VOGONS


My new 486-386 junk boards [IT gore]

Topic actions

First post, by Deksor

User metadata
Rank Oldbie
Rank
Oldbie

Hi folks !
This thread is gonna show you the journey of fixing old boards I got in a lot.
I've already show one of the boards from that lot, which was probably one of the least damaged ones here : Fixing GMB-486UNP

So let's begin with that 386DX board (abit FU3) :
N3sJmwul.jpg

Missing half a bank of ram with * quality * desoldering as you can see :
geOSlF0l.jpg
It's also missing the KBC, cache and has a good amount of corrosion. it might still be fixable though (maybe not that RAM bank tho ...)
rkaTanYl.jpg

Let's continue with that 386SX 16 (ECS NEATSX) motherboard using very old kinds of RAM
xPOgMqIl.jpg
As you can see it's missing RAM and BIOS ...

Corrosion is even worse here
ejPDP9Cl.jpg
I'm not sure if I can fix that, especially since I need very old kinds of RAM that I don't have to use it (except if I replace ZIP ram slots with simm 30 pin slots maybe ?) Edit : found out that the chip's left are enough to fill two banks, so the board only misses its bios.

I also have this CH-471A VER1.0B :
ulveoWgl.jpg
There's not much to say about that board, it's looking ok, but I'll need to replace that socket.

On the other hand, this ATC1415 is in a poor shape 🙁
eveciaAl.jpg

VyY7sd8l.jpg

Somebody even removed some pins in the RAM slot !!
7nobrlLl.jpg

(and on the other side, the metal things used to hold the RAM stick are missing). I'll need to find RAM slots in working order.

Same happenend to this PCChips M915i :
5hEySRyl.jpg

And the chipset is busted as well
IgwmRKsl.jpg

... except here somebody has already attempted at desoldering it with a soldering iron I bet ...
d07sZSYl.jpg
That's ugly.

And finally, here's a PAT48AV-1.40
RlVrlUyl.jpg

32cJ2a6l.jpg

That one is very corroded, I'm not sure if I'll be able to fix that. I fear traces inside the PCB could be impacted. The KBC is missing, and the socket for it is totally busted.

Thankfully, I already have dead 486s that will become great sources of parts 😀. One is a pcchips m912 which has RAM slots that should be salvage-able and a working socket. The other is a no-name one that have 30 pin RAM slots and a socket 3 as well. I also have other 486/pentium boards, but I'd like to fix these as well (though some might totally unfixable).

Thankfully I have a desoldering iron, it will work wonders against the poor desoldering job someone has attempted in several spots !

Edit : new board from another lot that also look kinda horrible :
Z0ae2xMl.jpg
a "TK-82C491/GP-4N-D28B1" google doesn't seem to know what that is.

Mobos status recap :

  • GMB486-UMP : Working (needs some new tantalum capacitors nearby the PSU connector)
  • ATC1415 : Working (Needs a new VRM and perhaps a better corrosion treatment)
  • CH-471A VER1.0B : DEAD (broken socket replaced, doesn't POST, voltage on VCC pins too low), parts will be salvaged.
  • Abit FU3 : Working, ready to be cleaned.
  • ECS NEATSX : No activity, new BIOS rom chips installed, need to remove corrosion
  • PAT48AV-1.40 : No sign of life at all, needs heavy treatment for corrosion
  • PCChips M915i : untested, needs new socket and a new ram slot
  • TK-82C491/GP-4N-D28B1 (aka SNB-M022) : Working (corrosion treated, burn mark in the socket seem to have no effect anymore now that I removed the melted plastic, also cache is fake, I need to fix that)
Last edited by Deksor on 2019-05-13, 19:20. Edited 13 times in total.

Reply 1 of 102, by Deunan

User metadata
Rank Member
Rank
Member

I wouldn't have bothered to be frank. I mean there's only so many hours in a day and most my days are busy, all that repair would quickly turn from a fun project to a grueling work.
That old 386SX board is basically a 286 one - that at the very least would be someting different so worthy of repair. But that's my opinion.

Reply 2 of 102, by Deksor

User metadata
Rank Oldbie
Rank
Oldbie

I have the tools to fix at least some of them and broken beyond repair boards that can be salvaged to fix some of these, so I don't mind ^^

Like I mentionned, the GMB-486UMP gives signs of life and 4 traces need to be fixed. If I can fix even one board, the lot was worth it.

Reply 3 of 102, by treeman

User metadata
Rank Oldbie
Rank
Oldbie

I recent repaired a ATC1415 much worse then yours, in the end had to removed the odin chip which has faulty and preventing post.

This is fun work but need to do only 1-2 hours at a time then it turns to gruelling like you said, and its very hard to stop and break it down into sessions, once we start want to keep going until its done

Reply 4 of 102, by Deksor

User metadata
Rank Oldbie
Rank
Oldbie

ATC1415 fixed ! (as well at the GMB, but I updated its own thread)

7amqOXQl.jpg

wxdJUsbl.jpg

I removed the ram slots from a broken PCChips m912 to transplant them on this one since the old ones were totally busted.

pzJBIC0l.jpg

nUsGB14l.jpg

I also improved the traces state around the battery
1imYkzJl.jpg

The original reason of the mobo not POSTin is THIS :
ulVLqI0l.jpg

As you saw earlier, the soldering job was just HORRIBLE (no idea why), either the circuit didn't work due to dodgy solder joints, or the previous owner burned it by attempting to make this terrible mess. To test the board, I just bridged the output of the transistor to the +5V thus letting me use the 486DX2 I use for testing. That'll work until a new regulator arrives.

However you may have noticed that the board looks a bit filthy. I didn't manage to clean properly the flux, I tried alcohol, but it leaves some goo-y stuff there ... Any idea of what I could use to properly clean that off ?

Reply 5 of 102, by treeman

User metadata
Rank Oldbie
Rank
Oldbie

alcohol is the way to go but you need a bit of work with it, a cotton bud and rub rub until it slowly comes off, sometiems I use tweezers and a piece of some material soaked in alcohol because you can press much harder and it doesn't bent like a cotton bud but you have to be careful and have good control that the tweezer doesn't rip through the material and scratch what's under

Reply 7 of 102, by treeman

User metadata
Rank Oldbie
Rank
Oldbie

yeah personally for this reason I use material like a piece of material I cut off from something like a phone screen wipe or those wipes to clean subglasses and use tweezers to hold a chunk soaked in alcohol, many people use cotton to clean then use a different material to clean all the fluff off the solder joints I think there isint really any other way

Reply 8 of 102, by retardware

User metadata
Rank Oldbie
Rank
Oldbie
Deksor wrote:

I tried alcohol, but it leaves some goo-y stuff there ... Any idea of what I could use to properly clean that off ?

This is normal. You need to submerge and brush carefully but thoroughly the board in the cleaning solution to make it look like new.
An analogy would be the difference of just wiping the face with a moist cloth or to take a bath.

Reply 10 of 102, by brostenen

User metadata
Rank l33t++
Rank
l33t++

Or submerge the thing in water with alcohol and scrub it there with a semi-stiff brush.

Don't eat stuff off a 15 year old never cleaned cpu cooler.
Those cakes make you sick....

My blog: http://to9xct.blogspot.dk

001100 010010 011110 100001 101101 110011

Jah ich will trynen... Die Leute wie macht scheisse in dem Grünen.

Reply 13 of 102, by brostenen

User metadata
Rank l33t++
Rank
l33t++
Deksor wrote:

Water in my area is chalky, won't that be a problem with corrosion ?

I knew I forgot a word. Sure I meant deminiralised water. 😁

Don't eat stuff off a 15 year old never cleaned cpu cooler.
Those cakes make you sick....

My blog: http://to9xct.blogspot.dk

001100 010010 011110 100001 101101 110011

Jah ich will trynen... Die Leute wie macht scheisse in dem Grünen.

Reply 15 of 102, by brostenen

User metadata
Rank l33t++
Rank
l33t++
gdjacobs wrote:

Distilled water is best.

Naaa.... Then you can mind as well, just use botteled water. Destilled still have all the minerals, like calcium.

Don't eat stuff off a 15 year old never cleaned cpu cooler.
Those cakes make you sick....

My blog: http://to9xct.blogspot.dk

001100 010010 011110 100001 101101 110011

Jah ich will trynen... Die Leute wie macht scheisse in dem Grünen.

Reply 16 of 102, by retardware

User metadata
Rank Oldbie
Rank
Oldbie
brostenen wrote:
gdjacobs wrote:

Distilled water is best.

Naaa.... Then you can mind as well, just use botteled water. Destilled still have all the minerals, like calcium.

There are big differences between aqua demineralizata, aqua destillata, aqua bidestillata and aqua tridestillata.
@gdjacobs did not talk about demineralized water.

But there is no real need to use treated water in place of normal tap water, as the last step should be submerging into IPA and wiggling in it anyway to displace the remaining water on the board (together with the minerals it carries).

Reply 17 of 102, by Deksor

User metadata
Rank Oldbie
Rank
Oldbie

Well I'll do that when I'm done with all these boards, I'll clean all of them in a row.

Now I've removed one of the broken socket and desoldered a socket 3 from the same pcchips (I feel like after this, the board will end up with litterally nothing left, it will just be a blank PCB 🤣)

The receiver board is ready to get its new socket ... but one thing is missing ... SOLDER 🙁 I don't have any left, I used a lot of it to tin the RAM slots I removed and soldered and to tin the legs of the sockets X)

So the repair of that board is halted for the moment.

In the meantime, I attempted to power on the 386DX board (without cleaning the corrosion as I can't remove the keyboard connector or the heavily corroded chip because I don't have any solder left). According to my POST card, there is some activity, but it's not POSTing. The code stays at "13" for a while and then turns to "00" forever. It seems like it does the same thing with or without any RAM ...

Have anyone a clue what that board actually is ? I'd like to know which jumper do what about the cache, I suspect the issue might be related to cache being missing, but I can't verify that.

Of course the missing RAM slots in the bank 1 will be replaced, but I think I'll do the same with all the RAM slots as they're not in a great shape (enough to hold the RAM firmly in place and making proper contact, but they feel a bit fragile and bulky now)

Another board I'd like to identify is the 386SX-16 board, because I'd like to write a BIOS eeprom (or two ?) for it so I can test it.

Reply 18 of 102, by treeman

User metadata
Rank Oldbie
Rank
Oldbie

personally I probably wouldn't try turning on heavily corroded boards things can probably short out and make it worse, I know the excitement to continue the work is there

Reply 19 of 102, by Deksor

User metadata
Rank Oldbie
Rank
Oldbie

That one isn't too badly corroded though. (I'm not sure corroded copper is even conductive, at least oxyded copper isn't, but maybe this case is different because the chemicals are the same).

Anyways.

Some traces will probably need to be fixed, as well as some parts might need to be replaced, but there was no magic smoke and the board gave signs of life, so this is very promising !

Also, I've added a status recap of all the boards at the bottom of the first post so you can keep track of the progress 😀

Edit : by the way, that board is a CH-471A VER1.0B

and here what it looks like as a whole (without its socket of course xD) :
WXAObIAl.jpg

Last edited by Deksor on 2019-03-18, 21:37. Edited 1 time in total.