Forgot your password?
typodupeerror
Bug

Bill Dog's Journal: more info on broken PC - what would you try now? 9

Journal by Bill Dog

Another busy weekend (my dog's dying), but I made some time to take my PC apart. I unplugged all the drive cables, from both the drives and the mobo, and took them out. I took all the cards out -- video, sound, ethernet, and modem. I took all the RAM out -- PC133, 2 DIMM's of 128MB each that came with the computer, and 2 DIMM's of Corsair 256MB that I later added years ago. I even took the brain out -- the CPU and heatsink and fan is all on a card of its own, that plugs into a slot in the mobo. I unplugged the cables to the case reset button and HDD LED's etc. The only thing I currently have plugged in to the mobo is the ATX power cable from PS to mobo, and the connector between the mobo's power-on pins and the case's power button. I.e. the minimal # of connections to still be able to turn it on. And I unplugged and replugged-in both. I also brushed out with a small and very soft paintbrush any dust on the boards and on any pins and in any part of any of the slots. Same deal, the PS fan spins up, it runs for about two and half seconds and then quietly shuts off. No click sound, like when I do the Windows shutdown and it shuts itself off. Instead, it sounds like it might if it just got unplugged or something. I even tried a different power cord.

To quickly answer the remaining questions from the last JE, I had vacuumed with a plastic attachment, and hopefully plastic's not capable of transferring electrostatic charge. (And it's summertime and humid, so *I'm* not zapping anything, at least as far as I notice.) And it's not running XP, it's the Windows 2000 Professional SP4 system that's been stable for forever. (And I don't use automatic updates from MS, I go to the technet section of MS's site and review them first and apply them manually, and I haven't updated in months (I'm behind).)

So, about all I have left I think are the main board, the power supply, and the CMOS battery -- everything else has been taken out.

And the two events of note, as far as I can think of, that coincided with the thing breaking, are unplugged for several hours, and vacuumed out.

The world wide intertubes seem to think that a failed or failing battery wouldn't prevent a boot. But corrupted CMOS data might. I'll try shorting the pins that reset the CMOS data. I think if the battery is dead it would still boot, and maybe just have to rediscover all the H/W parameters again?

Can a power supply be "dead" when it still starts up? My only (limited) experience with dead PS's is that they don't start up at all. (One even made a loud snapping sound and spat sparks at me, and a really strong smell, before it ceased powering up -- startled the crap outta me at work.)

And I'm hoping that the main board didn't just decide to die, or rather, some part of it -- it still spun up hard drives when connected, and the CPU fan, and LED's, etc.

So, with this narrowed down some more, anyone have any hunches? I figure a battery's pretty cheap, but I don't want to buy a new power supply (if I can still even get one for this thing -- a 230 W for an ATX board with it looks like some 3.3V and some 5V and some 12V lines, circa April of 2000) if it's doubtful that that's it. Any advisement is deeply appreciated -- I'm just a programmer so this stuff is not my area of expertise. :(

This discussion has been archived. No new comments can be posted.

more info on broken PC - what would you try now?

Comments Filter:
  • After resetting the CMOS, even with a dead battery, you would just get a warning about changed parameters, hit F1, bla bla bla. You don't to buy one. If this is your only box, I guess I can safely assume you don't have another power supply, which is what you need to continue narrowing it down before you get to the mainboard. Even though you don't feel the static, there is a small chance you zapped something during the first cleaning. And yes, plastic is like the balloon you rub on your head to get static. T

    • by Bill Dog (726542)

      Okay, I'll forget the battery (and it being unplugged for hours) as the culprit.

      If I messed up vacuuming, now that I think about it, I only dragged the plastic attachment across some solder joints on the cards, and I never touched the main board. And the cards are now all sitting out of the system.

      No I don't have a spare PSU (well, I have a really old Pentium I 90MHz system, that I'll look at tonight, but it might not be compatible). Can this be a bad power supply, even tho it starts up, and its fan starts

      • by iminplaya (723125)

        Can this be a bad power supply, even tho it starts up, and its fan starts up?

        Yes, I've learned the hard way not to discount anything. I've seen this just recently, only the fan wouldn't wouldn't even turn a full revolution before cutting off, and it was the PSU.

        If you already got most of the dust out it, don't worry about the air for now. Try the experiment in the link with the PSU. I'm almost sure the one from your 90MHz machine is an old AT supply with a real mechanical on/off switch on the front of the c

        • by Bill Dog (726542)

          I don't understand what the experiment in the link you provided will show? What that author seems to be getting at is don't let a lack of a PSU fan starting up let you necessarily think the PSU is bad. But that (a lack of the PSU fan starting up) is not a symptom I'm experiencing. (And I think I have that low voltage path he talks about, because my mobo has an LED on it that lights when the power cord is plugged in and the rocker switch on the PSU is in the "on" position (even tho the PC is "off").)

          But I th

          • by iminplaya (723125)

            I just got done trying it on a spare. The running fan just means that it's on. If it does stay on, then the PSU is not the problem... assuming the output voltages are ok*. That's all you need to know at this point, right? His symptoms were a bit different, but the experiment is to run the PSU without being attached to the mainboard so there is no load on it.

            * 20 pin [pinouts.ru]
            *24 pin [pinouts.ru]
            Caution: Russian site. Could have viruses. I don't know. I'm not running Windows. Use Linux or a live CD :-)

          • by iminplaya (723125)

            Need to add...

            ...because my mobo has an LED on it that lights when the power cord is plugged in and the rocker switch on the PSU is in the "on" position (even tho the PC is "off")

            That comes from the 5 volt standby. He's talking about grounding the green wire to any of the black ones(all the black ones are common) with the paper clip to turn on the machine. That's what the switch on the front of the machine does, through the mainboard, which keeps it latched to keep the machine on. The machine shuts off whe

  • As iminplaya said, plastic is not what you want in a nozzle. If you have to vacuum, a grounded metal nozzle would be better. Even better would be a can of air.

    Was the computer plugged in while you did the vacuuming? Despite what they often say, I *always* leave it plugged in when I am doing anything inside the case. The best protection against static damage is the ground wire. Even that soft paintbrush could have been a problem. When I was working for Texas Instruments, they put us through an ESD (ele

    • by Bill Dog (726542)

      Ugh, nope, the PC was not plugged in while I was sucking out the excess dust wabbits. I'm going to feel like throwing up if I hosed my system by trying to clean it.

  • A while ago, we saw that same kind of behavior at work: the machine would suddenly stop, then it would reboot and get a bit into the memory-counting, and then shut down again, restart, then get not quite as far into its memory counting as the last time. Eventually it would only manage a brief burst on power-up with no oscillation.

    When the power supply was replaced, everything worked OK again. The fact that the power supply behaves the same whether the load is light (motherboard only) and heavy (motherboa

The study of non-linear physics is like the study of non-elephant biology.

Working...