Old computer in a coma state - bx-6xp2

[Benny]

Probably this isn't the correct foorum to post this so tell me where I should move this discussion to.
Today I've dealing with an old computer that had no sign of life in it. After oopening it and sweeping the dust I decided that probably the PS was at fault. I then took an also old PS from a Gateway that gave almost all the voltages as the one I took off (except for -5V@150mA - as per the old PS specs) and I could hear some noises but no 'beeps' and just for "3-5" secs.
I noticed that there're a twisted pair of cables (Black & Red) cut at one apparent extreme (coming from the front panel at a led display that apparently shows the clock frequency - which, by the way shows signs of been broken at the plastic front at least). I'm not sure to where they might go but have a plausible idea which needs to be studied further.
The original PS is FSP 180-51NI(V) & the motherboard is BX-6XP2. I learned that this motherboard needs this -5V but I don't know for what and/or if the absence of it is what's preventing the machine to boot-up.
Apparently, the Floppy, Harddisk & the CD drives are in a working capable state (yet).
Does anyone has faced something like this puzzle & knows what is the best next few actions?

PS. I used to work as an electronic technician at Digital Equipment Corp., so facing electronics doesn't give me nightmares. Any of you could give me any advise no matter how 'risky' it might sound because I'm well prepared to decide for myself if it involves risks to the equipment and/or to myself and decide if I'm willing to accept it or not. In either case you will be notified of my decision & the reasons behind it.

[kiosk]

GIS shows that this is a Slot1 manboard. You need to reseat the cpu card, Slot1 machines were a nightmare to run (and keep them in the running condition). Also, clear the CMOS and replace the battery. Do the insides smell like burnt cocoa? These mainboards were manufactured at the height of the bad capacitor plague, and many, MANY of these manboards ended up in the tip due to failing capacitors.

[123456]

Slot 1's are a nightmare? I have a Slot 1 machine in my room from 1999 that's never given me ANY problems ever since.

[kiosk]

Originally Posted by 123456

Slot 1's are a nightmare? I have a Slot 1 machine in my room from 1999 that's never given me ANY problems ever since.

I went through around ten different slot1 configurations over the years, and only low-clocked celerons ran ok most of the time. Pentium 2s were pretty bad, and cpus clocking over 500 MhZ were very difficult to keep them running smoothly, especially if the fan gave off vibration.

[Benny]

Thanks kiosk for responding!
I thought that 'GIS' meant 'Google Internet Search' & decided to search (again) to see what it means to be a "slot1 mainboard or machine" but found no info over it (even tried Wikipedia). No big deal.
What do you mean by 'clearing the CMOS'? The machine has no signs of having something burned out (but the PS is surely at fault).
Would the 'capacitors' plague syndrome' have had an ill effect on the power supply (I mean damaging it)? Where can I obtain more info on the usage of those -5V by the mainboard?
I'm planning to experiment with three 1.5 volts dc AA batts to supply (momentarily) a voltage near the -5V range just to see if something different happens. But not before I take everything out, undust carefully everything & reposition the CPU... I'll be back with the outcome!

[kiosk]

Originally Posted by Benny

Thanks kiosk for responding!
I thought that 'GIS' meant 'Google Internet Search' & decided to search (again) to see what it means to be a "slot1 mainboard or machine" but found no info over it (even tried Wikipedia). No big deal.

Slot 1 is an example of a standardised interface between the processor and the mainboard. For example, you can put a Slot1 processor into a Slot1 mainboard, Socket 7 processor into a Socket 7 mainboard, etc. Thus, we have Slot1/Socket 7/Socket AM2/whatever mainboards and processors. They are designed to work with one another.

Originally Posted by Benny

What do you mean by 'clearing the CMOS'? The machine has no signs of having something burned out (but the PS is surely at fault).

Mainboards have a tiny amount of volatile memory stored on them - this memory contains user settings you enter in the BIOS setup; like the type of your disk drives, video card, time and date, etc. This memory is called CMOS memory, and is maintained by a coin-shaped battery which is socketed on the mainboard. The battery is CR2032 type and usually keeps powering the CMOS memory for a few years. What happens is that if a battery goes flat, or if something goes haywire inside the computer (usually something related to the power supply), the CMOS memory is going to get corrupt and the system will refuse to start. This isn't destructive; the mainboard simply looks at the weird values contained in the CMOS memory and doesn't know what to make of it.

Your first step should be to remove the CMOS battery - you can do it by pressing the small metal retention clip at the edge of the battery socket. This way, every time you power up the computer, the mainboard will power on in the "fail safe" state which should at least give you video output and let you figure out what's wrong with the board.

Additionally, you can force-clear the CMOS memory by moving a jumper cap over two pins which are usually marked "CMOS", "JBAT1" or something similar. Your mainboard probably has three pins, two out of which are capped by default. This is "normal" mode; i.e. "retain CMOS memory" mode. If pins 2-3 are capped, move the jumper to pins 1-2 for a few seconds, then put it back where it was. This should clear CMOS memory and your computer should be able to start.

Originally Posted by Benny

Would the 'capacitors' plague syndrome' have had an ill effect on the power supply (I mean damaging it)? Where can I obtain more info on the usage of those -5V by the mainboard?
I'm planning to experiment with three 1.5 volts dc AA batts to supply (momentarily) a voltage near the -5V range just to see if something different happens. But not before I take everything out, undust carefully everything & reposition the CPU... I'll be back with the outcome!

Capacitors don't usually damage power supplies; they usually go kaboom or leak electrolyte. Your mainboard has voltage regulators on it: if the capacitors on the mainboard have failed, the voltage regulators won't be able to supply ANY voltage to the processor/memory and the board will appear dead. This is very likely, and should be the first thing you should check.

Don't mess with AA batteries; you'll achieve nothing at best and build a nice bonfire at worst. Go get a proper power supply.

[kiosk]

Also, not all capacitors fail by leaking or exploding - many dry out and do absolutely nothing, and appear just fine from the outside. You should unsolder the capacitors around voltage regulators, hook a few up to the multimeter and watch the resistance across pins as the cap charges. If there is no movement of the needle (i.e. the cap is permanently open-circuit) it has failed and ALL capacitors of the same type need to be replaced.

[Benny]

Thanks kiosk for such a complete explanation, I'll need a couple of days, probably, to digest everything & carry out the corresponding tasks. Indeed, you spread much light on the subject!

[Benny]

All this time, kiosk, I've been assuming that the capacitor thing just involves 'electrolitic capacitors': am I wrong on this?
I passed my eyes over the board & estimated > 20 caps of this kind. If I'm right it will take me ½ - of - a - little - bit to deal with them! I've not given up, yet!!!
By the way, the coin-shaped-battery is dead & don't know if it would sustain some charge until I manage to give it a try... I'm suspecting that this -5V might have something to do with recharging that cmos battery, have to dig in more internally...

[kiosk]

Originally Posted by Benny

All this time, kiosk, I've been assuming that the capacitor thing just involves 'electrolitic capacitors': am I wrong on this?

You are not. Check out this thread for more info on bad electrolytic capacitors:
http://www.pcguide.com/vb/showthread.php?t=25482

Originally Posted by Benny

I passed my eyes over the board & estimated > 20 caps of this kind. If I'm right it will take me ½ - of - a - little - bit to deal with them! I've not given up, yet!!!

I found that low voltage, high capacitance electrolytics give most trouble. For example, a 16V/470uF capacitor is almost certainly okay, a 6.3V/1000uF cap should rise a few eyebrows, while 6.3V/2200+ uF caps are the nasties. The latter are physically large capacitors usually found around the CPU socket. The good news is that there aren't too many of them on your average mainboard; the bad news being that they are absolutely essential for the operation of the mainboard. Check if some of them are bulging or if there is brown goo coming out of either side of them. If just ONE capacitor of the same type looks off, it's safe to assume that all capacitors of the same type on the board share the same manufacturing defect and that all of them need to be replaced.

Originally Posted by Benny

By the way, the coin-shaped-battery is dead & don't know if it would sustain some charge until I manage to give it a try... I'm suspecting that this -5V might have something to do with recharging that cmos battery, have to dig in more internally...

CR2032 batteries are non-rechargeable lithium type cells. Computers did actually use rechargeable NiCd batteries for the CMOS memory back in the days of yore, but this whole thing died out with 486-class computers in the mid 90s. Having a soldered-in NiCd battery which LOVED to leak corrosive electrolyte onto the mainboard was a pretty bad idea anyways.

Do yourself a favor and get a known-good power supply. You can't properly diagnose a board if you don't know if it's not working because the board itself is defective, or if the board is actually fine but it's not powering up due to a slapped-together excuse of a power supply.