normal sixe? Through shuttle is my only option?
 
well if you've narrowed it down to the ram slots, you probably need a bios update, try flashing your bios to the newest version of your bios, hopefully that will fix any ram issues

as far as replacing your mobo, what socket is your processor 939 or AM2? im guessing your 6800 is pci-e? and your ram ddr2? once i know this stuff i can find a new mobo...

what type is it? im guessing its just a simple microATX mobo, there are two major sizes ATX (w/ 6 or so pci slots) and micro atx with maybe 2-3 pci slots, yours is probably microatx for that small of a case therefor a standard microatx motherboard should work, although you might want to consult shuttle on whether or not it is a standard microatx motherboard...
 
Aright, going to try to do my best explaining the computer.

As per CPU-Z, my motherboard is manufactured by Shuttle with model number RS480-M1573. The first (RS480) is the chipset, also goes by the name of Xpress 200 (by ATI). The second (M1573) is the Southbridge made by ALI, while the LPCIO or Northbridge is IT8712 made by IT.

The motherboard is a socket 939 and thus that is why I went with the 4800x2 (best upgrade for that mobo).

The RAM is two sticks of Nspire 1gb PC3200 DDR RAM. There are two slots on the motherboarrd for RAM, a black slot closer to the outside of the case and a blue slot further inside the case.

The blue slot (further back) will work fine with either of my sticks (through stress tests, RAM tests, video encoding, etc) so I figure both sticks are fine.

However, when I put either stick of RAM in the black slot (further out), the computer will load windows fine, however, once I try to do something RAM intensive (like encode a video) instant BSOD or black screen with no POST. Have to unplug, press power to discharge caps, plug back in and only then will it start up.

You are right, the Vcard is PCI-E 16x which is why both the 6800gt and the 7950gt works great. The fact that I removed the Vcard and the problem still appeared even on integrated graphics rules that out entirely

...and the fact that I can encode or do whatever with the blue stick only means that my process is fine, not overheating, etc. I learned that shuttle on theST20G5 does not provide accurate temps for the sensors, so my real CPU temp is probably upper 29's lower 30's when idle.(I could put my hand on the heatsink and it isn't close to burning. )

The PSU, though rated for only 240 watts can easily put out 325 total and probably more based on the max amperages listed on the different rails. This and the fact that the PSU is outputing close (within 5% on all rails with the 3.3 and 5 within a .1volt) is good enough for my computer. That rules that out too.

So to recap... (Only what I think but still)

*Video Card is fine (I took out my 7950gt and ran on integrated and same problem in same spot, no POST, etc.)

*PSU is fine (have encoded for over an hour yesterday using one good slot, both cores maxed, HD spinning, Vcard working, etc with no problems. The PSU is still touchable throughout) (not close to burning)

*DVD-RW and Floppy are fine, removed them and same problem with black slot (have to test everything ;))

*Not an external FW device or USB causing problem... disconected everything except LAN power DVI, keyboard, mouse...

*Not the case causing overheating, etc (tried with case off and on, same problem)

*Either RAM stick works in the second (Blue) slot no problems

*Neither RAM works in the first (Black) slot

*CPU should be fine... (unless something with memory bus, etc but everything is working fine, both cores work great, even at 100%)

*Forgot to mention: Flashed my Bios to the latest version before installing CPU... when problem began I loaded optimized defaults but have not tampered with it any other way. Should I reflash?

For Reference, my CPU is the following:

*ST20G5 Shuttle SFF about 1 1/2 to 2 years old, stock MB
*2GB (well... 1GB now) of Nspire 3200 DDR 400 ram. (Also 1 1/2 to 2 years old)
*EVGA 7950gt Graphics card at stock speeds (about 6 months old)
*AMD 4800x2 running at stock 2.4ghz (Codename Toledo) (About 3 weeks old)
*Stock PSU 1 1/2 to 2 years
*floppy drive 1.44Mb, came with it for ghosting... never really used it
*DVD-RW use for burning stuff...
 
hmm this is tough, thanks for all the info though...

you might want to check shuttle for some answers as far as BIOS updates go, i saw that you did flash when you put the new cpu in, thats a very good idea to reflash just in case, if that doesnt solve, you might need to buy some cheap kingston ram, im running 2gb pc5300 of kingston value select and its great ram with awesome price

but if flashing the bios doesnt fix the blue slot problem, this could be a motherboard problem and you might need to get a new motherboard, hopefully it wont come to that, but if it does that might be the safest road to go..

let me know how the bios flashing goes, and ill help from there
 
Back
Top Bottom