I can't back up my Windows 7 Home Premium

[entinggi]

I had the Battery in with the adapter.. Then I tried just with the adapter.. Same thing.. It happen for both the Seatools and the memtest+.. I ran the memtest through usb. is that ok or should burn a CD for it?

When I switch it on back after I ran the memtest+, BSOD appeared.

 

[F5ing]

Best if you boot memtest from USB or CD/DVD, not when running it via Windows.

 

[entinggi]

Best if you boot memtest from USB or CD/DVD, not when running it via Windows.

yeah I did boot it from usb.. probably the what night hawk suggested was true.. The laptop did heat up alot

 

[F5ing]

Figured you must have. I'm not sure it'll even run under Windows.

Review this, starting at Part 3:

http://www.sevenforums.com/tutorials/105647-ram-test-memtest86.html

 

[Night Hawk]

Did you note the error seen on the blue screen itself? That besides the memory dump can be a large if finding out if you are seeing a hardware or OS related problem like a corrupted driver set preventing a normal boot.

The next step now you can try would be booting up in safe mode and opening up the msconfig utility to disable almost every startup item in the Start tab. You can look over the Services tab as well checking off the "hide all microsoft services" box to reduce the unnecessary items for the time being.

The reduction of startup items even disabling all seen in the Startup tab and then suddenly seeing a normal startup without issue will point to software only as the main cause. If you still see BSODs however then you will need to boot live to run memtest from a bootable cd.

The lack of bad sectors being seen with the HDD Tune results is good but wouldn't rule out any possible drive fail. But that also wouldn't explain why the laptop simply shut itself off either.

One other thing to mention here is that memtest doesn't fix the causes for Blue screens. The program stress tests the memory installed on the system and reports on any defects found when left running for a few hours.

On one occasion I found memtest wasn't accurate while being one of the top rated memory testers. The problem seen with one old build was simply the wrong type of memrory not working out on one board. Memtest reported a list of errors. When the same pair of dimms went onto a new board a run of memtest found none! The next board accepted performance dimms while the previous would only run standard ram.

 

[entinggi]

Ok will do

 

[Night Hawk]

Hopefully it prove to a non hardware related problem where a simply drive wipe for full recovery will be all you will need. But a drive failing or some bad ram would need to be ruled out first. The main problem with OEM preinstalls to begin with is once something gets trashed you don't always have as many repair options available.

 

[F5ing]

On one occasion I found memtest wasn't accurate while being one of the top rated memory testers. The problem seen with one old build was simply the wrong type of memrory not working out on one board. Memtest reported a list of errors. When the same pair of dimms went onto a new board a run of memtest found none! The next board accepted performance dimms while the previous would only run standard ram.

Maybe I'm misreading this, but those seem to me to be accurate memtest results.

 

[Night Hawk]

Memtest displayed numerous errors on the first pass suggesting the memory was faulty when it wasn't. It was simply the wrong type of memory for that one board at the time.

Unfortunately there's no error chart to go by with memtest which would indicate the wrong type was installed. The number of errors seen on the initial run however pointed at something being wrong.

At first you think brand new memory was bad and needed to be sent back. Many have asked how do you read memtest results however where you typically won't find any specific charts to go by.

 

[F5ing]

Agree that one's first thought would be bad memory. But if the board cannot communicate properly with the memory it will obviously report errors.

It does take quite a bit of time analyzing memtest errors. I haven't found much info about it either. But I haven't really taken much time looking for it either. I have found you can sometimes pinpoint a bad stick if you know how the board accesses the sticks and observe the error locations. Then swap sticks around and make sure the error shows up in the area you would now expect to find it (rules out a bad slot and definitely pinpoints the bad stick). Really only good though if you reliably get errors on first pass each time. Otherwise it would probably be easier to do the one stick/one slot at a time method (taking longer to go through all the testing though).

 

[Night Hawk]

The usual method for finding out which module is bad news is to run memtest or another memory stress tester with only one dimm present at a time. The bad dimm will be flagged right away when the initial run provided the system posts normallly will see a number of errors.

A slight flaw however might take longer to discover and why many suggest leaving memtest running for a few hours to repeat each test over and over in order to reveal the one needing replacement. If no errors come up at all despite a good lengthy run then you are back to drive for hardware or drive errors as well as an OS problem.

This is why looking over the memory dump files as well as any other error logs as well as taking notes on the error codes seen on any BSOD are the most valuable for discovering where the actual problems are. Once you know the cause(s) then you can work on the solution(s).

 

[F5ing]

The usual method for finding out which module is bad news is to run memtest or another memory stress tester with only one dimm present at a time. The bad dimm will be flagged right away when the initial run provided the system posts normallly will see a number of errors.

A slight flaw however might take longer to discover and why many suggest leaving memtest running for a few hours to repeat each test over and over in order to reveal the one needing replacement. If no errors come up at all despite a good lengthy run then you are back to drive for hardware or drive errors as well as an OS problem.

That's exactly what I'm talking about. If the error shows up on first run and you can identify the stick/slot (based on the report and what you know about how the board accesses the memory) then you don't have to go through sticks/slots one at a time. All it takes is another single pass to confirm the bad stick (when placed in a different slot). Then remove it and do the final run for 7 or more passes to make sure there are no errors. Done!