Pinball Project

Hi,

I'm a total newbie, so please forgive every forum convention I unintensionally break.
I have just made my two first posts, presenting myself and a little of my bigger project plan in this following thread:
http://arduino.cc/forum/index.php/topic,149747.0.html
Still I haven't even recieved my first Arduino units, but I have downloaded and tried the Windows IDE. So far, so good!

In order to restore as much functionality as possible to the old, broken Spanish Solid State pinball, I have decided to try to use one or probably more Arduino units together with some TTL circuits and the boards that still are somewhat functional on the game. See link above for more details.

The first step involving an Arduino unit will be controlling the behaviour of the triple drop target bank. I have written some sketch code that I don't know if it will work, probably not becuse I obviously can't test it before the Arduino boards have arrived. But at least it compiiles fine without warnings or errors in the IDE, but I'd like to have some input on the code to make sure I got it all right. So, here it is:

// Drop Target Sketch rev 001 2013-02-19

const int alldropsswPin = 2;     // All Drop Targets Down Switch Pin
const int addBonusPin =  11;      // Add Bonus Pin
const int score1000Pin =  12;      // Score 1000pts/Hi-pitch Beep Pin
const int dropsolenoidPin =  13;      // Reset Drop Target Bank Solenoid Pin


void setup()
{
  // initialize the All Drop Targets Down Switch pin as input:
  pinMode(alldropsswPin, INPUT);    
  // initialize the Add Bonus pin as output:
  pinMode(addBonusPin, OUTPUT);    
  // initialize the Score 1000 Points pin as output:
  pinMode(score1000Pin, OUTPUT);    
  // initialize Reset Drop Target Bank Solenoid pin as output:
  pinMode(dropsolenoidPin, OUTPUT);   

  // Set initial state on output pins
  // All activated/triggered on pin going LOW
  digitalWrite(score1000Pin, HIGH);
  digitalWrite(addBonusPin, HIGH);
  digitalWrite(dropsolenoidPin, HIGH);
}

void loop()
{
  if (alldown())
  {
    delay(200);
    // Score 1000 pts + add 1 bonus
    // TO DO: Disable the 3 individual Target Going Down Momentary Switches!!!
    digitalWrite(score1000Pin, LOW);
    digitalWrite(addBonusPin, LOW);
    delay(40);
    digitalWrite(score1000Pin, HIGH);
    digitalWrite(addBonusPin, HIGH);
    for(int i=0;i<4;i++)
    {
      delay(100);
      digitalWrite(score1000Pin, LOW);
      delay(40);
      digitalWrite(score1000Pin, HIGH);
    }
    delay(200);
    // Reset Drop Targets
    digitalWrite(dropsolenoidPin, LOW);
    delay(200);
    digitalWrite(dropsolenoidPin, HIGH);
    // TO DO: Enable the 3 individual Target Going Down Momentary Switches!!!
  }
    
}

boolean alldown() 
{
  // Check if the All Drop Targets Down Switch is closed (ie grounded)
  // If so, double-check if it's still closed after 50 ms
  // just to eliminate the risk of false triggering and other error sources
  // Return TRUE only if switch is closed at both times,
  // Else return FALSE
  if (digitalRead(alldropsswPin) == LOW)
  {
    delay(50);
    if (digitalRead(alldropsswPin) == LOW)
    {
      return (false);
    }
    else return(true);
  }
  else return(true);
}

Many thanks in advance,
/Tore

How are the switches (to be) wired to the Arduino? Since you are not using the internal pullup resistors, you'll need external resistors. Using the internal ones is so much easier.

50 milliseconds is a loooong time for a switch to bounce. 10 milliseconds is usually long enough for bouncing to die out.

In the long run you have no chance at all of getting code like this to work as it uses delay(). Look at the "blink without delay" example , and then look at FSM's (Finite State Machines) in the playground.

Mark

I should also have pointer you at the use of shift registers (and other chips) the extend the number of I/O ports you have.

Mark

Thank you guys for the input. Wow, you guys are fast!

How are the switches (to be) wired to the Arduino? Since you are not using the internal pullup resistors, you'll need external resistors. Using the internal ones is so much easier.

50 milliseconds is a loooong time for a switch to bounce. 10 milliseconds is usually long enough for bouncing to die out.

In most of the cases, there will be TTL AND or NAND gates between the switches and the Arduino. But not in the case of the All Drop Targets Down Switch, my guess is that if it makes any difference at all, I'd go for pulling the switch up towards the Arduino's 5 volt level, rather than the pinball game's.

Changed the debounce delay to 10 ms. Thanks. The other delays will most likely be subject to change, too. I just have to hear it to know if it produces the right "feel".

In the long run you have no chance at all of getting code like this to work as it uses delay(). Look at the "blink without delay" example , and then look at FSM's (Finite State Machines) in the playground.

Mark

I will look at it, right after I've finished writing this post. I think I've already forseen that "blink without delay" issue if I got it right, I'm not sure.
I think there's no chance of using only one Adrino for all the tasks, even if the ports and the memory are sufficient. So this unit will handle three of the four instances where there just have to be delays. (The forth is the two Bumpers, but I will use a dual copy of Gottlieb's System 80 Bumper Driver board there, and the switch signal for 100 points score will go to the other Arduino unit, the one without or with very short delays.)
This one will only control the two pits and the drop target bank. While it works with its delays, the ball will hit the slingshots and other targets That would have posed big troubles if the same unit was supposed to take care of those signals, too. But they will not be sent to this busy unit. And while the ball is in one of the two pits, of course no other inputs are made from the rest of the playfield. (Sorry, my English skills is not 100%, so sometimes I need ten words to express something that should take two words...)

I should also have pointer you at the use of shift registers (and other chips) the extend the number of I/O ports you have.

Mark

Shift registers are currently beyond my knowledge. I think I got the idea in general, but no details.
Don't know if I have the brain capacity to figure that out. But for now, I think the I/O ports of the Uno will be sufficient for this part of the project.

SimLego:
I think there's no chance of using only one Adrino for all the tasks, even if the ports and the memory are sufficient. So this unit will handle three of the four instances where there just have to be delays.

I'll take you up on that challenge. Without knowing much about this particular pinball or driver board, it is much less demanding than - say - drive a 3D printer (3 steppermotors, LCD display, Serial I/O, monitoring switches, heatbeds etc.)

there just have to be delays

There is the Two LED without delay which shows you can have several, totally independent "delays" running, that is wait a short time, without using delay()

I'll take you up on that challenge. Without knowing much about this particular pinball or driver board, it is much less demanding than - say - drive a 3D printer (3 steppermotors, LCD display, Serial I/O, monitoring switches, heatbeds etc.)

Now I see that you're most likely right, so I can't accept your challenge. :slight_smile: It is probably possible with an Uno and some workarounds (multiplexing, shift registers...), but it's going to be quite easy and straightforward with a Mega. I thought that this "blinking without delay()" was some low-level special coding, but now I see it's uncomplicated with the standard function millis().

The next trick is polling basically your code should look something like

void loop(){
     check each input and get there current state;
     if an input changed (the ball hit something) then do something about it.
}

You need to look at "debouncing" - basic when you turn on a switch it does not just go to the on state. It "bounces" doing a few on - off- on - off before it ends up in the on state.

Mark

Just for fun - how about a photo of this machine.

Mark

Just for fun - how about a photo of this machine.

I will take some photos along the work, meanwhile you can have a look at a few quite low-res ones at ipdb:

(Not my game, but same model.)

My Uno arrived today from HK, much earlier than expected. :smiley:
Installation and writing/uploading 'Hello World' was totally hassle free, guess it took about ten minutes. I'm already in love! Discovering the non-standard gap between digital pins 7 and 8 was the only unpleasant surprise so far. Makes it much harder to make DIY shields on prototype boards.

Now I have to study the Multiplexer/Driver Board once more to see how much is still useful. This will surely take some time...

SimLego:
My Uno arrived today from HK, much earlier than expected. :smiley:
Installation and writing/uploading 'Hello World' was totally hassle free, guess it took about ten minutes. I'm already in love! Discovering the non-standard gap between digital pins 7 and 8 was the only unpleasant surprise so far. Makes it much harder to make DIY shields on prototype boards.

Now I have to study the Multiplexer/Driver Board once more to see how much is still useful. This will surely take some time...

Yes the non-standard gap can be a hassle. It was just a simple rush to market last min error that didn't get caught when they released their very first version of the arduino board. And once release into the wild they didn't want to correct it as it would make useless any shield made prior if they corrected the spacing.

However there is a pretty simple fix, use a bent 'offset header' that will align up with standard .1" spaced proto-boards:

https://www.sparkfun.com/products/9374?

Lefty

Great idea, Lefty. Only problem is their international shipping costs are not of this world.

SimLego:
Great idea, Lefty. Only problem is their international shipping costs are not of this world.

Well you are free to bend your own up from standard long pin headers, now that you know the trick. Many E-bay sellers have free shipping. Perhaps setting up your location in your profile settings would help others know from where you are posting and others can tell of possible local suppliers?

Bad news. Seems like two out of four functions on the Multiplexer/Driver Board are dead. The solenoid drivers are ok, but neither the multiplexer to the switch matrix nor the multiplexer/driver for the Bonus lamps work; some of the inputs are shorted. Remains to see if the other lamp drivers apart from the bonus array work.

Well well, I'll begin with making a new Multiplexer board just for the switch matrix. It's very simple and straightforward. Just a 7445 and some 1K2 pullup resistors for the switch lines in use. I don't have a 7445, but I can't see why one of my 7442's shouldn't do the same job.

Here's some code I have in mind for scanning the switch matrix. I'll use the same addresses, just to make it easier to refer to the manual. No debounching or any response to input yet.

// scan_switches.ino

const int bit0pin = 2;  //  MA41 (Black)
const int bit1pin = 3;  //  MA43 (Blue)
const int bit2pin = 4;  //  MA45 (Green)
// const int bit3pin = 5;  //  MA47 (Yellow) // Not needed?
const int Apin = 6;     //  JC09 Slate-White   or MA48
const int Bpin = 7;     //  JC10 Yellow-White  or MA46
const int Cpin = 8;     //  JC11 Red-White     or MA44
const int Dpin = 9;     //  JC12 Brown-White   or MA42

int ret;

void setup()
{
  pinMode(bit0pin, OUTPUT);
  pinMode(bit1pin, OUTPUT);
  pinMode(bit2pin, OUTPUT);
 //  pinMode(bit3pin, OUTPUT);  // Not needed?
  pinMode(Apin, INPUT);
  pinMode(Bpin, INPUT);
  pinMode(Cpin, INPUT);
  pinMode(Dpin, INPUT);
}

void loop()
{
  while(1)
  {
    ret = scansw();
    if(ret == 0) continue;
    if(ret == 1)  // Drain Pit Switch closed
    {
       // code...
    }
    // if(ret == ...) and so on...
  }
}


int scansw()
{
  // Scan Line 0
  digitalWrite(bit0pin, LOW);
  digitalWrite(bit1pin, LOW);
  digitalWrite(bit2pin, LOW);
  //  digitalWrite(bit3pin, LOW);  // Not needed?
  if(digitalRead(Apin) == LOW); return(1); // Drain Pit
  if(digitalRead(Bpin) == LOW); return(2); // Top Target
  if(digitalRead(Cpin) == LOW); return(3); // 50k Target
  if(digitalRead(Dpin) == LOW); return(4); // Top Pit

  // Scan Line 3
  digitalWrite(bit0pin, HIGH);
  digitalWrite(bit1pin, HIGH);
  if(digitalRead(Bpin) == LOW); return(32); // 500p Rollovers
  if(digitalRead(Cpin) == LOW); return(33); // 30p Contacts

  // Scan Line 4
  digitalWrite(bit0pin, LOW);
  digitalWrite(bit1pin, LOW);
  digitalWrite(bit2pin, HIGH);
  if(digitalRead(Bpin) == LOW); return(42); // 5k Rollovers, Bonus
  if(digitalRead(Cpin) == LOW); return(43); // Rollovers x2

  // Scan Line 5
  digitalWrite(bit0pin, HIGH);
  if(digitalRead(Apin) == LOW); return(51); // Left Drop Target
  if(digitalRead(Bpin) == LOW); return(52); // Center Drop Target
  if(digitalRead(Cpin) == LOW); return(53); // Right Drop Target

  // Scan Line 6
  digitalWrite(bit0pin, LOW);
  digitalWrite(bit1pin, HIGH);
  if(digitalRead(Apin) == LOW); return(61); // Left Bumper
  if(digitalRead(Bpin) == LOW); return(62); // Right Bumper
  if(digitalRead(Cpin) == LOW); return(63); // Left Slingshot
  if(digitalRead(Dpin) == LOW); return(64); // Right Slingshot

  // Scan Line 7
  digitalWrite(bit0pin, HIGH);
  if(digitalRead(Cpin) == LOW); return(73); // 30k Rollover
  if(digitalRead(Dpin) == LOW); return(74); // Top Slingshot
}

Seems to be OK. A few "cosmetic" changes
loop() gets called continously. Thus the while(1) { ... } construct is totally superflous.
The brackets around the return value are not needed. http://arduino.cc/en/Reference/Return

In your main loop, instead of lots of if ( ret == n ) do something ; use the switch. Same functionality, code more clearly states what is happening (code readability).

void loop()
{
  switch( scansw() ) {  // omitted the "ret" variable, too. (Optional)
    case 0: continue; break ;
    case 1: DrainPitSwitchClose() ; break ;   // Drain Pit Switch closed
    case 2: Othercall() ; break ; 
    case 52: digitalWrite(13,HIGH) ; if (Counter>10) FlashBonus() ; break ; 
    case 21: somethingelse() ; break ;
   default: Serial.println("Something wrong - impossible value") ;
  }  // end switch
}

Note that case do not have to be in order, the default case is optional (but usefull); you can just call a procedure, but any sequence of statements is OK in a case (including a nested switch) - and most importantly : remember to end each case with a break

Thank you, Msquare.

loop() gets called continously. Thus the while(1) { ... } construct is totally superflous.

I used while(1) to be able to use break/continue.

I used if statements instead of switch in order to get a new switch - electric switch, that is... - scan as quickly as possible if no swicth was activated in the last scan. Not that I know if it makes any difference since I don't know how long it takes for each instruction to be executed. A switch statement may even be faster than 20 if statements, I don't know.

The brackets around the return value are not needed.

I didn't know that. Thanks again!


The Multiplexer board for scanning the switch matrix is ready now. Just have to restore some connections to the switches on the playfield, some disconnected by previous owner, some hacks made by me when I didn't realize how easy it (hopefully) is to scan a matrix, especially using an Arduino! Then it's time to test if it rellay works, not just in theory. Exciting!

Allright then. Time to find out if the theories work in practice.
To the left: my Arduino Uno board.
In the middle: my homemade mulitiplexer for the playfield switch matrix.
To the right: the original multiplexer/driver board. I'll probably use only the solenoid drivers there.

I have a feeling it won't work. First, there were separate grounds for the playfield and the logic. Maybe that have to be kept apart or the electronics may be interfered? Also, that are many chances to make mistakes, even though I think I've doublechecked every connection.

I'll give you a report in a couple of hours.

/SimLego

What are you worried about - noise on signal lines (causing the wrong signal to be read/sent) or a voltage spike (all those heavy solenoids) ruining you electronics or a fatal short (like getting 24V on a 5V line)?

The ground has to be common - else there is no meaning to voltage levels. This is true also with several power supplies.

Unless you choose to galvanically seperate the two, in which case all signals have to go through optocouplers, and of course you have seperate power supplies.

The spikes should already been absorbed by diodes (the driver board contains electronics I see). Spurious signal on the input side can be compensated for in the software, it is akin to debounce a switch. To guard against a fatal short you need galvanic isolation, but if you have nothing more than 24V then you "only" need to replace the AVR microchip (or some driver chip you may have) in "the unlikely event" of disaster

Total failure. Nothing works, and I have no idea why. This means debugging all night...