A few months ago, I upgraded my computer arrangement, rebuilding to a 64 Bit WinXPPro OS, and an i7975 CPU. I found that I had a few things that did not want to work on the 64 bit system, so also put together a secondary computer with a WinXPPro 32 Bit OS for those purposes. Through WideFS and/or SimConnect the secondary is also implemented for FSC, PFE/Proflight2000, FSM, ASA, AISmooth, and FDC for FS9 and/or FSX.

I normally leave everything runing twenty-four hours a day, seven days per week.

Everything works swimingly with one odd occurance. Periodically, the secondary computer turns off. So far at least, while it is being used for any program or purpose, it keeps running fine, but when left idle, if I turn my back it may shut down all by itself. This may occur within minutes after being idle, or it may wait until the next day and then shut down. I have "disabled all" startup programs with msconfig and she still does the same thing. I also have searched and tried every combination in the Bios settings I can think of that might affect this with no success.

The Motherboard, CPU, and memory were replaced in the secondary computer as listed below. All other hardware remains from the previous setup which never had this problem.

Another curious item: I also replaced my wife's computer with the same type of motherboard and CPU for an upgrade, also running WinXPHome, and there is no sign of this issue. The two computers are essentially duplicates.

Temperatures are not the problem as everything is relatively cool and there are no demanding programs running.

The programs running on the secondary computer are as listed above plus Microsoft Office, Adobe, Nero, Britannica 7.0, Newsoft, PrintMaster16, Quicktime, WinTV, WinZip, and WordSearch8. Note that all of these programs were installed and running fine on the previous computer with similar hardware.

I also use Synergy for a common keyboard and mouse which of course is installed on both the primary and secondary computers.

For antivirus, the primary computer runs ESET Smart Security 4 (NOD32). The secondary computer uses AVG Free Edition. The issue persists with all firewall, antifvirus, and AntiSpam provisions disabled.

The only thing I have not tried yet, which will be next is to shut down my primary computer, and leave the secondary running and see if the issue continues or goes away. I wonder if possibly in some way the primary is shutting down the secondary if it is idle.

Any idieas?

Respectfully:
RTH

The following are my system specs. for each computer. PLEASE do not make cute "everything but the kitchen sink" remarks about my listings as has been done in times past. These documents are saved for obvious reasons and it is just much eaiser to copy them than to make a new condensed list for this post.

Primary Computer:
ASUS P6T Deluxe V2 LGA 1366 Intel ATX Motherboard
Intel Core i7-975 Extreme Quad Core CPU
Scythe SY124010L CPU Fam
6GB (3x2GB) 240 Pin DDR3 1600 OCZ3P1600LV6GK Memory
WinXP Pro 64 Bit
EGVA 01G-P3-1180-AR GeForce GTX 285 1GB 512 bigt DDR3 Video Graphics Card
21” Pro Series P815 Primary Monitor
23" HPw2338h Secondary Monitor
Navadia version 182.50 drivers for GTX 285 Video Card
1 Western Digital Velco/Raptor WD3000HLFS 300 GB 10000RPM WATA 3 GB System Harddrive
1 Western Digital Velco/Raptor WD3000HLFS 300 GB 10000RPM WATA 3 GB FSX Harddrive
1 each Segate 320GB SATA FS9 Harddrive
1 each Segate 320GB SATA Backup / Utility Harddrive
Std. Floppy Drive
2 Each CDRW/DVDRW
PlusDeck 2C Audio Tape Player/Recorder
WINTV
Konica Minolta magicolor 2400W Laser Printer
TI Microlaser Plus Laser Printer
HP 4315v Printer/Fax/Scanner/Copier
CH Pro USB Pedals
CH USB FlightSim Yoke
Microsoft FFB2 USB Joystick
Logitech Keyless Keyboard
Logitech cordless mouse
High Speed Cable Internet/email Service with Router
Corsair 1000 Watt Power Supply
SuperPower ATX Full Tower Case with sides Customized for more ventilation

Note: Keyboard and Mouse are shared with a secondary Core 2 Duo Computer for WinXP Pro 32 bit programs, WideFS6 (FS9) and Wide FS7 (FSX). Implementation is by Synergy Free Software.

Secondary Computer:
ASUS P5N-SLI Motherboard
Intel E6600 Core2 Duo CPU
Zalman CNPS7700-CU RT CPU Fan
4 GB KST DII800 Memory
WinXP Pro 32 Bit
NVIDIA VCG8800GTX 768MB Primary Video Graphics Card
21” Pro Series P815 Primary Monitor off of Primary 8800GTX (Monitor #1
Navadia version 182.50 drivers for 8800GTX Video Card
1 each Segate 250GB UIDE Harddrive
1 Each Segate 80GB SATA Harddrive
1 each Maxtar 40GB UIDE Harddrive
Std. Floppy Drive
CDRW/DVDRW
Epson Perfection 2480 Photo Scanner
MicroTech Z!O! Flash Card Reader
Audigy 2 ZS Platinum Sound to Creative Inspire 5.1 5200 Speakers
BFGR 1000 Watt Power Supply

Note: Keyboard and Mouse are shared with a primary i7-975 Computer for WinXP Pro 64 bit programs, WideFS6 (FS9) and Wide FS7 (FSX). Implementation is by Synergy Free Software.

Hello Alastair:

I have "System Standby" and "System hibernates" set to "Never" in my Power Options Properties.

Temperatures may not have been as good as I first thought.

My CPU temps are in the low 40's C which I would think are acceptable. However, my GPU temperature is running in the mid 80's C which is pretty hot. However, if I remember correctly, it ran this hot in the previous setup with no problem. From what I read on the NVADIA forum, there seems to be quite a debate as to what the limit is (90C if I read correctly). I may open the case and put a fan on it to see if that makes a difference. If this is the problem, the results seem to be reversed to me. Under load when the computer is being used I have no problem. It is when idle and being ignored that it shuts down (wonder if I hurt its feelings?).

Thanks for the help.
RTH

 

 

GPU in mid 80s at idle???   Thats NOT right!

 

 Under a load I would still consider that high but just under acceptable

 

Make sure power scheme is set to "Always on" with no timeouts listed

 

And set screensaver to NONE if any is applied right now

 

This rings of a timeout of some kind, or, a heat situation

 

When you say the tower shuts off.. it 'exits' Windows correctly and powers down? Or just blanks out and goes dead with no shutdown?

 

 

 

 

Update:

Thank you kindly folks for all of the good input and suggestions.

I do have the windows setup for automatic updates, with a bios setting for restarts upon a shutdown. So far, I have seen a number of updates requiring a restart. It always asks first and this requirement seems to have worked fine.

Nick, Murphy seems to be ruling so far relative to finding out whether Windows shuts down first and the power cuts off, or if the computer instantly shuts down with no warning. My back has been turned so far every time it has happened. I walk into the room and the computer is off, unfortunately I was not present when it shut down.

In the power scheme, as mentioned earlier I do have "System standby" and "System hibernates" set for "Never". I did have a screensaver set for 30 mins. "Turn off monitor" set to 45 mins. and "Turn off hard disks" set to 1 hour. I will also reset all of these later to "None" and "Never" if my temperature correction attempts do not cure the problem and see what happens.

As I earlier stated my suspicions concerning the GPU temp and confirmed by several of you, I will address that issue first and see what I come up with. To prove anything it will take a few days at least to be sure the same thing is, or is not happening. I find the following initially:

Note: Ambient room temperature during the following is 22 degrees C (71 degrees F).

Closed case without doing anything, GPU temp as displayed by both "Real Temp 3.00" and "OCT v3.1.0" is a steady 86 degrees C. To my surprise, the card is only moderately warm to the touch. I expected it to be too hot to leave one's paw on the thing. I am assuming this temperature reading is correct as the correct card is identified, "GeForce 8800 GTX".

With the case side removed, the GPU temp cycles between 76 and 77 degrees C.

With an additional reasonably large fan blowing directly on the card from the open case side, GPU temperature cycles between 66 and 68 degrees C.

The GPU fan is running, however I am curious about the speed. In a search on other posts regarding GPU temps, I do see reference to the capability of being able to set fan speed to run at 100% maximum speed in the NVIDIA control settings. Thus far I had assumed that only one speed was applicable and that it was running as fast as it could, but now I wonder. However, thus far I have overlooked any option for this feature in the "NVIDIA SETTINGS" (Taskbar Icon and "nView Desktop Manager"). What tool do I use in determing GPU fan speed, and set it to maximum speed, assuming the power connection on my MB has this capability?

How hot is hot, and what to do about it seems to be the first issue. It is a small case with only one inlet fan, but it is blowing air nicely, and there are no programs even installed on the computer that are demanding. In fact, I see no change in the GPU temperature (closed case) when I start up all programs as listed earlier on the computer. It reads a steady 86 degrees C idle, or with everything cranked up. There are only two cards in the computer, the Video GPU card, and a sound card. The sound card is in the bottom slot far removed from the fan side of the video card. There are no restrictions.

I may just opt for a lesser video card and see what happens. In fact, I think I have an old archived PCI video card in a box somewhere that would probably suffice for a test, but any idieas relative to the scenario above are greatly appreciated.

To be continued:
RTH

 

 

The 285 series has a few issues that tend to be consistent between the manufactures..

Some 285 cards have a bug whereby they display a completely black screen however this issue does not shut power down, at least I have never heard of that happening.. does not mean it wont.

The other problem is the heatsink.. there have bee a lot of reports of heatsinks improperly installed from the factory and I suspect it is possible that may be what is happening here

If a card overheats it can cause a system shutdown.

A 285 card being ID'd as a 8800GTX is not right either.. so there is a possibility this VC is defective

Download GPUz and check that please.. http://www.techpowerup.com/gpuz/

 

 

 

 

 Alright.. well, same deal.. if you are idling in the mid 80's on the VC that does not sound right at all and the 8800GTX is not a 285. They do not run mid 80's normally @ idle and have a lower heat threshold than the 285. They should not be run above 85c for long periods of time but can.. 90c+ and you are into a very bad zone where the 285's are not into that zone till about 100

I do not know how often this shows up but take the side off the tower and leave a fan running on low for a day or so and see if it stops. If it does then there is probably a heat related problem

 Past some strange intermittent issue elsewhere...

The only other hardware possibilities are the CPU and the memory.. or a defective motherboard or PSU (seen defective tower switches do that too). However given the symptoms whereby this never shows under load conditions, one can "assume' (<--- I hate that word) there is nothing wrong in the other areas and center the attention on what does not appear in line which right now is the temp on that 8800 at idle

Its possible the card is simply on its way out

Multiple cards? Well, for the current system it is kind of a yes and no.

You are correct in that my previous system used a combination of the NVIDIA 8800GTX, and an ATI PCI card. Nick, please never think that I am taking issue with your advice as I have the highest respect and regard for it, but ironically they worked well together for two or three years for multiple monitor displays on that previous system. I guess I just got away with it for a while (Old blind sow found an acorn).

However, with my present setup, the Case, Power Supply, Motherboard, CPU and Memory have all been replaced with new hardware. The only pre-existing hardware is the one harddrive, CDROM/DVD drive, and the Floppy Drive. The harddrive was reformatted with a new installation of Windows XPPro32. I now use only the single GTX8800 Video card.

I have not updated the bios on the motherboard from that which was the default version included on the motherboard when I received it. That might be a worthy consideration.

I downloaded and started "NTune". I discovered that my GPU fan was running on automatic and the slider was on 60%. I changed that to "Direct fan control" and manually set the speed to 100%. All temperatures I listed before dropped about ten degrees C. (i.e. Closed case 86 degrees C to 76 degrees C). One odd thing though: If I shut down the computer and restart it, the setting is back to automatic and 60%. I see no provision to save my setting with NTune. It has to be reset back to 100% every time the computer is restarted. A suggestion here sure would be appreciated.

I then added a small side case fan directly adjacent to the video card. It is a puny one, but hopefully it will help a little bit. I probably will order and install a higher capacity model later. Looks like it reduces the temperature at least another two to five degrees. I have not had time for a long time stable line out yet.

I can now crank up the computer from a cold shutdown. GPU temperature will climb reasonably rapidly to the low 80's C. I can then start NTune and set the GPU fan manually to 100% speed, and the GPU temperature will decrease reasonably rapidly. At the moment it has decreased to 72 degrees C and looks like it has lined out.

Now an added piece to this puzzle.

I finally saw it shut down. It was as if the plug was pulled out of the wall. No windows shutdown or warning of any kind. At the time I had both NTune and RealTemp displaying on the monitor and the temperature was between 72 and 73 degrees C and decreasing. I saw it again shut down two more times in progression. My estimate is that it ran about ten minutes before the shutdowns and the temperature was approaching 72 degrees each time.

With a stop watch, it has now been running for almost an hour with no problem and the GPU temperature is lined out and stable @ 72 degrees C. If the same scenario repeats that has been happening for weeks now, it may run until day after tomorrow, or may shut down before I finish typing this.

Obviously, this scenario is going to take some time to find a solution and there will be quite a bit of trial and error along with posteria scratching to isolate the problem. A different video card will probably be next (Probably a simple one that runs cool).

I will keep you posted, and thanks again for the kind interest and help from everybody.

To be continued:
RTH

 

 

I did not say it was not possible to run ATi and Nv together.. it is, and, I have done it myself.. but from someone who has worked with such setups in many different systems my advice to people who are not highly tech oriented is that it is not wise to do and can lead to issues, some of which the user may not know are happening without a 2nd system to baseline a compare, or, a very clear understanding that what they are seeing on the screen is/is not correct in performance and smooth operation.  

The success of 2 platform cards varies from chipset to chipset, and, issues related to installing Windows on the ATI or Nvidia card and then adding or swapping are real and are seen all the time..  case in point just recently:

 

A complete power drop is not going to have anything to do with the power button.. when you push the power button Windows shuts down normally before the system turns off assuming the power properties for "When I press the power button on my computer" is set to: Shut Down

It wont instantly power off

That suggests a failure at a main power connector to the motherboard, the motherboard itself, the CPU or the PSU. The video card is not overheating at the temps you are posting.. none the less, that temp seems very high and it sounds to me like the heatsink on that card is not working right. I assume the card heatsink is clean and not plugged up with dust.

85c with 60% fan at idle just does not sound right unless the ambient in the room is 85 degrees+, regardless that temp is not going to shut the system down unless there is a 'hard' failure of the card taking place.

So at this point, I would start doing a 'wiggle' test..  and do it very carefully. Start at the wall plug, move to the PSU AC input, open the tower and gently wiggle the main 20/24pin motherboard plug.. the CPU 4/8 pin plugs.. and gently see if a little movement of the memory or even the video card may cause a shutdown

also move the PSU wiring harness around a bit

The last item you touch and the system goes dead is connected to the culprit.

If the wiggle test passes then it sounds like a hard intermittent failure of a component and probably not the video card the way this reads now.

 

 

 

 

Update:

The shutdowns continued!

Removed the case termination block and connected the speaker, HDLED, Power LED, Power Switch, and Reset Switch leads directly to the motherboard. Same thing, the computer shut down after running just a few mins.

Tried underclocking the video card to minimum slider settings with the NTune options. The temperature went down to 60 degrees C. However, the computer still shut down.

The problem is accelerating and now shuts down repeatedly anywhere between bootup to about a half hour with or without programs running. I believe Nick was right and the video card is near death.

Removed the video card. Replaced it with an old ATI 3D Rage II Plus PCI card I had in a storage box. Everything else I had archived only fit an AGP slot and was not useable on the Express16 slotted motherboard. To my great surprise everything seems to work, and with display resolution I would have never believed. I feared that some of the programs would give me an error message saying they would not play with such an old dog but all programs seem to work. I may try it for a while as there is nothing too demanding on this computer anyway. I would imagine ASA has the most load and it seems to perform well. The only thing that looks unuseable is WinTV. (The thing only has 4MB Memory Size, boy what a slid show). Fortunately, the WInXP default drivers (2001) seem to work fine as there are none available from ATI for this antique.

For the similar computer I most recently rebuilt for my wife, I installed an EVGA nVidia GeForce 8400GS 512MB DDR PCI Express X16 video card that seems to have worked well for a few months now. The thing only cost $38.00. Internet/Email and Documentation are about her only interests. I may order another for this computer unless someone has a better suggestion.

After the first couple of hours there is no sign of my problem. I will hopefully let it run for a few days and then post a successful conclusion to this scenario if my luck holds.

Thanks again for the helpful interest, suggestions and redcommendations.

Respectfully:
RTH

 

 

If the system remains stable and does not power down then there are one of 4 possibilities here

a. Video card is in fact defective and there is a component on it hard failing

b. The PSU PCIe power rail is dropping, which would also heat the card up too. Low current to a device will drive its temp up. However in this case I would place this in a lower probability than A

c. The PCIe slot on the motherboard has a defect. Could be a mechanical failure of a contact in the slot, a solder joint on the PCB motherboard or a issue in the control circuits.

d. Moving things around has created a coincidence effect into the situation whereby you think the issue is resolved but something else being jarred has corrected the problem, such as in the case of C

 

More than likely since another card is showing no issues, A is the answer... but as I said, I hate the word 'assume'   LOL

 

 

 

 

 

Funny little OT story about the word ASSUME

 

About 2 years ago I was having some intermittent problems with a system which only showed up about once every 2 months, sometimes once a month. One day the issue happened again and I got fed up with waiting for it so I started the process of elimination which included just about everything we discussed above, including electronic testing. I did not find anything wrong but went ahead and swapped out the card.

Everything seemed to work fine and about a month later, BAM

OK.. So to get rid of any possibility of software issues after rechecking the internals I wiped the system and started clean. Two months later, BAM

Now at this point I was ready to trash the motherboard but then I started going back over in my mind everything I had done, including putting scopes on the motherboard and video card to find no issues.

Then it hit me..   the one thing I ASSUMED is that the card was in the slot strait. Visually it looked fine and I have replaced literally 100's of video cards over the years and 1000's of expansion cards when you take into account all the military and aircraft work I have been around.

So I took the card out and used a square to mark a line on the circuit card with respect to the slot depth which I used the standard to ID then reinstalled the card..   it went in strait and the line appeared sold at the plastic port encasement.

2 months later, BAM

 

OK.. so this has to be a bad motherboard at this point..  right? One more check,...

I opened the tower, this time without picking it up,. and looked at the card in the slot. The line was NOT strait. It has moved a solid 1/16+ of an inch out of square. Thats when my assumptions about the situation stopped and the real issue made itself clear.. the tower itself was tweaked and over a period of time the card was slowly sliding out of the front of the slot and creating the time bomb issue. Picking the tower up setting it on its side and opening the tower was tweaking it back somewhat which made the card visually correct and reset the timer.

So I made a little 'concave' adjustment to the tower in my workshop and sure enough, no more 1-2 month crash

So, there is far more to all this than just electronics. Something that simple when placed into a intermittent situation can cost a user hundreds of dollars to find out the problem had nothing to do with the hardware or software.

 

This is an area where big-headed "PC Techs" (especially internet PC techs) and sys admins make a lot of errors.. they ASSUME too much because "they know"  it cant be "this" or it cant be "that"  LOL

 

 

And things like this can get engineers too... its not the first time I have been snookered by myself so I have grown to really despise the word "assume"

 

 

 

 

 

 

Here is what is so bad about that situation with someone like me..

1. The visual corruption that occurred just before the crash I knew for a fact was indicative of losing signal on the first 6 pins of the PCIe slot. That alone, and #3 below, should have had the issue nailed on the first day it occured.

2. When I opened the tower for each check, with the exception of the last, the movement slid the card back into place and since I could not find anything wrong with tests and had removed/replaced./reseated the card several times I assumed that connection was solid.

3. This was the first ‘el-cheap-o’ 125 dollar tower I had purchased in years and that exact issue is one of the reasons I usually never skimp on a tower purchase. As I know those cheap metal towers are usually never perfectly strait unless designed to strict standards. In the computer component biz you are lucky to get one that is strait much less one that is not so far out of alignment issues like this wont happen.. they happen all the time with those cheap towers and I have never bought one since.

Knowing all of the above I should have nailed it the first time it happened without ever replacing the card or tearing it down to test.

Absolutely… Assumption is the mother of all duckups..   *quack*

 

 

That... and buying cheap. Since I broke my own personal rule when it comes to buying computer parts, I got exactly what I deserved in the end