Abstract:

This is a fighting game in which a player wins by activating an attack on the opponent with the most attack power. Game logic is done with op-amps. Audio needs to be implemented.

The circuit design was based on making a game loop that used 2d collision detection and a square wave formed by a button press.

The circuit was validated with an LTSPICE simulation. I built the project on a breadboard for testing the circuit and verified it.

Benefits of using an analog circuit for game logic.
+ Game can be preserved for a long time with maintenance & repair. Assuming parts available.
+ Easy repair. Replace easily available discrete components in board.
+ Upgrade video & audio modules without changing game logic.

Drawbacks of using analog circuit for game logic.
– Limited memory options. Analog circuit typically requires analog memory.
– A/D conversion bottleneck.
– Data transfer bottleneck from accessing A/D conversion results

Full journey of how the idea came to be is documented here. https://tinyurl.com/mrv8dzr5

How It Works:

This is a general overview of how the game logic circuit works.

Player Coordinates

Player coordinates are set by the joysticks. The vertical axis of the joystick sets the y value. The horizontal axis of the joystick sets the x value.

Collision Detection

Collision detection is observed by looking at whether the difference between player coordinates are the same or very little with the formula below. The formula is implemented with op-amps.

diffsum_result = | xdiff | + | ydiff |
Same coordinates: diffsum_result = 0 , Close-enough: diffsum_result = 0.1V.
Collision detection test. Collision detection signal goes to 2.7V when both joysticks are in the same position.

Player Attacks

Player attack involves tracking the button release of the button by taking the derivative of the square wave made by the button press. A differentiator circuit produces a negative exponential voltage RC discharge curve on the rising edge of the square wave and a positive exponential RC discharge curve that describe the attack value(player_it_value).

The difference between attack values of the players are measured and put into diff signal. Diff signal is positive if player 2 attack value is more than player 1 attack value, and negative if player 1 attack value is more than player 2 attack value.

Player with the biggest attack power has a certain switch activated indicating that they are “It” and have the biggest attack power. Player 1 becomes “It” and is indicated as having the biggest attack value in p1_it signal if the diff signal is negative. Player 2 becomes “It” and is indicated as having the biggest attack value in p2_it signal if the diff signal is positive.

If players are near enough and one player has a bigger attack than the other player, then, a winning player switch is activated. The switch is implemented using a NMOS transistor.

Rendering

The game is intended to be rendered on various systems e.g. PC, TV, oscilloscope. An example of a rendering that has been done is on the PC in which an stm32 microcontroller digitizes the analog output signals from the game(e.g. player coordinates, p1_it and p2_it status signals, etc.), passes these signals to PC via a USB-UART bridge, and then C++ code interprets the signals and draws what is happening in the game world.

The way rendering works on the PC is that it is split into 2 separate processes. The rendering of ADC values is done in the main thread and reading and storing ADC values from UART is done in a separate thread. Rendering and updating adc buffer must be done in separate threads or else rendering would slow down adc buffer update process which stalls rendering from drawing the latest adc buffer values. Mutex is used to lock the ADC buffer to give exclusive access to the ADC buffer when updating or reading from it.

Video Results