Windows 98 machine being a pill

[Tommy]

Alright, so I'm having major woes with my Windows 98 machine. Random issues have crept up and now it's almost unusable at times. Symptom 1: The machine boots and almost gets to the GUI, but then it's just a black screen and the monitor says no video signal. Symptom 2: Random BSoD mainly concerning 0028:-------, which generally are random numbers, but the 0028 is always the same.

 

Right now I'm working on a remote Malware Bytes scan and attempting an AVG scan but AVG isn't working out so well right now. I have no antivirus protection on my 98 machine but I only go to a handful of sites with it, mainly YouTube, here, and occasionally Facebook but that's only to check things out, I don't really do anything with it. So I don't know if I have a bug on the machine or not, so that's what I'm testing out.

 

Now, this machine does have 2GBs of RAM with the RLoew patch installed. I have done a Memtest on the RAM, but only with one pass, but the RAM was found to be clean. I have done no further testing with the RAM since, but this is one of my prime suspects as I didn't seem to have this problem until I installed this RAM.

 

I did have the processor slightly overclocked from 2.6GHz to 2.75GHz which since I've reverted because I thought it may be causing problems. It's a P4 Northwood and I heard those can be slightly sketchy. But so far I don't know what the issue is. The computer itself doesn't seem to have the problem, just Windows 98.

 

So if anyone has any ideas on what I can start troubleshooting on, all help is appreciated. Otherwise I'm just going to bring one of my older machines out and use it for the time being as I really don't want to format the drive if I don't have to since I do have some important data on it and I don't really want to deal with backing it all up just yet. Speaking of which, the hard drive doesn't appear to be failing either but of course it's always a possibility too. Today the BSoD happened while in the middle of a Markiplier video and I wasn't doing anything except watching it....

 

Edit: This machine also has the USP3 and KernelEx 4.5.2 installed on it.

Edited February 16, 2015 by Tommy

[loblo]

I bet on dying motherboard capacitors.

[jaclaz]

Being more optimistic , I would try with another (surely working)  PSU, over the years I have seen every kind of issues that in the end resulted being connected with a dying/weak PSU.

 

jaclaz

[netbookdelgob]

Try disabling ACPI, most motherboards that include ICH7 and later chipset have problems with Windows 98. It seems that 9x doesnt know how to work with acpi apic .

[Tripredacus]

May seem weird, but last time I had a similar problem (it was recently) I got past it by replacing the CMOS battery.

[Tommy]

Thanks for the tips, I'm not sure if the capacitors are bad or not as I did inspect the board when I moved it over to a different case because I needed the one it was originally housed in for a full sized ATX board. But I don't remember seeing any bad ones. This 'computer' was indeed going to be sent to recycle by the local computer shop and I saved it from its demise and for months it seemed perfectly fine.

 

As for the PSU, this is where I wish I could remember if it started before or after I started using a different one. It seems to me that it started when I was using the original PSU and even with a different one that I'm using now, it still has the same results. The CMOS battery would be the easiest to change and even test to see if it is low or not. I might even swap RAM modules to see if it could be doing something funky even though memtest said no. I know the computer has been sketchy with automatic shutdowns as well so it's possible ACPI could be doing something funny as well. It didn't always but it seems ever since I put in the new RAM, it starts saying error writing to device aux and then if gives the standard fail, retry, etc options. So it's hard to say exactly what's going on here.

 

Edit: Also, MalwareBytes found the machine as being clean, so that's not the issue.

 

Edit 2: I had another thread started that I didn't think really meant too much as Dencorso said probably not, but I'll throw it on the table here. Going into the BIOS and into the PC Health Section, the CPU Vcore -5V was being reported around -61.69V. So the machine is still working so it can't be right, but is it possibly something is being fouled up because of this, perhaps a sensor is giving off the wrong readings?

Edited February 16, 2015 by Tommy

[loblo]

Try checking your voltages with a tool from within Windows (such as Speedfan or Hardware Monitor) as the value you report just cannot be right but may indicate there is a problem. If your Vcore voltage fluctuates by more than a few hundredth of a volt you've most probably got a problem with some of the mobo's capacitors and this would account for the random crashes you report.. Dead or dying capacitors don't always show physical signs of damage such as bulging or leaking.

[submix8c]

This 'computer' was indeed going to be sent to recycle by the local computer shop and I saved it from its demise...

Capacitors have a "small thin space" between them and the MoBo board. If there's any particulate under them, that could couse a "short".

 

And I agree that you should check from within Windows, if possible, to be sure.

[dencorso]

For what it's worth, here's what I'd do:

1) Disconnect the current HDD (I remember you said there's just one on that machine), but leave it where it is.

2) Grab another PATA HDD of any size (say...  >= 20 GB) and partition it with a single active primary partition.

3) Install *plain vanilla* 98SE, add RLoew's RAM Limitation Patch *only*, tweak the desktop minimally for confort (without any third party app, except perhaps MS's own TweakUI), add Opera 10.10 for connectivity and super-pi to give the OS something to do, then run super-pi and let it churn some for 6 to 30h (unless it crashes before that) to decide whether I do have a hardware problem (and if so, then troubleshoot for it) or not. In case it crashes soon, reboot it some 3 to 4 times to make sure it's crashing consistently

 

 If it does not crash and behaves solidly, then disconnect the new HDD, reconnect the old one and troubleshoot instead the Win 98SE setup.

 

 

This is just my 2¢, of course... please feel free to ignore it completely. 

[Tommy]

Thanks for the help everyone. I'll certainly look at all possibilities. Since the computer itself cost me nothing when I got it, I'd say I had a great run with it while it lasted. I do think stressing the hardware would be a great place to start.

 

This is my board: http://www.biostar-usa.com/mbdetails.asp?model=U8668%20PRO

 

I do not know if this came from the time of bad capacitors or not, but it's likely. I had another computer given to me that had a Biostar board in it and died after 3 boot ups and when I thought I accidentally damaged the socket (775), I found it had two slightly bulged capacitors right by the CPU so I trashed it. I didn't think it was worth fixing anyway.

 

Edit: Alright, so I tried both speedfan and hardware monitor. I'm still getting flimsy readings. HWMonitor is reading -12v at around -4.30 and -5v around -16. Speedfan is reporting -12v at around -11.26 and -5v at around +3.84. So something is definitely screwed up.

Edited February 17, 2015 by Tommy

[loblo]

Let's assume HWMonitor doesn't  work well with your motherboard sensors. As for Speedfan values, 11.26 is OK for +12 but  3.84 seems really low for 5v which may indicate a power supply issue. If you wanna know if you've got capacitors problems check Vcore fluctuation over 10-15 minutes with the graph feature.

 

And in the vein of dencorso's suggestion but lazier you could always boot a Live CD, Linux or WinPE, and see if you get random lockups too which would definitely  point to hardware problem.

[Tripredacus]

I do not know if this came from the time of bad capacitors or not

There is no "time of bad capacitors" !!!

There was, however, a time of "worse than usual capacitors." Caps can go bad over time. It doesn't matter if it was made during that bad time, or a good time.

[jaclaz]

What exactly provides voltage in a PC (AND BTW contains capacitors as well)?

 

Let me think what I would try next ....

 

jaclaz

[dencorso]

loblo is right!

Stop everything and do this first!

It's easy and it will provide valuable info.

 

And in the vein of dencorso's suggestion but lazier you could always boot a Live CD, Linux or WinPE, and see if you get random lockups too which would definitely  point to hardware problem.

[Tommy]

Thanks for the suggestions! I like that you mentioned Linux because I have tried Linux on this machine a few months ago in fact and at times it acted weird, sometimes I'd have to hard reset the machine. I don't recall having problems with Windows 98 at this point but I do remember the occasional lockup on Linux. At first I thought maybe Linux was just too powerful for this machine and wouldn't work out well on it. Maybe something funny was going on then.

 

@loblo: Why would it be reading positive for -5v? I did put the positive sign in purposely because it was in the positive range, not negative, unless it's suppose to be like that but my main Windows 2000 machine I'm using right now isn't like that. I do have another power supply in stock but the linux problem happened with the other power supply, not this one. So it's likely maybe one of the capacitors are shot.

 

@Tripredacus: What I meant was the bad capacitors used in the older Dells like the Optiplex GX270, which I owned and it suffered bad capacitor fate which I was hoping I'd get lucky on but didn't.