Bally Twilight Zone Pinball

Repairs, Restorations, Tweaks and Insights

Introduction

May 5, 2013

Updated May 15, 2013

It was time to change up my pinball collection and I had decided that Star Trek the Next Generation would be the machine to go. As of the time I'm writing this, many pinball prices have gone through the roof. While there are pros and cons to this significant increase in price, one clear negative is that it costs a ton of money to get a "top tier" machine. Luckily for me, I owned Star Trek the Next Generation for a good 6 years and it had increased in price as well as everything else. For this reason, I set out to make a trade using my Star Trek the Next Generation. I was looking for an even swap for a Theatre of Magic or a Twilight Zone where I would add some additional cash.

I ended up finding a Twilight Zone locally and doing a trade. While this Twilight Zone had been cleaned up, it wasn't as nice as my STTNG (IMHO). It needed various fixes, both cosmetic and mechanical/electrical. This helped reduce the TZ vs. STTNG price delta. Additionally, since I enjoy fixing pinball machines, this was a bonus in its own way.

When I was checking out the Twilight Zone I noted several issues:

Overall this was no project machine and was mostly working. I was hopeful it simply needed some TLC and elbow grease and would become a very nice Twilight Zone example. It is turning out to be quite a nice machine.

What's Special about Twilight Zone?

Original May 11, 2013

It's hard to talk about the merits of Twilight Zone without first mentioning that it has consistently been a top rated pinball machine for 10+ years. I can completely see why Twilight Zone has been ranked so high for so many years. Simply put, it's a great machine. Here are some of my reasons:

Prevent Pinball from Bouncing Out of Bridge Diverter

May 14, 2022

I've rediscovered Twilight Zone over the past couple of weeks and have played it more in this time than the previous couple of years! During this resurgence, I noticed that the ball would occasionally hit the bridge diverter and bounce away. This is the diverter that is just to the right of the Power mini-playfield. After bouncing away from the bridge diverter, sometimes the pinball would jump off the wireform while other times it might sneak onto the Power mini-playfield.

I asked about this behavior on Pinside and was quickly reminded this is a very common situation. While I had forgotten, it took little time to find people discussing this over 15 years ago! There was always one common solution. This was to put a magnet on the diverter. The magnet would help grab the pinball, preventing it for bouncing away.

The trick is finding the right magnet. It needs to be strong enough to grab the pinball yet weak enough that it allows the pinball to fall down to the playfield when the diverted "dumps" the pinball. After a couple of failed attempts, I've found a $6 (USA, half that price online) "ceramic block magnet" from Home Depot that works perfectly. Here's a picture of the product. Note that it was originally in a plastic shell and had 2 magnets.

This ceramic magnet does not weigh very much, about 25 grams. For reference, a pinball weights 80 grams. For some scale, here is a picture of the ceramic block magnet and a pinball.

Adding the magnet to Twilight Zone's bridge diverter is very easy. The approach is to use double sided mounting tape and attach the ceramic magnet to the bridge diverter. The double sided mounting tape I used was "Scotch Mount" (which claims it can hold 15 lbs). This Scotch Mount is pictured below.

Attach the Scotch Mount to the ceramic block magnet and cut to size. The end result looks like the picture below. The Scotch Mount tape and a pinball are shown for scale.

I found attaching this specific cermanic block magnet to the upper outside corner of the diverter works best (when viewed from the player position). Remove the adhesive protector from the Scotch Mount then press and hold onto the appropriate diverter location for about 20 seconds. Below is a picture of how mine looked when complete.

I've rolled the pinball pretty fast at the bridge diverter and, with this ceramic block magnet solution, the diverter always catches it -- and correctly dumps it to the playfield. As people who did this approach 15+ years ago mentioned, it unfortunately does not help with the powerball bouncing out because the powerball is not attracted to magnets.

A quick comment about the ceramic magnet placement. As the picture shows, I mounted it as high up on the diverter as I could. At this position, it seems to have perfect strength. While I expect the Scotch Mount to last a long time, for additional mounting strength, it is possible to use the mounting tape on both the top of the magnet and the side. At this time I'm only using mounting tape on the side. Additionally, if your setup requires a bit more magnet strength, I suspect you could lower the magnet's position.

ColorDMD LCD For Twilight Zone

May 5, 2022

I installed a ColorDMD LED for Twilight Zone. Here are a few pictures of the colorized DMD.

Clock Only Runs in Reverse

July 8, 2020

I turned on my Twilight Zone pinball after having not played it for a bit and immediately knew something was wrong. From the instant I flipped the power switch, Twilight Zone's clock started running. It keep running... and running... and wouldn't stop.

That's not quite true. The clock stopped as soon as Twilight Zone displayed the "Testing" message on the DMD but as soon as that message ended, the clock started running and ran forever. I also realized that it was running in reverse.

I found I could get the clock to stop running in reverse if I went into the Clock Diagnostics. While in clock diagnostics, the clock worked perfectly for slow and fast reverse. While the clock arms moved, all the optos correctly reported the minute and hour. When I switched the diagnostic to have the clock run forward, nothing happened.

After doing some quick checks to verify connectors were attached, I started more structured debugging. The clock motor is controlled by the A-16120 "D.C. Motor Control Assembly". Generally, this board accepts 12VDC, an input for "move forward" (J1-1) and an input for "move in reverse" (J2-2). While running clock diagnostics, I verified the WPC was driving both the forward and reverse inputs pulsed correctly. When in clock diagnostics, the input to these pins would pulse and the rate would be fast or slow depending on the test. This was a strong indication the problem was related to the A-16120 board itself.

Below is a picture of A-16120. Notice the U1 chip (about 1/3 of the way from left and vertically centered). This is a 4N25 optoisolator. An optoisolator electrically separates 2 parts of a circuit (much like an isolation transformed). Motors are known to cause all sorts of electrical noise so the optoisolator isolates the motor side of the circuit from the rest of the pinball circuity. Generally speaking, the isolation is achieved using an LED emitter and receiver that are inside the 4N25.

At this point I decided to google the problem and found some discussion of Twilight Zone clock's running only in one direction. For clocks stuck in reverse, the typical solution was to replace U1 on the A-16120 board. This was consistent with my findings because the "reverse" input (J2-2) feeds the U1 chip. I decided to just replace U1 rather than spending additional time digging into the schematic and testing all the various elements. The theory here is that either the internal LED emitter or receiver went bad, causing the 4N25 to not pass the input to the isolated output correctly (and get stuck in one state).

The new U1 fixed the problem. Twilight Zone clock ran perfectly in forward and reverse (and didn't run immediately when power was applied).

Decal for Battle the Power Diverter

August 17, 2014

I'm not a fan of most pinball "mods". Far too often they are ugly, amateurish and unnecessary additions that the game designer wouldn't have put on the game if they had an infinite budget. When I saw a simple mod that Jeff at pinballpimp.com created, I thought that the designers would have included this in the machine. .

Twilight Zone has a unique diverter that takes the pinball off a metal ramp and drops it onto the playfield. The diverter lets the ball pass when it's time to "battle the power" yet drops it to the playfield when that mode isn't available. The diverter is pretty basic looking. If one doesn't think much about it, it's easy to ignore. However, once it grabs your attention it is pretty hard to say it's anything but ugly. A picture of this diverter is below.

The mod is a custom designed and cut decal that fits perfectly on the (ugly) diverter. After installed, the diverter looks as pictured below. To me, it really looks like something the designers would have put on the game if they had some extra budget.

Installation is fairly straightforward. While I'm sure there are many approaches, I did the following.

1) Start by cleaning your ugly (soon to be beautiful) diverter. I used isopropyl alcohol.

2) Leaving the backing attached to the decal, line up the decal and use a rubber tipped clamp to hold it in place. Make sure it's where you want it. Adjust as needed. Smaller clamps may be better than what I used, but these worked fine.

3) With the decal still clamped, remove a section of the backing and attach the exposed area to the diverter. This part should stick exactly where you want it because of the clamp. Press it down well and rub it to ensure good adhesion. Now that the decal is partially connected, you remove the clamp. The decal will stay in place because it is partially attached. Removed the rest of the backing and attach.

Clicking / Popping Sound Problem

August 9, 2014

My Twilight Zone developed a problem where there was popping/cracking in the audio. After performing some tests with the diagnostics, I quickly realized the cracking/popping only occurs during speech. It never occurs during the FM synthesis audio or the basic "DAC" computer generated audio.

Since Twilight Zone uses the "pre-DCS" audio board and uses technology from 1993, the speech isn't the highest quality in the first place. However, I wasn't confusing the "early 1990s quality" speech as cracking/popper. This was real cracking/popping that occurred somewhat randomly during speech samples.

I hooked the sound board up to my scope in an attempt to see if I could catch the popping/cracking. While connected to part of the output circuit from the HC55536, which is the speech generation chip, I captured a quick voltage drop that correlated to the popping/cracking. The image below shows the sound wave quickly jumping down to about -1.4VAC. (Note: the voltage scale is off by 10x in the image below)

I looked at the signals feeding the HC55536 and everything looked fine. I decided the HC55536 itself might be bad, so I ordered a replacement.

It arrived. I put it in. No difference! ARGH!

I decided I really needed to cut the circuit in half and see whether the problem was on the op-amp side or the HC55536 side. In general, the HC55536 generates the audio in DC space. There is a Vcc/2 DC offset. So instead of a sound wave centered around 0V, it's centered around ~2.5V. To remove the DC offset, there is a 1 uF cap at C37. This is connected to the audio output of HC55536 and marks the beginning of what I'm calling the op-amp phase. The op-amp side of the house actually has the signal going through 2 op-amp circuits before it is finally summed/mixed by a 3rd op-amp circuit. The summing/mixing is where the CVSD sourced speech is mixed with the DAC audio and the FM synthesis audio.

Side note: since the DAC and FM synthesis didn't have popping/clicking, I've always been targeting these 3 areas: 1) the HC55536, 2) the first op-amp circuit and 3) the 2nd op-amp circuit. BTW, I'm pretty sure the 2 op-amp circuits are actually 2nd order filters and don't actually amplify anything.

I desoldered C37, effectively isolating these two halves. Clearly there would be no speech on the speakers but I could still measure things.

With C37 removed, I hooked everything up and started "Twilight Zone" speech test. This is the test where the machine keeps saying "the Twilight Zone" over and over. I heard nothing, as expected, but I didn't hear any popping/clicking either. Hardly conclusive, but it was an indication the popping/clicking wasn't being randomly generated by the op-amp side of the circuit.

I hooked my scope up to pin 3 of the HC55536 (audio out) and was able to see the downward spike! The problem was happening even though I couldn't verify it with my ears. This clearly indicated the cause, still unknown, was on the HC55536 side of the circuit.

But I had replaced the HC55536 and checked the various inputs (Vcc, clock, etc.). Clearly I had missed something, so I went back and looked again.

For the 2nd time... Vcc looked good, clock looked reasonable, digital input looked reasonable, etc. After scratching my head, I figured I must be doing something wrong. Then it occurred to me...

But first, a little more background. The +5VDC comes into the sound board from the main driver board. Before it goes to the HC55536 Vcc input, it goes through a 1K resistor (R32) with a filter cap (47uF 10V C34). This 1K resistor causes a voltage drop of 0.4VDC or so. I'm going to ignore this for the rest of the writeup as it's not really important, but I wanted to acknowledge it.

I realized I had never tested the Vcc under load. By this I mean that previous measurements of Vcc going into the HC55536 were done with the game sitting idle in diags. I decided I should have the HC55536 "working" when I measured it.

That's when I found something odd. Using two scope probes, each on different sides of the 1K resistor R32, I saw that the +5VDC was very different. Before the resistor, it was very clean. After the resistor, it had lots of drops. Nothing huge, but downward spikes visible on the scope. I couldn't correlate these with the audio clicking (by moving one probe to the A-out of the HC55536), but this just felt wrong.

I believe the cap at C34 exists to smooth to voltage going into HC55536 and the resistor at R32 exists to ensure that not too much current is drawn from the main 5V supply. The cap at C34 should store enough charge to satisfy any short term need.

I wondered if C34 was bad. I didn't have a 47 uF cap. All I had was a 1 uF cap. Not really a good replacement, but I decided to use it. As a complete hack, I bent the leads of the 1 uF cap and connected it in parallel with C34. The +5DC going into the HC55536 immediately looked better! I ended up adding solder to my hack so it wouldn't move around while I took some more readings. This is shown below.

With the hacked 1 uF cap soldered in parallel with C34, I couldn't see any downward spikes in the A-out from the HC55536 also! This was a happy moment!

Side story: After replacing the HC55536 earlier this week and it having no impact, I got pretty tired of this. I reached out to one of the board repair guys and asked if they could repair the board. I listed the symptom, what I had replaced and so forth. The response I got back was along the lines of "no thank you, I'll pass". I can understand this... it wasn't an obvious problem and, from his point of view, who knows how hacked up the board was. I was beginning to wonder if I'd need to purchase a new one... and there aren't repros, so they aren't cheap. Ultimately being told "I'll pass" gave me enough energy to try fixing it one more time.

I couldn't tell for sure if this fixed the problem since C37 was still removed and no audio was being sent to the speakers. It simply looked fixed on the scope. I pulled the board, soldered C37 back in place...

Got everything connected ... went to diags... started the "Twilight Zone" speech test...

AND NO CRACKING/POPPING!

I was 99% there. I just needed to get the right part and remove my hack. It was late so I called it a night.

The next day I went out to find the right cap. The schematic says C34 is a 47 uF 10V but it's actually a 25V on the board. All I could find was a 47 uF 35V, which is fine although noticeably bigger. I was happy that I was able to find an axial version, which is when the cap has the leads on each end rather than both coming out the bottom.

I installed the bigger 47 uF 35V cap. It just fit in the space available. The first picture below shows the space for the cap (with the old cap removed) followed by a picture with the new cap installed.

While I know it's difficult to test caps (I don't have an ESR meter, caps don't test well in circuits, etc.), with the original C34 removed I could see something was wrong with it. Using the cap mode on my DMM, it never showed any capacitance. This probably explains why adding the hack 1 uF cap showed a difference. Basically, 1 uF is a lot better than effectively none. It also indicates that 47 uF is perhaps overkill since 1 uF seemed to get the job done. At least for the audio I was playing.

I put the correctly repaired board back in Twilight Zone and started the audio test. Again, no popping/cracking! I left it speaking "the Twilight Zone" for 30 minutes and never had any cracking. I pronounced this problem solved.

Bottom line: If you are hearing cracking/popping only in your speech on a pre-DCS machine, it could be related to your +5VDC supply. Either the supply is poor coming into the sound board or some aspect of the sound board's +5VDC handing may be failing or failed. In my case, it was C34.

Clock Repair

May 8, 2013

When I was checking out this Twilight Zone before I acquired it, I noticed the clock was broken. The hour and minute hands would move, even at different speeds, but the diagnostics yelled THE CLOCK IS BROKEN. Since opto problems, or even resistor problems, are common with the Twilight Zone clock, I wasn't sure whether it would be a cheap repair (replace resistor?) or an expensive repair (purchase a reproduction clock board for $90) or something in the middle.

Below is the Twilight Zone clock.

Taking the clock off the playfield isn't trivial, but it's fairly simple. You disconnect the 4 wires from below the playfield, pull them up to the top of the playfield, and then remove a single hex screw. The clock mounting then "slides back" to release itself from a clip. At this point you can pull the clock off the playfield.

Once off the playfield, I started removing the long screws that hold the front of the Twilight Zone clock in place. After removing the plastic front you need to remove an E clip that holds the minute hand in place. This is shown below.

Once the E-clip is removed, you can remove the clock hands. Inside the minute hand is a small rod that holds it in place. This is shown below.

Pull the rod out and then you can pull the hour hand off. At this point you can remove the plastic piece with the clock artwork.

Finally, the minute PCB is exposed.

When I ran the clock diagnostic, I could see the minute functionality worked fine. The problem was related to the hours. Occasionally the game would show the correct hour, but it would mostly display 0. This was a strong sign there was something amiss with the hour PCB. I was lucky to find the opts work fine but there was poor connectivity between the minute PCB and the hour PCB. These two are not connected with a ribbon cable. The hour PCB has several long pins that plug directly into a connector on the minute PCB. This is shown below.

The minute PCB's connector has metal contacts that press up against the connectors from the hour PCB. In my case, these contacts were not strongly pressing up against the connectors. I used a small (jewelry like) flat head screwdriver and was able to bend the contacts so they press more firmly.

Once I put everything back together, the clock worked fine. At power on it would search for noon. During attract it would display the current time.

Clock Repair Part 2

March 30, 2014

The clock in Twilight Zone is known for being problematic. Much of the time the optos start going bad, causing the game to report "the clock is broken." As mentioned in the earlier entry, I had previously worked on my Twilight Zone's clock and got it working. However, the months that passed showed that it was working pretty well, but not perfectly. It would always initialize correctly at boot (by finding noon), but once in a while it would flake out during gameplay and the credit dot would appear. The clock does get hit by the ball occasionally, which I suspect contributes to any clock problems. Recently the clock went from working most of the time to not working almost all the time.

The Twilight Zone clock has two PCBs. The front (when looking at the face of the clock) PCB has optos for the minute hand and the back PCB has optos for the hour hand. The front PCB has the connectors that provide power and logic from the game. The rear PCB has a connector that hooks into the front PCB to get power and access to logic signals. This is shown in the image below.

After doing some debugging, I started to suspect that the connector between the front/minute PCB and the rear/hour PCB was causing my troubles. This was based on having trouble with the connector before and noticing that very slight movements of this connector would cause the clock to work or stop working. After looking closely at the solder joints on this connector, I determining they looked fine. Because there wasn't anything obviously wrong, I decided to simply remove the connector and replace it with wires. No connector (using direct wires instead) means a guaranteed good connection.

Since there are seven pins on the connector, I cut 7 pieces of wire that were about 2" in length. My goal was to leave enough slack in the wiring so I could "fold open" the two PCBs if I ever needed to do some work on them. I desoldered the male and female connectors from each PCB and then soldered the individual wires in place of the connector. The result is shown below. The wires have an "S" shape and the "slack" is tucked between the two PCBs.

This repair made a HUGE difference. The Twilight Zone clock diagnostic test will automatically stop running when it detects the optos are reporting unreasonable values. Prior to this fix, I could never run the test for more than a minute or two. After this fix, I was able to run the Twilight Zone clock test (fast, forward) for over 15 minutes. It never failed, I only stopped it because I was getting tired of the constant beeping that occurs during the test.

Speaker Hum Solution

September 29, 2013

My Twilight Zone had a humming noise that was coming from the font left (when looking at the pin from the player position) speaker. Some pinball machines have a little hum, but this was much more noticeable. First I thought it might be bad connectors or grounding, but I quickly showed that wasn't the case by reseating everything that connected to the sound board and ensuring the mounting screws were tight. No change.

I then convinced myself the problem was with the sound board itself. After spending several nights looking at the Twilight Zone sound board with my scope, I couldn't find anything wrong. I was focusing on the +/- 50V that comes in and creates the +/- 25V which eventually goes through voltage regulators to create +/- 12V. These voltages are the ones used by the op-amps and the amp, so I thought I should focus there.

The WPC pre-DCS sound board supports 4 channels. They are called DAC, CH1, CH2, and CVSD. These channels are all summed though an op-amp, then pass through an attenuator (for volume) and finally go through the real amp. When I connected my scope to the amp output, I could see the hum was actually occurring at 120 Hz. This is shown in the below. The picture was taken when the game should be silent with the volume set at 10 or 11 (I don't recall).

Where was this 120 Hz wave coming from? I decided to check each of the channels. Before the channels are summed, each channel goes through an op-amp (CVSD actually has two op-amps in the path). Op-amps have two inputs, IN+ and IN-. Very informally, they amplify the difference between IN+ and IN-. For the WPC pre-DCS sound board, IN+ is always tied to ground and IN- is the actual source. For each channel, I shorted the associated IN+ and IN- together. Since this makes IN+ and IN- equal, the op-amp shouldn't do any amplification and should output a signal equivalent to ground (silence).

What I found was that doing this to DAC, CH1 and CH2 had no effect on the hum. However, doing this to either of the op-amps in the CVSD path caused the hum to go away! I worked my way back to the CVSD generation chip (a 55536). I immediately though it was interesting that this chip only used the 5V source and didn't use the 12V or 25V volt sources that I spent so much time checking out (and found nothing wrong). I hooked my scope up to the output of the CVSD and I could see a slight 120 Hz wave in the signal!

Initially I was worried because I thought the 55536 may be bad and wondered how easy it was to acquire a replacement. But I quickly decided to check out the 5V feeding the 55536. This is the same 5V that drives the CPU board and, since the game wasn't resetting, I expected the 5V to be good. I had previously measured the 5V with a DMM and it showed a solid 4.9V signal. Knowing this, I was quite surprised when my scope showed that the 5V was actually quite poor! Below is what the scope showed.

The 5V definitely had a 120 Hz "ring" and was not very solid at all. I took the driver board out of the Twilight Zone and checked BR2 and C5 for continuity ---they tested fine. Not having any better idea, even though the BR2 and C5 connectivity tested well, I decided they must not be connected well enough and decided to add the "Clay" jumpers. These are jumpers that directly connect the BR2 and C5 with new wires rather than relying on the traces on the PCB. After completing this modification, I put the driver board back and, upon powering Twilight Zone on, could immediately tell the hum was gone!

The scope supported my ears and showed that the hum was gone. The 5V signal looked MUCH better. A picture of the good 5V is shown below. Note that the X and Y scale is exactly the same between these two pictures.

Bottom line? If you have a 120 Hz hum on your Twilight Zone or other pre-DCS WPC pinball machine, you may want to check out your 5V supply and the associated BR2 and C5 connectivity.

(Note: While the hum has gone away, I'm surprised to see a close to 690 mV swing in the DC. I've been told this is normal and not to worry about it. I'm not so sure, but it's not causing any trouble so I'm not looking at it currently.)

Gumball VUK Broke

August 23, 2013

I noticed the gumball VUK was having a hard time shooting the pinball up into the gumball machine. There would be a "knock" and the pinball would pop up 2/3s of the way and go right back into the VUK. This would repeat a few times and eventually the pinball would make it into the gumball machine. Prior to this, everything was working fine so something had obviously just broke.

It turns out the ball popper cup (03-8561) and cracked into two pieces. I found one of them in the subway leading to the VUK and the other in the bottom of the cabinet. This isn't a part I had in my parts drawer, so I needed to order one (along with a spare). The image below shows the broken ball popper cup along with the new ball popper cup.

Replacing the ball popper cup is very easy. You need to remove the VUK mechanism, which only requires removing 4 hex screws. The image below shows the location where the ball popper cup should be attached. That screw head should hold the ball popper cup in place.

I unscrewed the ball popper cup screw, placed the new ball popper in place, and reattached with the screw. The image below shows how it should look when the ball popper cup is in place.

After reattaching the VUK mechanism to the playing using the 4 nut screws, I did a quick test and the VUK shot the pinballs into the gumball machine perfectly --- just like before the part broke.

Rebuilding GI connector (J120)

August 23, 2013

When I was checking out the Twilight Zone pinball machine before deciding to make the purchase, I noticed that nothing was connected to J120. This is one of the connectors for the G.I. (general illumination). After closer inspection, I realized the wires for J120 were going behind the driver board. Since the GI was working, I assumed some previous owner had soldered the wires directly to the driver board. It turns out my assumption was correct.

When I removed the driver board from my Twilight Zone, this is what I found.

I cut the wires off and rebuilt the connector (shown below). Now the GI is connected properly to J120.

J121 Burnt with Home Use

Original May 4, 2014

I noticed that the lower left playfield G.I. stand stopped working. At first I thought it may be a fuse, but I quickly realized that pin 1 of the GI connector had burnt. I wasn't expecting this because the previous owner had rebuilt the female end of J121. I thought it was good to go for a while. However, when GI connectors burn, both the female side and the header pins on the board should be replaced. Clearly the J121 header pins had never been replaced.

I rebuilt the J121 female connected and replaced both J120 and J121 headers. These should be good to go for a long time now. The rebuilt headers are shown below.

General Illumination will not Dim!

Original May 5, 2014

Bally/Williams WPC games have strands of lights known as general illumination (G.I. or just GI). Each strand is controllable, but not each individual bulb in the strand. For a specific stand, all lights will be off, on or dim. Dim means the bulb is not off but is also not at full brightness.

After rebuilding the GI headers on my Twilight Zone, I went into the diagnostics to test that GI. I was puzzled when the GI worked except for allowing the stands to dim. Each strand would correctly turn on and turn off, but when they should dim they would remain at full brightness. I couldn't think of any reason that rebuilding the GI connectors would cause this behavior and wondered whether the dimming had been broken for a while and I had never noticed. I started investigating under the assumption that the GI dimming had been broken for a while and I never noticed it.

The Twilight Zone manual and the WPC schematics do a good job showing how the GI dimming operates. A picture from the Twilight Zone manual (page 3-10) is below.

From this picture one can see the major players in allowing the GI to operate (including dimming) are the Zero Cross detection, Main CPU, LS374 Latch and the Triacs. I started investigating the zero cross detection.

AC (alternating current) has a natural cycle (60 Hz here in North America) which can be used to create an inexpensive timer. This timing information is provided to the WPC ASIC and the ASIC / Main CPU use it to determine how quickly to cycle AC power to the bulbs on the strands. Slower cycling means the bulbs are illuminate longer, so they are brighter. Longer cycles means the bulbs are off more and therefore are dimmer. This cycling occurs very quickly so your eyes don't see it when traditional incandescent bulbs are used. Note: if you have replaced your GI incandescent bulbs with LEDs, they react much quicker to voltage changes and you will likely see flicker.

The zero cross circuit works because the AC signal has one peak and one low each 60 Hz. The zero cross circuit generally detects when the peak and when the minimum occur, resulting in a 120 Hz timer since there is one peak and one minimum in the 60 Hz. An LM339 comparator is used as part of the zero cross detection circuitry. The WPC driver board has TP4 (test point 4) which allows easy access to the zero cross circuit's output. I connected my scope to TP4 and found the zero cross functionality was working just fine. See the picture below.

A 120 Hz signal has about 8.33ms between each peak. The scope measured between 8.2ms and 8.3ms. I happened to snap the picture when it was showing 8.2ms.

The Twilight Zone manual documentation mentions that the LS374 is part of the GI dimming circuitry, so I started looking at this part of the schematic and circuit. I quickly determined that a bad LS374 would likely cause serious GI problems, such as not turning on or off at all, rather than just preventing dimming. While there is some information on the internet that seems to state that replacing the LS374 can fix GI dimming issues, as of this writing I don't see how that can be the case. If you disagree, please don't hesitate to drop me an email.

Having ruled out the LS374, I started looking at the WPC CPU board. I verified that the WPC ASIC was receiving the zero cross signal from the driver board (the same signal available on TP4). I also verified that the TRIAC output would occasionally change, which indicated that the CPU board was reacting to the zero cross input in some form or another.

At this point everything I had looked at appeared to be functioning correctly, yet the GI simply wouldn't dim. Since I didn't want to randomly start replacing chips and hoping for the best, I decided to swap the Twilight Zone driver board with a driver board from my Terminator 2. This could rule out the driver board completely. When I put the driver board from the Terminator 2 into the Twilight Zone, the GI still wouldn't dim! This pretty much removed the driver board as a problem spot, but where was the real problem?

At this point I decided to posting a question over on Pinside, asking whether Twilight Zone will dim the GI in the test mode. I thought it was a crazy question, thinking that obviously it must, but I wanted to make sure I wasn't chasing my tail. A friendly Pinsider (kbliznick) responded that adjustment #25 ("Allow Dim Illum.") will prevent the GI from dimming during gameplay AND DURING TEST. When I read this, I thought "oh my... I think I may have disabled this option when I placed the LED alternative blinking bulbs in the backbox.".

I powered on the Twilight Zone, selected "Standard Adjustments" and went to option #25. What I saw is pictured below.

Sure enough, I had turned this feature off. (Note: it's funny the camera decided to show the NO very brightly, as it if was saying "you idiot." However, it's probably caused by capturing the image during the DMD refresh and not actually the camera expression an opinion.) Who would think that a feature would prevent the diagnostics from performing its stated job? I suspect this is actually a bug in the code. I changed the setting from No to Yes, as shown below.

With the "Allow Dim Illum." set to Yes, Twilight Zone tells me that this is the default factory setting. I then went to the GI test option and found the GI dimmed perfectly. I had spent several hours working on this issue only to find there was nothing wrong. Rather, my GI not dimming during test was an unexpected side effect of a setting. Still, I learned a bunch about the GI dimming during this process, so it wasn't wasted effort.

In summary: If your Twilight Zone will not dim its general illumination (GI) in both regular gameplay and GI test mode, there are several reasons this can happen. However, before starting to look for potential hardware problems, make sure that Standard Adjustment #25 ("Allow Dim Illum.") is not set to "No". The solution may be as simple as changing this setting to the default "Yes".

New Clock Housing (and clock orientation surprise)

August 18, 2013

My Twilight Zone clock casing looked ok, but had a certain dingy look no matter how much I cleaned and polished it. Cosmetically it was fair. Someone had previously drilled 3 holes in the top to help let the heat out from the incandescent bulbs. While I'm not a "pinball mod" guy, I decided to get a new Twilight Zone clock housing and decided the translucent blue would look very nice. I placed an order and was quite pleased with the new blue housing when it arrive. The image below shows the new blue casing next to the original casing.

While performing the housing swap, I realized something that had been wrong since I got my Twilight Zone. A previous owner had mounted the clock internals upside down. There is a rather large protrusion from the back of the clock. A little of this is the gears, but most of it is the motor. The motor should be on the bottom of the casing. In my case, however, it was at the top! See the image below.

The change casing projected turned into a little more. Now I was going to turn the mechanism right side up. This required adjusting the gears in the gear box so that noon was at the top rather than at the bottom. It wasn't a bad thing to do since the minute hand was slightly out of alignment. It always showed a minute or two ahead of where it should.

(Note: Hopefully I'll add a section about adjusting the Twilight Zone clock gears so that it shows exactly noon at power on)

Once everything was done and put back together, the clock casing looked great! The first image below shows that it now has the proper orientation (with the motor at the bottom). This is followed by a nice image of the clock in a darkened setting.

And here is one last picture showing more of the sides of the new Twilight Zone casing as well as some of the surrounding playfield.

Replacing the Slot Machine Target

May 18, 2013

The yellow target to the left (as you look at it) of the slot machine scoop is almost always bent toward the back of the playfield. The standard target simply wasn't strong enough to withstand being hit by pinballs over the years. While having the target bent back meant that it (probably) didn't register, a worse problem occurred. The pinball would hit the metallic side of the slot machine scoop and, eventually, break the scoop. It also dinged up the pinballs.

I can tell my scoop had broken in the past since it has been welded back together. It seems the welding done was good quality. Also, since much of the scoop isn't visible, no one would notice. However, my target was bent back causing the pinball it hit the scoop. I wanted to fix this, so I ordered the enhanced slot machine target from Pinball Resource. It is a sturdier design so it will not bend back. Below are some pictures of the enhanced target.

Replacing the target is a short 10 minute exercise. The target is slightly obscured by a light PCB. Removing this PCB requires only removing 3 nut screws. The target and the light PCB are shown below.

With the light PCB removed, the target is easily accessible. A picture of the old target with the light PCB removed is shown below.

The old target is held in place with two screws. Simply unscrew and pull the old target out. Put the new target in its place and reattached the screws. The new target is shown below.

Now desolder the wires from the old target and solder them to the new target. The Green/Blue wire goes to the center tab. The other wire goes to the tab without the bar on the diode. The soldered new target is shown below.

Reattached the small light PCB with the 3 screws and it's done. Here is the new target from on top of the playfield. When I play tested Twilight Zone after this repair, I immediately noticed that a ball hitting this new target may quick come right back. Be careful! The old target, which bent each time it was hit, would absorb some of the force. Now it appears the force is used to send the back right back at you.

Cleaning the Flippers

May 18, 2013

Twilight Zone has 4 flippers. While they were all structurally sound, they had "ground in" black rubber. I removed the flipper rubber and scrubbed them down with Goo Gone. After a little elbow grease, they looked quite good. A before and after picture are shown below.

Cleaning Gumball Subway

May 16, 2013

A subway is a ramp that is below the playfield. In Twilight Zone, when you shoot the right "gumball" orbit, the ball ends up below the playfield and uses the subway to get over to a VUK (vertical up kicker). The VUK shoots the ball into the Gumball machine.

Unlike many subways that I've seen, the gumball subway isn't a clear piece of plastic. Rather, it is tinted. Even knowing the subway was tinted, I could tell it was pretty dirty. Below is a picture of the subway.

Removing the gumball subway is very straight forward. There are 5 hex screws that need to be removed. Also, because the ramp is slightly captured by the VUK mechanism, the VUK mechanism must be removed. This only involves removing 4 hex screws. The subway has a microswitch that is connected, so you cannot move it far from the playfield. Even with the connected switch, there is enough slack to clean the subway. Below is the removed subway before cleaning.

I cleaned the subway and reattached everything. The picture below shows the cleaned (and tinted) gumball subway. This will definitely prevent pinballs from picking up dirt while in the way and spreading the dirt across the top of the playfield.

Replacing Backbox Blinking #545 Bulbs with LEDs

May 12, 2013

Twilight Zone uses 6 #545 bulbs in the backbox. #545 bulbs are just like #555 bulbs except that they blink by themselves with no CPU control. Another difference between #555 and #545 bulbs is that #545 bulbs are impossible to find in 2013. Bally/Williams was nice enough to use black bulb mounts to indicate where the blinking #545 bulbs should be placed. Below is a picture showing an empty black socket (where a #545 should go) as two regular white #555 sockets (with bulbs).

Since it's close to impossible to get #545 bulbs, Twilight Zone owners typically choose 1 of 3 solutions:

  1. Simply put #555 bulbs in the sockets. They have the same dimensions and consume the same power. This is a very simple solution, but you will not get that extra backbox attention to the various features that should be flashing.

  2. Do a small amount of electrical work and replace the #555 mounts with bayonet style mounts for #455 bulbs. These are effectively the same as the discontinued #545 bulbs but have a different mount. It's not a huge deal, but must be repeated 6 times

  3. Take advantage of the fact that Coin Taker has created an LED version of the #545 bulb. These run about $2 each with about $3 shipping, for a grand total of $15.

I decided my time was worth my than $15, so I went with the Coin Taker solution. I'm expecting these LED #545 replacements to last a lifetime. Below is a close up of the Coin Taker #545 LED bulb.

Below is a picture of the Coin Taker LED #545 bulb installed.

Finally, here is a picture of an illuminated LED #545 bulb.

No Audio at All!

May 11, 2013

After bringing Twilight Zone home, I plugged it in and found it had no audio! This was a big WTF moment. After pondering this problem, I was left asking myself if I ever heard audio when I checked it out before completing the trade. I really don't remember. I know I was very focused on the playfield, the electronics and the mechanisms. It's possible I totally glossed over an audio problem. Or, it's also quite possible that the audio broke while Twilight Zone was being transported. It's pretty common for something to break while moving a pin.

I opened up the backbox and checked the audio fuses. They looked good. I then opened up the playfield and looked down into the cabinet. There was another WTF moment when I realized there was no cabinet speaker. I had read in the past that the audio goes in series from the cabinet speaker to the backbox speakers and, if the audio isn't connected to the cabinet speaker, you don't hear anything.

The story ends well. After scratching my head a bit, I looked at the way back of the cabinet. There, in the darkness, was a speaker. It wasn't an original Bally/Williams speaker, but (apparently) a more advanced one. It looks like a previous owner had upgraded the cabinet speaker but never properly mounted it. At some point when Twilight Zone was moved, the speaker simply slid to the back and pull off one of the speaker connectors in the process.

I was able to move the speaker back to the proper place and connect the speaker while, which caused the audio to start working. The speaker is a little smaller in diameter than the original, so I have an outstanding task to build a mount or possibly replace it with a regular sized speaker.

The picture below shows the "Pyle Driver Pro" speaker that is in my cabinet. It's currently not mounted correctly, so it's just sitting there on the speaker grill.

Change Clock to use LEDs

May 17, 2013

The main reason Twilight Zone clocks break is because of the heat buildup. My Twilight Zone clock did not have any #86 bulbs inserted when I acquired it. I decided that I'd convert it to use LEDs rather than #86 bulbs. The conversion is fairly straightforward. First, disassemble the Twilight Zone clock mechanism. Then remove the #86 bulb holders and the diodes that are at D1, D2, D3 and D4. The original board is shown below followed by a picture with these parts removed.

Solder in 4 LEDs where the bulb sockets previously existed and then solder in 4 resistors where D1, D2, D3 and D4 were located. That's it!

Here is a picture of the clock with working illumination. Since I had never seen it illuminated before, I didn't know that the lights occasionally flash during some modes. Having a Twilight Zone clock that lights up does significantly improve the appearance of the clock and the playfield.

Switch 53 is Broken

May 8, 2013

Diags were reporting switch 53 was broken... after looking into it, I found it was simply not connected below the playfield.

Mail Order Widebody Glass

May 10, 2013

One improvement I always do to a pin that I've acquired is to get new playfield glass. It had been several years since I needed glass and I was surprised how expensive widebody glass has become. As of May 7th, 2013:

Couple of notes:

  1. My goal isn't to "ding" any store, so the more expensive options I've obscured.

  2. Web Store #1 looks really cheap initially ($43 a sheet), but they charge $20 to package and then over $30 to ship!

  3. pinballife.com charges $50 for a sheet and then just over $20 to ship

  4. Local Store #2 offered an additional 10% off if you follow them on twitter

What did I do? I ordered from pinballlife.com. I needed some parts anyway, which added just about $1 to shipping. I figured the shipping on the parts alone would be $8, so I can say $13 goes directly to the glass. $50 + $13 =$63, which is the cheapest price I could find. I don't need to spend gas money to pick it up either.

The glass arrived quickly (as quick or quicker than any of the local options would have taken) and was packaged very well. A couple pictures are shown below.

There is clearly some substantial savings possible if you do a group buy, but in my case I needed a single sheet and didn't want to wait (as well as organize) a local group purchase.

Twilight Zone ROM Version

May 10, 2013

My Twilight Zone came loaded with an old ROM, version L3. The L3 version is several releases out of date. There are 3 version to consider: 9.2, 9.4 and 9.4H. The 9.2 version seems to be the one recommended if you are still placing Twilight Zone on a route. Version 9.4 has several fixes, but Bally/Williams seems to steer operators away from it. The 9.4H ROM is the "home" version of 9.4. It adds a couple of features (such as the ability to pause the game), but unfortunately removes the ability to use quarters. I decided to try out the 9.4H ROM.

Below is the old L3 ROM followed by a picture of the DMD running the 9.4H ROM.

Ground Pin Broken Off Power Cable

May 10, 2013

Anyone who has ever got a pinball machine as almost certainly seen this:

Isn't that the nicest picture of an old grubby plug with a broken ground pin you have ever seen? I can only imagine that it is extremely common for operators to tear off the ground plug.

It was only after several years of fixing pins that I realized these AC power cables are easily replaced. They are the same cables that plug into a PC power supply. If you remove the 4 screws holding the cover in place (shown below), you can easily pull out the old cable and replace it with a new one.

Ugly Lockdown Bar

May 10, 2013

The lockdown bar on my Twilight Zone was pretty ugly. It had been welded at some time in the past and it showed through the top of the lockdown bar. While you cannot see it in the picture below, there is actually a hole there. I didn't want to bore you with a ton of lockdown bar pictures, but the bar also has problems on the other side as well as in the middle.

I purchased a new lockdown bar... It's purely cosmetic, but I (unfortunately) find a nasty lockdown bar to be very distracting.

Slingshot Mylar Removal

May 15, 2013

My Twilight Zone had mylar in front of the 2 slingshots. It was in good shape but I could never get it as shiny when polishing up the playfield. Because of this, I decided to remove it. I used my favorite method, the freeze method. The mylar came off fairly easily. As always, it left the glue behind and I needed to carefully scrape it off with a credit card and Goo Gone. The picture below shows the Goo Gone soaking for 10 minutes before I began scrapping it off with a credit card.

This pile of muck, shown below, is what the glue looks like when it is being scrapped off.

Below is the final result! This is after some polishing with Novus #2. Very sweet looking!

Bunch'o'Pictures

May 15, 2013

Here are a bunch of pictures of my Twilight Zone pinball machine. These pictures were taken after doing a basic shop job (I didn't remove any parts of the playfield, but I did remove Mylar, and I polished it up quite a bit). Unfortunately I had not cleaned the flippers when I took these pictures. Still, it's looking quite nice!

Skill Shot Scoop

Original May 15, 2013