Corrosion removal - Motherboard with leaky Ni/Cd battery

[CkRtech]

Ian.M. - Once again, I would like to thank you for the thought and time you put into your replies in this thread.

1 - Regarding copper foil: I haven't yet done trace repair replacement using copper foil vs just soldering jumper wires. I look forward to it. There are some copper foil tape rolls with conductive adhesive out there (such as This one), however I assume it is better to get a sheet of it, cut it to size with a razor, and apply to a clean, de-traced area using an epoxy of some sort.

2,3 - Excellent. Will do that after the repair work is complete, then. It lets me test prior to that step as well.

4 - OK. The manual for the motherboard says that a replacement battery can be hooked up to J1 (4 pin header that I am also replacing) and be from 3V to 4.5V. I considered that when I first saw the damage, but I thought about utilizing that empty space on the board for a 2032 instead. I do have plenty of space in the case to velcro a plastic 3xAA holder and use the 4 pin header as the motherboard manual recommends, so perhaps I will stick with that.

Regarding securing the repair wires to the board - It sounds like from what you said that the PCB lacquer will help with that as well? Was considering using this silicone conformal coating, but there is also an acrylic lacquer (keyword) here. I assume the acrylic is much more forgiving and the better option? (You did use the word "lacquer," afterall.  Just wanted to talk a bit more about it since I don't own either)

Although each comes with a brush cap, I prefer your Q-tip suggestion for a more rigid application with minimal spread. Do you typically use something different for securing wires during a repair when only wires are needed, or is conformal coating/pcb lacquer typically a multi-purpose coating/securing tool? Most wire repairs I have done to this point have been minimal length and just float. I would indeed prefer to do a neater job on this one.

[Ian.M]

It helps to dress the wires as neatly as possible so they lie flat naturally.  Use dots of superglue gel to tack wires down.  Cut a strip from a Polypropylene food container (Recycling code 05/PP) and notch its end by about 1/2 wire thickness to act as a tool to hold the wire down in the gel dot while you blow on it to accelerate setting.  Don't trust lacquer on its own to hold long wires.

I wouldn't use a silicone coating for anything I may need to work on in the future, but perhaps I am excessively prejudiced against it.  IMHO you risk serious adhesion problems if you ever need to recoat or stick anything to the treated area, but maybe MG have solved that chronic problem typical of older silicone formulae.

Hint: Work from a few drops on a sheet of glass, not from the bottle - you don't want to contaminate your bulk supply with a dirty Q-tip.   The Acrylic should have a very long shelf life if tightly closed (but beware of closing it so tightly the cap fractures with age).  MG claim the Silicone has an extended shelf life then seem to be very coy about how long its actually good for - I suspect its shorter than a hobbyist would find desirable. 

You can get copper foil tape from art & craft supply stores - you'll find it with stained glass supplies.  It wont have conductive adhesive, but you can work around that by cutting away the backing paper and cleaning off the adhesive where its going to overlap the track you are repairing so you can tin the ends before you peel the rest of the backing paper, stick it down and sweat the ends into place onto the pre-tinned track.  In any case you sould clean the ends a bit and not rely on conductive adhesive + a solder  bridge on top or you may not get the reliability and low resistance you need.   

Epoxy sticks poorly to copper and its alloys unless sanded into the surface.  Sand clean, solvent wipe, apply mixed epoxy, sand again with a fresh clean fragment of wet&dry paper, wipe off excess epoxy and sanding grit leaving a thin smear, apply an even film of epoxy and press in place - its a royal PITA, especially when epoxy meets soldering iron.

[CkRtech]

I was hoping to get a sanity check and a little assistance with some of the component replacement. It involves the parts 386 motherboard I use as well as potentially picking up some new parts from digikey.

Inductors
1: How concerned should I be with inductor values for the inductors near keyboard port? I have 4 on the 486 but only 3 on the 386. I could pull a random inductor elsewhere from the 386 board to slap it into the 4th spot, but I don't really know much when it comes to inductor values.

tantalum cap
2: I have a cap that reads 106 16K on the 486. I believe that means it is 10uf, 16V, and K is the tolerance. I have one marked 10u 16 on the 386. I assume this is a legitimate replacement, but I wanted to mention it here.



14069 hex inverter
3: That particular one by Motorola isn't made anymore. Datasheet page. The 386 board has a Toshiba TC4069 hex inverter on it. (I also have a bag of 74HC04 chips.). I can find tooth comb the datasheets, but if any of you guys say, "Don't worry about it. Slap anything pin compatible in there," I can do that.

Diode values
4: Unfortunately, I am uncertain as to what values I should use concerning the diodes. Diodes are also components that I don't typically have to order/replace/etc. I have a photo of the diodes prior to any cleaning, and I also have some available on the 386. One of the old diodes (CR4) has what looks like 48 on it. Any other writing has come off, but the original photo seems to have a 1N. So - 1N48? I probably won't replace the rechargeable battery, and CR5 is for voltage to go to rechargeable battery positive - So I guess I could omit CR5. Just thinking out loud.





Misc capacitors
5: I have some other caps marked "104 ze5." There are some similar ones on the 386 that are simply "104 Z." I assume it is an easy drop in replacement candidate, but I wanted to run it by here as well. I am not sure what the e5 indicates.



Thanks!

[james_s]

The diodes are almost certainly 1N4148, that is a very common small signal diode in a glass body.

The IC is an old CMOS chip, I'm a bit hazy on the old 4xxx series CMOS ICs, it's been a long time since I've worked with those. Used to be you couldn't just sub in a 74xx series as those were TTL but many of the modern 74xx chips are actually CMOS. Whether you can sub it likely depends on what it's used for, but you probably won't hurt anything by trying, provided the pinout is the same.

The capacitors are just 0.1uF bypass caps on the power rails. The suffix may refer to the dielectric type, voltage rating, tolerance or temperature range, I don't know off the top of my head but it should be no problem to sub them. Most stuff has enough margin that it will even work with some of the caps not present at all, but there's no sense inviting random behavior.

[Ian.M]

I cant think of any *GOOD* reason to use a 4000 series CMOS chip on a PC motherboard other than an UNBUFFERED inverter used as a quasi-linear amplifier for a crystal oscillator.  A 74HC04 is a multi-stage inverter, would have far more gain, and would probably increase the drive level to the point where it could damage the crystal.

I suspect the original part number was MC14069UB.  If so, you can substitute any 4069UB from a different manufacturer.

[hammy]

One of the old diodes (CR4) has what looks like 48 on it. Any other writing has come off, but the original photo seems to have a 1N. So - 1N48?

These are 1N4148 diodes.

Nice work! 

[CkRtech]

...an off-board 3xAA holder with a series diode to provide anti-charging protection and drop the voltage a bit would be the best option.

Saw elsewhere that someone used a 1N5817 for this. They mentioned using that diode vs 1N4004 because it had a smaller voltage drop. Open to diode recommendations if you think otherwise. My plan was to trim the legs down and wire it in series with + when I solder on the 4 pin female connector to it with some heat shrink. I suppose I could also get a 4xAA holder and put the diode in the 4th slot.

Side note - Was also going to use that fiberglass brush on a few of the legs of the keyboard IC to get some of the green off - or should I just dab some vinegar on it, rinse, and iso to make sure it is neutralized?

[Ian.M]

It doesn't need a very high current capability so you could also use the much smaller BAT43 Schottky diode.   Heatshinked over, in series with the +ve lead is the way to go.

I'd use vinegar + brush.   Rinse thoroughly with water before the IPA - same details as earlier, but don't leave chips immersed in water for more than a couple of minutes.

[cs.dk]

Cut a strip from a Polypropylene food container (Recycling code 05/PP) and notch its end by about 1/2 wire thickness to act as a tool to hold the wire down in the gel dot while you blow on it to accelerate setting.

Sorry for OT, but why PP? Won't superglue tack on that material?

[Ian.M]

Polypropylene has the lowest surface energy of plastics commonly used for food containers.  Its notoriously difficult to glue.  The only thing better for a 'push stick' for gluing would be Polyethylene, cut from a heavy HDPE sheet or pipe then smoothed by pressing against a hot highly polished metal surface.  (Don't flame polish - it increases the surface energy and makes many plastics glueable)

[CkRtech]

Both ICs have been cleaned and are drying.

When scrubbing the 14069 - an IC which I planned to socket and try out on the board first prior to using a replacement - the legs immediately showed what I assume to be a gold (copper?) color after a few brushes.

So now I am wondering if I should be concerned about mixing metals with an IC socket (Tin, Tin-Lead) - a concern I haven't had to worry about before but am slightly aware of existing.

OR I could just fork over the $0.48 toward my grocery list of small components I am putting together and pick up a TI CD4069UBE to go with that socket .

Either way, wanted to throw in the mixed metals concern for educational purposes.

Appreciate all the feedback you guys have provided.

[SeanB]

Would replace the 4069, as the gold colour is the kovar leadframe, and you will have has some electrolyte wick up into the package as it corroded off the tin there, so it will probably work short term till the corrosion reaches the bonding wires, when it will go intermittently high resistance on the connections till they eventually all erode away inside the plastic case material. Best buy a new one.

[CkRtech]

All replacement components arrived. I soldered in all remaining items, jumper wired three traces, and fired up to test progress (with no batteries). It fired up on first try. So I have been happy since that happened - thanks again for the assistance.

Tonight I fired it up again and actually only got a power supply to kick in but no response (beep codes or otherwise) from the computer. Definitely a little strange. Turned off and back on and it came up. Still no battery connected - not sure if it is related. Ran a stopwatch with the clock in CMOS, and they seemed to stay in sync - so I think the donor crystal works well.

I'll take the battery holder, wire up the diode, do some proper heatshrink, and plug it into the board so I can reset CMOS once more and test retainage of the values. Hopefully adding the battery will present a 100% boot success at any point I turn it on. I really hope I don't have a cold solder joint somewhere. Lots of soldering and patient continuity testing went into this project!

If everything looks good, I will do a more proper routing of the trace repair wires and use the conformal coating. I definitely went to town with the fiberglass brush. Should I stick with a Q-tip method across the large traces or also bring in some brushing near the top where I exposed all of that copper (that now has components back on it)?

By the way - what is the typical coating on a PCB fresh out of manufacturing?

[james_s]

You might get unpredictable results without a battery. Another thing to check is the Power Good signal from the power supply and the reset circuit on the motherboard. Reset generally should be low for an instant at power up and then go high, allowing the CPU out of reset. If the Power Good signal is low(IIRC) it will hold the CPU in reset.

[bitseeker]

It's alive! Nice work, CkRtech.

[lasworren]

Sorry for resuming a quite old post (very useful for me! thanks guys!)
What do you use for desoldering process? I'm referring in particular to Isa slot.
A desoldering iron like this on ebay:
https://www.ebay.it/itm/ZD-915-Desoldering-Station-with-LCD-Display-designed-for-lead-free-desoldering/222764412825
maybe with a preheating step?

Hot air/infrared station?
Only desoldering pump/wick?

Thanks