The Bi-Weekly British Backtrack – Disco Fever Pinball Restoration 05
by, 17-03-12 at 11:27 (2438 Views)
Welcome to part five in the thrilling serialisation of my electronic blunderings through the world of pinball repair. I do this so you don't have to.
My boards arrived, or should I say, a CPU board and a driver board arrived, not the same ones that left, but a CPU board and a driver board nonetheless, and I was eager to get them straight in my machine and fire it up, but there was a little work to do first. Scott had replaced my System3 CPU board with a System 6 CPU board, easily discernible immediately by the different placing of the battery holders on the board. This isn’t really an issue but you might recall that I mentioned one of the first things you should do when repairing a machine of this type is to move the batteries of the board completely and house them somewhere else in the backbox because the batteries can leak corrosive material over the CPU and driver board. Not only that but the corrosive fumes from the batteries can also corrode the ROM sockets and the 40 pin connector strip between the boards. This damage is not always obvious so it’s a good idea to just do this modification anyway and put them elsewhere in the backbox and simply run wires back to the original connectors.
That done I fitted both boards and crossed my fingers as I powered the machine on. Once a few seconds had passed I began to breathe again, but only slowly, as I was fully expecting to smell the burning metal or melting plastic any second, but luckily it didn’t come. The machine had booted cleanly and none of the coils had locked on, the machine had reset completely.
It seemed like a good idea to insert a coin and play a game then, so that’s just what I did, albeit a very limited game due in no small part to my utter incompetence at playing pinball but also due to that fact that some of the playfield items were not working as intended. More important than that though, it was a very quiet game because I did not yet have a working sound board so I’d not connected to the driver board and it wasn’t powered.
For now though, I was more concerned with the physical aspects of the game, some of which weren’t quite working as planned, most importantly of all both of the flippers were quite weak, one of them in particular, the left one, was so weak that it could barely lift when the ball was on it so something was clearly not right.
Bear in mind that about half of all Disco Fever machines (including mine) have curved flippers which are heavier than normal flippers, so I changed them for a pair of lighter, straight flippers to see how much of a difference that made, but with the left one it wasn’t much. Flippers problems aren’t normally very difficult to diagnose and the cause can be found fairly simply by looking at the symptoms.
If the flippers don't work at all then it is almost certainly a blown fuse, either fuse F4 on the power supply board or on earlier games a flipper fuse under the playfield. Once the fuses are checked then you can test for voltage at the flipper coil, if there is none at all then you need to start looking at the connectors on the driver board that drive the flippers, check for broken wires and check the connector that links the back box to the playfield.
If you have power getting to the flipper coils but the flippers still won’t fire then you should check the ground wires from the flippers to the playfield to make sure that you have a good grounding. Assuming you’ve already tested the coils and resistance much earlier then the coils themselves should still be sound.
If the flippers work but "flutter" and don’t stay up when the flipper button is held then usually the hold winding on the coil itself is broken. The hold winding is the thin wire that makes up the coil and if it is broken you can usually see it broken away from one of the coil solder lugs. It could also be that the EOS (End Of Stroke) switch is not adjusted properly.
If both flippers work but are equally weak then the most likely suspect is the bridge rectifier as it runs both flippers, but if only one flipper is weak then that points to one of two things, either mechanical binding caused by mal-adjustment of the mechanism or once again the EOS switches.
As the left flipper was far weaker than the right, it seemed reasonable to assume that the bridge rectifier had seen better days but that there was another issue with the left flipper. The best thing to do in this case is to rebuild the flipper assembly to make sure that there is no mechanical obstructions and that the mechanism can move freely and that’s what I did, stripped them both down completely and cleaned them. Bear in mind that you should never lubricate the flipper assemblies because the lube would attract dirt and before long jam the mechanism up rather than keep it moving freely. Mine had not been lubed but were very dirty so I gave them a good clean before putting them back together again, making sure that the coil and the coil sleeve are not distorted because of heat from the coils, and making sure that the inner lining of the coil is not distorted either.
Once it was put back together again the mechanisms moved freely enough when operated by hand but the left flipper still didn’t fire properly which pointed me towards the EOS switch.
The way that a flipper works is that the coil is powered by the power supply board with two different voltages and it switches between the two voltages during use. The power supply board sends both 30 volts and 6 volts and the flipper mechanism controls which voltage is used and at what time. To physically move the flipper from rest to an up position needs a relatively large voltage, so when you press the flipper button the coil is energized with 30 volts and the piston is pulled into the coil by electro magnetism and the flipper mechanism moves until it hits the adjustable stopper where the flipper should be at its furthest extent.
Once the flipper is at its furthest extent, it no longer needs 30 volts to hold it there because holding it in place requires much less power than firing it, so as the mechanism hits the stopper it also triggers the EOS switch which is a leaf switch that breaks the 30 volt supply and switches it to the 6 volt supply. If this wasn’t the case then the coil would constantly have 30 volts running through it and would get very hot and wear out much more quickly. As we’ve seen elsewhere there are always steps taken to try and reduce the heat generation inside a pinball machine, either by design from the word go or by making modifications later, such as fitting more economic light bulbs and fitting more efficient transistors.
The reason I suspected that my EOS switches were at fault is because what often happens is they become dirty or pitted and no longer make a good connection, so if the current doesn’t flow through them it is as though the switch is open and only 6 volts is ever used. This is obviously not enough to fire the flipper from rest to the up position and can affect either one flipper at a time or both, rather than the bridge rectifier which affects both flippers equally. Once again I stripped down the flipper mechanism on the left flipper and cleaned the EOS switches before re-assembling and trying again. Lo and behold, both flippers now flipped equally, a little weaker than I would like, and eventually I will replace the bridge rectifier to improve this, but for the time being they are perfectly usable and I could have a few more quiet games on my improving machine.
The last major component to get working was of course the sound board, and to that end I built a simple device to test it with.
Testing the sound board is relatively simple because to all intents and purposes it is a complete computer with its own RAM, ROM and CPU (a 6808) as well as a single 6821 PIA like those found on the other boards. It also has its own Power Supply Unit and takes 18 volts AC directly from the transformer which it rectifies to 5 volts DC for the IC logic circuit and 12 volts DC for the CPU reset circuit. The Williams sound boards have several test points on them where you can check that those voltages are present, and on mine, guess what? Not all of them were present. The Williams sound boards also have a test button on them, but on two of their games this test button does not work and as you might have guessed one of those games is Disco Fever, the other is Phoenix.
I knew the voltages were fine going to the board but they were not showing up on certain parts of the board so it was going to be one or more of the components or one or more of the traces that were at fault. You can also get these voltages from a normal PC Power Supply which I fitted to a board along with a speaker to test the outputs.
The ROM chips on a Williams sound board are numbered by type, either Williams' Sound ROM 1, Sound ROM 2 or Sound ROM 3, and these ROM chips don’t store the different sounds as digital samples or WAV files but as mathematical strings which have different parameters such as attack, frequency, decay and echo to determine the way they sound when triggered.
The way that they are triggered is the same way that the coils in the game are triggered, by a corresponding transistor pair on the driver board grounding them. You can manipulate this process in one of two ways, either by using diagnostic mode or firing them manually. Diagnostic mode on a Williams machine has various stages that fire different parts of the game in turn so you can tell if anything is not working as intended. You can also fire them manually in the same way that you can fire the solenoids, by momentarily grounding the correct TIP120 transistor which is just mimicking what the game does anyway.
To do that you need a short length of wire with one end connected to a good grounding point and with the other end you can touch the end of a transistor when you should hear either a solenoid activate or a sound fire. If your game has special solenoids controlling flash lamps then you may also get a lamp flash. Be careful not to ground the transistors for longer than is necessary to prevent them from overheating and causing damage.
Here are the transistors for each part of my game other than the dead sounds of course.
When I tried to trigger the sounds manually none of them worked, and when I tried diagnostic mode the same silence followed, but very occasionally when the game booted and I played it, I would get a single droning sound which seemed to suggest that the sound board had tried to boot but had not done so cleanly and was playing a corrupted sound. It also seemed to play it at random times that had no bearing on what was happening in the game. Even out of the game on the test board the sounds were not firing, so bearing in mind the state that the other boards in the game had been in, I took what you might call the easy way out at this point.
If I’m honest, I’d had just about enough of trying to fix annoying faults on the machine, and I considered going through the whole process again of testing ROMs, PIAs and CPUs and I saw only frustration ahead, and because I was so close to having a working machine I decided to let somebody else do the leg work again and I found a member of the UK Pinball Group to help me. The owner of the site www.robotron-2084.co.uk, Dave Langley, had a fully working sound board and would swap it for my old one along with some spare boards I had, so that’s what I did. Dave sent me his board that was ready to drop straight into my game.
Once it arrived it was a case of fingers crossed while I booted the game, and would you Adam and Eve it, but the sounds worked perfectly which meant that I finally had a fully operational chain of ICs and transistors;
- power transistor TIP120s
- 2n4401 pre-driver transistors
- 7408 logic ICs
- 6821 PIAs
Not to mention the sound board itself of course.
Tune in next time for the final part.