Saturday, April 30, 2011

It's ALIVE!

"Imagine if every Thursday your shoes exploded if you tied them the usual way. This happens to us all the time with computers, and nobody thinks of complaining."

-- Jeff Raskin, interviewed in Doctor Dobb's Journal

Last weekend, I had a little "computer incident" that immediately reminded me of Mr. Raskin's observation. I climbed the stairs to my attic, as I usually do, pressed the power switch on my Dell Dimension 9200's console, and was rewarded with nothing but a blank text screen with a blinking cursor at the top left-hand corner. This is not abnormal, except that it normally lasts for only a second or so before the computer proceeds on to the usual BIOS startup blurb, memory test, etc. This time, this did not occur. The blank screen just sat there, impudently blinking its cursor at me.

I wasn't concerned. This sort of thing happens once in a while. Some part of the boot sequence doesn't quite kick in at the right time or in the right way and the whole thing just hangs. I call it a "hiccup". The normal solution is to just kill the power and try again, and that's just what I did. I pressed the power button, held it until the power was cut, waited about 30 seconds, then pressed the power button again.

The console lights came on, the hard disk indicator blinked briefly, I could hear the whirr of the cooling fan, but the monitor stayed black. This time, there wasn't even a cursor. You know how most monitors go into a sort of low-power "sleep" mode when they don't detect any signal coming from the computer? Mine does that. You can tell when it's in that mode because its power indicator light changes from green to orange. Well, the monitor's power indicator light stayed orange, which told me that the computer wasn't talking to it at all, in spite of the fact that it was running. Hmmm... curious.

I tried cycling the power again, with the same result. Now I was becoming concerned and just a bit frustrated as, without any kind of monitor display, diagnostic options become very limited...

...but not non-existent. The Dell Dimension 9200, like most desktop computers, has an array of diagnostic lights on its front console for when something is really out of whack with the hardware. In the case of the 9200, these lights consist of the numbers "1" through "4". When all is well, all four numbers are dark. When a problem is detected, different number combinations light up, depending on the problem of course.

That's why it's a very good idea to keep that owner's manual, folks! Oh, sure you could call Dell Support, but we all know what that's like, don't we? First of all, Dell usually publishes their call center number either on-line (to which one hardly has access if one's computer isn't working) or somewhere in the owner's manual which all too many people discard or lose.

Even if you do have the call center number, then you're going to "talk" to an automated attendant, press 113 buttons on your touch tone number pad before the automated attendant finally understands that you want tech support, then sit there on hold for at least 68 minutes listening to Mantovani's instrumental rendition of "Dirty Deeds Done Dirt Cheap" while periodically being told (again by the automated attendant) that "we are experiencing a higher than usual volume of calls" (Really? You mean half of North America's Dells have decided not to boot this morning? Maybe it's sun spots or a big solar flare) and being assured that "your call is important to us" and heaven help you if you hang up in frustration because the next time you call back you're gonna have to do this all over again, bucko!

If and when you finally get a live tech on the line (most likely with an accent that convinces you that Mahatma Gandhi is alive and working at a Dell call center) and you explain to him that your computer won't boot and there's nothing on the monitor, he'll ask you probing questions that would never in a million years have occurred to a lay person such as yourself, such as "Is the computer turned on?" and "Is the power cord plugged into an electrical outlet?" Eventually, he'll finally get around to having you look at those diagnostic lights on the console and, if your lucky, he'll know what they mean (only because he still has his copy of the manual there in his cubicle). Trust me, it's a whole lot faster and easier to just keep that owner's manual and look it up yourself!

I don't mean to denigrate Dell's support staff. I'm sure that most of them are knowledgeable professionals with a sincere desire to help and a voracious appetite for curry. To be fair, being a computer-literate techie myself, I probably know at least as much as most of the people I'd get on the other end of that phone line. For those of you out there who think that a POST is something that holds up a fence, maybe you'll find it helpful to talk to these people. I'm usually better off looking into it myself.

Anyway, sure enough, two of the diagnostic numbers on my console were lit up; "3" and "4" to be precise, and I did happen to keep my Dell owner's manual. Unlike all too many owners' manuals, whose idea of in-depth technical information is to tell you where the power switch is, my Dell manual actually has a lot of good, useful information in it, including a trouble-shooting section that covers those diagnostic lights, what they mean, and even what to do about it when they light up. The "3,4" combination, as it turns out, means "Memory modules are detected, but a memory failure has occurred". Damn! It looked like one of my memory chips had died on me.

The manual went on to suggest, very sensibly in my opinion, that I should remove all the memory and then re-insert the modules, one by one, until I found the bad one. My PC has only two memory modules; each with a capacity of 1 gigabyte. So I shut down the computer, opened the case and pulled the second memory module, leaving only the first.

Here's where those less technically-adept and adventurous than myself might get a bit skittish and, indeed, I would not recommend ripping things out of your computer's motherboard without at least a rudimentary knowledge of how things are put together. It should go without saying that, if you're going to try this sort of thing, always kill the computer's power first. In fact, you may want to unplug the electrical cord, just to be safe. Unplugging and plugging componets into a powered circuit board may well fry the component, the circuit board or, in a worst-case scenario, you. Always remember the Seventh Commandment for Technicians:

Work thou not on energized equipment, for if thou dost, thy fellow workers will surely buy beers for thy widow and console her in other ways.

That having been said, removing and inserting memory modules on a motherboard is a relatively simple affair. They tend to be held in place by two plastic clips; one at each end. Just gently bend back the clips, firmly grasp the chip, either in the center or at both ends, and pull it straight out. To re-insert it, line it up carefully with its slot (there should be a notch that prevents you from accidentally inserting it "backwards") and, applying even pressure at both ends, press it straight down into the slot. The locking tabs should click into place on their own.

You should also know that memory slots need to be populated in a certain order and memory modules need to be paired in certain ways. I won't go into all that here. My owner's manual explained that very nicely. Hopefully yours does too, or you can look it up online. Again, if you're not sure what you're doing, it might be better to spend some quality time with Sandeep and the automated attendant after all.

Having removed my second memory module, I restarted my computer. The "3,4" diagnostic code remained and the computer still didn't boot. Okay, it looked as though the first module was the bad one. I'd have to replace it. Bummer. For now, though, a gigabyte of RAM should hopefully be sufficient to check my e-mail.

I shut down the computer again, pulled the first memory module and replaced it with the one that had occupied the second slot. Then I powered up the computer again ... and watched in dismay as the numbers "3" and "4" lit up yet again. This would suggest that the second memory module was also bad but I "knew" that the situation was actually much worse than it seemed. It's extremely unlikely that both memory modules would die at the same time. And there was this nagging question about why the computer wouldn't talk to my monitor. Even if the memory was bad, there's no reason why the BIOS shouldn't at least display an error or something to the screen. I concluded that the worst-case scenario had come to pass. The memory was probably fine. It was the motherboard that had fried. That's why it was unable to properly communicate with either memory module or the monitor. My computer was dead.

At this point, I could have acquiesced and called Dell support after all but I figured they'd just take me through the process of doing exactly what I had already done and reach the same conclusion in the end. My computer is well out of warranty. They might be able to refer me to a service depot but, if the motherboard is gone, you may as well just buy a new computer.

That's what I decided to do. I was happy enough with my Dell that I decided I'd order another one. I like ordering computers from Dell. You can configure them with exactly the hardware and options that you want and they custom-build your machine especially for you, only after you've ordered it. They're like the Harveys of computer retailers. The only problem is the inherent wait while Dell custom-builds and then delivers your system. It looked as though I was computer-less for at least a couple of weeks.

If you don't see how that's a big problem, then you obviously don't know me. I live on my computer! It's how I spend most of my leisure time. I seriously wondered what I would do with myself for the next couple of weeks. Pathetic, isn't it?

Then I got to thinking, maybe this was a blessing in disguise. Maybe "unplugging" for a while might actually do me good. Maybe I could, I don't know, go outside or something. Maybe I might wind up like the Springfield kids in that video that I put in a recent post. I wondered if anyone in my neighborhood would know who I was, or realized that I lived among them. Maybe they knew exactly who I was. Maybe I had become a local legend, like Boo Radley in "To Kill A Mockingbird". Maybe they'd all point at me and speak in hushed whispers as I walked by.

The next day I was trying to think of something to do, but apparently really thinking about my computer, because the thought occurred to me, for no apparent reason, that I hadn't tried booting my PC with no memory modules at all. Now, I realize that this concept may seem a bit daft at first. Even if it boots, a computer "sans memory" is hardly going to do anything useful. It would be sort of like trying to drive a car whose engine runs perfectly but that has no wheels. However, it occurred to me that doing this might at least persuade the diagnostic console to display something other than "3,4". And so, because I had nothing to lose, and because, apparently, even tinkering with a dead computer is more enjoyable for me than, you know, getting a life or something, I went back into my attic, opened the computer, pulled both memory modules, and turned it on.

Sure enough, the diagnostic console had changed. Now, only the number "1" was lit up. Going back to my owner's manual, I looked up the code and, unsurprisingly, learned that "1" means "No memory modules are detected". Well. It had that one right, anyway. Now I began to second-guess my earlier conclusion that my motherboard must be dead. I mean, it remained "lucid" enough to know when it had no memory.

For no logical reason, I plugged one of my memory modules back into the first slot. I know, I know ... at this point even the not-so-computer-literate out there are thinking to yourselves "Well, wouldn't that just make the '3,4' code come back?" Yep. You'd think that, wouldn't you? But I tried it anyway, turned on the computer ... and it booted. The monitor came up, the computer went through its POST routine (which means "Power On Self Test", by the way; no, I didn't throw it through a fence) and then beeped a couple of times and displayed a message something to the effect of "Oh my! The amount of memory appears to have changed!". That's two it had got right. This machine was getting smarter by the minute. I, on the other hand, just stood there with my arms outstretched, shouting maniacally, "It's alive! It's ALIVE!!!"

I could have just pressed "Enter" at this point and let it boot with half of its former memory, but now a sort of inkling was growing in the back of my brain, like some kind of alien parasite, ready to burst through my skull like in that movie. Instead of continuing, I shut down the computer again and inserted the second memory module. I turned the computer back on, and now the code "3,4" returned. Normally, this would suggest to me that my second memory module was bad. However, my earlier diagnostic work had suggested that both memory modules were bad, and I now knew this to be untrue.

Growing ever more suspicious, I once again pulled both memory modules, rebooted the computer back to the "1" code, then shut down again and this time inserted the second memory module (the one that had caused the "3,4" code to reappear) in the first slot, leaving out the other module for now, and turned on the computer again. By this time, I wasn't even all that surprised when it booted.

Now it appeared that both memory modules were fine on their own, but caused problems together for some reason. Perhaps the second memory slot was faulty? That would actually be a reasonable hypothesis at this point, but I had another idea. I booted "sans memory" one last time, then shut down the computer and reinserted both memory modules before powering up again. Sure enough, the computer booted, and has been running completely normally ever since. I even ran a full hardware diagnostic scan on both the memory and the motherboard, using Dell's built-in diagnostic tools, and both checked out 100% error free. Go figure.

I don't know what caused the original failure. I suspect that my PC may have temporarily overheated, as it was warm up in my attic last weekend and I did notice, when opening my computer's case, that the air intakes had gotten clogged with dust (a situation which I quickly remedied, but which did not, in itself, revive the computer). Whatever the cause of the failure, it seems that the computer needed to be started without any memory in order to clear the problem before it would boot normally again. This makes very little sense but, then again, neither do shoes that explode every Thursday when you tie them in the usual way. Like Jeff Raskin says, computers are just like that.

One last thought; I don't want to seem blasphemous or anything, but I find some significance in the fact that my computer died, and then came back to life, over the Easter weekend.

Hallelujah!

3 comments:

George said...

Congrats! In some ways, being able to fix these problems and build systems is gratifying in it's own right.

When I was building my last PC, I found out, after much investigation, that Windows Vista 64 would not, in fact, install with 4GB of memory. As a known problem, Microsoft's instructions were to: Install 2GB of RAM, install Windows, install a patch for Windows to now recognize more than 2GB of RAM, then power down and install the other 2GB. O.o

Things they don't tell you in the system manual! Even more of a mystery and frustrating was Windows would begin the install procedure and carry through to almost completion before displaying a blue screen with no mention of this problem or how to resolve it.

Very frustrating at the time, but actually quite fun to diagnose and fix.

Then there was the cheapo loose-fitting HD cables.... but that's a story for another day ;)

Anonymous said...

I do not know if it's just me or if perhaps everyone else encountering issues with your website.
It appears like some of the written text within your content
are running off the screen. Can someone else please
comment and let me know if this is happening to them too?
This might be a issue with my browser because I've had
this happen previously. Thanks

Look at my weblog :: shadow fight 2 cheats (http://lottokorean.net/)

Halmanator said...

Well, I just viewed this post with both Internet Explorer and Google Chrome, and I saw no issues.