Repairing a Rusted Chip

[tpw_rules]

I've got this vintage laptop (Toshiba T3100) with a Matsushita gas-plasma display MD400F640PD2. Unfortunately, there must have been some screwup at the factory and now most of them are rusting. Specifically, the plasma driver ICs get rusty pins which stop conducting, causing lines on the screen.

Given that replacing the screen or the chips does not seem to be especially feasible, is there any way to repair the ones I've already got? Fortunately, the rust is on the chips and not really on the flex board itself. I've considered dremeling the sides with the legs to remove them and expose fresh leadframe, but I'm not sure where to go from there. Pictures of some of the chips are attached.

[helius]

Neither the packages nor the FPC should be ferrous in nature. What is the liquid in the picture? Are there electrolytic caps leaking from up above? Is anything in the assembly glued?

[tpw_rules]

Neither the packages nor the FPC should be ferrous in nature.

Fair enough. But it's been corroded to the point that the lead metal has been eaten away in several areas.

What is the liquid in the picture? Are there electrolytic caps leaking from up above? Is anything in the assembly glued?

I'm not sure regarding the shininess. It's not liquid, it's some kind of hard conformal coating or lacquer. There are no leaking electrolytics. There is glue in the assembly, but it's to keep the board folded over. You can see the black foam remains of the glue pads on the left edge.

Whatever happened to mine, it happens to all of them. I've even seen threads on this very forum about their screens getting lines, from what I assume is the same problem. All the ones on ebay have lines as well. Everything else in the machine is rust-free. Rest assured I didn't dig this out of a rainforest and wonder why there was rust; I'm convinced there was some kind of incorrect process during manufacture (maybe corroding flux) or design (the entire module sits at -200V) that has doomed all of these screens.

[JFJ]

Neither the packages nor the FPC should be ferrous in nature...

Copper IC leads are often iron plated - prior to receiving their outer plating of nickel (or other non-ferrous finish). That is why many IC's can be picked-up with a magnet:

[Paul Moir]

I'm sure you'd be able to grind into the plastic a bit to expose a bit of lead, then solder on a wirewrap wire.  I used to do that all the time as a kid when I accidentally broke leads off.

[james_s]

I've used some stuff called Evaporust to clean various things with success. Never tried it on an IC though.

[helius]

Whatever happened to mine, it happens to all of them. I've even seen threads on this very forum about their screens getting lines, from what I assume is the same problem. All the ones on ebay have lines as well. Everything else in the machine is rust-free. Rest assured I didn't dig this out of a rainforest and wonder why there was rust; I'm convinced there was some kind of incorrect process during manufacture (maybe corroding flux) or design (the entire module sits at -200V) that has doomed all of these screens.

I wonder if there is an electrolytic process at work between adjacent pins? It looks like the conformal coating either failed or was inconsistently applied. It may have been required to prevent electric creepage at such high voltages.
Another issue could be electromigration inside the flex PCB, causing CAFs that subsequently corrode the pins—I can't see under the chips.
Either way it looks like a challenging repair.

[coromonadalix]

I would try to slip a fine needle to remove anything between the ic legs / pins, try to very gently brush them with an horse hair tooth brush, just to remove rust excess, and try an power up, if its fine, i would coat them with an colloid clear to cut the oxygen ...

my 2 cents

[T3sl4co1l]

More likely the lead frame is steel (or invar, especially in the case of ceramic packages?), then copper and/or tin plated.  Dunno how common that is versus solid copper.

Copper can also corrode with a brown oxide, being Cu2O (cuprous oxide, the mineral cuprite).

Tim

[Rasz]

ultrasonic cleaner would be able to shake most of that crap off, except its all connected to the glass in your case?

I would leave mechanical solutions (scraping with needles etc) for last, those are prone to irreversible damage.
Removing those ICs from flex wouldnt hurt either, but that would depend on your hotair skills.

Maybe start with vinegar on a cotton bud, it will eat al the rust, but monitor the process and remember to clean it off

[kizmit99]

I've used some stuff called Evaporust to clean various things with success. Never tried it on an IC though.

Evaporust is great for removing only the rust and not affecting the base metal.  I've never tried it on any electronics, but have used it many times on machined (steel) parts and other than removing the pre-existing rust it did no further damage to any of the parts (doesn't etch the surface or anything like that).  I've also never seen it damage plastic, so I would expect it might be worth a try on these (even without removing the chips from the flex-board).

[Gyro]

Remember that plastic packages are not hermetic, anything you put on the leads is going to find its way to the [Edit: probably Aluminium] bond wires and the Silicon at some point. I would go carefully with anything chemical. Stick to mild things and remove any moisture afterwards with IPA. Mechanically removing as much contaminated 'crust', dry, as you can is probably going to be less contaminating than letting it get back into solution.

[james_s]

That's an interesting point, has anyone done testing to see if an IC is affected by long term liquid immersion? I've certainly washed plenty of boards, and I've seen old nasty crusty ICs that still worked. I've also seen plenty of bad ICs though unfortunately it's rarely possible (with my resources) to determine the cause of failure.

[Bashstreet]

It is best first just clean everything with alcohol and toothbrush to see the real damage under the surface rust.
It might be many pins will be beyond repair and you might as well stop there.

If there is hope and if you can bodge pins that have gone you need neutralize further damage from happening.
Vinegar is one option but it is corrosive and not really useful in this case as it can do damage.
Better use rust converters that convert the rust and prevent further damage (If legs are ferrous)
You can get rust converters from most hardware shops.

[Gyro]

That's an interesting point, has anyone done testing to see if an IC is affected by long term liquid immersion? I've certainly washed plenty of boards, and I've seen old nasty crusty ICs that still worked. I've also seen plenty of bad ICs though unfortunately it's rarely possible (with my resources) to determine the cause of failure.

I'm not sure about long term liquid immersion, but there have certainly been lots of plastic package humidity reliability studies, eg. http://web.cecs.pdx.edu/~cgshirl/Glenns%20Publications/ASQ%20Webinar%20Package%20Reliability%20CGS%20Rev%203.pdf


P.S. Most manufacturer sites have some information, but this mostly relates to package cracking due to boiling of absorbed moisture in non-dessicated parts during reflow.

[helius]

Inside the package are the lead frame, the gold bond wires, and the chip itself. Gold is pretty hard to damage. The chip is passivated with what is essentially glass and is also quite resistant to all but the most caustic chemicals. The pads to which the wires are bonded are not passivated: they are made of aluminum or aluminum oxide.
The most susceptible thing inside the package is the lead frame, which as mentioned earlier may be ferrous. The references I was able to find state that the lead frame is either a copper stamping, or a photoetched coupon of Alloy 42 (58% iron, 42% nickel).

[james_s]

I recently came across a handful of very old ICs, as in late 70s which had been stored in ESD safe foam. Unfortunately the foam had broken down and seemed to have attacked the leads, they were covered in black oxide and rust. I tried brushing off the bits of rotted foam and several of the pins fell off too. Not being super rare valuable parts I threw them away, whatever the pins were made of certainly seemed to be ferrous.

[coromonadalix]

I would try to clean the leads on place without trying to de-solder them, you may add thermal stress in them, they may break ... as i wrote earlier, if it still work, add an sealing rosin with high dielectric isolation on top of them ..

Trying to remove them with hot air may work, but the flex pcb maybe damaged in the process add kapton tape as a shield ?

[james_s]

I agree, I would also avoid desoldering them unless absolutely necessary. I think I would try Evaporust first, maybe lay the board chips-down in a glass cake pan and pour in just enough Evaporust to immerse the PCB. It shouldn't hurt the plasma panel to get wet but no reason to tempt fate.

[amyk]

What is the marking on those chips?

Maybe there are replacements or equivalents available.

[helius]

You can find them for sale by chipmongers with minimum quantities of 40. Group buy opportunity, perhaps.

[JacobPilsen]

but the flex pcb maybe damaged in the process add kapton tape as a shield ?

Flex PCB is made of Kapton.

[stj]

what a mess, that's what happens when pcb's are left covered in flux.

the tin-plating / tin in the solder has rusted because moisture has become trapped in the flux.

[metrologist]

I need to see cleaned pins, immediately. 

[coromonadalix]

loll   we do use very strong flux at my job, we have at least 24-36 hours max to remove it, afterwards it will attack the parts and the pcb plating, because it's an eco friendly flux we can flush the drain  lolll

[Bashstreet]

what a mess, that's what happens when pcb's are left covered in flux.

the tin-plating / tin in the solder has rusted because moisture has become trapped in the flux.

Tin do not rust 

[stj]

yes it does.
you must be pretty young to have never seen a rusty tin can before they switched to aluminium.
even so, you should have seen rusted pins on leadfree soldering that was dumped outdoors.

[helius]

Tin is an expensive metal. When canning was originally developed in the early 1800s, steel was rolled with tin to make tinplate. The cans were not ever made of solid tin—that is absurd.

If you really believe that cans were made from solid tin, you should go out and buy all the razor blades you can, since they say they are platinum.

[T3sl4co1l]

Tin corrodes as well as any other metal, though it's not usually called "rust".  Tin is rarely used pure, almost always as a plating.   Copper has an oxide that may be confused with rust.

Tim

[james_s]

As far as I know, "rust" refers only to iron oxide, but oxygen is a very reactive element and most metals will oxidize readily. The biggest difference is in the properties of the oxide. Iron oxide crumbles and falls off exposing fresh iron which then oxidizes and this is what happens when something rusts away. Aluminum oxidizes more readily than iron but aluminum oxide sticks quite firmly to the surface forming a barrier that protects the metal below. I don't know as much about the properties of tin oxides but I know they will form.

[Bashstreet]

yes it does.
you must be pretty young to have never seen a rusty tin can before they switched to aluminium.
even so, you should have seen rusted pins on leadfree soldering that was dumped outdoors.

No Tin do not Rust

The cans are not pure tin it would be too expensive and pure tin is too soft.
They are just tin plated soon as tin gets knocked, way to the iron/steel is exposed and they rust out 

[stj]

so why does lead-free soldering rust???

[james_s]

If it rusts then it must have iron in it. I don't recall ever seeing solder rust but I suppose there could be an alloy with iron, or maybe the flux resembles rust?

[Bashstreet]

so why does lead-free soldering rust???

No idea ? why would anyone add iron on solder.
I have never witnessed such solder nor seen it sold 

[KL27x]

I would guess the iron is in the chips. And that additionally, these boards were soldered with a water soluble flux containing ammonium/zinc chloride. It's probably the steel in the lead frame producing the rust, and the epoxy case of the IC was not adequate to seal it from chloride ions and air getting in there. So tinning can't cover the steel and stop the rust, because it is leaking out from where the pins come out of the epoxy. It's maybe older encasing tech or some failure in processing that damaged the epoxy.

If OP could not buy new chips, I would remove them all, clean the flux residue thoroughly, and dunk them in any acid but HCl to eat away any rust. Rinse in water and dry with alcohol. Dunno how many chips that is. It's certainly a lot of work. But mechanically removing the visible rust probably can't stop the source of the problem. This probably can't either, but it theoretically has a chance... of slowing future problem. If the chips are already acting funny, it's probably too late. It's not just presence of rust you can see, but there is maybe rust in the chip pushing on and breaking the wire bonds.

If you can buy new chips I bet you have no problem in the future, provided you are using halide-free flux. I would wash the FPC thoroughly in hot water and a brush before putting on the new chips. I imagine they would be a horror show on the inside, already, by the time rust is pushed out of the chip.

* scratch all that.
If you have device(s) that still work, I'd clean what I can see with a fiberglass brush, then coat thoroughly with some suitable lacquer/paint to slow any more oxygen from getting into the chip.

I have some PCB I soldered with plumber's flux. Maybe this will happen to them in 30 years. 

[PKTKS]

Nobody said the most plain solution which BTW I have done
a lot with decent results.

I just clean what is possible (the  just metallic dust part) with ISOPROPANOL
As the chip may be very fragile a soft brush in a strong solution should do.

Then if not wasted completely I would cover the parts with SOLDER MASK (UV curable)
with some thin tip a portable UV.

The really wasted parts I would consider a JUMPER  -  really not that hard w/today
tools and optical assist. And BTW I would use LEADED solder.

LEAD-FREE profile is just too complicated to bond that crappy left overs.

It usually works cheap and easy
Paul

[Bashstreet]

Nobody said the most plain solution which BTW I have done
a lot with decent results.

I just clean what is possible (the  just metallic dust part) with ISOPROPANOL
As the chip may be very fragile a soft brush in a strong solution should do.

Has been recommended by many.

Then if not wasted completely I would cover the parts with SOLDER MASK (UV curable)
with some thin tip a portable UV.

Unless the rust is neutralized it will continue corroding under the solder mask as it has bonded with sufficient oxygen.

The really wasted parts I would consider a JUMPER  -  really not that hard w/today
tools and optical assist. And BTW I would use LEADED solder.

LEAD-FREE profile is just too complicated to bond that crappy left overs.

Using leaded solder to rework lead free boards is wrong way to go.
Use lead free solder to repair lead free boards 

Cheers.

[helius]

A device made in 1986 is very unlikely to use lead-free anything.
But in general, unless there is a very good reason, rework should use leaded solder no matter what the factory used. It makes better, more consistent joints, removes the risk of tin whiskers, and avoids unnecessary thermal stress to the board and components.

[VK5RC]

Have those leads corroded right through? - looking at the photo - it appears as though something is joining the pins - perhaps rust but other oxides possible - do you have access to a fibreglass pencil brush? If you do, wear gloves, sealing safety goggles and work outdoors - little bits of fibreglass in you fingers take weeks to come out (don't ask) - can't imagine what would happen if you got some in your eye.

[PKTKS]

Once properly covered the amount of corrosion will be very low.

If you manage to clean enough a still usable part the protection cover will last
last enough to not worry for a long period.

The tip on LEAD-FREE is just to make that sure that by some convenient
you pick the wrong solder.  LEAD-FREE today is as common as LEADED

IMHO  for hand soldering BENCH REPAIR - LEADED just works better
LEAD-FREE or paste are very specifics.

In that case I would do all in LEADED with a very very minimum temp
to avoid stressing the corroded parts. They will get worst (worse?!) with the heat...

Paul

[Bashstreet]

A device made in 1986 is very unlikely to use lead-free anything.
But in general, unless there is a very good reason, rework should use leaded solder no matter what the factory used. It makes better, more consistent joints, removes the risk of tin whiskers, and avoids unnecessary thermal stress to the board and components.

It is better repair lead free boards with lead free solder as it is not good idea to mix alloys and different melting temperatures can cause bonding issues.

You can repair lead free boards with leaded but as guide it is better use same alloys.

Modern lead free solders are very good and there is not need to continue using leaded solders except to repair very old equipment or in special circumstances where leaded solder will be required.

Some people seem be very fond of their leaded solders that's fine but times move on and no point get too nostalgic about solder

[james_s]

I haven't found a lead-free solder that works as well as the standard tin-lead stuff. With good old 60:40 rosin core I get nice clean shiny joints every time.

[KL27x]

There are several very different lead-free alloys. Without knowing what alloy and what flux has been used on a board, do u still rather use any random lead free? Im asking, not judging. I am jot voting, here, but I have been reworking lead free boards with leaded, and I'm curious, now.

Through the years, several new products have hit the market. Is there even a standard lead free alloy, today? With leaded you are talking a few percent here, maybe a dash of silver. pretty much it.

[PKTKS]

I haven't found a lead-free solder that works as well as the standard tin-lead stuff. With good old 60:40 rosin core I get nice clean shiny joints every time.

Neither do I.

For repairing rusted or oxidized parts LEAD-FREE just can't do it

A good tin bond requires plain LEADED and 99% cases I use a flux
proper to OXIDIZED MATERIAL - not the regular "normal" flux.

The thing i use has several drawbacks:
- contains AMONIA - very crappy smoke fume
- it is HIGHLY CONDUCTIVE and if not properly cleaned it will left conductive residues.

But it works like nothing more. It justifies the drawbacks.
Once applied it will  VANISH all corroded material and BOND the LEAD SOLDER almost
perfectly in minimum time. It works but there are drawbacks using the stuff....
Normally this flux is used in HIGH THERMAL mass chassis and parts. Not micro-soldering

But i haven't found anything that works near close that ...
for rusted oxidized parts. No matter how bad the rust
the crappy flux vanish and bond the tin like magic.