Teenage Mutant Ninja Turtles

This restoration was for a customer who had just come into possession of a Teenage Mutant Ninja Turtles (Data East, 1991) pinball machine from a family member who originally wanted to sell it. Luckily, he convinced them not to sell it and instead get it restored. Of course, I was happy to help, and was keen to get some more experience with Data East machines. Early 90s Data East machines are generally not known for their fantastic gameplay, but the games are simple and fun, which is all that was really needed in an arcade in the 1990s.

Initial condition report (click on sections below to view details)

Cabinet

Good condition overall.

• Timber in good condition.
• Side artwork in good condition.
• Backglass in good condition.
  
• Legs in average condition.
  

Above playfield

Average condition overall.

• Playfield was very dirty.
• Consumables (rubbers and lamps) dirty and in poor condition.
• Playfield artwork worn in some areas.
  
• Mylar peeling up and dirty in some areas.
  
• Plastics in good condition.
• Playfield mechanisms/toys in good condition.

Under playfield

Average condition overall.

• All mechanisms and assemblies in average condition.
• Consumables (coil sleeves, flipper parts) dirty and in poor condition.
  
• Ball shooter assembly broken.
  

Electrical

Poor condition overall.

• Machine would not boot up.
  
• Significant battery leakage on MPU board.
  
• Playfield and backbox wiring in good condition.

The major issue with this game off the bat was the battery leakage that was evident on the MPU board, which explained why the game was unable to boot. Otherwise, the game was in average condition physically, which wouldn't be too hard to fix up. As the game couldn't boot I wasn't able to test anything in terms of gameplay or mechanical function, but coil plungers seemed to move freely and there was no evidence of any coils having caught fire (always a good sign). The customer didn't want the game fully restored to as-new condition, but wanted it fixed to a point it was playable, and given a quick once-over shop job. So this restoration focused on the electronic issues first, followed by the playfield work.

Disassembly

Like most early Data East games, TMNT has a pretty simple playfield with just two ramps. The rest of the playfield is a single level with one subway (and reportedly the first use of a subway in the pinball industry). So, once the ramps and TURTLES light assembly are off, the rest of the game is quite easy to take apart. There's nothing to take particular care with when disassembling the playfield, apart from the tendency of tee nuts to destroy screw threads when you try to remove them. I found two broken tee nuts on this playfield, which were replaced. Below are some images of disassembly.

After disassembly, the game went through my standard restoration process to get it playing and looking like new. During the restoration process, I dealt with a number of issues, described below.
​
​Tips & Troubleshooting (click on sections below to view details)

Game would not boot

When turning the game on, the general illumination lamps would light up, lights on the MPU board in the backbox would flicker, and the "Teenage Mutant Ninja Turtles" startup jingle would play through the speakers. But that's all that happened. The display did not turn on, a game could not be started, and the machine seemed to hang in a perpetual half-booted up state.

As with any game that has been in storage for a long time, the first thing you need to do is check the condition of the boards in the backbox. Unfortunately, once I checked the MPU board, the problem causing the game not to boot became clear. The PIA led on the MPU stayed lit, whereas it should go out after a second or so according to the normal LED flash sequence listed in the first page of the manual. I looked over towards the 6821 PIA chips on the left side of the board and realised that this entire half of the board was affected by battery alkali leakage. The caustic fluid had dripped all the way down the board to the transistors on the bottom edge. There are some images of the damage below:

The batteries looked to be the original batteries installed at the factory and there was no expiry date or lot number printed on them to determine how old they actually were. After the damage was fully assessed, it was clear that it was not only the 6821 PIA which was affected, but a bunch of other components responsible for the game booting up and running. Unfortunately, with leakage this extensive, the most cost effective fix is to replace the entire board. This board may be repairable but it will never be as reliable as an undamaged board and there is no guarantee that corrosion abatement procedures will be 100% effective. Not to mention, this kind of damage would take at least 10 hours to properly repair. My advice to the customer was to buy a Rottendog replacement board, which he agreed to do.

After waiting a week or so for the board, it arrived. The MPU004 board from Rottendog is not plug-and-play for Teenage Mutant Ninja Turtles. This board was purchased sans processor and game ROMs, so these were transferred from the old board and were thankfully in good condition despite the alkali leakage affecting the processor socket. Ninja Turtles does not use a 512k game ROM, so the jumper at J5 had to be removed and a jumper at J4 installed. After that, the board was good to go.

Thankfully, the alkali leakage had not progressed to the connectors or other boards in the backbox, so these were safe to continue using. So the board was plugged back into the game and turned on. Success! The game booted and was able to start a game as normal. The booting issue was resolved.

New Rottendog MPU board installed.

Turboboost kicker not working

The turboboost (ball launch) kicker has a short but tumultuous history. It was used in a few of Data East's games including Checkpoint (Data East, 1991) and Batman (Data East, 1991). However, the kicker was poorly designed. The kicker arm was made to sit in front of the shooter rod, which would collide with the rear of the kicker arm, transferring energy to the ball and sending it up the shooter lane. In theory, this makes sense, but in practice the mechanism just isn't very reliable. What you often get is exactly what happened on this game - a loose assembly with incorrect screws from when someone had hamfistedly tried to reattach the assembly to the playfield. When the kicker arm tried to launch the ball, it would move only a fraction of a centimetre and couldn't move the ball at all.

Action Pinball has a great article with information and parts advice for replacing this troublesome assembly in a variety of Data East games. However, Teenage Mutant Ninja Turtles is an exception to the rule, and it is recommended to leave the assembly as originally intended as upgraded Stern parts do not fit on the Ninja Turtles playfield. So, I began by cleaning the coil assembly and getting rid of the incorrect screws. I filled the holes in the underside of the playfield with Selley's Knead It, and left it to harden. Once solid, the kicker arm and baseplate were realigned on the playfield so that the kicker arm was in the centre of the shooter lane. Then the assembly was reattached to the playfield with the correct screws.

I removed the shooter rod rubber tip, as it pushed the kicker arm too far forward when the shooter rod was at rest. This robbed the kicker of power when it tried to launch a ball automatically. The rubber tip really isn't needed here, as the shooter rod never contacts the ball. Instead, it hits the rear of the kicker arm, which doesn't need to be protected like the ball does. Check for smooth and reliable action of the kicker arm when you fix this assembly both when using the shooter rod as well as when the autolaunch coil is fired. It is important the kicker arm doesn't collide with the shooter housing.

Turboboost kicker assembly with incorrect screws and blown out holes.

Most flashers broken

Data East games use a lot of flashers. I don't quite know why, but they do. When I first tested all of the flashers I thought I may have had an issue on the board or with the wiring, as only a couple of them in the whole game actually worked. I've seen plenty of abused games with plenty of lamps out, but there are usually a fair few flashers still working. As it turns out, there was no weird issue, most of the flashers were just broken and in need of replacement. I used about thirty flashers in total on this game (a mix of 906s and 89s). In particular, the lamp board with eight flashers on it below the ramps had no working flashers at all. Then, there are four flashers behind a large red insert on the back panel attached to the playfield. There's another four flashers in one spot. So, whenever you work on a Data East game, make sure you have plenty of flashers to spare!

Check the flashers on the lamp board under the ramps - most of them will probably be dead!

No power to flippers

This was a pretty simple issue and one that is very common on Data East games. The flippers were not working at all, so I made sure that the flipper switches were working and that there were no loose wires anywhere. As it turns out, one of the flipper switches was not working, but this was due to an unrelated problem (see right flipper switch not registering section, below).

Instead, this problem was on the playfield power board. One of the fuse clips for fuse F5 had cracked and was no longer making good contact with the fuse. Therefore, no power was getting to the flippers. The fuse clips used on Data East games are of poor quality and crack easily. So if you begin to get intermittent coil or flasher problems, suspect the fuse clips! I was able to grab a small pack of new clips (Jaycar) to install into the board straight away. Once the new clips were in, power returned.

Broken fuse clip at F5 on the playfield power board.

Even though I now had power at the flippers, they were so weak that it was almost as if the flipper coils were not being supplied their full voltage. After the fuse clip replacement, they were definitely getting their correct voltage, but the worn flipper parts robbed the flippers of any power. I grabbed a flipper rebuild kit (RTBB) and installed it and the difference was night and day. The flippers suddenly had the power to make the ramp and orbit shots at the top of the playfield.

Freshly rebuilt flippers with new parts.

Knocker coil stuck on

After installing the new MPU board, I was keen to get the game playing again. Unfortunately I didn't get too far, because every time I turned the game on, the knocker coil would lock on. I checked the schematics and the knocker is controlled by a drive transistor on the MPU as well as a TIP36 transistor at Q4 on the playfield power board. I tested the transistor on the playfield power board and found that it had developed a short internally. I checked the diode and resistor associated with the transistor and they were both fine. Replacing the TIP36C was all that was needed to restore the knocker to normal.

Replacement of the TIP36 transistor at Q4.

New transistor installed.

But, that wasn't the end of the drama with Q4. Some time after I had fixed the transistor, the knocker locked on again. This time, the failure was more spectacular. Resistor R16, which is in series with transistor Q4 for the knocker coil, decided to catch fire. The aftermath looked pretty nasty, but it actually cleaned up really well. Once there is a failure this spectacular in the circuit, it is generally a good idea to replace all affected and nearby components as some of them may have been damaged by the event. Now it was time to install a new transistor at Q4, new resistor at R16, and for good measure, a new knocker coil.
​

Catastrophic failure of R16, resulting in fire!

Scorch mark at R16 after cleaning the board.

With the circuit entirely refurbished, the knocker started to work like a charm. 

String of general illumination lamps not working

This was an easy one. The general illumination lamps on the right side of the playfield would not light up even after I replaced all of the lamps. This was due to a failure of one of the general illumination fuses. Replacing the fuse brought the lights back to life. According to Pinwiki, problems with general illumination are more often related to burned connectors on the power supply board or the playfield power board, but both of these connectors looked totally fine. So if a fuse isn't the problem, suspect the connectors and header pins.

Sewer upkicker not consistently detecting balls

When balls are sent to the sewer, they roll down a subway ramp and into a super vertical upkicker (VUK) assembly (part no. 500-5320-00). Then they get shot up through the manhole cover and onto the playfield. I found that about 10% of the time, balls sent into the sewer were not getting kicked out. This would eventually trigger a ball search, and more often than not, the game would serve another ball into the shooter lane as well as ejecting the one in the sewer. This would confuse the game no end. The Data East ball search algorithm is really no good; it gives up and serves up another ball way too quickly.

Fixing this just required a simple adjustment of the switch on the super VUK assembly. The switch mounting screws cannot be moved around, so you need to bend the switch blade itself. I adjusted this for better contact with the underside of the VUK plunger cap and it started to register the ball 100% of the time.

Switch on the vertical upkicker assembly.

Right flipper switch not registering

Data East was the first company to introduce solid state flippers. The way they implemented the flipper circuitry is a little different to Gottlieb or Williams. The flippers are controlled by a solid state flipper control board (part no. 520-5033-00) mounted on the left side of the cabinet. The flippers are, in a way, separate to the rest of the MPU-controlled switches in the game. A press of the flipper button will trigger the relevant flipper regardless of whether that switch is connected properly to the MPU. That's because the flipper board provides the flipper coil's path to ground before the signal is sent to the MPU for processing.

This issue proved difficult to track down. The right flipper activated properly when the flipper button was pressed, according to the theory of operation described above. However, the right flipper switch was never actually being recognised by the MPU as a switch closure. As long as the flipper works, this shouldn't really matter, right? Wrong! This has the annoying effect of making it impossible to scroll right when selecting high score initials, and also making it impossible to scroll rightwards through the test menus.

Initially I suspected there may have been a fault with the Rottendog MPU board I had just installed. However, I manually jumpered the switch column and row associated with the right flipper (column 2, row 8) and could see the MPU register the flipper switch. This eliminated the MPU being faulty. I then tested the wiring for the right flipper switch and made sure there was good continuity from the switch, to the flipper board, to the MPU. There was a good connection. Next, I reflowed solder on the flipper board header pins and checked all of the fuses and fuse holders. These were all good.

At this point I was stumped. So, I checked all of the components on the flipper board. One of them gave odd values when testing - the transistor at Q5. I had a sneaking suspicion this transistor may have been involved in the right flipper switch circuit. According to the manual, it did! It was connected to switch drive 2, and switch return 8. Switch return 8 was where the right flipper button switch sat. I wanted to replace the transistor but couldn't find an exact replacement in the toolbox. I then discovered that Jaycar sell a catch-all PN100 NPN transistor that can replace 2N3904s and a bunch of other transistor types. I grabbed a few of these, installed one in Q5, and the right flipper started registering properly. Nice!

Q5 after replacement with a PN100 transistor.

Ball feed assembly did not serve balls into shooter lane

The ball feed assembly was not able to push balls from the trough into the shooter lane. The balls would get pushed upward a little, but not enough to get them over the raised playfield wood and onto the playfield surface. I tried actuating the ball feeder arm manually and found it a little stiff.

Despite what most people would tell you, there are actually some assemblies that you should lubricate with oil or grease. Any assembly with a pivot point (such as this ball feeder assembly) should be lubricated as a regular maintenance activity. While not mentioned in the Data East manual, the assembly is the same as the Williams ball release mechanism, so the same lubrication requirements apply. Below is the excerpt from a Williams manual:

After a few drops of 3-in-1 oil and working the oil in, the assembly started to loosen up nicely. Testing the ball feeder coil in coil test showed that it now had plenty of power to serve the ball into the shooter lane.

Reassembly

As I said at the start of this post, Ninja Turtles has a simple playfield so there aren't many tricks to reassembly. Note that the Michelangelo figurine is meant to sit on top of the bottom pop bumper cap and his legs are secured to it with two screws. Michelangelo on this machine had literally lost his legs - they had disconnected at the hip socket - from the vibrations of the pop bumper. His legs simply snap back into his body, so make sure they're tight.

A white rubber ring kit was installed to replace the broken and perished rubbers originally in the game. Note that you'll need two 4-inch rubber rings for the very back of the playfield, behind the pop bumpers. New flipper rubbers were also installed. New lamps replaced any broken ones, with the majority of broken bulbs being flashers under the playfield rather than controlled lamps or general illumination. I opted to stick with incandescent globes. The foam backings had perished from the standup target brackets, so these were replaced with small strips of adhesive weather strip.

Other than that, the playfield, plastics and ramps were given a good clean before being put back together. One ramp protector for the left ramp was found in the coin box and was reattached to the ramp with double-sided tape.

Conclusion

Working on Teenage Mutant Ninja Turtles was a good introduction to early Data East dot matrix display games. Some of the issues on this game were troublesome but none were insurmountable, and after all was said and done, the game finished up playing very well and looking good. The flippers were so weak you could barely make it up the ramps, but after a rebuild, they were as strong as ever. The game is much more fun to play when you can hit all of the shots! The sewer shot is advertised as the first ever subway shot in pinball on the game flyer, which is a pretty cool thing when you think about how ubiquitous subways became on pinball machines in the early 1990s. That said, not all Data East innovations were a good idea. That turboboost kicker is a horribly designed mechanism and it is no surprise it causes so many issues on other games. It was not fun to work on, but it is good to know that it can be fixed with some effort.

All in all, the customer was very happy with the end result and was eager to get the game home and finally give it a play after so many years. Cowabunga!