3a. Updating the Game Code.
There are 3 ways to update the Pinball2000 game software.
- Change the complete PRISM card to a one with newer Software on it.
- Update the game via Laptop, serial cable and Pinball2000 upgrade manager.
- Program a PUB (prism update) Card in your PC and plug it into the Pinball2000 game.
Updating a Pinball 2000 Game using a Computer.
A new system has been designed to update
the game and sound software in your Pinball 2000 System. Software updates
are now stored on the PRISM card in your CPU box, no erasing and programming
of EPROM chips is necessary.
The following is required to update your Pinball 2000 machine:
- A PC running Windows 95, Windows 98, Windows XP with
approximately 10 megabytes of disk space.
- A serial "null-modem" cable (with at least one DB-9 male end) to connect
the PC to the pinball machine. Unfortunately the newer style USB and USB2 ports
will not work with the Update Manager program.
A "Null Modem" connector defines a connector
that crosses the wires between pins 2 and 3 (that is, pin 2 of the first connector
is connected to pin 3 of the other connector).
- Update software, available from Williams (Pinball.com).
- Note this process ONLY works if your PC has a conventional
(old school) DB-9 or DB-25 style COM1 serial port. Unfortunately
the new style high speed USB will not work with the Pinball 2000
Update Manager (or with the game, since the game mother board does
not have a USB port).
Step 1: Get the Update Manager.
The Update Manager handles the transfer
of the game program from your PC to the Pinball 2000 system. You only need
one copy of the Update Manager for all Pinball 2000 machines and all game code versions. If you do
not have the Pinball 2000 Update Manager, just download one of the game code
links below (they all have the Update Manger code inside the corresponding ZIP file).
Step 2: Get the Game Code files.
Again, see below and download the appropriate game code ZIP file to your local PC.
Save the file into a temporary area on your
PC. Do not change the filename when saving these files to your PC!
Also you do *not* need the "Pub" version of this software (this is for
use with a "Pub Card", which we are not using here). Not that it hurts to have
it, but it's just for the Pub Card.
Step 3: Run the Update Manager.
From the Windows Explorer, locate the Pin2000_UpMgr_x.exe
file on your PC and double-click to start the Setup program. Follow the
instructions and prompts to install the program on your PC.
Using a laptop running Windows XP, connected to a RFM using a null modem cable.
|
|
Step 4: Connect the PC to the Pinball 2000 Machine/
With your Pinball 2000 system turned off,
connect the serial port "null-modem" cable from your PC to the serial port
inside the coin door. If your cable has thumbscrews, tighten them to make
sure the cable does not become disconnected during the update process.
Once the two machines are connected, power up the Pinball 2000 system.
If the local PC and the P2000 game don't connect, check the serial cable
inside the game. Often it's not attached at the computer side.
Running the Pinball 2000 Update Manager on the PC. Here the status window shows
the program has connected to the RFM.
|
|
Step 5: Select the COM port in the Pinball 2000 Update Manager on the PC.
If the Pinball 2000 Update Manager is not currently running,
double-click the Pinball 2000 Update Manager on your local computer.
Under the 'Connection Port' box on the
Update Manager window, select the serial port on the PC that you will be
using (COM1 or COM2. On most PCs, the external serial port is COM1).
Once the proper serial
port is selected, confirm that the Pinball 2000 game is powered up. Click on the
'Connect to Game' button on the upper right of the Update Manager window.
The program will connect to the pinball machine and report back the model
number and software revision found (example: 50070 = Revenge From Mars,
0.82 = Software Revision 0.82).
In the middle box on the upper right, select
the drive letter that is holding the software update file that you downloaded
from the website. Click on 'Scan Drive for Updates' to locate all the software
updates related to this model number. This will scan your entire C: hard drive
for the Pinball 2000 software (this is why the filenames of the update software
can *not* be changed!)
Click and highlight the update version
to upload to your Pinball 2000 system, then click the Update button to
begin the upload process.
Step 6: Make sure the game is ready.
Depending on the status of the Pinball 2000 game, you may have to
navigate some menus to get the PC and the P2k game to connect.
For example, if the game had it's "update disabled", you will most
certainly have to do access the following game menus (see pics below):
If updates have been disabled previously, you will need to use the coin door
diagnostic buttons, enter the System Menu, and go to the Utilities menu.
|
|
From the Utilities menu, select "Update Game Code".
|
|
Game is now ready to connect to the PC and start updating the game code.
|
|
Step 7: Do not disturb!
The update process may take up to 10 minutes
to complete. A bar graph will be shown on the pinball screen and in the
Update Manager window to show approximate progress (green bar moves from left to right),
but do NOT disconnect
once the bar has reached the end. The system needs to reset itself before
the update is considered complete. Once the machine is rebooted and running
in attract mode, the Update Manager will reconnect to verify that the update
is successful.
The game code is being updated by the PC.
|
|
Status window shows game code being updated.
|
|
Step 8: Disconnect Cables
Congratulations, all done with the software update!
Update is finished.
|
|
After the update process is finished, the game should reboot automatically.
|
|
Game Code.
All game code ZIP files below include the update manager program. Older versions of the
code are archived here, but generally speaking, you will always want to use the
newest version available for your game.
3b. Motherboard Repair/Replacement Info.
Pinball 2000 uses a "baby AT" personal computer
mother board with a Cyrix Media GX 200, 233 or 266 mHz processor
(most Pinball 2000 games were shipped with 233 mHz processors,
but some early games had 200 mHz and some later games had 266 mHz processors).
This Cyrix processor and the Cyrix bridge are required for Pinball 2000 and are not
replacable by any other PC mother board or processor!
There were at least three different mother board brands/types used in Pinball 2000 production,
but all shared this common Cyrix Media GX architecture
(for example, the GXM/GCT/7520 by Semi).
At one time you could order a complete motherboard from a Williams distributor,
part #04-12604, for around $300. Now often these motherboards can be found on Ebay
(not in the pinball section!) and through other distributors like
DMS computer
(www.dmspinballs.com/p2k.html).
Other sources for finding these is at local computer shows,
computer recyclers, or online auctions.
A 200, 233 mHz or 266 mHz Cyrix processor used in Pinball 2000 motherboards.
|
|
The Cyrix 5520 companion chip used in the Pinball
2000 motherboard.
|
|
In order for a PC motherboard to be a replacement candidate, it has to have the
Cyrix MediaGX 200, 233 or 266 mHz processor. And it must also have the 5520 companion chip,
and one of the following types of super I/O chips: SMCFDC37C932, SMCFDC37C931, NSPC9731.
Note Williams used several different style of PC motherboards in their Pinball 2000 games.
These vary in color and layout. But they all have the above chips in common.
The Cyrix Media GX board system was choosen by Williams because it was an "all in
one" CPU board, and it was inexpensive. Unfortunately Cyrix went out of business,
and it's assets were purchased by National Semiconductor. Via then turned around
and bought all the Cyrix assets from National, except for the Media GX line (Via
had no interest in this system).
Since National did not have the right to use the "Media GX" name, they renamed the line
"Geode". National still makes the Geode line of chips and boards, which are using the 5530A
chipset (an update to the 5520; Williams was porting Pinball 2000 to the newer
5530 chipset when they closed the pinball division). But unfortunately the new National Geode
boardsets usually lack PCI slots (and
some other required features), and the newer 5530A chipset is not Pinball 2000 compatible.
So the new National Geode boards are not usable for Pinball 2000.
The original boardsets used for development of Pinball 2000 were 150MHz with the
Media GX 5510 chip. Sometimes these old 5510 chip boards with
150 MHz or 200 MHz CPUs can be found on Ebay or at computer "junk" sales.
Though somewhat risky,
all of the code for the 5510 chip is still in the RFM and SWE1 code,
and it autodetects on startup which one to use (the 5510 or 5520 chipset).
However, the older 5510 code
was obviously not as extensively field-tested as the 5520 code,
and RFM at least was running into speed limits even at 233MHz
(SW:EI was not as bad in that regard). So if buying one of these
older 5510 CPU boardsets, be aware it may not work properly.
In theory it should work, but it could "bog" down on some
display routines, or the display could be out of sync with sound.
A companion NSPC9731 Super I/O chip used in some
Pinball 2000 motherboards.
|
|
Motherboard Chips NOT Needed.
The motherboard choosen by Williams was clearly an "all in one"
type board. That is, the motherboard had an on-board video,
sound, and other support circuits/chip. The audio stuff is however
not needed. Williams had a off-board sound card, making the on-board
sound chip(s) unnecessary. Some boards will have these sound chips
installed, and some will not. So an empty socket or two on an otherwise
fully populated motherboard can be OK.
An unneeded audio chip often found on Pinball 2000 motherboards.
|
|
Motherboard Memory.
The memory used in pin2000 motherboards is SDRAM PC100
memory in a DIMM package (commonly known as "168 pin PC100 SDRAM 3.3v DIMM").
Any amount 8 meg or higher can be
used (typically 8 meg is seen in most games). PC133 memory can
probably also be used (the only difference between PC100 and PC133 is
that PC133 can run at a higher bus clock speed).
P2000 motherboard memory.
|
|
3c. Game does not Boot or Re-boots.
Here is a list of things to try when the game will not turn on.
Pinball 2000 power switch behind the backglass on the computer case.
|
|
Power On.
- Switch on the machine.
- If backglass does not light up go to next section (Main Power)
- If backglass lights up but after 30 seconds or so there is no bong,
no monitor pic, and no attract mode flashing continue...
- If bong can be heard and/or attract mode lamp flashing arises,
but no monitor pic, go to monitor diagnostics.
Main Power.
- Check the power line to the game.
- Check the main fuse.
- Measure the 110V which come out of the transformer (all games, regardless of the
wall voltage). Measure the 110V that goes into computer case's power supply.
- If no power there, check main fuse again and wall plug.
- If power there, go to next secton (Check Computer Case).
The power plug from the power supply to the motherboard.
|
|
Check Backbox Computer.
- Verify the red button in front front of the computer case is pressed down (on).
- Check the power cabling on the rear left side of computer case.
- Are the fans (CPU and power supply) running? If not, go to the next section.
- Check the parallel port cable and monitor cable on left side of computer case (as facing the game
from the front).
- With the game off, open computer case and check whether PRISM seats firmly within it's PCI slot
(it can come loose).
- If game still does not boot, check the power supply voltages with a DMM,
where the power supply plugs into the mother board (red=+5 volts, yellow=+12 volts,
black=ground).
The +12 volts should be in the 10 to 14 volt range.
The +5 volts should be 4.9 to 5.2 volts. If either the +5 or +12 volts
is not within these ranges, replace the power supply, as the game
will not reliably boot with voltages that are outside these ranges.
- If everything looks OK but still nothing works, go to next section.
Boot up of a Pinball 2000 computer, outside of the game, Prism card removed, a
keyboard attached, and hooked up to a regular computer monitor. Since there is no
hard drive or Prism card, the computer tries to boot from the missing floppy disk drive.
|
|
After the inital boot message, pressing DEL on the keyboard should bring up this screen.
You can change the game's date and time in the Standard CMOS Setup options, if desired.
|
|
With the Prism card installed, most newer monitors won't "sync" and you'll get this
error message. If you have an older monitor from the 1990s, you may be able to get
some game picture on the monitor (though it will be blurred and look odd).
|
|
Backbox Computer Problem.
It's a lot easier to diagnose problems with the Pin2000 computer if it's
removed from the game and hooked up to an external computer monitor.
- Remove all cables from the computer case's left side (as facing the game from the front).
- Slide the computer case out of the game.
- Open the computer case.
- Remove the PRISM card from its internal slot inside the computer.
- Connect any PC VGA Monitor to the Monitor output on the outside of the computer case
(also connect a PC keyboard if available).
- Connect the PC box power supply directly to a 110 volts line (if in US)
or switch the power suply to 220V
and connect it to 220 volts (don't forget to switch back later to 110 volts).
Use a standard PC power cable for connection to power.
- Switch the computer on via the red switch on the front panel.
- The pin 2000 computer should now boot like a standard PC. Video should be seen
on the computer monitor. Note a BIOS Message will be seen ("No boot device found").
This message is Ok and should be seen (after all, there is no hard drive or floppy
to boot from). If this is OK, then go to the next section.
If not, go to the "power supply or mother board error" section).
Note if a CMOS (BIOS) checksum error appears on the screen, the
motherboard battery can be removed
(this will reset stored BIOS settings/memory).
A new CR2032 battery should be installed.
Pinball 2000 Specific Boot Problems.
- Switch off the computer case power switch.
- Re-install the PRISM card.
- Switch the computer on again.
- Pin2000 should boot up now. Although the image on the monitor may look ugly,
it should display something (the monitor will not "sync", and never monitors
may display a "does not sync" error message). If nothing happens,
PRISM may be faulty. Try to reseat the PRISM card in another PCI slot.
Power Supply or Mother Board Error.
- Check whether Fans are running (power supply and/or CPU fan).
- If not, power supply seems to be bad. Replace with another AT style computer power supply.
- If power supply OK, probably one of the mother board components is bad.
Check the CPU processor (MEDIA GX 233 or 266 mHz), RAM (minimum 8 Meg, 168 pin PC100 SDRAM 3.3v DIMM or better),
mother board (Media GX board), or power supply.
- Measure the -5, +5, +12 volt output from the power supply with mother board attached.
If erronious it still might be the mainboard, but don't measure power supply output without mainboard
attached (the lack "load" may kill the power supply).
- Try to exchange the easiest available thing (RAM, power supply), and reboot.
- Replace the mother board with a working spare board.
When Entering Diagnostics/Bookkeeping, the Game Re-boots.
The game boots and plays fine, until the ENTER button is pressed
on the coin door. This causes the machine to reset and reboot.
To diagnose this problem, a laptop computer running a terminal emulator
(Windows Hyperterminal, using settings "9600 8N1")
was connected to the serial terminal port of the pinball 2000. With the
computer on, the pinball 2000 game is booted. When the game's Enter button
is pressed (to enter diagnostics), a Fatal error was being generated and
the game reboots. But the fatal error can be seen on the laptop computer's
terminal emulator program, "*** Fatal: Got invalid time stamp!"
This can be further diagnosed
with the pinball 2000 game on and the laptop connected and running
a terminal emulation program. The XINA command "enter" can
be issued on the laptop to enter the pinball 2000 diagnostics.
This Xina command may cause the game to reboot (just as if the "enter" button was
pressed on the coin door). Again, during the transaction, the
laptop terminal received the error message.
The error contained this key information:
% BAD: 4, 12, 29, 128 (22)(00)(45)
*** Fatal: Got invalid time stamp!
The numbers after the "BAD" are the time stamp's: weekday, month, day, year, hour, minutes,
seconds. In this case, the year seems to be out of the acceptable
range of 1999 to 9999. Since pushing the ENTER button causes a time stamp to
be generated and checked, the checking code (for reasons unknown) does not
just report and fix the problem, but instead generates a Fatal error,
which then resets the machine.
When "time" command was entered on the laptop, the pinball 2000 echoed back,
"Sorry, clock is bad".
Fortunately, there is a fairly simple cure to this problem. With a laptop computer
running a terminal emulator (Windows Hyperterminal)
attached to the serial port (9600, 8N1), the
following command can be entered to set the date and time, after the pinball 2000
game is booted:
time set <year> <month> <day> <hour> <minutes> <seconds>
(for example):
time set 2002 12 31 08 53 00
(which sets the game to Dec 31, 2002 8:53am)
The battery on the motherboard (CR2032) and the PRISM card (BR2325 Li-battery, or
in a pinch a CR2025 can work)
should probably both be checked and probably replaced prior
to performing this fix, since the CPU battery
is what keeps the time when the game is off. The Prism battery keeps the high scores
and the game audits (and some audits are also written to the Flash memory, in case
of a complete battery shutdown).
Swapping from SWE1 to RFM, and Game No Longer Boots.
If you have one Prism card, and are switching SWE1 and RFM rom boards
to this Prism card, you may have a problem when going from SWE1 to RFM.
What happens is there is a code mis-match between the flashed code
in the Prism card, and the ROM board code. On SWE1, this code is smart
enough to see the two different versions (RFM on the Prism, SWE1 on the
ROM board), and disable the update (this can be seen on the white screen,
where it say "code mis-match, updates disabled"). This allows the Prism card to
boot with the factory ROM .4 version of the software. So taking a Prism card from
RFM to SWE1 is usually not a problem.
Unfortunately, RFM code is not as smart. If the RFM rom board is installed
on a Prism card that was previously installed with a SWE1 rom board, sometimes
the boot process does not disable the update. The game will boot
(white screen, blue messages), and show no errors. The playfield lamps
will come up for a couple of seconds, and then go out, and the monitor
screen will be black. The game will essentially hang in this mode,
where the the coin door diagnostic buttons do nothing.
The easiest way to fix this is to update the game with a PUB board.
But since most people don't have a PUB board, there is another way.
Hook up a keyboard to the computer, and boot the game. Again, the playfield
lights will go on for a few seconds and turn off, and the monitor screen
will go black. At this point, type the following command on the keyboard -
note this command will NOT be seen on the screen, so don't make any typing
errors!
After pressing ENTER, turn the game off, remove the keyboard, and turn the game back on.
The game should now boot with version .4 of the RFM code.
At this point your can run the update manager with a
null modem cable, and update the firmware to RFM version 1.5.
P2K Won't Boot from Bad Motherboard Capacitors.
If your Pin2k computer box will not boot, this may be due to the 1000 mfd
electrolytic capacitors on the motherboard. The symptom is simple - turn the game
on, and the monitor stays black, with no boot-up messages. This may not happen
all at once. I have seen situations where a P2000 machine would not boot at
first, but a quick flick of the power switch got the game to boot. Eventually
though this would not work, and the machine would never boot.
There are five 1000 mfd caps surrounding the Cyrix processor,
which gets quite hot. In addition, if the CPU fan has failed, this only makes
the heatsink on the Cyrix chip hotter. This heat bakes these five 1000 mfd
capacitors, making them prone to failure.
Also Pinball 2000 machines use computer motherboards made around the time there
was corporate espionage in the Asian capacitor market. An
incomplete/incorrect electrolytic formulation was stolen and used by some Asian capacitor
makers without knowing the formula wasn't good. I'm not sure this was a
huge problem with these motherboards, but it may have contributed.
There are at least two kinds of caps in the Pinball 2000 motherboards in the
five spots numbered in the picture below. One has green caps (bad), and the other
light blue (better). The green caps are radial 6.3v, 1000uf, 105c and made by TAYEH.
These are the ones that you will usually find as bulged or leaking.
If you see light blue 10v 1000uf, 105c caps from XICON, these seem less prone
to problems. But even if the caps are not bulged or leaking, they can still
cause the motherboard to not boot.
I recently fixed a motherboard that would not boot by replacing these
five 1000 mfd capacitor. In my case, the caps were the Xicon brand, and
showed no sign of failure. I put them on my cap tester meter, and they
read about 850 mfd. That doesn't seem to be "that bad", but after replacing
these five caps, the board booted up fine.
The P2k CPU (mother) board. The location of suspect caps is shown.
Picture by Victor Ireland.
|
|
Another look at a suspect P2k CPU (mother) board, with the five 1000 mfd caps removed.
|
|
Look at the five locations in the picture. They
are all the same capacitor. I recommend replacing all five capacitors.
Mouser carries suitable replacement caps. Check for part number
647-UHE0J102MPD. Or, if you want to use a 10v radial cap (647-UPW1A102MPD).
Just make sure they're 1000uf, radial, 105 degrees, and the same form factor (size).
Caps on the P2k motherboard.
|
|
When you're removing the existing caps and replacing them with the newer
ones, be VERY CAREFUL. There are a ton of fine traces near the pads for
these caps and one "oops" will make this basic repair much more painful or
even impossible. If the solder isn't coming out, do not overheat the pads.
Note that the caps are marked with a "-" and a stripe on the negative side.
Just make sure you put the positive back on the positive side and the
negative on the negative (the board is marked with a "+" or a dot where the positive
lead goes, and usually has a big white band for the negative side).
Close up of the main processor and the five 1000 mfd caps removed.
|
|
Getting the old caps out is not too tough. Just heat the solder pads from
the solder side of the mother board, and gently pull the old caps out.
Use a low-wattage soldering station to do this work, and go slowly.
Getting the old solder out of these holes is tricky. Use a GOOD quality
desoldering station. These motherboards are usually a 4-layer board.
This makes desoldering very difficult. And the fine traces going around
the caps on the board are very easy to damage, and very difficult to repair.
So go slowly and use a good quality desoldering station.
Once all five caps are replaced, put the machine back together and power it up.
Chances are you've just rescued the motherboard in
your pin2k machine for about an hour labor and $1.00 in parts.
3d. Video Monitor Info, Fixes & Replacements.
Pinball 2000 uses a 19" CGA color video monitor from
Ducksan (CGM-1901CW) or
Wells Gardner (19K7302), model number WGZ1973-U3GS35J.
This monitor is *not* compatible with Personal Computer monitors
(though I have heard of some early VGA monitors that may work with p2000).
P2000 needs a "negative sync" VGA style monitor (the negative sync
part is unusual for the Personal Computer world).
In addition the vertical deflection is inverted on the tube or chassis (for the mirroring effect).
Another new Wells Gardner monitor that works is model number WGE1972-HOGS39L.
Some cabling changes will need to be made with this monitor too.
However the p2000 CGA (lo-res 640x240) RGB monitor used is a standard video game
monitor. The Pinball 2000 monitor is lower resolution than standard computer VGA (640x480) monitor or
even med-res video game monitor. Other than that, any standard
RGB low-res video game monitor, with inputs for positive horizontal and vertical sync,
can be used. Also there is no need for a separate isolation transformer for
the monitor, since the monitor power already comes from the P2000 transformer in
the bottom cabinet. The mounting brackets for the Pinball
2000 monitor are custom though (but can be transfered to another monitor).
Exact replacement Ducksan and WG 19k7302 monitors are no longer
available. But you can use a Well Gardner 7200 monitor. The only
change needed is in the cabling - change the monitor cable pins around on the
header to use positive sync instead of negative sync. This is described below.
On the original WG7300 the sync pins used did not matter,
rather the negative vs. positive sync was controlled by a jumper of the
monitor chassis pcb. More info on this below.
WG7300 color neck board monitor adjustments (board removed from monitor neck).
|
|
Monitor Color Adjustments.
Pinball 2000 color monitor adjustments are on the neck board.
There are six adjustments, two for each color (RGB).
Monitor controls behind the backglass, to the bottom right of the monitor
(as playing the game). Yes, this board is mounted upside down!
|
|
Brightness Adjust.
Remove the backglass and you'll see six small knobs (pots)
mounted to the bottom right of the monitor. Four of the knobs
adjust the picture position, horizontally and vertically.
The two knobs on the right are the brightness and contrast.
Also on the back side of the monitor on the flyback
transformer there are two knobs. The knob closest to the circuit
board controls the brightness (the other knob controls the focus).
Be careful - only turn knobs with a HV insolated screwdrawer.
Turn the brightness knob (on the little separate adjust board) to a lower value.
Then carefully turn the SG knob for more brightness - if you turn it too much,
then you'll get flyback lines on the screen.
When you have reached a better brightness you should do a brightness adjustment
with the brightness button.
Monitor adjustments (brightness, focus) on the flyback
(as playing the game).
|
|
Monitor Cabling.
Unfortunately, the cabling (the cable from the monitor to the computer)
used for the Ducksan and Wells Gardner 7300 are different.
Different connectors are used because the cabling goes directly to the monitor chassis
header, rather through a standardized connector. So plugging a Wells Gardner 7300
into the Ducksan monitor cable will not work.
| 15 Pin PC Connector Pinout |
|---|
| Pin # | Description |
|---|
| 1 | Red Video |
| 2 | Green Video |
| 3 | Blue Video |
| 4 | Sense 2 (Monitor ID bit 2) |
| 5 | Self Test (TTL Ground) |
| 6 | Red Ground |
| 7 | Green Ground |
| 8 | Blue Ground |
| 9 | Key - reserved, no pin |
| 10 | Logic Ground (Sync Ground) |
| 11 | Sense 0 (Monitor ID bit 0) |
| 12 | Sense 1 (Monitor ID bit 1) |
| 13 | Horizontal Sync (HS) |
| 14 | Vertical Sync (VS) |
| 15 | Sense 3 - often not used |
| Ducksan Monitor Cable Wiring |
|---|
15 pin PC
Connector |
Signal |
Monitor Board
Connector |
| 1 |
R |
pins 1/2 |
| 2 |
G |
pins 3/4 |
| 3 |
B |
pins 5/6 |
| 5 |
GND |
pin 10 |
| 10 |
GND-1 |
pin 9 |
| 13 |
H/V |
pin 7 |
| 14 |
V |
pin 8 |
Note: GND-1 is not labeled on the monitor board,
but this pin is located between the GND and V pins.
Ducksan 1900 monitor cabling.
|
|
| Wells 7300 Monitor Cable Wiring |
|---|
15 pin PC
Connector |
Signal |
Monitor Board
Connector
Negative Sync |
Monitor Board
Connector
Positive Sync** |
| 1 |
R |
pin 1 |
pin 1 |
| 2 |
G |
pin 2 |
pin 2 |
| 3 |
B |
pin 3 |
pin 3 |
| 6 |
GND-1 |
pin 4 |
pin 4 |
| 6,10,11 |
Key* |
pin 7 |
pin 7 |
| 10,11 |
GND-2 |
pin 8 |
pin 8 |
| 13 |
HS |
pin 9 |
pin 6 |
| 14 |
VS |
pin 10 |
pin 5 |
* Yes there are wires connected to the .156" monitor
board header "key" position. I have no idea why, and
they don't need to be connected.
** Positive sync can be used instead of negative
sync if the monitor will not adjust properly
horizonitally. There is a switch on the monitor that
can swap between negative and positive sync.
Wells Gardner 7300 monitor cabling.
|
|
Converting a Ducksan Monitor Wiring to Wells WG7300.
The Ducksan monitor is the worst of the two different monitors
that were sold with Pinball 2000. Because of this, there may
be some need to convert a Ducksan monitor cable so it can be
used with a Wells Gardner 7300 monitor. Below is that conversion.
| Converting Ducksan to Wells 7300 Monitor Cable Wiring |
|---|
Ducksan
Monitor
Connector |
Wire Color |
Signal |
WG7300 Monitor
Connector
Negative Sync |
WG7300 Monitor
Connector
Positive Sync* |
| 1 |
Red/Wht |
R |
pin 1 |
pin 1 |
| 2 |
Red/Blk |
Gnd |
pin 4 |
pin 4 |
| 3 |
Grn/Wht |
G |
pin 2 |
pin 2 |
| 5 |
Blue/Wht |
B |
pin 3 |
pin 3 |
| 7 |
Wht |
HS |
pin 9 |
pin 6 |
| 8 |
Yel |
VS |
pin 10 |
pin 5 |
| 9 |
Blk |
GND |
pin 8 |
pin 8 |
Connector pins 4,6,10 are not used.
* Positive sync can be used instead of negative sync if the monitor
will not adjust properly horizonitally. There is a switch on the
monitor that can swap between negative and positive sync.
Wells Gardner 7200 Monitor.
To use a Wells Gardner 7200 monitor (currently available from Wells
19 inch K7200A Horizontal CGA Monitor part #WGE1972-HOGS39L),
the cable from the Wells Gardner 7300 (or from the Ducksan)
will *not* work. If the 7300 cable is used,
it will cause the 7200 monitor to scroll horizontally. No amount of
adjusting will correct this either.
The solution is that the 7200 has different pins for negative versus
positive sync. On the 7300 there is a switch setting on the chassis
board that is set to either negative (the most common, and how it is
set for Pinball 2000), or positive (there for backward compatability for
older video games). Because of this, the 7300's H/S-2 wire need to be
moved to +HS on the 7200 monitor. Likewise, the 7300's V/S-2 wire must
be moved to the 7200's +VS. Also the 7200 should be set to negative sync.
The corrected cabling for the 7200 Wells
Gardner monitor (set to negative sync) is shown below.
| Wells 7200 Monitor Cable Wiring |
|---|
15 pin PC
Connector |
Signal |
Monitor Board
Connector |
| 1 |
R |
pin 1 |
| 2 |
G |
pin 2 |
| 3 |
B |
pin 3 |
| 6 |
GND-1 |
pin 4 |
| 13 |
+HS |
pin 6 |
| 14 |
+VS |
pin 5 |
| 10,11 |
GND-2 |
pin 8 |
Wells Gardner 7200 monitor cabling.
|
|
Terry's mounting instructions:
In addition, when mounting the new 7200 monitor,
you'll to re-use the original P2000 monitor mounting bracket.
The picture tube itself transfers fine, however the new K7200 board
physically will not re-mount where the old Ducksan board was.
Discard the Ducksan metal pan that held the old monitor board and mount
the new K7200 monitor board using 4 large zip ties to the monitor mounting
bracket (it sounds hokey but the mount is strong enough and clean). Also
when mounting the new monitor board, cut a piece of cardboard (like the kind on
the back of a large notebook pad) the same size as the new monitor board and
attached it to the back of the new board. This provides an insulator to the back
of the new monitor board. Additionally since the new monitor
board is not grounded to the monitor chassis (because of the zip ties arrangement),
add a wire from the monitor board ground to the monitor mounting bracket.
If you don't want to go the zip tie route, you'll have to
fabricate a mounting bracket(s) for the new monitor board.
Wells Gardner 4600 Monitor.
The WG4600 was a very popular video game monitor from the early 1980s,
and was standard equipment in many Williams video games such as Stargate
and Robotron. It is a good example of how an early video game monitor with
inputs for positive and horizontal sync can work in p2000. This info thanks
to Russel Willoughby.
| Wells 4600 Monitor Cable Wiring |
|---|
15 pin PC
Connector |
Signal |
Monitor Board
Connector |
| 1 |
R |
pin 1 |
| 2 |
G |
pin 2 |
| 3 |
B |
pin 3 |
| 6 |
GND |
pin 4 |
| 13 |
HS |
pin 6 |
| 14 |
VS |
pin 5 |
Wells Gardner 4600 monitor running P2000 SWE1.
|
|
Upside Down or Reversed (Left/Right) Video.
When installing a new video monitor, sometimes the video can appear
upside down or reversed (left to right), while playing the game.
Probably the easiest way to fix this is to issue a fb flip
command to the Xina operating system. This will require a
keyboard to be plugged into the Pinball 2000 computer (I am not sure if
this command will need to be issued everytime the game is rebooted).
The fb command will only fix an upside down picture (it does not change
left to right picture reversals).
A better (more permanent) way to fix the problem involves reversing
the yoke header wires (on some monitors there is
a switch for this). Check for a switch near the yoke, or on the
edge of the neck tube circuit board.
If there is no switch, find the yoke wires. The Yoke is the coil of
wires on the neck of the picture tube. There should be four wires coming off of
the yoke. Swap the blue/red wires if the picture is reversed left to right.
Swap the green/yellow wires if the picture is upside down.
Be careful to get the right two voke wires though,
as the monitor chassis can burn up if this is done wrong!
Degaussing.
The monitor in a P2K degausses automatically everytime it does a "cold
start". You can hear this as a short buzzing sound (only a fraction of a
second) at the moment the machine is turned on.
If you are repairing the monitor on a workbench and power on, and then
move it to the game and power on, it may need to be 'degaussed' manually
(or just wait a few hours and power on, so it does a cold degauss its self).
Sometimes when moving the game, a light colored "spot" (varying in size, usually
about 1" to 3" diameter) on the monitor may appear.
If this is the case, there is probably an external
magnetic interference from speakers or nearby lighting transformers (anything
with a magnet or electro-magnet). Check for anything like this near the machine.
If there is no hum at game start up, there is usually
a varistor that burns out. If the monitor gets a reasonable
magnetic field, it may take several power cycles to
clean up (look out for vacuum cleaners and speakers,
these often have strong magnetic fields). Check for
what is causing the magnetic field before just
powering on and off to degauss. Five or so power cycles
should be all that is needed. If the purity doesn't
come back, refer to a shop manual
for the monitor that describes purity adjustments for
the monitor. Even the earth's magnetic field is strong
enough to effect purity of many color monitors.
If the spot remains the degauss mechanism is malfuntioning.
Often the monitor can be degaussed by taking a powered-on soldering gun,
and moving it in front of the monitor glass. The magnetic field from
the soldering gun can often degauss the monitor.
Ducksan Monitor Problems and Solutions.
Much of this information is from Roy Fash and Ken Layton.
The original pinball 2000 Ducksan monitors seems to be the
more problematic. For example, symptoms could be
no picture at all, tube was dark, stopped working suddenly
in the middle of a game. Or smaller problems like the pictures
is fuzzy or changes sizes as the game warms up.
A web site is available for tech support with Ducksan monitors.
The company is Merit Industries and
the site is http://www.meritind.com.
Click on Support, then on FAQs, then
on Technical Information, then on Monitor/Picture. Also check out
http://www.monitec.co.kr/english/product-1301-chassis.htm.
Look for information
related to Ducksan problems. It seems that a number of the Ducksan monitors
must have used similar chassis boards. Merit considers a "Type 59" monitor
to be the Ducksan CGM1901 monitor found in RFM. Also there is a schematic
of the Duckscan monitor available
here.
Installing a "cap kit" (a set of new electrolytic capacitors that
commonly fail in a video monitor) is often a good idea for one
of these monitors. Merit Industries has a
parts list for a cap kit for the monitor (frequently failed parts). The
list contains proper replacement values for the caps (sometimes different
than the original value) as well as the Mouser part numbers to make it easy
to order the parts. C804 appears to be one of the most important caps to
replace. There is also a monitor suppliment page available
here.
Roy reports that in addition to replacing the caps he had to replace a
resistor that metered open (R348), and also replaced the vertical output
chip U201 (mounted to a large aluminum heat sink on the
monitor chasis board). Chip U201
is labeled DBL2054D on the chip and schematic, but is really a
TDA-1675A (Mouser part# 511-TDA1675A). This chip shorts a lot and kills
the B+ supply, which in turn might cause R348 (56 ohm 1 watt) to
go open. If resistor R348 is open, replace both U201 and R348.
Touched up many bad solder joints on the main monitor board,
including those at B+ that were particularly bad. Be careful when soldering
on the foil side of the chassis board as the traces lift easily on this
board.
Ducksan also appears to have used at least four different
neck boards with their monitor: DS2905, DS2906, DS2907, and DS1301. The
part numbers in the schematic below match the DS1301 neck board, but they do not
match the DS2905N neck board (don't know about the DS2906 or DS2907). So
the schematic below may not be correct for the neck board even
though it is correct for the monitor chassis board. Accordingly the
portion of the parts listed in the cap kit that relate to the neck board may
not be correct for some Ducksan monitors, even though it is correct for the
neck board shown in the schematic. It is pretty easy to tell when looking
at the neck board and schematic if a particular Ducksan monitor has the neck
board shown in the schematic. Also the neck boards have a part number
printed on them. Luckily most repairs don't relate to the neck board anyway.
One other thing to mention is that you
must connect the monitor to the wiring harness from the RFM body in order to
power up and test the monitor. This is because the power transformer is in
the body of the game, not in the head.
Cap Kits for Ducksan and WG7000 Monitors.
Bad Capacitors for the DUCKSAN model CGM-1301 (Merit type 55) & CGM-1901 (Merit type 59)
VGA monitor repair. The test in parends is the problem often seen with the associated
capacitor.
- Ducksan Main board:
- C319 1 mfd @ 250 volts (off value)
- C334 10 mfd @ 160 volts (off value)
- C804 2.2 mfd @ 50 volts (open) Change to 10 mfd. @ 50 volts
- C337 470 mfd @ 16 volts (off value)
- C336 470 mfd @ 16 volts (off value)
- C806 100 mfd @ 160 volts (off value)
- C809 1000 mfd @ 35 volts (off value)
- C208 47 mfd @ 25 volts (off value)
- C210 1 mfd @ 50 volts (open)
- C209 220 mfd @ 35 volts (off value)
- C306 1 mfd @ 50 volts (off value)
- C203 1 mfd @ 50 volts (off value)
- Ducksan Neck board:
- C405 1 mfd @ 50 volts BI-POLAR (open) Upgrade to 100 volts
- C445 1 mfd @ 50 volts BI-POLAR (open) Upgrade to 100 volts
- C425 1 mfd @ 50 volts BI-POLAR (open) Upgrade to 100 volts
- C406 1 mfd @ 160 volts (open)
- C446 1 mfd @ 160 volts (open)
- C426 1 mfd @ 160 volts (open)
Here's a list of Mouser (www.mouser.com) part numbers for the capacitors (about $3.00 total):
- (1) 140-XRL250V1.0 - 1 mfd @ 250 volts
- (1) 140-XRL160V10 - 10 mfd @ 160 volts
- (1) 140-XRL50V10 - 10 mfd @ 50 volts
- (2) 140-XRL16V470 - 470 mfd @ 16 volts
- (1) 140-XRL160V100 - 100 mfd @ 160 volts
- (1) 140-XRL35V1000 - 1000 mfd @ 35 volts
- (1) 140-XRL25V47 - 47 mfd @ 25 volts
- (3) 140-XRL50V1.0 - 1 mfd @ 50 volts
- (1) 140-XRL35V220 - 220 mfd @ 35 volts
- (3) 140-XRL160V1.0 - 1 mfd @ 160 volts
- (3) 647-UVP2A010MDA - 1 mfd @ 100 volts BI-POLAR
Wells Gardner K7000 series monitors cap rebuild kit should include the following
15 pieces. You may use a higher voltage but never less voltage.
- (3) 1 mfd @ 50 volts
- (2) 10 mfd @ 63 volts
- (2) 22 mfd @ 25 volts
- (1) 33 mfd @ 16 volts
- (1) 47 mfd @ 25 volts
- (1) 100 mfd @ 35 volts
- (1) 1000 mfd @ 16 volts
- (1) 1000 mfd @ 25 volts
- (1) 2200 mfd @ 35 volts
- (1) 22 mfd @ 160 volts
- (1) 47 mfd @ 200 volts
Increasing Monitor Constrast (Getting a Better Picture.)
(Thanks to Mickey Johnson for this info.)
For the best combining effect in Pinball 2000 games,
you need the monitor to have a very brilliant
monitor display, but not overbright (in that it has poor black level). What it
amounts to, you need the graphics extremely high contrast.
On many games, the original Ducksan or Wells Gardner monitor had decent color,
good focus, no screen burn, but did not have high contrast. To get the
colors to look bright enough, the constrast needs to be turned up
on the monitor. Unfortunately the downside is the blacks would look grey.
The problem is, this really kills the effect on a RFM, as the black
needs to be black, and not bleed through to gray.
When the graphics appear, you want them to kill what
is behind the image. You want the virtual targets to be vibrant and clear
so you can see all the detail. And the inverse is true also, when the
target disappears, you want the area to be completely black so that you
don't see the monitor reflection. You want to see the playfield.
The Monitor Amplifier which can be used in Pinball 2000 games.
|
|
Now you can do a cap kit, replace the monitor, or have the tube rejuved,
but this most likely will not fix the problem. The problem actually is in
the computer. I learned this from years of doing Mame videogame projects
where I would be interfacing computers to arcade monitors. The problem
stems from the fact that the video output of the computer is too low for the
arcade monitor (the two devices were actually never meant to be used
together). It would be like having way too low of a pre-amp audio output
and when amplified you end up with a poor quality sound. There is an easy
fix to this problem, using a
Monitor Amplifier
(scroll down this page a bit).
This is a standalone video amplifier converts the 1 Volt output
from a VGA card into 4 Volts required by most arcade monitors.
The amplifer contains four channels, Red,Green,Blue and Sync.
The Sync is a buffer rather than an analog amplifier, and this
is to avoid the "fold-over" at the top of the picture on some
monitors that are sensisive to DC input levels.
To implement the monitor amplifier, the original cable will need to be
modified using a short jumper that goes from the amplifier to your monitor input.
This will have a drastic improvement on the contrast. I went from a washed
out picture to a vibrant one. I could see detail that I never saw before.
Also the explosion graphics really get punched up by this mod (since they
are mostly white/yellow). I was also able to lower the brightness and
contrast (that were previously maxxed) on the monitor controls and had a
wide range to adjust from.
Note there are two different models of the amplifer.
The new version is powered by VGA output, and the old version requires
connecting to the amplifier's 5 volts to the power supply drive connector of
the pin2k computer. The problem is that not all VGA card outputs have
that 5 volt line. My RFM did, and my SWEP1 did not. Now, if you have
the new style amplifier and don't have the 5 volts on the VGA output,
you will not get a picture (because the amplifier is not being powered).
There is a easy fix to this, you have to solder a wire to the new amplifier
pcb (its documented at the amplifiers website), and run it to the 5 volt
power supply drive connector (which is what I did for my SWEP1). Note
this entire issue is documented at
http://www.ultimarc.com
Now as far as the results of adding the amplifier to the SWEP1.
Wow, what a difference. I thought the picture was good before, but
now I see so many small details in the picture that I never saw before.
My machine now has a very good black background (I have always been picky
about black levels on TVs), and the graphics just pop up bright and vibrant.
Even the stars in the background are bright as can be, lit against the deep
black background.
3e. When things don't work: Checking Transistors and Coils (stuck on/not working coils/flashlamps)
A transistor is a small electronic device that can cause changes in
a large electrical device (a solenoid), by small changes in a small
input signal (the driver board logic circuit). That is, a weak input
signal can be amplified (made stronger) by a transistor. Essentially, this allows
a small +5 volt logic signal to control a large 50 volt solenoid coil.
Occasionally driver transistors fail.
If a coil is "stuck on" (energized) when the game is turned on, a shorted driver
transistor is usually thecause. This section will help diagnose this.
When the game's top glass is removed, the Pinball 2000 backglass can be used
as a temporary reflective mirror to read the monitor during diagnostics. The
backglass conveniently fits under the monitor, and the backbox head tabs stop
the backglass from sliding forward.
|
|
What do the Driver Transistors Do?
Basically, a driver transistors completes each coil's path to
ground. There is power at each coil, all the time. The driving
transitor is "turned on" by the game's software, through a TTL (Transistor to
Transistor Logic) chip. When the transistor is turned on, this completes
the coil's power path to ground, energizing the coil.
Driver transistors also work the CPU controlled lamps and flash lamps,
causing a lamp to "lock on".
Sometimes these driver transistors short "on" internally. This completes
a coil or flash lamp's power path to ground permanently, making it "stuck on",
as soon as the game is turned on.
Coil Diodes are Back!
On the earlier WPC games, Williams moved all the coil diodes off
the coils and mounted them on the driver board. This was done because
coil diodes often broken from coil vibration. Or when operators installed
a new coil, it prevented them from hooking up the power wire to the wrong
coil (and diode) lug.
But with Pinball 2000, the coil diodes are back on the coils! This was
done to make the driver board more flexible. With the coil diodes off
the driver board, the driver board transistors (FETs) can drive any
device (it doesn't have to be a coil). But the down side of this is
the coils now *must* have a coil diode! Also the power wire
(the thick wire going to the coil) must be installed to the coil
lug with the *banded* side of the 1n4004 coil diode. If the coil wires are reversed,
a fuse will blow and possibly damage the driver board FET.
Transistors Used in Pin 2000.
There are basically three types of transistors used on a pin2000 driver board:
- FET STP20N10L (N-Channel logic level, NTE2987): used as the driver transistor for all
solenoids and flash lamps.
- TIP102 (NPN, NTE2343): used for the lamp matrix rows, to
switch ground on for any particular lamp row.
- TIP107 (PNP, NTE2344): used to drive the CPU controlled lamp (columns) on the playfield.
The TIP107 switches the +18 volts on for any particular lamp column.
FET Transistors on the P2000 driver board.
|
|
FET Transistors (New for Pin 2000).
One area that really has been really changed for
the better in Pinball 2000 is in the solenoid driver circuits. Williams has
switched from using conventional "bipolar" transistors to MOSFETs in the
solenoid and flash lamp driver circuits. A MOSFET (Metal Oxide Semiconductor,
Field Effect Transistor) looks exactly the same as a regular coil driver
transistor. It comes in the same TO-220 package as the TIP102 that is familiar
to just about everyone that works on pinball machines. But this is where the similarity ends.
In fact, even the names of the three component leads are different. In
the transistor, the three component leads are called the emitter, collector
and base. In the FET they are the source, drain and gate.
Although the component lead names are different, their
functions are much the same. For example, the major current flow through
a transistor is between the emitter and the collector. It's called the
"collector current". The emitter/collector connection is used to control
the current through the solenoids. In the MOSFET, the same job is handled
by the source and the drain. To energize a coil, the FET closes the source/drain
connection, completing the circuit. The main current flow in a FET is called
the "drain current.
The remaining component lead is the controlling
element of the device. In the transistor, the "base" lead controls the
flow of current between the emitter and collector. A small voltage on the
base will turn the transistor on. In the MOSFET, the controlling element
is called the "gate". Pinball 2000 uses MOSFETs that are controlled by
standard logic levels (about 5 volts) on the gate. Instead of using a system
of integrated circuit latches, pre-driver transistors and driver transistors,
the MOSFETs are driven directly by data latches (a 74HCT574 chip). This reduces the number
of components on the circuit board, simplifying troubleshooting and reducing cost.
For example, when Williams was using a TIP36 to drive a flipper coil,
it required (in addition to the TIP36), TIP102 and 2N4403 pre-driver transistors,
which were driven by the octal latches (a 74LS374 chip).
The MOSFETs used in the solenoid driver
circuits are type 20N10L (regardless if the FET controls the flippers, an upkicker,
or a flash lamp). The first pair of numbers refers to the current
rating. In this case, it's rated at a maximum drain current of 20 amps.
This is more than enough for any coil with plenty of headroom. The "N" indicates
an N-channel FET (some FETs are of opposite polarity, known
as P-channel FETs). The final 2 digits indicate the voltage rating. The
20N10L is a 100 volt MOSFET. And last (but not least!), the "L" suffix
indicates that a "logic" level is used on the gate to turn the FET on
(this is very important, as some FETs do not have the "L", and hence
will not work in Pinball 2000).
|
|
|
|
Coil 41 on SWEP1: Exception to the Rule.
On Star Wars E1, the Neon light is not driven by a FET like the other coils and
flashlamps in the game. Instead a ULN2803 and a 74HCT574 on the driver board are used as
for the neon lamp transformer driver.
In this repair, it's fairly easy to see if the neon
transformer is the problem or the driver board circuit. Notice the two
wires going to the neon transformer (one red, one black). The red is 12 volts DC,
and the black is ground. Using an alligator test clip, ground the black transformer
lead to the metal side rails on the game (or any other ground point). If the neon
tube glows, the driver board circuit is the problem. I say this because the neon
transformer can fail, so it's best to test this before digging into the driver board.
(You may want to check the neon glass too for any damage.)
The next thing in-line to the neon transformer is a ULN2803
chip at U35 on the driver board. In my repair scenerio, this was not the problem with the SWE1 neon.
Instead it was the 74HCT574 at U34 which had failed. At first I changed the ULN2803 at U35,
since this is first in-line to the neon transformer, but this changed nothing. Next
I looked at the 10k bussed resistor network RP10. Testing this in-circuit showed a problem,
but the true issue was a short in the U34 chip (which was making the resistor network
look bad, but it really wasn't the problem). Changing the U34 with a new 74LS574
fixed the problem, and the neon light worked again.
TIP Transistors used in Pin 2000.
The only TIP style transistors used in Pinball 2000 are for the lamp
matrix. This includes TIP107 and TIP102 transistors. The lamp matrix
circuit is nearly identical to the previous WPC circuit. That is,
TIP107 transistors are used to provide power to the lamp matrix columns.
TIP102 transistors are used to complete the ground for the lamp matrix
rows.
P2000 lamp matrix TIP transistors on the Driver board.
|
|
I have a Stuck-on Coil (or Flashlamp), What should I Replace?
The following procedures will test the driver transistors in
question. If bad, it will need to be replaced.
Inside the front cover of the game manual is a list of each coil used
in the game. Also listed are the driving transistor(s) for each coil.
Use this chart to determine which transistors could potentially be bad.
Also use the schematics.
If after replacing the driver transistors the coil/flashlamp is still
stuck on, then replace the TTL 74HCT574 logic chip.
A Coil just Does Not Work - What is Wrong?
Driver transistors can go "open" too. This means the logic prior
to the open transistor could be working fine, but the coil will not
energize. If there is power at the coil, this is something to consider
(but first see the test procedures below to make sure the coil itself
is actually OK).
Do the Transistor Test Procedures work 100%?
In short, no. But they do work about 98% of the time, and are an
excellent starting point. But yes, a transistor can test as "good",
but still be bad. The DMM test procedures test the transistors with
no load. Under load, a transistor could not work.
Transistor Testing procedures using a DMM.
If the driver board is out of the game for some reason,
test all the transistors. It only
takes a moment, and will ultimately save time.
To test a transistor, a digital multi-meter (DMM) is needed,
set to the "diode" position. NOTE: testing transistors with a DMM
is not 100% fool-proof. A transistor can test as "good" and still
be bad (rare, but it does happen!).
Testing Transistors INSTALLED in the Pin2000 driver board.
- FET 20N10L: Put the black lead of the DMM on the
metal tab of the transistor. Put the red lead of the DMM
on the right outside leg of the FET (as facing the transistor). A reading
of .4 to .6 volts should be seen. Then move the red lead of
the DMM to the left leg of the FET. A reading of 1.2 to 1.4 volts
should be seen. Put the red lead on the center FET leg,
and a zero reading should be seen. Any other value, and the FET is bad and will need
to be replaced.
Testing a FET on the P2000 driver board.
|
|
- TIP102: Put the black lead of the DMM on the
metal tab of the transistor. Put the red lead of the DMM on
each of the two outside legs of the transistor. A reading
of .4 to .6 volts should be seen. Put the red lead
on the center transistor leg (collector), and a zero reading should be seen.
Any other value, and the transistor is bad and will need
to be replaced.
Testing a TIP102 on the P2000 driver board.
|
|
- TIP107: Put the red lead of the DMM on the center leg
or on the metal tab of the transistor. Put the black lead of the
DMM on each of the two outside legs of the transistor.
A reading of .4 to .6 volts should be seen. Put the black lead
on the center transistor leg (collector),
and a zero reading should be seen.
Any other value, and the transistor is bad and will need
to be replaced.
Testing a TIP107 on the P2000 driver board.
|
|
If a coil is not working, the following approach is a good one to take.
It starts with the easiest test first; using the internal pin2000 diagnostics.
Then the tests moves to the coil itself, and goes back towards the driver board.
This makes the chain smaller, and gives a very systematic approach to finding
the problem.
Information shown in the solenoid test diagnostics. The top row of information
(J102) is the power to the coil. The bottom row is the drive (J110), completion
to ground back at the driver board transistor.
|
|
Testing Transistors/Coils, Driver board installed in a (near) WORKING game,
using the Diagnostics Test.
If the game powers on, the diagnostics can be used to test most
devices.
- Press the "Begin Test" button inside the coin door.
- Select "MAIN MENU: TESTS".
- Select "TEST MENU: SOLENOID TEST".
- Use the "+" and "-" buttons to move the test from coil to coil.
Each coil should fire.
- Remember the coin door interlock switch must be held in.
Otherwise the coil 50 volts will be turned off, and the coils won't fire.
Also make sure the "REPEAT" portion of the test is used. This can
be changed using the "Begin Test" button.
Solenoid Doesn't Work during Diagnostic Tests.
If a solenoid doesn't work from the diagnostic tests, here's what
to check. Turn the game off before doing this.
- Check all the fuses on the driver board. A non-working
solenoid could be as easy to fix as just replacing a fuse.
- Find the solenoid in question under the playfield. Make
sure the wire hasn't fallen off or become cut from the coil
(a very common problem). Remember, the power wire is "daisy chained"
from coil to coil. If one breaks "upstream", all coils "downstream"
will not work.
- If the above is correct, make sure the winding of the coil
haven't broken off from the solder lugs. If one has broken, it can be
re-soldered. Make sure the painted enamel insulation is sanded
from the wire before re-soldering. This is easy to check using a DMM
set to DC volts. There should be voltage on *both* lugs of the coil
(with the coil door closed). Voltage on only one lug means a coil wire
has broken.
- Check the coil diode (for any other pinball game, this would
be the next step). The coil diode for all games (except Pin2000 and WPC)
are attached right to the coil, with the banded side of the
diode connecting to the power side of the coil.
On Pin2000 and WPC games however, Williams
moved this diode to the power driver board for all coils but
the flipper coils. This increases reliability
as the diode is not subject to the jarring and heat a coil can
produce. It also eliminates the need for the operator to know which
coil wire goes to the banded side of the diode when replacing a coil!
These coil diodes are mounted on the driver board
near the transistor that drives each particular coil.
The Coin Door Interlock switch.
All pin2000 games have a coin door interlock switch. This turned off the power
to all the coils when the coil door was opened (for safety
reasons). On games with this interlock switch,
make sure the coin door is closed when testing coils!
Failed Coin Door Interlock switch.
Yes it does happen. The coin door interlock switch can fail,
or does not get pushed in enough when the coin door is closed.
This will prevent voltage from getting to the solenoids. If
none of the solenoids work, and the fuses are good, check the
coin door interlock switch for problems. A sure sign of this is
the Driver board solenoid power LED's will NOT be lit if the
coin door interlock switch is not closed! The interlock switch
opens the coil power coming from the transformer, which is way
before the power gets to the Driver board's fuses and power circuits.
A Systematic Approach.
1. Test for Power at the Coil.
Most pinball games (including pin2000) have power at
each and every coil at all times. To activate
a coil, GROUND is turned on momentarily by the driving transistor
to complete the power path. Since only ground (and not power) is turned on and
off, the driving transistors have less stress on them. With this in
mind, if we artificially attach a coil to ground, it will fire
(assuming the game is turned on).
- Turn the game on and leave it in "attract" mode.
- Lift the playfield.
- Close the coin door interlock switch.
- Put the DMM on DC voltage (100 volt range or higher).
- Attach the black lead of the DMM to the metal side rail.
- Touch the red lead of the DMM on either lug of the coil in question.
- A reading of 50 to 80 volts DC should be indicated. Switch the red test
lead to the other lug of the coil, and the same voltage should be seen again.
On flipper coils, test the two outside lugs of the coil.
If no voltage reading is shown, no power is getting to the coil.
On a two lug coil, if there is only voltage at one lug, the coil winding is broken.
- If no power is getting to the coil, a wire is probably
broken "upstream" or there is a bad fuse.
2. Testing the Coil and the Power Together.
This test will show if the power and the coil are indeed working
together:
- Game is on and in "attract" mode, and the playfield lifted.
- Close the coin door interlock switch.
- Connect an alligator clip to the metal side rail of the game.
- Momentarily touch the other end of the alligator clip to the GROUND lead of the coil
in question. This will be the coil lug with the thinner single wire attached.
On flipper coils, this is the middle lug (the power wire on most coils is
usually the thicker violet or red wire).
- The coil should fire (if the alligator clip is accidentally touched
to the power side of the coil, the game will reset and/or blow a fuse,
as the solenoid high voltage is being shorted directly to ground).
- If the coil does not fire, either the coil itself is bad,
or a fuse is blown, or the "daisy chained" power wire has broken
"upstream".
3. Testing from the FET or TIP102 Transistor to the Coil.
If the coil fires in the above test, this test will check from the
FET's or TIP102's output, the wiring to the playfield, and the coil itself.
- Game is on, and the "test mode" button is pressed once to enter
diagnostics.
- Close the coin door interlock switch.
- Find the transistor that controls the coil in question
(refer to the manual, inside the front cover).
- Attach an alligator clip to the game's metal side rail or some other ground point.
- Momentarily touch the other lead of the alligator clip to the metal
tab on the FET or TIP102 transistor in question.
- Its associated coil, flash lamp or lamp matrix row should energize.
- If the coil/flashlamp/lamp row does not energize, and it did
in the previous tests, there probably is a wiring problem.
A broken wire or bad connection at the connector would be most common.
It is also possible there is a bad transistor. Continue to the
next step, or use the DMM meter
and test the transistor on the board (see
Transistors Testing Procedures for details).
Testing the coil by touching the metal tab of the
FET to TP6 (ground) with a jumper wire.
|
|
Testing a lamp matrix row by touching the metal tab of the TIP102 to ground with
with a jumper wire. The entire lamp matrix row associated with this transistor should
light.
|
|
What if the above tests worked, but the coil still does not work
during game play?
If all the above tests worked, there is probably a driver board
problem. Everything has been tested from the FET back to the coil
itself. That only leaves the FET itself, a resistor,
and the 74HCT574 logic chip that controls the FET transistor. It has to
be one of these devices that are causing the problem.
4. Testing the FET (simulating the Logic Chip's Output to test the Coil).
The next thing to check is one step further "upstream". This is done
by simulating what the 74HCT574 logic TTL chip does to turn on a coil.
The TTL chip turns +5 volts on, which goes to the resistor, which then turns
on the FET, completing the ground path for the coil (hence the coil
fires). By simulating the output of the logic chip, everything in the
chain is tested except for the logic chip itself. If this test works,
it is nearly certain the 74HCT574 has failed. If the test does not work
(and the previous tests did work), there is probably a bad FET (most common),
an open resistor, or a broken board trace.
First find the resistor that connects to the FET to be tested.
Refer to the chart below to determine this resistor.
| FET # | Resistor # | 74HCT574 # |
|
FET # | Resistor # | 74HCT574 # |
| Drive Bank A |
|
Drive Bank C |
Q59 | R160 | U31 |
|
Q43 | R143 | U29 |
| Q60 | R161 | U31 |
|
Q44 | R145 | U29 |
| Q61 | R162 | U31 |
|
Q45 | R146 | U29 |
| Q62 | R163 | U31 |
|
Q46 | R147 | U29 |
| Q63 | R164 | U31 |
|
Q47 | R148 | U29 |
| Q64 | R165 | U31 |
|
Q48 | R149 | U29 |
| Q65 | R166 | U31 |
|
Q49 | R150 | U29 |
| Q66 | R167 | U31 |
|
Q50 | R151 | U29 |
| Drive Bank B |
|
Drive Bank D |
| Q51 | R152 | U30 |
|
Q67 | R168 | U32 |
| Q52 | R153 | U30 |
|
Q68 | R169 | U32 |
| Q53 | R154 | U30 |
|
Q69 | R170 | U32 |
| Q54 | R155 | U30 |
|
Q70 | R171 | U32 |
| Q55 | R156 | U30 |
|
Flipper Drivers |
| Q56 | R157 | U30 |
|
Q35 | R136 | U28 |
| Q57 | R158 | U30 |
|
Q36 | R137 | U28 |
| Q58 | R159 | U30 |
|
Q37 | R138 | U28 |
| |
|
Q38 | R139 | U28 |
| |
|
Q39 | R140 | U28 |
| |
|
Q40 | R141 | U28 |
| |
|
Q41 | R142 | U28 |
| |
|
Q42 | R144 | U28 |
Access +5 volts on the driver board by connecting an alligator clip to
the top side of resistor R190.
|
|
- Connect an alligator clip to +5 volts. This can be found on
the top side of resistor R190.
This resistor lives just above the +5 volt LED #17 (see picture).
The alligator clip *must* be clipped only to the top side of
this resistor (the side closest to the fuses).
- Turn the game on.
- Close the coin door interlock switch.
- Momentarily touch the other end of the alligator clip to the
above noted resistor on the *top* side (the side closest to the fuses).
- The coil should fire.
Momentarily touching +5 volts to the *top side* of the resistor
which connects to the FET to be tested. This should fire the coil.
|
|
What if the above tests worked, but the coil still does not work
during game play?
If all the above tests worked, the 74HCT574 chip has probably failed.
Everything has been tested from the FET back to the coil
itself. That only leaves the 74HCT574 logic chip that controls the FET transistor.
End of Systematic Approach.
Coil Diodes.
On all electronic pinball games, each and every CPU controlled
coil must have a coil diode.
This diode is VERY important. When a coil is energized, it
produces a magnetic field. As the coil's magnetic field
collapses (when the power shuts off to the coil),
a surge of power as much as twice the energizing voltage
spikes backwards through the coil. The coil diode prevents
this surge from going back to the driver board and damaging components.
If the coil diode is bad or missing, it can cause several problems.
If the diode is shorted on, coil fuse(s) will blow. If the diode
is open or missing, strange game play will result (because
the driver board is trying to absorb the return voltage
from the coil's magnetic field collapsing). At worse a missing
or open diode can cause the driver transistor or other
components to fail.
On older games (pre-1990 Williams games), sometimes a diode lead breaks on the coil
from vibration. Also, when replacing a coil, the operator
can install the coil wires incorrectly (the power wire should always be
attached to the coil's lug with the banded side of the diode).
To prevent this, Williams moved the coil diode to the Driver
board. This isolates the coil diode from vibration and
eliminates the possibility of installing the coil's wires in reverse.
This was done on all coils except the flipper coils.
The coil diodes on a Pin2000 flipper coil.
The red (right) wire is the "hot" wire.
The yellow (middle) wire handles the initial
hi-power "flip", and the orange (left) wire
handles the flipper's "hold".
|
|
Installing a New Coil.
Many replacement coils will come with a diode soldered across
its solder lugs. On Pin2000 and WPC games, all coils except the flipper
coils have the diode mounted on the Driver board.
For all coils except flipper coils,
cut the diode off the coil before installing.
Then put the coil wires to either coil lug.
The diode can also be left in place, but the
coil wires must be installed correctly. The ground wire MUST go to the
lug of the coil with the non-banded side of the diode. The power wire
attaches to the lug with the banded side of the diode. If
the wires are reversed, this essentially causes a shorted diode.
Though the Driver board mounted diode is still present as protection,
damage can occur to the coil's driver board transistor.
Coil Doesn't Work Check List.
If a coil doesn't work in a game, here's a check list to
help determine the problem.
Before starting, is the coil stuck on? (Hint: is there heat, smoke and a bad
smell?). If so, the coil's driving transistor has probably failed.
Turn the game off and check the driving
transistor, and replace if needed. See Transistors
Testing Procedures for more info.
If the coil just doesn't work, here's a list of things to check:
- Have the power wires fallen off the coil's solder lugs?
- Is the coil damaged? Has the internal winding broken off the
coil's solder lug?
- Is there power at the coil? See Testing for
Power at the Coil for more details.
- If there is no power at the coil, check its fuse. Use
the internal diagnostics and the "help" button to determine
which fuse controls the coil. See Testing
Transistors/Coils using the Diagnostics for details.
- Check the other coils that share one of the same wire colors.
Are they working too? If not, suspect the fuse that handles these coils.
- Power to coils are often ganged together. If the power wire
for this coil has fallen off a previous coil in the link, power may not get
to this coil.
- Using the DMM and its continuity test, make sure the coil
connects to the correct connector/pins on the driver board.
This information can be seen from the Diagnostics solenoid test.
- Check the driving FET transistor. Usually this transistor will short on
when it fails, but not always. Also check the 74HCT574 chip that drives the
FET transistor.
3f. Switch Matrix.
3g. Lamp Matrix.
Lamp Matrix Introduction. by Randy Fromm.
The circuit and operation of the lamp matrix
is the same in both Williams' WPC-95 system and the new Pinball 2000 machines.
The lamp matrix consists eight columns
and eight rows, with a lamp connected between each column and each row.
There is also a diode in series with each lamp. In a matrix, this is known
as a "steering" diode. It is needed to keep the current flowing along the
proper pathways.
Each of the columns is a separate source
path, individually controlled by each of eight TIP107 PNP darlington transistors.
These transistors are turned on and off in rapid succession, one after
another. This is known as "strobing".
It's easy to understand how the strobe
circuit works. The emitter connections of all eight TIP107 transistors
are connected directly to the +18 volt DC power supply. The collector connection
of each transistor becomes the strobe output that drives each of the lamp
columns. On the schematic diagram and in the lamp matrix table (conveniently
located on the inside of the back cover of every Williams operations manual)
they are labeled column 1 through 8.
In order to control the column outputs,
the base of each TIP107 is connected to the output of driver IC U11, a
ULN2803. Actually, this device only looks like an integrated circuit. It's
really an NPN darlington transistor array of eight individual transistors
that take TTL level inputs. Driving the ULN2083 is an "octal latch" (74LS374)
at U10. This chip grabs data off the rapidly changing data bus and holds on to it
for as long as the strobe lines need to be active.
One at a time, each output of U10 (74LS374) is activated
for a period of approximately 1 millisecond (1/1000 second). As one would
expect, it takes 8 milliseconds to complete the cycle which then repeats
and runs continuously.
When the output of chip U10 (74LS374)
activates, the associated TIP107 transistor is switched on. This allows
the +18 volt power supply to pass through the transistor, sending a +18
volt DC pulse down the column.
The columns are only half of the lamp matrix
circuit. Each strobe line is used to send a pulse to a group of eight lamps
but all eight lamps don't light up at once. How do we light just one lamp?
Individual lamps are selected by energizing a separate "lamp return" transistor
that is connected to each lamp in the group. Each lamp return is a "row."
In fact, the lamp return transistors are really our old friend the ground
switch!
The lamp return transistors are the same
type as those used to drive the solenoids, type TIP102. The lamp return
transistor completes the ground circuit for each lamp. To light a single
lamp, the lamp return circuit for that lamp is energized at the exact moment
the column is sending out its pulse. The column provides the source path
to the lamp while the row provides the return path and Viola! We have a
complete circuit and the lamp lights up. The computer then moves on to
the next column, energizing it while simultaneously activating the appropriate
row in order to light the lamps it wants to illuminate. The process repeats
for all eight columns.
With the strobe circuit to drive them,
the lamps are "refreshed" every 8 milliseconds. During each refresh time,
the selected lamps are given another pulse of current at +18 volts DC.
Although the lamps are standard 6 volt lamps, the +18 volt pulse does not
blow them out or cause them to glow too brightly. Remember, each pulse
lasts just 1 millisecond. This short "duty cycle" prevents the filament
in the lamp from getting too hot.
Also, despite the brief pulses they receive,
the lamps do not appear to flicker. With a refresh rate of approximately
120 cycles per second (120 Hertz) the filament in the lamp will not cool
enough to dim between pulses. Additionally, the human eye cannot perceive
any flicker faster than about 50 hertz. Even if the bulbs flickered a bit,
we wouldn't be able to see it.
The advantage of the lamp matrix over the
"one ground switch for each lamp" method is obvious. The lamp matrix allows
us to control 64 lamps using just 16 transistors and 16 wires connected
to the lamp driver circuitry on the printed circuit board.
|
|
P2000 lamp matrix TIP transistors on the Driver board.
| |
|
Lamp Short Circuit Protection.
Shorted lamps and lamp sockets aren't uncommon.
Without some type of protection circuitry, a short circuit in a single
lamp could take out an entire column or row. To prevent that from happening
(or at least reduce the likelyhood), Williams has included a simple but
effective over-current protection (OCP) circuit in the return path circuit.
The emitters of the TIP102
transistors in the row circuit aren't connected to ground directly. Instead
they are connected through a low value resistor of just a fraction of an ohm (.22 ohms).
That's not very much resistance. What could that possibly
do? Well, normally nothing. It's practically the same as grounding the
emitter directly.
But if a lamp or lamp socket short circuits,
there will be a dramatic rise in the amount of current flowing through
the emitter of the TIP102 and through the .22 ohm resistor to ground. This
will create a voltage across the resistor due to increased IR drop (V=IR).
This voltage is applied to the inverted (-) input of the voltage comparitor
U16 or U17, a LM339 chip.
The non-inverted (+) input of all the comparitors (LM339)
is connected to a +1.4 volt reference source. When an LM339 is connected
this way it has a very simple operation. The device compares the two inputs
to see which one has the highest voltage. If the non-inverted input has
the highest voltage, the output of the LM339 is +5 volts DC.
If the voltage at the inverted in the higher of the two, the output goes
to ground.
So when a lamp or lamp socket shorts, the
extra current causes the comparator's output to go to ground. This is,
of course, a logical "0" or "low." This signal is then used to "set" a
flip-flop (chips U12-U15). Since the base of the lamp return transistor is connected
to the inverted output of the flip-flop, "setting" the flip-flop causes
the transistor to turn off. This breaks the return path, opening the circuit
and preventing damage to either the column driver transistor or the row
transistor. Pretty neat eh?
All of the other lamps will continue to
function as normal. It is only when the shorted lamp or lamp socket is
energized that the OCP kicks in.
Testing the Lamp Matrix Rows.
This test will check from the TIP102's output, the wiring to the playfield,
and any lamp matrix row.
- Game is on, and the "test mode" button is pressed once to enter
diagnostics.
- Close the coin door interlock switch.
- Find the transistor that controls the lamp matrix row in question
(refer to the manual).
- Attach an alligator clip to the game's metal side rail or some other ground point.
- Momentarily touch the other lead of the alligator clip to the metal
tab on the TIP102 transistor in question.
- Its associated lamp matrix row should energize (all the bulbs in this row).
- If the lamp row does not energize, there probably is a wiring problem.
A broken wire or bad connection at the connector would be most common.
It is also possible there is a bad TIP102 transistor. Use the DMM meter
and test the transistor on the board (see
Transistors Testing Procedures for details).
|
|
|
|
3h. Sound Problems.
Problem: Sound was static noise, and then nothing.
Answer: Changed prism cards from my other one and it works fine.
Check to see if the game passes the POST for the sound section (it'll flag
it in the error report if not). If it does pass, the DSP chip is probably Ok and
look onwards from there for other faults. It's likely a very simple fix -
either a simple little opamp or a bad connection. If no error report, then the
DSP chip is likely okay, so suspect something in the output.
Unfortunately, there also isn't anything *between* the DSP and the output
connector, so it may be the DSP chip after all. (Jonathan Deitch)
When swapping RFM to SWE1 the sound might no longer work and you might
see the message U109 error! This is no hardware defect but a Star Wars
problem. You need the latest SWE1 sound software to be loaded and the
the sound works fine again.
Problem: Static comes on all three speakers during attract mode, particularly when there
was no background sound and white text was displayed.
Answer: There is "crosstalk" on the cable group between the PRISM board and the amplifier
board. This cable group is several cables, which are tied together with nylon
tie wraps. Cut and discard the nylon tie wraps (separating the cables),
and the static problem should go away.
If the static is still present, changed the 04-12621 audio amplifier board. The audio
amp is a small board inside the backbox's computer box, which connects to the
PRISM card. Sometimes the old amp, when failing, can cranking out way too much volume even at
the lowest settings.
3j. NuCore's Pinball 2000 Computer System.
The people at NuCore
have developed a software adaptation to Pinball 2000 that will allow any new
computer to be used inside any Pinball 2000 game. This has lots of great advantages.
First is the Prism board is NO longer needed! This is great news as the Prism board
is really difficult to repair, and replacements are rare/expensive. The other great advantage
to the Nucore system is if you swap your P2k game between RFM and Star Wars E1, you don't
have to swap the Prism ROM board (because there is no prism board). The Nucore software
auto-detects which playfield is installed in the game. Pretty neat stuff. Also the
animations and graphics are much smoother using the Nucore system. See their website
for more details on the Nucore system.
http://www.bigguyspinball.com/whatis.shtml
Frankly the Nucore system is the best thing to happen to Pinball 2000. It takes away the
arguement that Pinball 2000 games are "dangerous to own", because replacement boards and
computer parts are expensive or non-existant. One other neat thing about the Nucore system
is the Jukebox software that was demonstrated at the Pinball at the Zoo show in April 2008.
This allowed a Revenge from Mars to be used as a MP3 jukebox (when it wasn't being played
as a pinball machine!) Talk about cool stuff, it has a software jukebox interface allowing
the user to select songs by the flippper buttons and play them through the RFM's speakers.
When you want to play pinball, just press the game's "start" button, and the RFM comes up
and the jukebox turns off. Talk about making a Pinball 2000 game versatile.
3j. Misc Problems.
Problem: The game automatically recycles the balls,
on power up or whenever the coin door is opened and closed.
The game works perfectly but recycles upon each new powerup.
Answer: a bad trough eject optic (number 41). Replacing this optic
fixed the problem.
Problem:My RFM machine has on three occasions failed to boot.
When turning on the power, the screen is completely blank on no
lights appear on the playfield (only the backglass
powers up). I've looked over everything in the PC case and nothing seems loose. The
only "fix" to this problem was a good whack on the outside of the PC case.
Not a comforting solution, especially since this game is new.
Answer: The couple of times I have seen this it has been the same thing: the
Prism card is not seated properly in its slot. It looks like it's in,
the screw holding the one side to the bulkhead is in, you press on the
card and it feels like it's in, but it's not. With the power off,
open the lid to the PC box and with the heel of your hand press VERY
FIRMLY on the free end of the Prism card (the end away from the
bulkhead. I'm betting you will feel it THUNK down in another millimeter or two.
Also try removing the screw, pulling the card out and
putting it back in square to begin with (since if it's cockeyed like
that, the high corner of the card may be catching on the housing of
the PCI slot connector.) Also trying another slot in the mother board
is a good idea.
Problem: During gameplay sometimes the pictures on screen looked like
they were recorded during an earthquake.
Answer: Reaseating the connectors to the computer-case (at least the 15 pin
VGA/CGA conector) resolved the problem.
Problem: Pin 2000 Software Update program will not connect to the game.
Answer: Make sure that the filenames are EXACTLY as indicated by the source
files on the web site. If the update manager still can't find them,
move them to the "Pinball 2000 Update Manager" subdirectory under
"Program Files\Pin2000" (assuming that is the install directory).
Remember after connecting to the game, the update manager will
scan all the drives/directories for those exact filenames.
Now check the serial port cable on the Pinball 2000 game, which
runs from the coin door to the computer in the backbox. Remember on
Star Wars E1, this serial cable "splits" (so the backbox can be easily removed),
and the splits much be connected at the back of the game. Lastly check the small
ribbon cables inside the computer box. These connect the box's outside serial plugs
to the motherboard. Sometimes these are connected *backwards* from the factory!
Try connecting to the back of the computer directly, instead of through the
coin door serial connection.
Also make sure the communication is over the COM1 port, and doesn't have any other
program active that tries to grab the COM1 port on the connecting computer.
If "fupdate" still continues to fail, check the null modem cable and power
off/on the game after the game is connected to the connecting computer, and try
again. Try not to have other DOS sessions open (I've had fupdate fail
with other DOS sessions open)
If all else fails, run FUPDATE directly:
fupdate -b115200 -debug COM1
Also check to see if things like Dial up Networking or modem drivers are "claiming"
a serial port and locking it out on the connecting computer.
Problem: Light from the florescent tube is "bleeding" out around the monitor.
Answer: Add another layer of weather stripping around the corners and
monitor edge nearest the front of the backbox. The original weather stripping's
glue had come loose, allowing light to bleed past the monitor. This light
reflects off the top glass, decreasing the 3-D affect of the computer animation.
What happens is that the strip got stretched around the monitor when
originally installed. Eventually the glue lets go and the ends
shrink back away from each other to release the tension. Putting it on
with less tension, or even (for goodness' sake!) putting it on with the
gap on the other side of the monitor (so when it comes apart you don't
have light in the worst possible place) would have been great solutions.
You can buy similar stuff at any hardware store. Strips of
sticky-backed foam, used for weatherstripping, usually comes in a roll
in a plastic bag. I believe some pinheads already have some around as a
replacement for the foam on older EM lockdown bars. Just cut a few
lengths of it and stick them in the gap. You might need to stick some
pieces on top of other pieces to block all the light, but you shouldn't
need more than a couple of inches total.
Problem: The translight's florescent light is not working.
Answer: The florescent light in pinball 2000 uses a standard 120 volt
.35amp "ballast" (a small florescent transformer), a 120 volt FS2 starter, and a
F15WT8/35 18" (15w=15 watts, "T8" is the diameter where T8=8/8=1",
35 is the color degree [also seen is "CW"=cool white or "D"=daytime], 18" long)
florescent tube. Regardless of where the game is operated, the florescent light
runs at 120 volts (just like the PC computer). So even if the game is
jumpered and running in a 240 volt country, the florescent lamp is still
running at 120 volts.
When a florescent tube immediately blows on powerup, it is usually caused by a
shorted ballast. Since there is no ballast resistance and hence no way
of limiting the initial current flow to the elements in either end of the
tube, bang the tube is blown.
A ballast should always be warm when in operation, in confined spaces
with no air flow it will get very warm. But the construction of the
ballast should have a large enough surface area to allow heat
dissipation.
Replacement 120 volt ballasts can be purchased at any hardware (Home Depot)
or a lighting store. But note, if a game is operated overseas in a 220/240
volt 50 Hz environment, the replacement ballast should be capable of
running at 50 or 60 Hz. If a 60 Hz ballast is used in a 50 Hz environment,
the ballast will run hot and die quickly. So for European users a good replacement
is available from Robertson (1-800-922-9226) who makes a 50/60 Hz ballast, number SP1556.
Fortunately the cost of this ballast is low, but the shipping and tax cost
to the U.K. is quite high (perhaps as much as twice the cost of the actual ballast).
Thanks to C.King for this information.
Problem: None of my keys will open the backbox - either the
front or back. I have both sets of keys, the coin door and the
playfield security lock, and they work fine, but neither of them fit in
the backbox locks.
Answer: The key that unlocks the back of the
backbox will indeed open the backglass lock. The problem is the locking
arm gets jammed over the CPU box and the lock won't turn. Try
sticking your arm back there (through the back) and freeing it
manually. Since my game was already in place against the wall, I gave
the lock an extra twist with a pair of pliers before resorting to moving
it and the lock turned finally. Of course, this approach risks a key
broken off in the lock, so watch it!
Problem: My SWE1's neon lamp is not working.
Answer: First check that 12 volts is present going *into* the neon
lamp's transformer (is fuse 108 blown?) The neon transformer is mounted inside
the black handle of the light saber. The easiest way is to check for 12 volts at the Molex
connector going to the transformer (under the playfield), or at the power driver board
connector J111 pin 9. If 12 volts is getting to the transformer, next make sure the
return path to ground is being completed (this is handled by a ULN2803 transistor array on
power driver board). Ground the black wire on the Molex connector going to the neon
transformer. Does the neon tube turn on? If not (is the neon tube itself damaged?
very unlikely), the neon transformer itself is probably bad. If the neon does light
when the black wire is grounded, then the MosFET transistor and/or the ULN2803 chip
at U35 on the power driver board have most likely failed (or the wire going to the
power driver board has broken.
If the 12 volts is present at the neon transformer, and grounding the black
wire does not light the neon tube, then the neon transformer is probably at fault.
The neon transformer takes 12 volts DC and converts it to a very high voltage
(about 1500 volts, at low current). Because of this, to get the UL rating,
Williams was required to rivit close a plastic case around the transformer! To access
the transformer, the rivits will need to be drilled out with a 1/8" drill bit
(or grind off the heads of the rivets). On SWE1, do not try and remove the decorative
plastic "light saber handle" from half of the plastic transformer case! (they
uses silicon to attach it, and it does not come off without destroying the
decorative plastic!)
Once the rivits are removed, the transformer can be removed and checked.
Is there any high voltage (1500 volts DC) being output? If your DMM
does not go this high, just replace the transformer. The cheapest
way is to buy a car neon license plate transformer.
If needed, wire the automobile neon transformer under the PF
(if it doesn't fit in the ramp housing), and run
the high voltage wire up to the ramp and bulb. Note if you do this to be
sure to use wire rated for at least 2000 Volts (it'll have thick insulation; look
at the wire already on the bulb if you need some reference).
Specs for the original neon transformer are
here.
The original Williams SWE1 transformer (part number 04-10947)
may also still be available.
The original transformer for SWE1 (and Cirqus Voltaire) was a Ventex model VT12D5,
but they seem to have changed their model numbers so now it's
VT1510-12. A replacement is Ventex model NPS-12D5 and it fits and works fine.
Key specs are input 12 volts DC at 0.6A, and output 1500V 5mA.
You can find it at
www.ventextech.com/gen4.htm.
Note the output connector will need to be changed to a Molex connector.
Another transformer source is
www.sunsupply.com/transformers/winind.html.
Testing the neon tube itself, without using the high voltage transformer, and not
that easy. There is no to test a neon tube with a DMM - basically
the gas inside the tube conducts electricity. So basically a DMM can't generate a big enough
voltage to test it. They make little inductive testers - the tube will glow when this
thing is held near the neon tube, if the gas is still in there. Also try taking
the neon under some high voltage power lines at night to see if it glows
(and to scare yourself about how much energy is leaking out of them!)
Interestingly on Star Wars E1, the Neon light is not driven by a FET like the other coils and
flashlamps in the game. Instead a ULN2803 and a 74HCT574 on the driver board are used as
for the neon lamp transformer driver.
Again it's fairly easy to see if the neon
transformer is the problem or the driver board circuit. Notice the two
wires going to the neon transformer (one red, one black). The red is 12 volts DC,
and the black is ground. Using an alligator test clip, ground the black transformer
lead to the metal side rails on the game (or any other ground point). If the neon
tube glows, the driver board circuit is the problem. I say this because the neon
transformer can fail, so it's best to test this before digging into the driver board.
(You may want to check the neon glass too for any damage.)
The next thing in-line to the neon transformer is a ULN2803
chip at U35 on the driver board. In my repair scenerio, this was not the problem with the SWE1 neon.
Instead it was the 74HCT574 at U34 which had failed. At first I changed the ULN2803 at U35,
since this is first in-line to the neon transformer, but this changed nothing. Next
I looked at the 10k bussed resistor network RP10. Testing this in-circuit showed a problem,
but the true issue was a short in the U34 chip (which was making the resistor network
look bad, but it really wasn't the problem). Changing the U34 with a new 74LS574
fixed the problem, and the neon light worked again.
Problem: Tell me about Pinball 2000 serial numbers.
Answer: Because Pinball 2000
shipped in two separate containers (body and head) and the heads
were universal (not country-specific), they decided to give them
each their own serial number, in the normal random fashion, but
all out of the same pool of numbers. So twice as many serial
numbers got used up as there were games. This created all sorts
of problems, but they hadn't come up with a solution by the time
of the SW:EI MEL (Mechanical Engineering Lab) games.
So my game, at 000051, is the 26th MEL
game (my head has serial number 000050). But for production they
came up with some other scheme - I believe they gave the heads
numbers in a different range (the beginning digits were somehow
different). I'm not sure what number they started on, could have
been 000052, or it could have been 000100. No attempt was made
to make the body and head serial numbers always be the same way
odd/even, or to be sequential, so your head could have just about
any number!
Problem: Tell me about country specific differences in Pinball 2000 games.
Answer: The Pinball2000 tops are 100% identical all over the world - no
country specific packages.
The Pinball2000 bodys have been packed country specific with following differences:
- the coin door (e.g. german Pin2000 have an electronic Mars coin acceptor).
- the transformer jumper, which for worldmarket can be easily jumpered for different voltages.
- the country DIP switch settings on the powerdriver (mainly only for default language
and coin setting).
- the different power cord and main fuse (e.g. half value for 220V compared to 110V countries).
- the manual package, which might cointain translated parts.
Problem: Tell me about the differences between RFM and SWE1.
Answer: There seem to be several little technical and mechanical
differences between Star Wars and Revenge.
- There is a Backbox connector for the serial cable on SW:E1,
which makes the game assembly much easier. On RFM, the serial cable must
be fished from the back of the game to the coin door.
- There is a clamp at the end of the SW:E1 playfield, which fixes the playfield when raised,
so it does not lean against the backbox.
Problem: I have a SWE1 playfield, and put the SWE1 PROMS
in my current PRISM card, and swapped the playfield into my RFM. During the boot sequence it had
the following message: [UPDATE MISMATCH].
When the game started there was no sound except when the ball would hit
bumpers and slingshots. I went to the test and read the test report it
had the following entries:
- DCS Checksum test failed (U109)
- Check Lamp Column 6B
- Check Lamp 73B left sling GI Lower.
I reseated PROM U109 and rebooted, but had the same problem.
What am I doing wrong?
Answer: Since the PRISM card had a RFM update image in the flash ROMs with RFM
PROMs installed. It says [UPDATE MISMATCH] because the update (RFM) doesn't
match the PROMs (SWE1). All the other errors listed are probably also
caused by this.
To fix this, boot the game with the PRISM board and RFM PROMs installed
in it. Once it's up and running, go to the Utilities Menu and
disable the update. This will force the machine to boot from the PROMs.
Now put the SWE1 PROMs back in. Once the system boots, install the
latest flash SWE1 update.
It's supposed to effectively do that for you: when you fire up the
game with ROMs from one game, but the latest update of the other
game is still in Flash, it should say "update mismatch", disable the
updates, and then the Flash updater can be run without problems.
What it is doing is getting confused and sort of running the
mismatched update anyway. But unfortunately this PREVENTS
the Flash updater from running!! So you have to manually go
into the menu and disable updates first...
If you update with a PUB card it doesn't happen, because of the
way the PUB card takes over the machine.
Disabling Updates from the System Menu/Utilities/Update Game Code menus.
|
|
4a. Networking Pinball 2000.
Pin2000 uses PC-Xinu as the basis of its core OS, although WMS added a lot
of functionality to it. This was the decision of Tom
Uban who was the chief software engineer on the project and
all-round hardcore superstar programmer in general. PC-Xinu
already includes a TCP/IP stack, and he had already written a packet
driver for one kind of Ethernet card because all the
code and image downloads for development were done via ethernet.
A really simple web server wasn't too hard to write on top of
that - all the statistics it reports are already collected
by the game and displayed on-screen in the administration menus.
Louis Koziarz described the decision to use the PC-Xina operating system:
"XINA (the Williams written Pin2000 operating system)
is an application layer on top of PC-Xinu, which is a multithreaded operating system.
Originally created by Douglas Comer, he
documented it in a very well known two volume set of books (see
public.ise.canberra.edu.au/~chrisc/xinu.html
for more details).
PC-Xinu won over other operating systems for a few reasons. For example, making
Linux into a real-time system with thread support would involve modifying the kernel.
This would then obligate Williams into releasing the kernel modifications to the plubic,
under the GPL. Williams legal didn't like that. Comer's Xinu license was much
more friendly for the lawyers; WMS could do anything they wanted to it with no obligations.
Xinu also turned out to be quite easy to turn into a real-time system. And in the
end, proved to be as easy as Linux or BSD to work with in an embedded application."
Pinball Expo 1999.
The TCP/IP stack was demo'ed at Pinball Expo in 1999, where there was a tournament
automatically running. Barcode badges were made for entrants. This allowed them
to walk up to a game, swiped the badge in a barcode reader, play, and their
score was recorded. Their picture was also taken with a webcam and printed on the
badge, and the games showed the current
high score list including their digitized pictures
on all the tournament games during their attract mode
(while they weren't being played).
Why add an Ethernet card to your pinball 2000 game?
Good question! Maybe because you saw the SW:E1 machines network-linked together at Pinball Expo '99. Maybe because with an Ethernet card installed you can access your RFM's internal commands via the Internet (to check earnings, diagnostics, etc). Or perhaps you own more than one PB2K machine and you're going to have your own tournament.
Ethernet Card Specifications
Pinball 2000 only supports one brand of Ethernet Card,
an SMC (brand) ETHEREZ 16BIT ISA 10MBPS RJ-45 network card (mfg part# SMC8416T, typically goes for $31 USD at places like
www.buy.com or Ebay, but this card is becoming increasing
difficult to find). This network card needs to be installed
into the extra ISA slot in the RFM (v1.4 or higher) or SWE1 (v1.4 only) pinball machine motherboard.
After the settings are configured for the network adjustments, you can connect to the XINA command shell via telnet or alternately you can connect to the machine using a web browser (a simple httpd server is built in!). To all trying to add networking to a Pinball 2000 game: The networking capabilities
of P2000 were added and activated about a week or two before Pinball Expo 1999.
And one day after the Expo, Williams pinball was shut down! Tom Uban rushed to get
the networking added for the Pinball Expo 1999, so he standardized on a network
card that he had access to, and that was easy to write into the system. There
were no plans to "lock" into a certain brand of network card. But because Williams
closed down pinball, that's essentially what happened.
Ethernet Configuration Instructions.
Note that depending on the motherboard in the RFM machine, you may have
to adjust the BIOS setup setting prior to getting the network card
working in order to avoid an IRQ conflict. This can be determined if
there is a problem by configuring the network adjustments, setting an
IP address/mask, and then using the keyboard (or serial port) to access
a command prompt and run 'ping'. If the 'ping' works (the lights on the
card blink or device on the network responds to the ping, etc.), then
you don't need to change the BIOS. If you get some kind of "ez" error
message, then you will have to do the following:
Remove the PRISM board, plug in a keyboard and power up the box.
If you are lucky and the monitor "syncs up" to the BIOS screen,
then you can see what you are doing. If not, you may want to temporarily
plug in a PC monitor instead of the Pin2000 monitor.
Disable the 'Built in OnBoard Audio' in the 'Integrated Peripherals'
menu. On some RFM motherboards, the onboard audio uses the IRQ that
the default setting of the EtherEZ network card uses. Here are
the key strokes that should effect the change:
- Remove the PRISM board/plug in PC monitor and keyboard
- Power up (you will see the normal PC BIOS startup because the PRISM card is out).
- Press DEL to enter setup (the BIOS setup should appear).
- Press the RIGHT ARROW key (move to the INTEGRATED PERIPHERALS menu).
- Press the ENTER key (enter the INTEGRATED PERIPHERALS menu).
- Press the RIGHT ARROW key (move to the right hand column of menu items).
- Press the DOWN ARROW key (move to the BUILT IN ONBOARD AUDIO menu).
- Press the PAGE UP key (change the adjustment from ENABLE to DISABLE).
- Press the ESC key (leave the INTEGRATED PERIPHERALS menu).
- Press the DOWN ARROW key (move to the SAVE AND EXIT menu).
- Press the DOWN ARROW key.
- Press the DOWN ARROW key.
- Press the ENTER key (SAVE AND EXIT).
- Press the Y key (confirm).
- Press the ENTER key (do it).
- Replace the PRISM board.
The default username / password are set to "PIN2000" and "MANAGER" (respectively),
which you may want to change since all RFM machines are initially set
to the same values (but who would want to hack into a pinball machine?)
XINA commands.
The Xina operator system has its own commands, which are available at
marvin3m.com/pin2000/xina.htm.
Below are some notes from other users and their own networking experiences.
This weekend, I got my RFM up and running on our LAN using the
information at http://members.home.net/ratherplaypinball/rfmnotes.htm
(dead site). We're now attempting to write a frontend for the shell
commands, with the ultimate goal being a frontend that can start a game,
launch balls and use the flippers. This would be combined with a good
webcam and MS Media Server stream to facilitate someone playing my RFM
over the Internet.
Regardless of the outcome of this project, the ability to telnet into
the game while it's being played is great, if only for the world of
opportunities one has to mess with their (unsuspecting) friends while
they're playing . The game tilt, scenemgr and flip
commands alone can wreak havok on an unsuspecting player :) "Hey! These
flippers are possessed!! Tilt?!?! I never touched it! And why is it
starting Attack Mars already??" Muhahahaa :)
Another user: Installed a SMC 8416 Ethernet Card in my RFM. After disabling the
onboard sound, the ethernet card should work fine.
Afterwards I took a PC with WIN98 and plugged in an ethernet card, installed Internet Explorer
and started setup. A static IP adress has to be setup in the PC (normally
ISPs need dynamic IP adresses). I selected 168.0.0.2 for the PC.
In the communication options of RFM the own IP adress has to be
set and I selected 168.0.0.1. After that the RFM was rebooted.
Then TELNET could connect to the RFM and also Internet Explorer could
display the PIN2000 Webpage on RFM.
The communication options of RFM offer lots of settings.
I detected this when entering the PC's IP address as 'Tourney Address'
and activate tournament mode, the game hangs up at game over
(and seem to try a webserver update). An IP observer tool finds 'attacks'
on my PC Port 2069 when the RFM boots or ends game. The Port 5001
of Pin2000 wants to establish a UDP connection to Port 2069 of my PC.
The tournament functions already seem to be implemented on RFM 1.4!
I also find in the game info display, that there is a card_no display for the current player,
also some info about tournament functions.
End of Pinball 2000 document
* Go to the Pin Fix-It
Index at http://marvin3m.com/fix.htm
* Go to Marvin's Marvelous
Mechanical Museum at http://marvin3m.com
|
