weird silver leaf fuzz crap found on PCB? wtf , microscope pic included

[coppercone2]

So I was cleaning a HP 8350B mainframe, I fixed the problem, washed it out, dried it *usually im successful with this*, tested it last night (and it was broken, giving me e15) and today I had another look and I found some kind of weird 'blob' of fluffy highly electrically conductive stuff on one of the screws near the PCB. No clue what it is. I extracted it with a pair of tweezers, but it falls apart easily.

It's kind of like if you drop solder on ceramic while its molten and it makes a thin sheet, but imagine this stuff is about 1 thousandth thick, usually when you drip solder its fairly thick. This reminds me of gold leaf but silver colored. Like if you took gold leaf and crumpled it up good, or similar to the material on the inside of bubble gum wrappers.

Now all I really did on the main board that I found this on is pour a little dilute simple green on it, scrub around with a bamboo brush gently and then rinse off with water for a good while then DI water then light air compressor then thermocouple controlled bake for a while. I found this stuff 'accumulated' on one of the screws that screw an aluminum bracket on the main board.

Now I can't describe this any other way then really thin really fragile conductive foil material (probes to 0.1 ohm when stabbed with dmm). I found a 'blob' of this shit that is around 1cm in diameter near the power supply section.

I tried to photo it.


WTF is this? I never saw something similar in a meter, and I don't have any thin foils or anything like that. I never seen solder spatter so thin. I literary did nothing but slight agitation and water spraying. It looks like if you spread it out, it would be a high surface area thin foil.

[Gyro]

It sounds like Tin whiskers (google). I believe you can get it with other plating metals too, like Zinc and (hopefully not) Cadmium.

[jjoonathan]

Is it silver or white? If it's white, I suspect salt, which can form readily from metal and acid (or metal, water, and electrolysis). Hydrated salt will conduct, and many salts will self-hydrate if left in the open air. But if it's silver (and the term "foil" suggests that it is) then yeah, sounds like tin whiskers or some loose foil that blew in.

[coppercone2]

Hmm, I figured that would be more like a hair ball, this stuff looks like pieces of foil..

well a coin sized pile of tin whiskers is real bogus

Foil aint this fragile, even real thin foil, its like damaged foil. Its shiny reflective and has the conductivity of metal.

There must have been a pile of tin whiskers somewhere that the water during spray down washed into that area, because it was not present before cleaning IIRC, pretty sure I would notice that spot. This is like a odd shaped piece the size of a penny.

It is NOT the aluminum/zinc crystals, I know what those look like. This is looks like someone shredded a piece of foil that is on the back of gum wrapper and pressed it into a ball. Super thin. It also feels like.. shredded aluminum foil.

[coppercone2]

I might get the microscope out for this. I thought tin whiskers are supposed to be little rods.

[Gyro]

Silver can get pretty bad too!



https://nepp.nasa.gov/whisker/photos/pom/2003sept.htm

[coppercone2]

what on earth is this, its metallic.

that's as good a picture as I can get.

[coppercone2]

I guess its some kind of whisker? it looks too wide though, those all look kinda hairy

I thought maybe some crap flew out of a hidden capacitor or something, it reminds me of capacitor foil. There is one bracket I did not get to because it was a PITA and it looked clean in there, I guess it must have formed there, or maybe on the mains switch, I heard of that too IIRC.

[coppercone2]

I almost thought it was milling machine residue, because I do have a little mill that was doing aluminum and copper on the other side of the basement, but I picked some of that up and the texture is much grittier and the chips are larger (from 1/8 inch endmill).. this stuff I found feels very much like damaged aluminum foil (i.e. burned). I don't think I contaminated it with anything I have.

Maybe someone was doing something near that instrument with tools and it got in, I can't figure another explanation.

[coppercone2]

well with whatever residue that may have been left behind by unauthorized machining operations near the damaged roof set aside, I started removing all the structural brackets and shit, it looks like the residue the escaped specimen that hatched from those eggs stored in those long crimped specimen tubes left behind was spread out, it looks like it ate through some sub flooring and did some damage the signal distribution bus cover. No wonder the command center is not getting proper function. When I rubbed a wet finger on it, whats left of it came out slightly blackened. Must be manganese and zinc based life body fluids. Maybe the mop up operation spread that shit around a little and wakened life in the substance after rehydration. I never had to clean up a facility this large.


janitor's personal log: when dealing with xenobiotic fluid residues, a full disassembly is best, regardless of company timetables and management wishes. I wonder what kind of surprises lurk under neath the large water storage tanks and the facilities transformer. Remember to investigate auxiliary facilities in LV-8510 corporate park for infestation. Ask Robo Joe for cybernetic graft operation on whats left of hand. Try to borrow some napalm from the marine security detachment, supposedly it dries into a good plastic seal, maybe those bus bars will stop arcing over and I can get my microwave working again.

[coppercone2]

well it looks like the RF module broke too, the display was flashing in the new unit that works. But I resocketed the cards a bunch of times (at least 20 cycles per card*) and it looks like maybe something is up with the cinch connectors, it works. Maybe a spot of hardened nearly invisible oxidation. I guess I should make a cinch test stick

If you see flashing numbers on a 8350B module , it might mean bad cinch connectors. Could be part of the problem with the broken 8350B itself, but I might as well fix the solder mask before I trouble shoot the cinch connectors. Maybe weakened brass contacts or something. Pulled a green spec off one of the ribbon cable contacts too, even after ultrasound and blow out and alcohol and everything

One of the cards was really stuck in the other unit, but I did not see a sign of ripped spot weld or something like that.

this crap is like dealing with drift wood and rusty bolts after battery leak lol

*sometimes you just gotta turn your brain off and ignore Einsteins wisdom, its not insanity, its repair work, keep plugging it in and out and turning it on and off until it chooches. I also started banging on the side of the units with pipe pliers while it was turned off and on to see if it would do something.. and the microprocessor heard a rap a tap tapping and banging and decided to get off its ass and do something

[coppercone2]

ok, calm down, its just a old radio. The reflector parameters of the old cable you plugged into the machine are not important enough to have to fix it right now, its not like the planet will explode if you can't get the dicky module connector fixed tonight. if you poke it at the right angle on the front panel it works just fine, maybe you can fix it with a piece of shim some time 

anyone ever feel like they are on a broken submarine under Antarctica with these technical issues? I said I will just work for 10 minutes then I found myself standing there 3 hours later like no time had passed. movie night went bye bye

[coppercone2]

well, its the A3 board. I swapped them out between chassis and both instruments work on a good board.

No idea why it would break. oscillator is fine (10.7MHz). Interesting.

the u9 IC that is referenced so much is fine too, both units work on the IC from the broken board.

I wonder if I managed to screw up a bond wire in the processor or something

ultrasonic damaging a PCB seems rare, especially if its not the crystal oscillator. whats with my luck lol

[coppercone2]

god damn, the other HP 8350B has the same problem with the battery. The PCB is much less damaged but you basically gotta strip it to get the corrosion, because corrosive battery fluid gets under the structural brackets that are between the PCB and the chassis. Blue/green shit. It also gets the connector that the module plugs into. The bulkhead module connector also has some crystals in it and its intermittent.

What a mess I got myself into with these damn things. Fully loaded with 3 leaky cells, floods that PCB with corrosive  

at least its worth restoring.. but yeah expect 20 miles of bad road if you get a 8015C lol

and I gotta figure out the handles, I don't think I trust a braze repair, gonna have to bend some out of steel and weld them together, that thing is too heavy for the wimpy handle that HP put on it. Is all this worth it to see if a photon bounces off a circuit??