Advertisement
If you have a new account but are having problems posting or verifying your account, please email us on hello@boards.ie for help. Thanks :)
Hello all! Please ensure that you are posting a new thread or question in the appropriate forum. The Feedback forum is overwhelmed with questions that are having to be moved elsewhere. If you need help to verify your account contact hello@boards.ie

Build of retro Pentium 4 machine - OS installation fails and unreliable booting

Options
  • 09-03-2020 12:29pm
    #1
    Closed Accounts Posts: 3,292 ✭✭✭


    Hi,

    I am doing another retro build on some near vintage hardware and I am having problems.

    Pentium 4, DDR ram, AOpen MX46-533GN MoBo, PATA HDD. Trying to set up a dual boot of Me and XP on one disk and hopefully Win 95 & 98 on another.

    I find that the OS install sometimes fails due to various error messages and when I can get it to install, sometimes the booting is unreliable/fails after the Windows screen or I get registry errors.
    The odd time when restarting the screen stays blank and there is no POST until I reset again. Sometimes it works a few times after I have reset using the CMOS jumper.

    To rule out components by elemination. I have tried swapping RAM, different sticks in different slots, tried a different hard drive, re-burned the install media to a fresh CD-R, even swapped out the build's CD/DVD drive to rule that out. I have tried XP and ME installs and they sometimes work but booting is hit or miss.

    I had all four OSs ( 95/98/ME/XP) working stand alone on it before so I know they can work on it, but I started running into complications with multiple bootings. Now it seems I cannot even get a stand alone OS to install or run stable even with cleanly formatted and partitioned HDD.

    Could all the many repeated restarts, reinstalls, disk reformats and CMOS resets possibly have damaged the any of the hardware in some way?

    Given that it does successfully boot sometimes, I am assuming that the ISOs and install media are good.

    I am still having problems with fails and unreliable boot.

    Where is my next step here?
    Considering I have ruled out most of the other hardware at this stage, is the PSU or the MoBo more likely to be the culprit in these circumstances?

    I hope it is not the MoBo as ones with two IDE channels are scarce and quite expensive even now.

    Advice and experience is appreaciated. Thank you.


Comments

  • Moderators, Society & Culture Moderators Posts: 15,752 Mod ✭✭✭✭smacl


    Doubt it would damage the hardware but you could have a duff boot partition on one or both disks. Try booting to a command prompt and using diskpart or fdisk to get rid of these before re-installing your operating systems.


  • Closed Accounts Posts: 3,292 ✭✭✭TheBoyConor


    Oh I think I would have tried that with the disk, started from scratch and clean wipe a few times. As in deleting all existing partitions, recreating them and then formatting and attempting an install. I've tried multiple primary partitions and primary with extended partition and logical drive. Is that what you mean?

    I did 2 ways, using the old style Windows ME boot floppy disk with FDISK and then with the bootable XP DVD that has a partitioning and formatting utility as part of the install process.
    DISKPART isn't a recognised command on the ME version of DOS.

    I have done that with 2 HDDs now and still having problems.


  • Closed Accounts Posts: 3,292 ✭✭✭TheBoyConor


    I did more fiddling tonight and I'm starting to come to the conclusion that it is in some way temperature related.

    When turned on first from cold it booted to ME just fine 3 or 4 times. Perfect. I left it run for an hour and then tried more restarts. Either failed to boot and froze on the ME slash screen, went black or blue screen and unresponsive or booted to desktop but with a registry error message.

    In BIOS all CPU temperatures look to be normal. Fan is blowing fine and case panels are all off so it can't be a setup or fan issue. I had also removed and thoroughly cleaned the heatsink initially and applied new paste.

    Will temperature affect a PSU performance or a motherboard in general?


  • Moderators, Society & Culture Moderators Posts: 15,752 Mod ✭✭✭✭smacl


    On older motherboards blown capacitors used to be a common enough issue, I've replaced a few myself on older machines some years back. The capacitors are the small metal cylinders on the mobo that typically have a mauve or dark green plastic sleeve. If one of these is bulging, it is most likely blown and needs replacing.

    Assuming you've cleaned all the connectors, vacuumed the bare mobo, and re-seated all the components. Another problem could be a dry joint somewhere. Basically as the temperature rises the mobo warps slightly and a loose connection become intermittently disconnected and causes a crash. Dry joints are a pain to find. If you take the mobo out, flip it over and look for any soldered area usually near larger components that is either loose or looks duller than everything else, this could be the problem. Re-melting the solder around any suspect areas with a needle tip soldering iron might solve this but can be a lot of messing for no result.

    Also from a DOS prompt, try CHKDSK /r /f on both drives to check for physical errors.


  • Closed Accounts Posts: 3,292 ✭✭✭TheBoyConor


    Yeah I've run chechdisk several times, it runs as part of the installation, and I've also tried 2hdds so I've more or less ruled out the disk. I could try a different IDE cable as i have a spare lying around.
    I have already tried the current IDE cable plugged into both the primary and secondary PATA sockets, to no avail.

    I think I've no option left but to dismantle everything, take out the MoBo and check everything physically in fine detail and clean everything and have another go.

    I might start from scratch again, trying a bench top build so everything is easily accessible for testing and checking and try out a different IDE cable and a different PSU to see if it makes any difference.

    After that checks out, I'll try a spare CPU I have here, and after that if it still is giving trouble it just has to be the Mobo, right?

    I switch on again this morning before I left and it succesfully booted into ME 3 times.
    Now almost certain that it is something temperature related.


  • Advertisement
  • Moderators, Society & Culture Moderators Posts: 15,752 Mod ✭✭✭✭smacl


    After that checks out, I'll try a spare CPU I have here, and after that if it still is giving trouble it just has to be the Mobo, right?

    I switch on again this morning before I left and it succesfully booted into ME 3 times.
    Now almost certain that it is something temperature related.

    Maybe, but not necessarily. An older PSU on the way out could also cause this type of issue, I've had a few die on me over the years, one with an intermittent issue. If it was me, I start by taking the whole thing apart, vacuuming the mobo and all sockets, cleaning any edge connectors with an alcohol wipe, putting it back together again and trying again. After that try a different PSU if you have one handy, after than a different mobo.

    I see from your sig you're into retro computers so it makes sense to hang onto a bunch of spares for these type of occasions. You mentioned PATA cables which were also way more liable to damage than SATA when re-used due to loads of connectors and crimp/IDC assembly.


  • Closed Accounts Posts: 3,292 ✭✭✭TheBoyConor


    Thanks for all the advice. Came home and same story. Cold boot is fine. After a half hour running, rebooting starts to get iffy. I've it all taken out now. I've cleaned down the dust. My plan is to rule out further components by substitution in the following order: PSU > IDE cable >CPU > MoBo. Luckily, I have spares to sub in.

    When physically examining the MoBo in notice this. Board looks to be very substantially warped when I lay a rule across it. This can't be normal, can it? I know the heatsink levers exert a decent force downwards but surely it should not be warped to this extent. What do you think?

    505243.jpg


Advertisement