Computer Stuttering - Voltage?

My computer has been acting funny for a while. For the last year or so the mouse has been intermittently wonky... hanging and dragging at seemingly random intervals, with the cause disappearing as randomly as it appeared. About a month and a half ago, it starting getting a lot more frequent, and my entire system started experiencing slowdowns. It had been a LONG time since I had formatted and I was about to leave on vacation, so I just lived with it and went away. After getting back from vacation, the problem was still there. About last Friday/Saturday, things got much, much worse. It suddenly went from being an occasional problem to a constant problem that I couldn't get to go away no matter how many times I rebooted. The computer was also suddenly taking about 5+ minutes to boot when it normally takes 60 seconds.

So, I sucked it up and formatted over the weekend. I'm now running a fresh install of XP with all updates and not a whole lot else. The problem cleared up immediately, and my computer boots very quickly again. However, the mouse was still acting wonky... so just for the hell of it I tried another mouse (yes, yes... I know I should have tried that about a year ago). Well, turns out the problem isn't the mouse. The other mouse I tried experienced the same stuttering, but I use that one regularly with my laptop and it definitely is not faulty.

I then started surfing the internet for info on computer stuttering issues. A saw a lot of talk about temperatures and voltages, so I checked Speedfan, with the following results:

https://img23.imageshack.us/img23/401/voltagel.jpg

The temperatures all look fine to me. The GPU has a 'hot' icon next to it, but that's a GF 8800 and those things run hot. That temp looks normal to me for an 8800. However, there's something on there that concerns me greatly. The +12V field is showing less than 2 volts. I watched it for a while, and that's the stable result for that field.

I know absolutely nothing about computer voltage. I'm not even sure what that +12V is measuring. So, I could use some help here.

1) What does the +12V field measure, and what does my sub-2 volt reading mean?
2) If the reading is abnormal, could it be causing my stuttering issues?
3) If the reading is abnormal, what (presumably piece of hardware) is causing the problem?

(1) +12V, -12V, +5V, -5V, etc. refer to ‘rails’: http://en.wikipedia.org/wiki/Power_supply_rail
The names refer to input voltages, but the CPU should never, ever, run on 12V since it will stir-fry itself before you can boot. Just look: on ~2V it is already as hot as 57°C: you are proposing 6 times more is 342°C ‘worth of termal load’ (and that is presumably on the outside [most CPU's do not actually allow you to read core temperatures] where it is in fact much cooler). In many ways a CPU is exactly like the heating element of a electric boiler, a very efficient boiler at that... :juggle2:

The reason why there is a 12V input voltage is that it requires less amps to deliver the same amount of power as it would with 5V input voltage (roughly 12/5 times less amps) -- apart from the fact that with less amps running through a circuit you will also suffer less power loss as a side effect of the fact that the circuitry itself has a tiny bit of resistance and the power loss throughout a stretch of wire is a function (depends on) in that resistance and the amount of amps running on the wire.

(2) ~2V is the actual volts on the CPU. I don't really know much about what these readings are supposed to be for normal conditions; but I always thought that anything over 1.5V was kind of ‘much’.

(3) Software readings are terribly inaccurate. You may want to cross-check with other programs.

(4) Maybe your mouse port is damaged.

EDIT: A quick Google search yields this surprising post: http://www.tomshardware.co.uk/forum/....html#t1748496

Quote:

---

I've noticed a number of posts that indicate a problem with speed fan on reporting voltages. It appears that it is looking at a incorrect address and when it gets an invalid Nr, it reports it's rail limits, ie 0 for min and a very high valule for the upper.

If the computer boots, and is normal - IGNORE them.
as Several have pointed out, use the bios, better yet if you have a voltmeter, use it (always better than software interrupted)

Most PSU's have a tollerance of Plus/minus 3%

My system, speedfan reports:
+12V is 1.28V
-12 V is -16.97
Remainer (Vcore 1 and 2, +3.3, Both +5 abd VBat appear corect.

---

Never trust software voltage readings in any program, they were inaccurate even back in the days of MBM. Go into your BIOS and look at readings there. Voltage will drop on rails depending on how many devices you have attached.

I'm going to guess that you don't have spare parts laying around to swap in and out?

You might try and burn a cd with this image, boot and let it run for 72 hours. Yeah, it's a long time, but if it comes up with errors, that could indicate other weirdness.

Voltage shows as normal in BIOS, so that's a relief. The mouse issue was intermittent after the format; rebooting often made it disappear. It was definitely not a port issue, because the two mouse I tried were plugged into different USB ports.

There's something new now though. Shut down my computer last night without any problems. I was happily playing games on it without any issues. I wake up this morning, turn the computer on, and windows fails to boot. First I see a screen saying "invalid boot.ini" followed by a screen saying that the \system32\hal.dll is missing or corrupt and needs to be reinstalled. The only things I installed yesterday were Speedfan and GIMP (graphics software), so I doubt that's the issue.

My current suspicion is that the HDD is going bad. Thoughts?

[edit] I'm currently using the XP install/boot disk, planning on using it to repair the missing dll. The boot HDD isn't even listed in the available partitions... it's like it's not even plugged in, but it is and BIOS sees it. If there are compelling reasons to think it's something other than the HDD, please let me know because based on what I'm seeing I'm going to pick up a new drive immediately after work this evening.

[edit2] Opened up the case and fiddled with the cables to see if they somehow came loose. The computer still refuses to boot with the same errors as before, but the HDD is showing up again with the windows install disk. However, it shows the drive as ~75gb, and all of it as freespace. The drive is a 100gb drive, so it's not seeing the ~25gb, which approximately corresponds to the space that's already got data on it.

hal means hardware abstraction layer and is a key component of your OS: without it your OS is unable to ‘talk’ to devices (including your harddisk).

Apart from that my guess would indeed be the HDD (second guess would be badly behaved software [deliberately] breaking your boot.ini/hal). Plus, the capacity of the disk (100GB) makes it sound like its an old one too.

Quote:

---

Originally Posted by Tellos Athenaios

Plus, the capacity of the disk (100GB) makes it sound like its an old one too.

---

Yeah, it's an old one. I don't remember how old, but its the last remaining IDE device in my box. If I had to guess, I would say 5 years old.

I have to say, I'm very disappointed with the durability of HDDs. This is the FIFTH HDD I've had actively go bad on me in the last 8-9 years (across a total of three computers). Two of those drives were in a box I use as my house server and which runs 24/365, so that's a bit understandable... but two of the drives have gone belly up inside my personal box, and the other one did so inside my wife's box. Both of us turn our computers off while we're at work and asleep, so they don't clock up a lot of hours. Five HDDs is also too many for just bad luck... these things just burn out often and it annoys me greatly. :furious3:

Still... I'm grateful it looks like it's just the HDD. Of all the devices it could have been, that's definitely the easiest one to replace (now that I've already formatted, that is).

Hmm. I have had one disk fail on me in 3 years, but that disk had been running about 24/7 in the past last year - last year and a half of its life. By comparison it took Windows less than a year to fail on me twice over some stupid software update. Suffice to say it has not been re-installed on the new disk.

EDIT: Also if you have more than one disk in your PC you may want to try the following:

1. Re-configure your BIOS to boot off the second harddisk (or which one you select to be your ‘new’ boot disk)
2. Unplug the broken/faulty disk and any other disk that isn't the ‘new’ boot disk (because Windows installers are rubbish); we need to trick the installer into believing that you are installing to the ‘first disk on the system’.
3. Use some kind of partitioner (possibly the one delivered with the Windows installer) to free up some space in the second harddrive so you'll have room for installing a minimal XP there.
4. Install a minimal XP to that harddrive.
5. Test if you can use that minimal XP ... and see what it left of your old harddrive; you will probably run a fsck or what the NTFS equivalent may be on it if it turns out you can still ‘mount’ it.

Hmm, it seems to depend entirely on the HDD, my trusty old 80GB Maxtor is serving as a replacement for the second time now and it has served me for some years as a main HDD before I turned it into an external. About two years ago I had a failure and used it as a replacement and now I seem to have had a bad HDD again and the Maxtor serves me as a replacement until I get a new one.
It's kinda ironic as well because I often heard Maxtors are cheap while the ones which failed were both Samsungs and their SpinPoint series has a pretty good reputation AFAIK. :laugh4:
I've got to say I kinda love that old 80GB thing, it just keeps working so far, even in that external case that is just metal and can get pretty hot. I almost bought a new HDD this time and then remembered I had that old Maxtor lying around. :2thumbsup:

That's just my experience, hope you can get your PC working again, TinCow.

At least with SSDs you can usually say roughly when they will fail (provided you have a tool that keeps track of the write cycles etc. I think Intel wanted to deliver one for their SSDs), too bad they're still rather expensive.

Quote:

---

Originally Posted by Tellos Athenaios

Hmm. I have had one disk fail on me in 3 years, but that disk had been running about 24/7 in the past last year - last year and a half of its life. By comparison it took Windows less than a year to fail on me twice over some stupid software update. Suffice to say it has not been re-installed on the new disk.

EDIT: Also if you have more than one disk in your PC you may want to try the following:

1. Re-configure your BIOS to boot off the second harddisk (or which one you select to be your ‘new’ boot disk)
2. Unplug the broken/faulty disk and any other disk that isn't the ‘new’ boot disk (because Windows installers are rubbish); we need to trick the installer into believing that you are installing to the ‘first disk on the system’.
3. Use some kind of partitioner (possibly the one delivered with the Windows installer) to free up some space in the second harddrive so you'll have room for installing a minimal XP there.
4. Install a minimal XP to that harddrive.
5. Test if you can use that minimal XP ... and see what it left of your old harddrive; you will probably run a fsck or what the NTFS equivalent may be on it if it turns out you can still ‘mount’ it.

---

I've always run two HDDs for many years now to make formatting easier. The smaller HDD is the OS/boot drive and the larger one is the data/install drive. My current setup is 100gb IDE/250gb SATA, but I just ordered a 640gb SATA off newegg to replace the 100gb IDE. So, that'll make my current data drive the smaller one, which will require me to re-shuffle anyway.

This is possibly one of the reasons I have so many drives go bad... I pretty much use them until they die in hand-me-down fashion. My wife and I both run two HDDs each in this configuration. When a drive gets replaced in our computers for reasons other than going bad, it gets slotted into the server box, which runs 4 HDDs. The server HDDs generally just stay there until they go bad or another hand-me-down drive becomes available from our two main PCs. In fact, every single last part of the server is a hand-me-down, with the exception of the case. It's actually running a GF 6800, and it doesn't even have a monitor! :laugh4:

Tin Cow: The fact that your SpeedFan is registering a 21.2 CPU usage is disconcerting. In my case under normal circumstances(i.e if I donot click on my keyboard or mouse) after a couple of seconds it should show 0-.95 usage at the most. (Current OS:XP-SP2)

This is generally indicative of a serious glitch with one or more applications. In my case I was suffering from a similar problem, but was more fortunate. I used PROCEXP and found out that all my problems were due to NERO which was using upto 40-45% of my CPU/GPU and was the root cause of my problem, forcibly deleting it brought things to normal.

The CPU usage isn't surprising, because the computer was actively 'stuttering' when I took that screenshot. Last night I installed a temporary OS on my data drive and used that to salvage a few important files from the system drive. I expect a new drive to be sitting on my doorstep when I get home this evening, so hopefully it won't take long to get it all reconfigured. I do strongly believe that the HDD was causing the problem. I'll probably start pulling my hair out if the errors keep occurring after the new drive is in.

Firstly if you have even the slightest doubt, then download and run PROCEXP, it's nothing but a advanced prototype of your normal windows task manager. Just see which application if any is being a resource hog, if anything turns up don't install it in your new build.

Secondly what I had done was done purely to avoid going through the hassle of formating the HD and reinstalling the OS, since you are already determined to do so you should have no problems in the future in any case.:crowngrin: