yes I follow your logic totally. I re tinned the connections on the big umc chip, then it got me thiking about what you said about the signal from the ram to the chipset. So while I was at it I thought ill strip all the traces that head from the chipset in the direction of the ram and cache. Not a big job to tin them up while got eveything out.

a pretty quick job after scrubbing the traces my concentration is fading and unfortunately still stuck in the same roller coaster

Sometiems 64k sometiems 640k

Next step go further with the traces, I won't be taking off the ram slots but ill take off the cache chips and strip more traces that are in the area and retin

It makes sense that the signal coming from the ram to the chipset or chipset to bios gets corrupted somwhere.

I can't find the bios image for this board anywhere so the best I could do is save my bios and write it to a eeprom chip which gave the same results.

Few days rest and ill do more traces

wow I just bent one of those long black things, not sure what it is called and found hidden treasure... hmm need inspection number 100 and possibly a small bath

That's a resistor pack, possibly a line terminator of sorts for SIMM slots. But it's very surpising to see all that gunk there, the mobo looks clean otherwise. Either it's water corrosion, say mobo was washed but all the dirt went into these nooks and was still wet then power was reapplied. Or it was insects, I've seen that before.

The resistor packs are important for keeping the signal lines working properly.

A broken connection on them could easily cause intermittent problems.

So I got really excited and had to work on this.

Those resistor packs are so close together and between isa slots it was easy to miss this mess without looking for it. upon inspection I found 2 rusty corroded ones both right next to the simm slot. At the start of this post I suspected some spilled soft drink on this board as I saw blotches on a few smd capacitors, this confirms my theory.

Anyway removed both the resistor packs, it was pretty messy. Here is before and after pictures of the pins.

After cleaning

At this stage the resistor packs also look corroded so I don't know what to think, one of them lost 3 legs on the way out because I had to rock it around alot while using a solder pump in a tight space. Nothing to lose reinstalling them but questionable parts now

I made new legs for this one reinstalled and still facing the same problem.

I think the next step is to find replacement resistor packs, it seems a big coincidence I just uncovered this issue and like you guys say these resistor packs could cause a problem like this

That is a shame.

You did check all those traces. In a few of the photos, it did look like a few got a bit messed up, but hard to tell without doing a continuity test.

Are my eyes playing tricks on me or is there some sort of corrosion inside that ISA slot?
This could be shorting out signals and causing havoc as well. Inspect and clean if necessary.

It is just the light from a desk lamp I was using to take the picture.

When I took the resistor packs off I thought the traces were destroyed too but everything came off with iso alcohol, I was too excited to check continuity.

Do you think the actual resistor packs could be damaged from the corrosion and having power going through them?

I can't really find replacement ones using the codes printed on them anyway so I might retouch the solder on them again incase.

Hopefully there is no monsters hiding under the simm slots because I don't have it in me to take those off

Honestly, the resistor pack is probably fine.

You can test it by testing the end pin (pin 1) to each other pin. Should see the same resistance to every pin, with none open.

ok, thanx for the tip. Im going to take a break for now

Is it me or is there something with the first VLB slot?

Something you haven't shown to us is the back of the board. The UMC chipset has many pins going nowhere on this side, so maybe they go to the other side to connect stuff ?

Did the board look like in the photos you posted when you got hold of it, or are the pics taken after an initial cleaning/scrubbing? Because it seems to me that it was powered, maybe used for a while after being contaminated with soft drink/gunk or whatever that was, and your problems might be caused by corrosion, perhaps in a spot where you can't see it because it is covered by a slot or a chip or another component.

Unfortunately, without proper tools, your efforts might be in vain (and sometimes this is true even if you have the tools; I suppose an intermittent fault inside the UMC chip cannot be ruled out). At this point, I think it's a must to check under the first ISA slot, under the memory slots and under the ICs in that area, plus checking those ICs on a tester (TL866CS can do that). BTW, I see some corrosion on at least one pin of the chip that's right above the resistor pack that you worked on (74F244).

the first picture in the thread is the board how I got it. Nobody told me it was water damaged only that it doesn't post and sometiems gives beep codes, so when I got it to boot to bios it was already a win. however I didn't notice the small sticky stains until later so it was powered on no doubt many times in that state.

I think it might of been the light on the 74f244 but I took them both off anyway since now a game of hide and seek started. Nothing eye catching really, there was a bit of something around a few holes after the solder came out but maybye dust or flux not necessarily corrosion.

I will put them back in soon but like you mentioned it will be good to test them while they are out. I have no idea what this chips function is, some buffer? Any ideas how to test it? I do have a 866.

After this the next logical step would be taking off the simm, I do want to do it but I don't but my curiosity grows each time I can't fix it, so maybye.

I didn't post pics of the back because there is nothign interesting happening there at all. Yeah somwhere in between the original posts I did fix a trace on the back of one of the 72 pin simms, which seemed too easy and didn't fix nothing. I didn't really bother updating on it, was a quick jump wire and no change. Im guessing it shorted and blew when the original water damage took place

yeah no doubt there is multiple or was multiple issues

For testing the chip, in the MiniPro software choose "Logic IC", then 74244, the choice will be 74LS(HC)244, but in my experience many other series (ALS, F, HCT) could be tested just as well. Also, the broken trace that you repaired - was it burned? If so, check any chips that it connects to, because they might not have survived.

Removing the SIMM slots might damage them (the plastic doesn't take temperature well), or the mobo. Before you do that, have you tried pulling the L2 cache chips out, setting the size to zero and testing if you still get these memory size issues?

Ok thanx for that info, I tested both the chips and they pass, there is a shop near my that sells electronics and should have cheap sockets so ill socket the chips since they are out, might as well.
perhaps at some time in the future when im really out of ideas ill remove the remaining 3rd buffer chip which is a bit lower.

The trace on the back of the 72 simm slot was burned black when I replaced it but one side goes direct to the ram slot and the other side offshoots to a via which I traced on the front and not sure where that goes

I have tried this board with all the cache removed before with same results and disabled in bios. however on the cache settings I am not sure how to set jumpers to 0, have them all open? lowest setting is 64kb

I will hunt around for different cache chips to try from my other boards too, won't hurt

yes removing the simm slots is something I would like to leave for very last I have read other threads about plastic burning too but that was with a heatgun, I got a desolder station with a gun so it is pretty quick without leaving the joint heated for too long

Also I took out the 9 cache chips and did a test on tl866 under the code of the chip w24257ak-20 and they all pass.

But I am bit lost here 9 chips means one is a tag ram chip but they are all 32kb chips.

so first 8 is 32kb = 256, what the board is setup for, but the last chip 32kb should be a 256kb chip as the tag ram?

This board did boot with 256kb cache enabled at times, then other times external cache is always disabled in bios no matter how many times I enable it still is disabled on reboot.

Yeah this is normal.

Settings don't save ? Hmm makes me wonder if the CMOS is working ok

sorry I wasn't very clear, all the options in the bios save and keep hold, date, memory test above 1mb, interval cpu cache... however external cache option is a mind of its own when I set it to enable it will be disabled or post with no cache.

Few times it did boot with cache and saved in bios, I am putting that to the board being 100% stable at the time.

Then when all the erros with ram and parity happens the cache functionality goes with it.

I am also not 100% sure of the bios, I backed up the original bios with my 866.
Erased the original chip with uv light and reprogrammed it from the image. The board worked ok for a bit like that, then failed again. It was most likely to pulling out the bios chip and flexing it which is near the ram slots.

I got some eeprom chips and programmed one with the original bios image and still same problem. So maybye its a corrupt image maybye not, I can't find a original on the internet.

Because its intermittent I follow the same logic as mentioned before cracked trace, via or corrosion in a hidden spot with possibly a failed ic somwhere.

I will get some sockets and reinstall the buffers later today see what happens