CAN MY ACID DAMAGED BALLY MPU BE REPAIRED?
It is amazing
how far the electrolyte can travel around this 10” X 10” area. I have seen
boards with only a little damage on the ends of the battery but the electrolyte
had created a path of damaged pins under all three 40-pin sockets in the center
of the board. The fumes from the decaying NI-cad had embrittled the metal pins
of the 3 sockets, shutting down the MPU. When the covers were lifted from these
sockets, the broken pins just fell out. The damage was 7 or 8 pins wide on all
three sockets, with none of the usual signs of battery corrosion.
Another board
will show evidence of major copper erosion on the areas near the battery with
just a little traveling. Still another will show that the damage has followed
the copper up both sides and across the top of the MPU. To determine how much is
too much, it is the expected results from the cleaning process that will
decide which board gets repaired and which one gets sent back.
This can be a
tricky decision because once it’s started I’m committed to finish. Even after a
thousand MPU repairs I’m never sure of what I’m going to find. This is why other
factors help me to decide. Factors such as general overall look of the rest of
the board. Let’s look at two examples:
Board “A” has
heavy acid damage covering the bottom inch of both sides of the board and
extends up under connectors J4 on the left and J3 on the right. Major sections
of copper are eaten away along the entire bottom ground rail. The pins under the
U8 socket are reduced to little green stumps. Remainder of board is clean and
original.
Board “B” has
only a little acid damage along the bottom ground rail and the corrosion is just
starting up under the U8 socket.
Up to now,
the choice of which board’s the best candidate is easy and obvious. BUT, a
closer look at the rest of board “B” shows that someone has been working on this
slightly acid damaged MPU. They have replaced several IC’s but neglected to put
down any sockets first. In addition, they have ‘reflowed’ every pin on all of
the five connectors in an attempt to get board to run. These repair attempts
have ruled out any additional work on this otherwise salvageable MPU. I return
20 % of boards sent for repairs. 10% were too heavily acid damaged while the
other 10 % were heavily damaged from ‘repairs’.
Board “A” was a better
candidate.
It is not
unusual to find a CLEAN MPU that has had all of the (original) sockets replaced
or all of the IC’s replaced in an attempt to fix the MPU. This is the reason
that I ask several annoying questions before telling anyone to send in their
board. Yet, somehow, it seems that the board suffers additional acid damage in
transit because the MPU described to me over the phone is suddenly dripping
electrolyte when it arrives here.
Some of the
worst advice is free advice. The help found on some newsgroups can be beneficial
BUT it can lead to major board damage too! These Bally MPU’s are 20 years old
and (even if non-acid damaged) are getting fragile. To practice repairs on these
tired old boards is foolish. Those POW (Previous Outside Work) MPU’s are the
ones that I send back.
I charge
$69.00 for my Bally/Sterns MPU overhaul and this is about as MUCH as anyone
should spend on any MPU repairs. It is possible to fix just about ANY board if
enough time and $$$ is expended in the effort. But WHY? I tell my customers
that anything beyond the $69.00 should be put into finding another MPU. I hear
of MPU repairs in the $120.00 to $150.00 range and I wonder what kind of work
was being done to get the repairs that high. Why wasn’t the customer told that a
few more dollars would have gotten them into a clean (used) Bally MPU? (or a
MUTHA.PCB).
The purpose
of this series is to show exactly what is involved in the repairing of a
‘typical’ Bally MPU. The MPU shown below was sent back due to excessive damage
over large areas of the board.
Some of
the damage is obvious.
The entire upper
right section is coated with a corrosive film...
The normally
bright solder pads have turned dull gray and the corrosion has seeped under the
dip switches. This will create chaos with the I/O of the switch matrix. It is
obvious that this MPU will require hours of parts removal, bead blasting, and
rebuilding. Even if all that work were done, there is NO guaranty that elusive
shorts won’t show up in the finished board.
The smallest
particle of corrosion left on the board or under a connector will allow for
crossover signals on the switch lines. The size of these particles is so small
that they cannot be seen even with heavy magnification. Even the slightest
amount will short out switch lines and it is often necessary to lift up the
covers on connectors, to allow for several applications with the bead blaster,
before the switches will show the proper numbers in test.
This is not
the worst MPU that has been through here but it is typical of the type that is
being sent back as unrepairable.
The next
series of photos will show how an MPU (with average acid damage) is brought
back from the ‘dead’. This the board “A” described earlier, with areas of heavy
damage but otherwise a very ‘original’ Bally MPU.
The main tool
required for the restoration of any acid damaged board is the bead blaster
(commonly known as a dry blaster). This is simply a miniature version of the
more common sand blaster used in AutoBody shops. The cleaning agent is a fine
grained silicon bead (100-170 mesh).
In this
smaller version, the bead blaster stands about 3 feet tall and is about 16”
square. This requires a fairly good sized work area and with the required air
compressor added in, it can take up a lot of shop space. But, there is no
better way to clean the surface of an acid damaged MPU. No amount of scraping,
sanding or picking will match what this machine can do in seconds.
A closer look
of the inside of the bead blaster will explain how it works, what it can and
can’t do for a damaged MPU.
Shown in the
bottom (of the cone shaped chamber) is the copper tubing. This pickup tube has
a 1/4” opening at the end and several additional smaller holes, used for drawing
in the beads by suction. This suction is created when compressed air is forced
thru the nozzle. A venturi effect is created in the nozzle chamber, as the
compressed air goes thru. This pulls the beads up from the bottom of the
chamber, through the copper tubing and mixes the beads with the air blast.
The rods
across the hopper support the wire mesh platform shown standing in the upper
left corner. The square pad on the bottom of the hopper is part of spring loaded
drain valve, to help empty the chamber for cleaning. Keeping the beads clean is
a constant battle. It only takes a small amount of silk-screen paint chips to
shut down the pickup tube.
The above
photo shows the bead blaster in action. The nozzle directs the flow of the air /
bead mix, as the operator controls the area being cleaned. How long and how
close the nozzle is directed at a particular area is critical. Just the right
distance is combined with the right amount of time and the bead blaster will
clean (but not damage) the copper surface.
The area
being cleaned (in the above photo) is on the back of a Williams Driver Board.
The burnt section is opposite the row of eight 5-watt resistors in the lamp
section. These resistors are famous for heating up and burning, both the top and
bottom surfaces of the board. The area is not acid damaged but has
surface oxidation from continuous heating. It is difficult to solder, if not
completely clean. Bead blasting is a natural for this job. The rubber gloves are
worn to stop the high velocity beads from giving the operator skin abrasions.
After the
front and back of the Driver Board are bead blasted, the results are apparent
with these before and after photos.
BEFORE
AFTER
Also revealed
by the blasting process are the traces and connectors. Now, it is an easy job to
examine all connections for solder cracks and for breaks in the copper traces
too.
I now use
this cleaning process on all Power Supplies, Driver Boards and CPU’s / MPU’s.
Actually any surface that has heat or acid damage can benefit from this cleaning
procedure.
Now we’ll move on to the repair of the acid damaged Bally
MPU.
PHOTO SHOWS DAMAGE TO
LEFT SIDE
Most of the ground bus, on the left side of the board, is involved. Affected as
well, are the leads and pads on a row of capacitors. This is just surface
corrosion and will not affect the running of the MPU, once it has been bead
blasted clean. But, because of the location of the area involved, it is
necessary to determine whether or not any connectors are hiding potential
shorts. The covers of J1 and J4 will be pried up and the area around the pins
will be blasted cleaned. To accomplish this much cleaning with ordinary tools
would take hours and still not result in the pristine surface needed for
accurate soldering.
Photo below
shows the most common areas of damage on a Bally MPU. The U8 socket and
neighboring parts are usually compromised. Some of the resistors along the
bottom of the MPU are directly related to the signals to the U9 (MC6800)
Processor. If corroded, this group of resistors can shut down the reset and
interrupt signals needed to get the MPU running. Most will have to be replaced,
even after they were blasted clean. Resistors are sensitive to corrosion and can
display inaccurate values. It’s just common sense to replace them.
This close-up
of the U8 socket shows that the corrosion has damaged most of the pins in the
bottom row. Before the board is bead blasted, it is impossible to determine
how much damage there is on the copper traces and pads surrounding the U8
socket. The eyelet shown, just above the bottom ground rail, has been completely
consumed by the electrolyte. Before any blasting is begun, all of the pins and
remaining solder will be removed from the U8 socket.
The next
photo is the first in a series that show the bead blasted MPU and what the
cleaning has
revealed.
The area
around the U8 socket is now corrosion free and the traces below the socket have
been cleaned. But these traces show signs of being eroded. There are several
gaps where the copper has been eaten through. The bead blaster only revealed
this condition, it didn’t cause it. Although it is possible to remove the copper
surface, (if the nozzle were aimed at one spot for a long enough time), when the
copper plating is healthy and adhering to the surface, it won’t be adversely
affected by normal bead blasting.
The photo
below shows the same area around U8 socket after some additional steps have been
performed. The first step is the light sanding of all exposed copper surfaces,
in preparation for soldering. Next, the missing sections of copper traces are
filled in, using small gage wire for filler or solder braid for the replacing
the wider traces.
An opened
frame machined pin socket has been installed and soldered to both the top and
bottom copper pads. This will insure that any signal going into the pad on the
bottom and exiting out a pad on the top will not be interrupted. If a cheaper
closed frame socket were used, the pads and traces would be hidden and difficult
to ohm out.
Both the
vertical trace and the horizontal ground bus were replaced with solder braid.
These traces provide the ground path for the U8 socket. The capacitor for the
5-volt logic has one leg tied to this rebuilt ground circuit. Most devices
located in this area have ground traces going down to this bottom rail. After
bead blasting, many of these small ground traces are missing and will stop the
MPU from running, if not replaced.
Finally, the finished MPU, with all of the cleaning, replacing and updating
completed.
Tom Callahan 21 Heather
Dr. Plymouth, Ma 02360