Skip to main content

1971 Voice of the Mummy

with 1972 Seance and 1981 Dark Tower Board Games
Voice of the Mummy 1971  Mummy Box Hieroglyphs  Seance 1972  Dark Tower 1981  Sky Battle & Galaga  Mummy Boris Karloff  Mummy Blog  Mummy Contact Me   
Dark Tower Game Links > Dark Tower Online > Dark Tower Slideshow > Keypad Repair > Dark Tower Circuit Boards >  
Updated 07-04-2017 pm
Dark Tower Board Game Circuit Boards

This will be an ongoing page with details about the different versions of the  circuit boards used in the Dark Tower Board Game.

04-27-2017 Recently a friend sent a Tower that has an alignment problem when showing the Warrior, Food and Beast column.   The column was slightly to the left and not centered.  I've read on another site that a quick fix is to angle the clear diode down a bit and test until it centers better.  As I was making the adjustments, I noticed this tower was possessed by a sort of free electron demon.   It was very weird power on behavior.  On power on, the display would occasionally show an 8 only in the left display and do nothing else or a left digit eight with a rotating tower.  This spinning tower may stop on it's own and eventually display the 2 dashes or just keep going.  The lights would all be on and very dim also.  Power off and back on randomly would clear the problem.   I was challenged to find, what I thought was, a faulty part!

So, I removed the main board and speaker, motor and keypad and connected the parts on the test bench.   The battery holder with cable was attached and the switch used to turn the board off and on.  The power on problem was very random.  For example, it will power on with a normal selftest, left digit zero then dashes, 5 times with no issues then go crazy for 10 times in a row.  I replaced the battery cable wires, checked the crimps, removed, cleaned and polished the battery terminals.  After all of that work,  the tower continued to display the random power on problem.  The board also consumed a heavy amount of current.  A new set of "D" batteries were quickly depleted.  The amp meter displayed a 1.3A draw when the lights and motor are spinning.  I also could not connect my regulated power supply no matter how I changed the settings or current amount.  It would switch from voltage control back to current control and then cut back.  I could not get enough power for the board to power on at all.  The supply seemed to see a short and would turn down the current.

So, I used my perfect working tower as a donor to aid in quickly finding the problem part.  The idea was to swap parts until it's fixed.  First, I swapped the big items, the 3 chips,  no fix!  Still random weird power on issues.  Swapped transistors, no fix.  I swapped everything on the boards!  Still not fixed.  

The 2 boards use the wide DIP 28 pin processor chip N1LL to be compatible.  I noticed before that the circuit boards were different versions.  The bad board was A01 Rev B.  My tower board was a B03 RevB.   I assumed the same circuits also.  I'm thinking something else is up here!

So, I started looking around at the different sites with pictures of their boards and the parts used.  So far, I've seen 6 different versions of the main circuit boards and 3 chip versions.  My guess is that the answer lies in the idea that they made numerous changes to fix issues during production of the Dark Tower Game.  The early versions may have had known weird power on problems.  Read the Dark Tower instruction booklet in chapter 3.  There is a very detailed description of weird behavior and what to do.  If there is a problem or erratic behavior, turn it off and wait 10 seconds. Then turn it back on.  That tells me that there were issues from the start of production and there were ongoing fixes made.

UPDATE 04-29-2017  The weird power on issue with A01 RevB board is fixed.  I happen to notice a wire change on an A03 Rev A board that I just received for repair.  I tried the change on the A01 Rev B and the problem was solved.

Here are the details:  This A01 Rev B board seems to be an early version. It included a circuit change already.  The change used a 9112 transistor, a .47 uF electrolytic cap, a 1k and 1.5k resistor. They were installed at the white connector for battery power/motor/speaker.  Then this circuit was verified on the A03 RevA board that this change is incorporated into this version.   The traces and connections were compared with this which matches the external circuit change on A01 Rev B.  So that circuit change has stayed in the design.

So, on the A03 RevA, it was noticed that there is a wire connecting U1 Pin 4 of the processor to U2 pin 9 of the 75494 chip.  That was another nice clue to say, yes they made changes because of some problem.  On the A01 Rev B board U2 pin9 was connected to Batt-.  I made the change in the PCB by adding the wire and removing the wire jumper.  It worked!  Fortunately, this circuit change was about the exact issue I was having with my A01 Rev B.  The circuit change involves: removing the wire jumper from U2 pin 9 solder pad to the other connected solder pad that connects to Batt -.  Then add the wire from U1 Pin 4 to U2 Pin 9 solder pad near pin 9.  The board was then tested with several power on and off cycles including random off times in between power on.  Then completely tested with a few quick games at level 4.   I then tried connecting my regulated power supply and it worked fine.  Great, no more drained batteries. No more weird power on problems! The Free Electron Demon is gone!

UPDATE 05-01-2017 Another tower came in the mail today.  It was another A01 Rev B board.  Exact same issues.  Removed the jumper, added the wire.  The transistor circuit was already added. So, tower complete.

The next question is.  Why pin 4?  Pin 4 and 5 were originally left unused on the early version boards.  Then the wire was added from U1-4 to U2-9.  The D version boards used Pin 5 .  Then back to U1-4 on the E version boards.   It's seems to have started with pin 4 then to pin 5 then back to pin 4 that is used to route U1 Pin 4 or 5 to U2 Pin 9.   I decided to test this change on the A01 RevB,  I moved the white wire on U1 pin 4 to pin 5.  I fully tested with the same good working results...   BTW the U1 pins 4,5 are just above - voltage.  At about +.3V.   I don't have an answer to that besides it works also...   I have not seen an internal diagram of this chip.  Just the obvious, follow the traces and create a schematic based on it's working and that's what the pin are connected to.  The only way to answer is to made a complete schematic to start understanding the circuit better.

Do the later versions boards after A03 Rev A that use the U1 28 pin SDIP package size labeled N2LL have the same internal configuration as the original wide DIP 28pin N1LL chips?  Moving the wire from U1 pin 4 to pin 5, it still works but why was Pin 5 chosen over pin 4 in the earlier Rev boards with this similar change?

That was answered by Sean Riddle.

Richard-
 
              I think it’s just the packaging; I know N means plastic and J means ceramic.  There was a little discussion on the MAME board, and the thought was that L means DIP and LL or 2L means SDIP, but I’ve never seen those specified in an official TI datasheet.
 
Sean

So, the next feat is to document these boards with pics and schematics to show differences.  So far we have the boards front and back to compare.  They all function the same. Same chip program.  As long the circuit updates are made to the 001 Rev B, A01 Rev B, A03 Rev A boards.  So now there are 7 different PCB's and 3 package changes to the U1 Processor chip.

001 Rev B, A01 Rev B, A03 Rev A, B03 Rev B, D01 Rev C, D02 Rev C, E01 Rev A

There are 3 chip types with the same program code. MP7332-N1LL,  MP7332-N2LL  and maybe the last batch of chips labeled DCRS MP7332-N2LL.  The N1LL is the first released as a wide 28 pin DIP package and the N2LL and DCRS N2LL are a smaller SDIP package still at 28 pins.  The same internal pinout for the N1LL and N2LL.   The DCRS N2LL is using another internal pin layout.  Still the same program code but just shifted the pins internally.

Question: 05-01-2017 There is something confusing about the E01 boards and the processor chip PCB pinout.  Were these chips wired incorrectly?   Were the internal pins changed?  I'm just referencing the obvious unused pin4 which appears to be on the wrong side.  It looks like the pinout got flipped around in the chip...  I'll have some pics up to show what I mean.

Answered: 05-02-2017 An excellent source of TMS1000 info I have found in Sean Riddle.  The puzzle is answered.  The chips were manufactured using a internal pin layout of the TMS1100 chip.  Normally the chip was wired internally according to the TMS1400 pinout.  

Hmmm, still not sure why this was necessary as this chip change caused a new PCB layout to be created. $$$  Sounds like a CF with the chip manufacturer.  E01 PCB boards are reconfigured for this chip internal pinout change.  There aren't any functionality differences.  The program is exactly the same in all versions.

TO DO'S Add more pictures of the different versions of boards with Fritz schematics and maybe some theories to the changes.

A01 Rev B without U1 and Bulbs
This is Sean's Board used to create the very interesting snapshot of the processor die.
Transistor motor circuit added from factory. 
The PCB needs the U1-4 to U2-9 wire added as the jumper is still there.


A01 Rev B
Transistor motor circuit added from factory. 
The PCB needs the U1-4 to U2-9 wire added as the jumper is still there.



A01 Rev B
The TMS1400NLL is labeled on this date code.  The U1-5 U2-9 wire was added.
Note: I've noticed that the weird power on problem still occur every so often 
with U1-5 connected. I went back and moved it from U1-5 to U1-4.


A01 Rev B
PCB is missing the Transistor Motor and Power On Circuits
05-22-2017 am On first inspection, the jumper U2-9 to -Batt was present and the transistor motor circuit was not!  The Power On Circuit Mod was added from U1-4 to U2-9.  Then, the tower finally powered on and finished selftest.  So, now it's possible to move onto finding out what this transistor motor circuit actually solved.  With the board, motor, speaker removed from the housing,  placed on the test bench and powered up, it was obvious that the motor continued to spin a bit longer after self test had completed.  Next, was to run an L4 P4 game and watch the motor spin duration.  The same results, the motor spun a bit longer when inventory was checked.

To verify what happened when installed in the tower, the board, motor, speaker were installed back in the tower to test the timing with the carousel installed.  The problem that the transistor motor circuit solved was very obvious!   On power up, the tower continued to rotate a micro bit after completing self test.  Placing the carousel alignment a bit to the left instead of centered.  Then after starting an L4 test game,  it became even more obvious.  The tower continues to rotate a little more after each column is displayed during inventory.  So, in an L4 game, you have all the items to display so the tower carousel spins displaying your inventory.  The picture cells drift left more each time a new column is displayed.  By the time the last item is displayed, the sword,  the carousel has drifted past the sword.  The alignment fix of adjusting the LED's didn't help as this problem is much more severe with each column being displayed.

Then, the transistor motor mod circuit was installed using a 2N2907 or 2N3906 transistor.  These common PNP transistors are very close to the hFE of the original 9112 transistor.  On the test bench it's obvious that the motor turns on and off much faster.  So, the board and components are installed back in the tower.  Another problem cropped up though. On power on self test, the carousel is very misaligned to the right.  Whaaaaat?  So I compared this carousel with other units laying around.  I didn't notice this at first!  They are different!  The carousels are labeled 1 and 2 in the clear plastic.  That might be a mold number though.  The black rings used for the column alignment are different.  There are no markings at all on this back ring.  The spaces are the same size but one is shifted compared to the other.  This may be a huge problem as this carousel version may make the motor circuit mod incompatible!  Still working on this...  An idea would be to shift the LEDS further in or out from the board.   The motor mod circuit proves it value in stopping the motor faster for a consistent alignment throughout the columns.

05-22-2017 pm While watching the carousel turn to see the slot timing and when the light appears through each slot to when the motor shuts off, it was determined that the leds need to be on the other side of the board.  Measuring the difference of where the black slot ring was mounted in each carousel was about 1/4".  This should get the alignment close then tweak the LEDS to fine tune alignment. The means that the LEDS need to be mounted on the back side of the board.  Done, tested and it works great.  It took a bit of time bending the leds to get a perfect alignment but it works great. Watch bending the leads as they will break.  Use new LEDS.   Then, self test worked fine with an L4 game checking inventory. Perfect, all the columns lined up correctly.

Then I wanted to brighten up the LED display a bit so the 7 segment resistors were changed from the 620 to 470 1/4 watt resistors.  That brightened them up a bit.

So, this A01-B board has the power mod and the motor mod circuits with custom installed LEDS to perform as well as the later released tower boards.  90 hrs later...  but a fun problem!  

Can you imagine what the game designers went though.  ... a long time ago... in a far away... boardroom?, the MB marketing managers had a big meeting with the game designers about Dark Tower returns and angry merchants!  "These towers are not working correctly and we're getting too many returns...  fix it! waaaa!"  And so they did...

06-01-2017 pm Once again, I ran another question about the transistor motor circuit and it's purpose to Sean Riddle.

Rich-
 
                Originally R9 from the TMS1400 was connected through a resistor to an NPN transistor’s base; the emitter was connected to battery negative and the collector went to one of the motor leads.  The other motor lead was directly connected to battery positive.  That way, R9 turned the motor on and off.  Nice and simple. 
Then they added another resistor to the base of a new PNP resistor that has its emitter connected to battery positive and its collector to the motor lead.  The base is also connected through a 0.47uF electrolytic capacitor to battery positive, and there’s a new resistor from R9 to ground.
 
The circuit is now similar to the motor driver in Big Trak: www.robotroom.com/Big-Trak-3.html (except no split power supply and only one control input). 
                When R9 is used to turn on the motor, the new transistor is turned off.  When R9 is used to turn off the motor, the new capacitor charges up and turns on the new transistor.  I think that must be to help stop it more quickly than just turning off the original transistor.  I’m not sure if the new transistor ramps on, or turns fully on immediately.
 
Sean
The U1-4 to U2-9 Power On Mod wire was added.  I've noticed that the weird
 power on problem still occured, every so often, with U1-5 connected.  
I went back and moved the wire from U1-5 to U1-4.

The "LEDS" are mounted in the back to match the early version carousel alignment.
  The leads were bent 180 to keep polarity correct but failed later.  
The housing to the LEDS had cracked from bending and alignment and failed soon after. 
 New LEDS are needed.

A cell phone camera captured the infrared light below

BELOW: An early version carousel released with boards without the Transistor Motor Circuit.  The black ring opening on these early released carousels is shifted CCW compared to the later carousels.   A ruler is used as to show the shift in the black ring.  The ruler is lined up on the left edge of the larger slot in the ring.

BELOW: A later version carousel used with boards with the Transistor Motor Circuit built into PCB or added to the parts surface.  The LEDS can stay on the front on the board.  Later revision boards boards, A03-A and later, have the motor circuit embedded into the PCB and all use the later released carousels.

An easier way to tell the early version from the later version by looking at it from this angle. 
Early Version on the left with single slot.  Later Version on the Right with double slot.
The early version carousel, on the left, can be used with any board version as long as the transistor motor circuit is added or is already built into the board.  You will have to install the LEDS on the back to get the picture timing alignment correct.  A new set of LEDS are needed as the original LED leads will break when you reverse the bend in the leads. (BTW, these "LEDS" are a transistor set consisting of an Infrared LED the Transmitter and a transistor Receiver)

The later version carousel, on the right, can be used with any board as long as the transistor motor circuit is added or already built into the board.  The LEDS need to be installed on the front for correct carousel alignment.


001 Rev B - The Earliest Known Version
07-01-2017 Now the earliest known PCB version.  Changes needed are: 
Change the Elec Power Cap from 680uF 16V to 470uF 16V.  Add the missing 
solder pads for the 10K pullup resistor.  U2-9 jumper to Batt- changed to U2-9 to U1-4.
Transistor Motor Circuit needs to be installed.


installed the 470uF 16V cap, removed jumper, added wire

still needed to install the transistor motor circuit

A 10k pullup resistor was installed so the left digit is brighter. It seems like the board designers went back and forth on the need for this resistor.  If you remove it, you'll notice the left digit is 50% dimmer than the right.

installed the motor circuit and LEDS on the back for correct alignment with an early version carousel.



A03 Rev A
Power Mod U2-9 to U1-4 wire already added from the factory.
The transistor motor circuit is embedded into the board.


B03 Rev B
The Power Mod wire has been embedded into the PCB connecting U1-4 to U2-9
The transistor motor circuit is embedded into the board .


D01 Rev C - Without U1 and Bulbs 
Sean Riddle's cadaver board. This is a D01 board version with the SDIP package U1 Processor Chip.  I can assume that is the reason for the D version PCB.  This version also wires the Power Mod wire to U1-5 to U2-9.  U2 is flipped around.
The PCB layout changed in the D and E boards.  The .47uF cap is removed from the motor circuit.


D01 Rev C
The PCB layout changed in the D and E boards.  The .47uF cap is removed from the motor circuit.


D02 Rev C - Overseas Atlantis Game
I'm not sure what was changed in D02. The PCB traces and parts/locations appear the same.
The PCB layout changed in the D and E boards.  The .47uF cap is removed from the motor circuit.


E01 Rev A - Latest Known Version
This is the latest known PCB version.  The U1 chip internal pinout and PCB changed.  
U1 and U2 have been turned around.  The power mod circuit is changed again to 
U1-4 to U2-9.  Error or intentional?  The PCB traces are thicker with extra solder pads in 
different locations for an unknown purpose.
The PCB layout changed in the D and E boards with little change to the circuit design.


A01 Rev B with broken PCB traces
07-02-2017  PCB Repairs. Here is a board purchased off ebay with PCB traces corroded away from dead batteries.  There was a huge white wire jumpering the damaged foil to make connection.  Once the corrosion was removed, a new trace and solder pads were added and the parts soldered back in place.  The standard circuit updates were applied to an A01-B PCB.  Green enamel paint was then added to prevent further oxidation.  It works.

The connectors were removed to show the corrosion staining to board. Cleaned with alcohol and sanded with fine sandpaper and alcohol to remove stain.  This board had a 470uF cap at 25V.  I went ahead and swapped it out with a 16V.  Plus added 10K pullup resistor.

Scrubbed clean with alcohol.  Green solder mask scraped away to show  extent of the corrosion.

Completed PCB trace repair.  The bare conductive foil was not wide enough to match
 the total width on the missing copper trace, so it was added at an angle as best as possible.
At the keypad connector, one pad was added and the rest were touched up.

Completed foil.  The bare foil was not wide enough so it was added at an angle.

Enamel paint was added to cover the exposed copper traces

Added the standard upgrades to the A01-B board and new connectors.  It works...