Start a Conversation

This post is more than 5 years old

Solved!

Go to Solution

26007

December 27th, 2014 15:00

Dimension 8300 Freeze-Up

I am running a Dell Dimension 8300 that has served me well for years.  It was running yesterday with a pretty heavy load – Task Manager was showing the CPU memory usage as between 50 and 100 %.  This level of load has happened many times in the past, and the response was just to slow down.  But yesterday, it froze up.  Nothing worked!  So I turned it off by pressing the power button for 5 seconds. 

 

Diagnostic lights A, B, C, D were yellow-yellow-green-green, respectively.  According to the user’s manual, this pattern of diagnostic lights indicated a memory problem.  I performed the suggested trouble-shooting, and I determined that the pair of memory modules (2 x 512 MB hynix) in DIMM 2 seemed to be fine, and the computer would run well with just that pair.  But the other pair (2 x 1024 Corsair) was faulty – the computer would not boot with just those two.  So I removed the bad pair, replaced the good pair, and verified that the computer still operated normally.  This morning I purchased four new memory modules (4 x 1024 MB Patriot) to replace the bad modules and also increase the total RAM from 3 GB to 4 GB.  I removed all of the older modules and installed the new modules.  The computer seemed fine. 

 

When I returned to heavy use – doing some video work, downloading off the internet, several web pages running simultaneously – everything seemed ok until I tried to play a video on my DivX player.  Once again, Task Manger showed heavy memory usage.  The video player never seemed to turn on, and pretty soon the system froze up again, like it did last night.  This time the diagnostic lights formed a different pattern:  green, yellow, green, green.  I could not find this pattern in my user’s manual, so I thought it was time to get some help from the Dell Forum.  Does anyone know what this diagnostic code means? 

 

When I restarted the computer, I noticed that Internet Explorer, by itself, raised the CPU memory usage up to 280,000 kB.  This is when I was just sitting at my home page.  This behavior is very unusual. 

 

The earlier memory modules were hynix 512 MB SDRAM, and Corsair PC3200 1024 MB SDRAM.   The new memory modules are Patriot PC3200.

 

Any suggestions on how to address this problem?

 

Thank you,

 

laserguy

 

Dell Dimension 8300

Windows XP Home, SP3

Intel Pentium 4 CPU, 3.0 GHz with HT technology

4 GB SDRAM

500 GB primary internal hard drive, Seagate

250 GB internal hard drive, Seagate

500 GB external hard drive, Seagate, quantity = 2

Samsung optical drive

NVIDIA GeForce4 FX 5200 graphics card

SoundBlaster Live! sound card 

10 Elder

 • 

44.3K Posts

December 28th, 2014 18:00

Don't know who claims GYGG means a power supply problem. I don't see how BIOS can test the power supply status. BIOS can only indicate when a component fails to respond correctly but power supplies are dumb little boxes.

There are at least 3 error codes in the manual that indicate possible RAM problems: YYGG, GYYY, GYGY.

I'd be more inclined to look at the video card before looking at the PSU, but that's only my opinion. Have you tried reseating the video card in its slot? And blow it clean with canned air. If video card has a fan, make sure the blades are clean and turn freely.

Have you tried uninstalling and reinstalling the DivX player?  Do you have any other player that you could test?

Reboot and press F12 before Windows starts to load. Go to diagnostics partition and run extended hard drive tests. Copy down any error messages. If no error messages, reboot to desktop and then:

  1. Click Start>Run
  2. Type in: chkdsk c: /f
  3. Click OK
  4. Accept offer to run chkdsk on next boot and reboot.

BTW: What color is the power button and is it steady or blinking when the system goes GYGG?

When was last time you replaced motherboard battery?

You could try running free SUPERAntispyware.

Only you can decide how much time and $$ you want to put into this old system...

10 Elder

 • 

44.3K Posts

December 27th, 2014 22:00

I would hazard that G-Y-G-G is a RAM error. Any beeps from the tower?

Have you checked that the new RAM is correctly, securely installed in the slots? You might want to remove and reinstall them.

What specs does the new RAM have? The D8300 uses DDR 333 or DDR 400 NON-EC RAM. And just because RAM may have the "correct" specs, it still might not be fully compatible with the D8300. Dell PCs can be very fussy about RAM. And keep in mind that 32-Bit XP won't use more than ~3.2-3.5 GB of the 4 GB you installed.

When was last time you scanned thoroughly for malware, keeping in mind that XP is no longer supported by Microsoft so there are no new security updates for any new holes that hackers may find.

33 Posts

December 28th, 2014 16:00

ROHE,

Thank you for your comments.  I have some new information on the diagnostic codes, and also on the RAM.

According to the Dell manual that came with the PC, a RAM problem is YYGG.  I searched online for information on the more recent diagnostic codes.  A web page at pctechbytes.com says that GYGG means there is an issue with the power supply.  I also took the online diagnosis at Dell, and they concluded that the RAM was fine.  In fact, they couldn't identify any specific issue.  Instead they made a number of suggestions, including (1) running Microsoft Fix It to address any slow-running issues, (2) searching the Dell knowledge base of ways of determining how some software might be slowing things down, and (3) doing a thorough anti-virus and anti-malware scans (like you suggested).  I will start to go through those steps later today.  By the way, I have Norton Internet Security.  In your opinion, does that constitute a "thorough" scan, or should I use something else? 

You mentioned the beep as a diagnostic on the RAM. Good suggestion.  At the beginning of the story, when I determined that two of my RAM sticks had failed, I heard the beeping sound.  I didn't know what it was, but it was definitely correlated with trying to run with bad RAM.  And when I ran with the new RAM, there was no beep.  So the combination of having a beep with the old RAM and no beep with the new RAM, the fact that I no longer get the YYGG diagnostic lights, and the online Dell diagnosis formally blessing the functioning of the RAM, I am inclined to believe the RAM is ok.  So what else can freeze-up under a relatively heavy load?  Maybe the power supply?

I am puzzled by the suggestion of the diagnostic lights that there might be an issue with the power supply.  I thought it might be overheating, and that might limit the current generation and the ability to run under high operational loads.  But when I had another "freeze-up" late last night (same GTGG diagnostic lights), immediately after I turned off the PC I opened the box to feel if anything was not.  Nothing was hot.  I felt the RAM, the CPU heat sink and the copper tubes in the CPU cooling -- just slightly warm.  Nothing on the power supply box was hot (not even warm), but of course I could not find out if anything inside the box was hot.  Everything appeared normal.

Do you have any suggestions on how to assess the power supply?  I thought about measuring some voltages while it was running, but that means running with the box open, and then it is not clear that any issues relating to cooling would be the same as when the box is closed.  I looked online, and I looks like I can get a replacement power supply for about $ 30 (not sure if that is new or refurbished).  At some point, it might be a good gamble to see if that fixes the problem.   

Thank you again, and I look forward to your further suggestions.

laserguy

33 Posts

December 30th, 2014 19:00

Ron,

First let me thank you for your help.  I tried all of your suggestions, and now everything seems to be working again.  It is not clear whether any one change made the difference.  And, in fact, after I completed all of the changes, the PC still crashed a few more times, but it hasn't now for about 4 hours, even with a heavy CPU load.  But my suspicion, like yours, is that the memory wasn't working properly.  I also think that the DivX player might have been contributing to the problems.

For completeness, here is a summary of all the things I tried. 

  1. Early today I called the Dell tech support.  While they could no longer directly support my XP machine, the guy did confirm that the diagnostic light pattern GYGG also points to a RAM problem.  So following your suggestion and his, I removed all the new RAM and tested them one-by-one, all in the same slot.  They all tested ok.  I also verified that a good stick worked equally good in all four memory slots.  As a final check, I also installed one of the RAM sticks I had previously determined was bad, and it failed again.  This gave me confidence my troubleshooting was reliable.  I replaced all the new RAM after carefully wiping off the pins with lens tissue (like I use for cleaning my camera lenses), and then making sure they were in tightly. 
  2. I uninstalled, and then reinstalled, the DivX player.  It seemed to have some difficulty running with good stability and reliability at first (not sure why) but during the day it settled down and worked reliably for many hours.  Not sure how this happened, but this is what I observed.
  3. I downloaded the information on my new memory, and here is what I found.  The brand is Patriot, PSD2G400K, I GB, 400 Hz, PC3200, non-ECC, 2.6V, 7.8 usec refresh interval. 
  4. I did the disk tests you suggested, first by pressing F12 at start-up, and then by running chkdsk c: (I didn't run the /f -- I first wanted to see if it found any errors).  In both cases, the disks were given a good bill of health with no problems being identified.
  5. I ran Microsoft FixIT, for a total of four separate tests that seemed appropriate to my set of problems.  While several of them identified things to change or delete (which I did), it was not clear whether these changes made any difference, because I still had a crash afterwards.
  6. I downloaded and ran Superantispyware.  If found a lot of junk, all of which I removed.  Thanks for referring me to this web site -- looks like a real nice tool.
  7. I hadn't been checking the color of the power button in the past.  But since you asked, I looked each time the PC crashed today.  It was always steady green -- no amber or flashing.

So like I said above, I am cautiously optimistic that my problems are over, but since I never found a "smoking gun," I don't know for sure that the problem is fixed.

Nevertheless, I greatly appreciate your advice and guidance.  Somewhere along the line, as I followed your suggestions, I found a place where I could once again do the same work I could do before I first had to replace the malfunctioning RAM sticks.

 

10 Elder

 • 

44.3K Posts

December 30th, 2014 20:00

Glad things are settling down.

If SAS found/removed a lot of junk, that doesn't necessarily mean your system is now totally clean of garbage and malware.

You might want to register at spywarehammer.com for free malware removal advice from the experts. Some of the people over there also contribute to this forum so they know Dell systems well. If you decide to go there, be sure to read their instructions about posting a request for help and include your PC model, version of Windows, symptoms and a link to this thread so they can see what you reported and what you've done to get this system back to good health.

33 Posts

January 1st, 2015 12:00

Ron,

One final chapter.  The morning after my last message, the PC was back to its old tricks.  I could not do anything, even very light processing, without a freeze-up.  I opened the box to have another look, and I noticed that the heat-exchanger fins for the processor chip were almost totally clogged.  This obviously was limiting the effectiveness of the cooling system.  After I cleaned out the fins, I looked at the air-intake region at the front panel.  From the outside it looked somewhat clogged, but when I looked from the inside, I found that it was totally clogged with coagulated dust.  I thoroughly cleaned out that area as well, and then restarted the computer, anticipating a successful run.  There was a noticeable increase in the air flow at the output of the exhaust fan on the back side of the box -- good news.  However, there was essentially no improvement in performance.  I successively tried increasingly less aggressive processing throughout yesterday and today, and it kept freezing up after a half-hour or so.  For the final test this morning, I turned on the machine, but I didn't open any applications.  I just let it run at its start-up conditions.  It froze up in about 40 minutes. 

I can't prove anything, but my best conjecture right now is that, over the 10 years of operation, as the cooling passages became clogged up, the resulting thermal load may have systematically permanently degraded the capability of the CPU chip.  And now it is incapable of operating longer than 30 to 60 minutes, even under a minimum operational load.  If this is the case, the best next step is to give up on this machine (no point in buying a new CPU chip for a 10 year old PC) and buy a new one. 

What do you think?  Is this a possible failure mode for an overheated CPU chip -- a gradual and permanent degradation in capability?

Thanks again for your support and guidance.

laserguy 

    

10 Elder

 • 

44.3K Posts

January 1st, 2015 13:00

You may want to remove the heatsink, clean both surfaces and apply a fresh, thin coat of thermal paste, eg Arctic Silver.

Check the manual for instructions to remove/reinstall the heatsink. But don't remove the CPU from its socket...

10 Elder

 • 

44.3K Posts

January 3rd, 2015 18:00

There are quite a few sites that show you how to remove a CPU that's stuck to the heatsink, like this youtube.

I don't recommend any specific method, but I think soaking in alcohol is too risky. So you'll have to decide what will work for you. Whatever you do, be gentle and careful not to damage the CPU or scratch the surface of the heatsink.

And when you reinstall the CPU, be sure to align pin 1 on the CPU with "hole 1" on the socket (they should both be marked) and press gently / uniformly on the CPU so all pins seat without bending...

And I take no responsibility, so if this gets to be too much, perhaps a geeky friend or local PC shop can help...

33 Posts

January 3rd, 2015 18:00

Ron,

Ok, so I read up on how to remove the heat sink.  I followed the instructions and it was easy to remove the heat sink.  Unfortunately, the CPU stuck to the heat sink and came out with the heat sink. 

The obvious next step is to remove the CPU from the heat sink, clean up their respective thermal contact surfaces, then return the CPU to its socket and lock it in place with the small lever arm.  I thought about trying to replace the CPU while it's still stuck to the heat sink, but that seems too risky to me:  high probability of either inserting the CPU into incorrect holes (because I won't be able to see what I am doing), or bending the pins, or both! 

Before doing something I might regret, I thought I would check with you and get the benefit of your advice.  In particular, any special suggestions of how to remove the CPU from the heat sink?  I probably don't want to pry it off for fear of breaking the chip, or marking up the mating surfaces.  Perhaps I can slide it off sideways, but the thermal compound may make this difficult. 

What do you suggest?

Thanks,

laserguy

33 Posts

January 11th, 2015 12:00

Ron,

Sorry to take so long to respond.  I removed the CPU, cleaned the heat sink interfaces with alcohol, etc. and was attaching the heat sink when one of the green plastic clips that hold the heat sink broke.  I didn't want to run withoug the clip, so I ordered some online and it wasn't until this weekend that I could resume. 

I ran the PC continuously for 9 hours, and I thought everything was ok -- previously it wouldn't run for more than 30 to 40 minutes.  But then it crashed.  So I repeated the process with the CPU (this time I didn't break the plastic clip) and ran again.  This time I turned off the screen saver so I could view whatever was on the screen if the computer crashed again.  It did crash, and I noticed a message in the frozen image stating the Norton Internet Security had just started some background work.  I recalled that I had seen the same message after one of the other crashes.  I went online and discovered lots of problems of late versions of NIS on XP machines.

I uninstalled NIS and successfully ran the PC for more than 12 hours without any problems.  I went online on my other computer and determined that Vipre internet security is supposed to be better than NIS, particularly because it does not load down the system so much.  So I purchased Vipre, installed it, and now I will continue operation and see what happens.

So my problem may have had both hardware and software origins.  I will return with additional comments once I see how my luck goes in the coming week.

Thanks again for your help.

laserguy

10 Elder

 • 

44.3K Posts

January 11th, 2015 14:00

Microsoft is still providing support and updates for Microsoft Security Essentials on XP machines, at least for a while longer, and it's free.

Those heatsink clips get brittle after a while from all the heat and break easily. Sometimes they bend out of shape and don't keep the heatsink firmly attached to the CPU which results in overheating and crashes..

Don't know how much more time or $$ you're willing to invest in such an old system, but it may be time to start saving pennies for something new once Win 10 gets released this spring. That way you'll get security updates from Microsoft and the new systems are incredibly faster than the 8300.

No Events found!

Top