System unstable under load: MB to blame?

[redzo]

Some other thing: you could use this little program to comfortably view and navigate through BSODS.
http://www.nirsoft.net/utils/blue_screen_view.html

 

[Ketchup]

Get a battery backup with AVR and see what happens.

 

[dbr1]

Get a battery backup with AVR and see what happens.

Ordered, fingers crossed.

Plan B: build a new, low power system...

 

[BonzaiDuck]

I've never seen anything like this personally, but I've run into occasional situations wherein household wiring is at fault.

Even if you correct some "site-wiring fault" -- if one exists -- I'd get a UPS for your system. I've known "mainstreamer" folks -- friends -- whose computers "went south" in the last three years, and I discovered they were just using surge protection in a wall socket.

That may be fine for some, but after an electrical storm in Virginia during 1993, I began equipping every computer I own with a UPS.

We could trace down all the error codes you discovered, but your own persistence in testing wall-sockets seemed to uncover a likely problem.

Ultimately, if you're not "electrically savvy" yourself, I'd call an electrician and have him test your household wiring -- especially where this trouble occurred.

EDIT: ALSO -- you cannot count on a UPS to resolve a site-wiring problem. The UPS itself is likely to indicate such a problem with an LED indicator or warning.

 

[dbr1]

EDIT: ALSO -- you cannot count on a UPS to resolve a site-wiring problem. The UPS itself is likely to indicate such a problem with an LED indicator or warning.

My thought is that perhaps there are sporadic voltage drops that when the PC is requiring a lot of power will cause system failure. A UPS maybe would stabilize/buffer the power. We'll see.

Also, I don't know if any random electrician would know enough about this type of question, I will probably try to find whatever I can about this type of problem first.

 

[Ketchup]

My thought is that perhaps there are sporadic voltage drops that when the PC is requiring a lot of power will cause system failure. A UPS maybe would stabilize/buffer the power. We'll see....

After posting (I still think the battery could help) I was pondering this. How interesting that the system has time to write a minidump while encountering a low power situation. Hopefully, we can all look back at this with a good laugh soon.

 

[dbr1]

After posting (I still think the battery could help) I was pondering this. How interesting that the system has time to write a minidump while encountering a low power situation. Hopefully, we can all look back at this with a good laugh soon.

Found this somewhat interesting article from 1999: http://www.copper.org/publications/newsletters/innovations/1999/03/power.html

It's usually a software problem, maybe accompanied by an ominous message like "YOUR SYSTEM HAS PERFORMED AN ILLEGAL OPERATION " or "INSUFFICIENT MEMORY TO RUN THIS APPLICATION". But sometimes it's not the software's fault. It could be a question of inadequate power quality. Power quality? Since when does power have "quality"? It's either on or it's off, right....

​

...Voltage variations are another common source of problems to home computers and other sensitive electronic equipment. Voltage variations can be positive (higher than normal) or negative (lower). Variations can be huge, reaching thousands of volts, but even small ones can cause problems. ​

 

[BonzaiDuck]

Found this somewhat interesting article from 1999: http://www.copper.org/publications/newsletters/innovations/1999/03/power.html

It's usually a software problem, maybe accompanied by an ominous message like "YOUR SYSTEM HAS PERFORMED AN ILLEGAL OPERATION " or "INSUFFICIENT MEMORY TO RUN THIS APPLICATION". But sometimes it's not the software's fault. It could be a question of inadequate power quality. Power quality? Since when does power have "quality"? It's either on or it's off, right....

​

...Voltage variations are another common source of problems to home computers and other sensitive electronic equipment. Voltage variations can be positive (higher than normal) or negative (lower). Variations can be huge, reaching thousands of volts, but even small ones can cause problems. ​

I'm assuming that you use a surge-protector in lieu of a UPS. If you didn't, you should know better. If you did (most likely from the quality of your posts) -- then you possibly moved the surge protector to the other power source? Or did you use a different surge protector at the new power source?

As much as I remember, some surge-suppressors/protectors use components called "varistors" which eventually wear out.

Also (again from my hazy 65+ memory), what else is connected to the power-line that gave you trouble? I'm guessing that many households have two or more "trunks" -- each of which powers its own string of wall sockets. In the old days of Commodore 64 (etc), a light-bulb turned on and attached to the same power-line would stabilize electrical power. But if your computer was connected (especially without surge protection) to a line powering a refrigerator, when the refrigerator powers on, it can cause a "hiccup" which would crash a computer (of that day, in any case.).

Shooting in the dark here! But if you're sure that your system suddenly ran stable under stress at the alternate power source, I'd think that says something.

 

[dbr1]

I'm assuming that you use a surge-protector in lieu of a UPS. If you didn't, you should know better. If you did (most likely from the quality of your posts) -- then you possibly moved the surge protector to the other power source? Or did you use a different surge protector at the new power source?

As much as I remember, some surge-suppressors/protectors use components called "varistors" which eventually wear out.

Also (again from my hazy 65+ memory), what else is connected to the power-line that gave you trouble? I'm guessing that many households have two or more "trunks" -- each of which powers its own string of wall sockets. In the old days of Commodore 64 (etc), a light-bulb turned on and attached to the same power-line would stabilize electrical power. But if your computer was connected (especially without surge protection) to a line powering a refrigerator, when the refrigerator powers on, it can cause a "hiccup" which would crash a computer (of that day, in any case.).

Shooting in the dark here! But if you're sure that your system suddenly ran stable under stress at the alternate power source, I'd think that says something.

Did not move my surge protector to the other outlet for my little test, but I did first just plug straight into the wall in the outlet that I was already using to see if the problem was the surge protector. It made no difference.

On that same circuit there are only lights and my PC and peripherals, no large appliances.

 

[BonzaiDuck]

Did not move my surge protector to the other outlet for my little test, but I did first just plug straight into the wall in the outlet that I was already using to see if the problem was the surge protector. It made no difference.

On that same circuit there are only lights and my PC and peripherals, no large appliances.

Just FWIW -- Maximum PC in a recent issue featured a letter from a reader who wanted to donate his computer to the local boy-scout troop. It works fine at his home; at the boy-scout building/facility, he had the same problem you do. You might want to track that down and take a look at it.

I really hope you get this squared away. I personally made one unnecessary RMA this year for an HDD, and an unnecessary purchase for RAM sticks I'd wanted for some time but don't need right away -- all as part of a troubleshooting effort that dove-tailed with planned hardware upgrades.

It's the prospect of troubles such as yours that keeps everyone and anyone on pins and needles until they're sure their system is stable.

 

[westom]

Did not move my surge protector to the other outlet for my little test, but I did first just plug straight into the wall in the outlet that I was already using to see if the problem was the surge protector. It made no difference.

Voltage can drop so low that incandescent bulbs dim to 40% intensity. Voltage that low is ideal for every properly constructed computer. Because a power supply's job is to input massive voltage variations while outputting rock solid and stable DC voltages into electronics.

After all this, what was accomplished? What subsystems are now known good - without doubt? If none, then nothing has been accomplished. Any testing that ignores perspective (numbers) results in "I think it might be" answers. IOW nothing accomplished.

If a home's foundation is failing, do you spend time and money planing interior doors so they close? Of course not. Always start by inspecting or fixing the foundation.

Same applies to computers. Its foundation is a power system (more than just a PSU). Establish integrity by obtaining numbers. IOW get a digital meter, ask for instructions, perform the full minute of labor, and then get replies that say what is good and what is suspect. Without any more speculation or "it could be" answers.

A defective power system will still boot and run a comptuer for years. Then slowly get worse so that many other subsystems acts as if defective. Power system is the foundation. Only way to establish a good or suspect power system is numbers from a meter.

However a second alternative exists. Simply keep replacing good parts until something works. Those are your only two options.