[OT] rust

Reply to Thread

My old computer is starting to rust. There are great withe patches on
metalic parts. Someone told me its oxide. It seemed like fungus so I
heated it with hair dryer. Then it started changing from white to red
(like rust).
What is the best way to remove that white/red thing from my computer?
(don't wanna use sand-paper or similar)

How to prevent computer from further rusting/oxidising ?

Please HELP me, this computer means much to me.
If possible, reply to:


TIA

 

Wow! Neat problem. Where do you live? generally?

I live in a humid location and hang on to computers long past their
prime and haven't seen this . . .

My guess is you have a problem with fungus. I get a white powder over
certain plastics (notably floppy discs - how are your floppies?)
Inside where the disc itself is . .

Circuit board makers sometimes coat the entire board with a spray of
varnish material to protect it from condensed moisture (important when
the traces are close together). If the varnish resin supports the
growth of mold - there's your problem . . .

Sand paper won't hurt anything (or help anything except to remove rust
_which probably ain't your problem). Sand paper has silica or
aluminum oxide for an abrasive (both non-conductive) Emory cloth,
however, sheds particles that are conductive - avoid it.

I'd stay away from any cure until I knew what it was. You could sand
it off and wind up with a lot of hydroscopic mold particles that could
conduct and the sanding would expose circuitry to the conduction.

What does it look like? Is the problem more or less universal over
the entire board - and how do the sister/mother boards look. If the
entire board is coated it is more than certain it is mold. If it is
splotchy it is more than likely it is mold. If only certain parts
have the problem - that may be different.

I've had a few power transistors rust being out in the weather for 20
years. They started brown red and have stayed the same color - they
still work.

White is mold with maybe some iron fixing bacteria that turns red when
you kill the mold.

If you decide it is mold you could sand blast with aluminum oxide, or
pecan hulls, or other non-conductive media then spray it with varnish
(protecting any sockets on the board or connectors). Or just ignore
it and replace the computer when it's life is up. (don't fix if it
ain't broke - philosophy) (the corollary: if it is broke, you have
nothing to lose by trying to fix it)

 

Are you talking about any specific parts *inside* your computer ?
Or maybe you mean the case/box itself ? inside or outside ?

 

Serbia (obviously?)

Power supply is all covered with white. Aluminium(?) box is not in
such a bad shape, but still has a lot of that stuff on it.

Plastic is fine, nothing on it.

The problem is not in the board, only metallic plates inside computer,
and the box from inside.

Could you explain that to me (my English is not-so-good)

It doesn't work, so I'll take that last advice.


Thanks for a great reply.

 

snip

I guess it would be obvious if I left the headers on the posts, sorry.

If just the box and power supply have the problem can't you just wipe
the stuff off? Do you have any fiberglass type scouring pads? Will
those attack it? I think machinists call the stuff "glass paper," but
it is sold as kitchen scouring pads here. There's a new brand out
that has an unknown, and more aggressive abrasive imbedded in the
glass - I'd avoid that.

Sand blast? Air is sent through a venturi (nozzle) that entrains
(sucks up) particles of sand or other abrasive. The sand hits the
parts (at high speed) removing everything to the bare metal and
removing a layer of metal (with sand, less with softer media like
small bits of nut hulls). Similar to shot peening if you are familiar
with that.

Remove the boards and treat only the affected parts.

Now my question is why are you concerned with the cosmetic appearance
if it is only the inside of the box? Improving the appearance may not
solve the problem.

I ran into a problem where my floppy drives were dying and no amount
of the service (people) that came with the computer I bought could fix
it.

Turns out the real problem was in the floppies themselves. Anything
older than a year that was stored away would have mold on it and the
mold would transfer to the read/write heads. The operating system
couldn't be loaded and the poor service people gave me a new hard
drive, new mother board, several floppy drives, and finally a new
computer. In the meantime I was out of a computer for months.

When the new computer also started having problems reading floppy
discs, I bought a floppy cleaner and some new discs that I store in
desiccant (now).

That was 15 years ago - I no longer buy computers with service
contracts.

Try some logical troubleshooting on it.

Does anything work? Fans come on? Fans work you can assume the +12
volts is present so the power supply might be OK.

Power on Self Test (PC's should beep once if everything is working -
the BIOS (Basic Input Output System) checks. More than one beep
indicate a problem and where the problem is.

Follow the symptoms . . . .

Abrasives to avoid are anything conductive - steel wool would be bad,
carborundum and silicon carbide (emory cloth) are conductive (if my
memory serves me) Garnet, and aluminum oxide are non-conductive
(typical sandpaper abrasives)

Good luck

 

Almost anything metallic inside computer, especially box of power supply unit
case is slightly affected, too. Have a look here:


http://home.drenik.net/orange/photo/rust/Set48_01.jpg (~50kb)

http://home.drenik.net/orange/photo/rust/Set48_02.jpg (~50kb)

http://home.drenik.net/orange/photo/rust/Set48_03.jpg (~50kb)

Bye.

 

Hi Orange

That looks like just common rust to me. My guess is that the white
stuff is a zinc plating or a "zinc wash" that was applied to prevent
rust. Only good way to fix is to sand (or sand blast) it to the bare
metal and repaint.

One picture is worth a thosand words.

 

no, i'm sorry, its not obvious at all.

I wouldn't like using any abrasive at all. The best thing is probably
electrolysis. The only problem left is how to protect metal after
removing rust.

Yes, it didn't. But I'm worried that rust will "eat" metall faster
until everything is gone. (if not cleaned) Besides, it smells bad (but
that probably isn't coming from rust).

Where can that desiccant be bought?

No, the fan is *trying* to start and some clicking is heard. I've
opened the computer and removed PSU from it, but the same thing
happens (clicking). At one moment fan started to work slowly but it
stopped. When I unplugged it from AC it started rotating again for
short time. After opening PSU, it looks like electrolytic capacitors
have leaked, but I'm not sure. (never before seen leaked capacitors,
only heard about them). Do you think they should be replaced?

Probably the best thing is to take it to service. (hope it wont cost
too much)

BTW, the computer is Amiga 2000.

And just one more thing. It seems that motherboard is affected by
moisture, too. After trying to clean it with alcohol, the cotton was
green. Whats best for cleaning PCBs?

Thanks. I'll need it.

 

That looks like just common rust to me. My guess is that the white

I'm sorry for (accidentaly) posting those pictures twice.

That white stuff seems to appear on places someone touched by hand. It
seems there are some fingerprints visible.
I've tried spray called Kontak 60 from CRC. It's great for removing
that white stuff, but not good for red. On the side of spray, its
written that it removes oxide and prevents corrosion, just what I
need. But a LOT of it is required to clean all of the metal. Besides,
maybe it "attacks" metal, not just oxide.

How much do you thin repairing power supply should cost?

 

I see mail order ads for it from time to time. It lasts forever, so I
haven't bought any in a long time. They pack it in cartons arriving
from overseas and I save what comes my way. It comes with all new
electronics shipped here. Small envelopes of the stuff (marked
"discard - do not eat" - like that would be my plan . . .)

Sold for and used in chemical labs to preserve reference standards.
They buy the loose crystals and it has an indicator added so the
crystals turn from blue when dry to red when wet.

You can bake the crystals in an oven or microwave to drive off the
moisture and make them more effective.

Very handy stuff for preserving gummed envelops or stamps in a humid
climate. Put the stuff in a plastic bag with a some desiccant.

The chemical name is "silica gel."

Clicking can be a symptom of a shorted load or open load (no load
plugged in). Switching supplies use a type of over current protection
that clicks when the supply is over loaded - in the case of no load
the over voltage protection kicks in and shorts the supply. It is
normally a rapid clicking ("Hiccup") several times a second on PC
supplies.

The internal fan on PC supplies usually is run from the +12 volt
output and in between clicks it might turn.

Electrolytics can vent and leak. There is very little actual liquid
electrolyte. The ones I've seen have had a white deposit around the
leads and between the cap and board. Some also had corroded leads
that would break off the cap. I wouldn't want to guess about
replacement unless I saw them - but if they really did leak they may
need replacing. The vented leaked caps have lower capacity. They
may be "low ESR" type capacitors if it they filter the output of a
switching supply.

There's a lot of differing opinions on cleaning PCB's. One place I
worked at used a trichloroethane solvent in a vapor de greaser for
machine produced boards, on the human produced boards they used
acetone. The manual technique was to lay a paper towel against the
board and work a brush wet with acetone into the paper - the idea was
to let the solvent dissolve flux and let the paper towel soak it up
and trap it. (good technique for spot cleaning rugs) They also soaked
boards in acetone.

Whatever solvent you use may damage the board. What you really want
to avoid doing is finding the solvent that dissolves the "mask" a
green varnish applied to the traces to protect the copper and prevent
solder from getting where it doesn't belong during production. That
mask also protects the board from condensed moisture.

At home I use acetone with the towel technique and will sometimes also
use distilled water. I bake the board in my oven at low temperatures
to dry them immediately after. Acetone vapor is explosive - so that
stays out of the oven until the solvent evaporates. When using a
solvent that evaporates rapidly (like acetone) the temperature of the
board drops and if it is humid, moisture from the air can condense on
the board so even acetone can wet the board with water under the right
circumstances. Alcohol frequently contains some water.

The worse thing to do is apply power to a board that is wet.

Those pictures you had up on the web with the case don't look
encouraging . . . You ever heard the expression "beating a dead
horse?"

 

Did you post the pictures twice? "One picture is worth a thousand
words," is just an idiomatic English expression.

That white stuff is just some zinc oxide. Under that you have rust in
places. Phosphoric acid will attack rust, but it needs to be washed
off and neutralized. Something sold here called "Naval Jelly"
contains phosphoric acid as a rust remover. Deadly to electronics
however, since it does attack metal as well. Sandpaper is safer.

The cost of repair includes too many variables to just hazard a guess.
I don't think you can say the supply is the problem. Plugged in it
may be overloaded by a problem somewhere in the computer - unplugged
it may not work correctly since most switching supplies have a minimum
load they need to be happy. The symptoms may be the same in either
case - the clicking.

A bad capacitor could cause the symptom - but so could a lot of other
things.

You'd have to put a dummy load on the supply outputs and check the
voltages to be sure.

Here a PC power supply can be had cheaply in the surplus market so
replacement is probably a better option.

My computer has a fan going out in the supply, so I just looked up
supplies. New supply is ~$40, surplus supply is ~$20, fan was $1
(surplus) I bought five awhile ago.

 

This clicking is not so rapid. I'll put some pictures of PSU on net.
That might clarify the problem, OK?

If they leaked, could that cause clicking?

I dont understand this; could you explain to me meaning of word "work"
in that context?

Even water takes some green off the board, so there probably isn't
anything that won't damage the mask, right?

Hmmm, yes, you might be right about that. Still, I don't want to think
about it as "dead", but rather as "ill". Besides, rust shouldn't
affect functioning that much, or does it?

 

OK I've never fixed anything by looking at it - laying on hands
doesn't work either. Not for me, anyhow.

If they leaked, they could be bad, and bad caps can cause problems -
but so can many other things. If it were me and I was sure the leak
was from the cap I'd probably change them - but I would approach the
problem of a bad computer with logical troubleshooting.

Do you have any electronics experience? Have any test equipment?

One can tell a lot more with a few pieces of test equipment, and some
checking, than any amount of guessing.

We would take a brush with stiff short bristles and wet it and "press
it against the paper towel and rub it around." (work it) The liquid
and action of the rubbing get the flux loose and dissolved and the
paper towel absorbs the thinned flux.

I don't know. The masking varnish I'm used to seeing, withstands
water, rubbing and acetone without dissolving. Even some of the cheap
consumer electronic junk I've worked on withstands common solvents.

I've never seen a varnish mask dissolve but there is probably a
solvent that will do it . I worked in a pharmaceutical lab - alcohol,
water and acetone are relatively mild solvents.

Rust should not affect the functioning of your Amiga.

What is it about some Amiga users that make them so die-hard dedicated
to their machines? If there is a rational reason for it, someone
ought to capitalize and write an OS that will run on an Intel computer
to emulate the Amiga (or has that already been done?)

Have you tried some of the Amiga newsgroups? They seem more game
oriented to me, but there may be some technically competent people
there.

http://www.amiga.com/ any help?

 

I've taken it to service. The repair-man told me that he repaired some
high
voltage protection that was blown up, but the output is "crazy";
instead of 5v he gets 12v, etc. He will still try to repair it, but
chances are slim. (((

I've got a good voltmeter and soldering iron.

There is a good manual on the net called "PSRepair.pdf". But I don't
understand it.

This mask is a bit strange. It's not smooth like ordinary one. There
are stripes with rough surface (like some dots on it).

There is UAE and Amithlon, but nothing compares to real Amiga.
Correct me if I'm wrong: even the fastest Intel/AMD processor has to
be compatibile with 8088 (whatever the first one was named), and 68000
was a lot better than that one. The thing I like most is that most of
Amiga OS is in ROM, so it's fast (except 3.5 and 3.9).
All of this might be wrong; still there is something unexplainable
about "good old times" that makes us love them.

No way. They have stopped supporting "real" Amiga users long ago.


BTW, have a look what one of the (formerly) greatest Amiga hardware
company tells us:

http://www.phase5.de/

 

(Orange) wrote:

(snip)

I just had my PS apart yesterday and the computer was down most of the
day (blowing out the dust and fixing minor problems like the fan and
spare modem). Looks like all the protection it has is a small
capacitor across the line and a fuse, from there the line goes to the
input rectifiers and filters as shown on the repair pdf. My supply
was made in China and incredibly cheaply made. Your Amiga is probably
different.

I don't see from looking at the schematic how the 5V supply could
output 12 volts, but if that is the case you can figure it destroyed
all the 5 V logic on the main board. (presumably the repair person
knows that a minimum load may be necessary to check the supply with
the correct voltages)

Taken from the repair PDF:
". . . connect its power plug to the dummy load shown in
Figure 2. (I don’t recommend running a PC power supply
without a load.) . . ."

If the supply you have is like the PSrepair paper, the five volt
supply is closely regulated and determines the outputs of the other
supplies - so the 12 volt supplies would be proportionately higher as
well.

Sorry about the manual. I downloaded it. If you don't understand
that, your chances of repairing your supply is pretty slim.

I don't know about compatibility between processors. My forte is
industrial and laboratory electronics. (mostly analog, and discrete
digital hardware)

It would seem to me that the "Internet Two" model, and Digital Rights
Management as being developed by Intel and Microsoft will preclude the
operating system compatibility we enjoy today. Microsoft is intent on
maintaining their monopoly on PC Operating Systems.

My computer interest is a matter of necessity rather than preference.
I only had to watch the gorilla they sent out to fix my machine a
couple of times to realize I'd better learn something myself.

Having the OS on ROM would be an idea that PC makers ought to copy.
With all the flash memory available at reasonable prices it would seem
natural that someone would do it. It would also make it possible to
save more energy in "standby."

It looks grim.

 

That service man told me he blew up some transistor but he'll replace
it.

I tried it without load and nothing happened.

But you are here to help me, right?

Is there any old design that you prefer more than newest
mega_super_turbo stuff?
I've heard about analog computers in university, they are cool; what
is discrete digital hardware?

People would rather shift to Linux than sacrifice their freedom to
some DRM.

I dont see why they couldn't even use ordinary RAM with some battery
back up for refreshing it? Would that "eat up" battery too quickly?

 

I sent you an email

 

I would prefer acid because it removes less metal than sandpaper, or
does it?
Can acid be washed off with diluted baking soda (how much diluted
should it be)?

I took it to service, but the repair-man couldn't fix it completely.
He repaired something called high voltage protection, and fan works
now. BUT, the voltages are "crazy", where it should be 5v, its 12v;
instead of 12 its 20.

Its not so easy to replace Amiga power supply with PC because it needs
"tick" signal. Still, that is a good idea. If all else fails, I'll
have to replace it.

 

I'd recommend against acid unless you separate the metal from the
electronics. I'd also recommend against acid where you have that zinc
coating. Sandpaper is much safer.

With acid, you could always check the PH of the rinse water to see if
it is neutralized. With acid there's always the concern that you
didn't remove it all. It can lurk where the metal parts overlap or in
pores in the metal.

You might also use a paint coating designed to be applied to rust.
"Rustolium" brand here in the states. You burnish/brush the loose
stuff off and apply the paint. (many good oil-based paints will also
work)

If the area you live in is industrialized, there may be plating shops
that can clean and plate the parts for you.

That is believable: the + 5 volt is regulated and sometimes the -5.
The ±12 are raw supplies that base their output on the turns ratio of
the switching supply transformer.

If the feedback loop to the pulse width modulator (control circuit) is
open (a fault that would cause that symptom) it will sense a low
supply voltage on the +5 and attempt to boost the output voltage to
compensate. The ± 12 is along for the ride - tracks the +5.

BUT that may not be the case if there is no dummy load on the supply.
Some supplies (especially switchers) require some minimum load or the
supply voltage drifts up. Presumably your repair person already knows
this . . . Check the repair pdf for a dummy load that is cheap and
easy to rig. Ask your repair person if he is using one - show him the
pdf file if necessary.

The tick signal shouldn't be that hard to rig. Assuming a bad supply
with replacement as the only option.

A supply that is outputting voltages is probably reparable - the only
thing that may be impossible to fix would be the transformer, and even
that may be reparable or replaceable. Too high an output is probably
not a transformer problem.

Is that a valid email address? @mail.ru (or) @drenik.net

 

I can remove metal from electronics completely, but zinc coating is
everywhere (except for rust).

But doesn't it evaporate completely ?

I have something caled wash-primer and zinc spray.

I don't know where to look and it is probably too small for them to
clean.

Hmm, easy... I can simply steal the light bulb from nearest car.

Some good news, at last.

Both of them are valid. I use @mail.ru at office.