An I/O controller for virtual pinball machines: accelerometer nudge sensing, analog plunger input, button input encoding, LedWiz compatible output controls, and more.

Dependencies:   mbed FastIO FastPWM USBDevice

Fork of Pinscape_Controller by Mike R


This is Version 2 of the Pinscape Controller, an I/O controller for virtual pinball machines. (You can find the old version 1 software here.) Pinscape is software for the KL25Z that turns the board into a full-featured I/O controller for virtual pinball, with support for accelerometer-based nudging, a real plunger, button inputs, and feedback device control.

In case you haven't heard of the concept before, a "virtual pinball machine" is basically a video pinball simulator that's built into a real pinball machine body. A TV monitor goes in place of the pinball playfield, and a second TV goes in the backbox to serve as the "backglass" display. A third smaller monitor can serve as the "DMD" (the Dot Matrix Display used for scoring on newer machines), or you can even install a real pinball plasma DMD. A computer is hidden inside the cabinet, running pinball emulation software that displays a life-sized playfield on the main TV. The cabinet has all of the usual buttons, too, so it not only looks like the real thing, but plays like it too. That's a picture of my own machine to the right. On the outside, it's built exactly like a real arcade pinball machine, with the same overall dimensions and all of the standard pinball cabinet hardware.

A few small companies build and sell complete, finished virtual pinball machines, but I think it's more fun as a DIY project. If you have some basic wood-working skills and know your way around PCs, you can build one from scratch. The computer part is just an ordinary Windows PC, and all of the pinball emulation can be built out of free, open-source software. In that spirit, the Pinscape Controller is an open-source software/hardware project that offers a no-compromises, all-in-one control center for all of the unique input/output needs of a virtual pinball cabinet. If you've been thinking about building one of these, but you're not sure how to connect a plunger, flipper buttons, lights, nudge sensor, and whatever else you can think of, this project might be just what you're looking for.

You can find much more information about DIY Pin Cab building in general in the Virtual Cabinet Forum on Also visit my Pinscape Resources page for more about this project and other virtual pinball projects I'm working on.


  • Pinscape Release Builds: This page has download links for all of the Pinscape software. To get started, install and run the Pinscape Config Tool on your Windows computer. It will lead you through the steps for installing the Pinscape firmware on the KL25Z.
  • Config Tool Source Code. The complete C# source code for the config tool. You don't need this to run the tool, but it's available if you want to customize anything or see how it works inside.


The new Version 2 Build Guide is now complete! This new version aims to be a complete guide to building a virtual pinball machine, including not only the Pinscape elements but all of the basics, from sourcing parts to building all of the hardware.

You can also refer to the original Hardware Build Guide (PDF), but that's out of date now, since it refers to the old version 1 software, which was rather different (especially when it comes to configuration).

System Requirements

The new config tool requires a fairly up-to-date Microsoft .NET installation. If you use Windows Update to keep your system current, you should be fine. A modern version of Internet Explorer (IE) is required, even if you don't use it as your main browser, because the config tool uses some system components that Microsoft packages into the IE install set. I test with IE11, so that's known to work. IE8 doesn't work. IE9 and 10 are unknown at this point.

The Windows requirements are only for the config tool. The firmware doesn't care about anything on the Windows side, so if you can make do without the config tool, you can use almost any Windows setup.

Main Features

Plunger: The Pinscape Controller started out as a "mechanical plunger" controller: a device for attaching a real pinball plunger to the video game software so that you could launch the ball the natural way. This is still, of course, a central feature of the project. The software supports several types of sensors: a high-resolution optical sensor (which works by essentially taking pictures of the plunger as it moves); a slide potentionmeter (which determines the position via the changing electrical resistance in the pot); a quadrature sensor (which counts bars printed on a special guide rail that it moves along); and an IR distance sensor (which determines the position by sending pulses of light at the plunger and measuring the round-trip travel time). The Build Guide explains how to set up each type of sensor.

Nudging: The KL25Z (the little microcontroller that the software runs on) has a built-in accelerometer. The Pinscape software uses it to sense when you nudge the cabinet, and feeds the acceleration data to the pinball software on the PC. This turns physical nudges into virtual English on the ball. The accelerometer is quite sensitive and accurate, so we can measure the difference between little bumps and hard shoves, and everything in between. The result is natural and immersive.

Buttons: You can wire real pinball buttons to the KL25Z, and the software will translate the buttons into PC input. You have the option to map each button to a keyboard key or joystick button. You can wire up your flipper buttons, Magna Save buttons, Start button, coin slots, operator buttons, and whatever else you need.

Feedback devices: You can also attach "feedback devices" to the KL25Z. Feedback devices are things that create tactile, sound, and lighting effects in sync with the game action. The most popular PC pinball emulators know how to address a wide variety of these devices, and know how to match them to on-screen action in each virtual table. You just need an I/O controller that translates commands from the PC into electrical signals that turn the devices on and off. The Pinscape Controller can do that for you.

Expansion Boards

There are two main ways to run the Pinscape Controller: standalone, or using the "expansion boards".

In the basic standalone setup, you just need the KL25Z, plus whatever buttons, sensors, and feedback devices you want to attach to it. This mode lets you take advantage of everything the software can do, but for some features, you'll have to build some ad hoc external circuitry to interface external devices with the KL25Z. The Build Guide has detailed plans for exactly what you need to build.

The other option is the Pinscape Expansion Boards. The expansion boards are a companion project, which is also totally free and open-source, that provides Printed Circuit Board (PCB) layouts that are designed specifically to work with the Pinscape software. The PCB designs are in the widely used EAGLE format, which many PCB manufacturers can turn directly into physical boards for you. The expansion boards organize all of the external connections more neatly than on the standalone KL25Z, and they add all of the interface circuitry needed for all of the advanced software functions. The big thing they bring to the table is lots of high-power outputs. The boards provide a modular system that lets you add boards to add more outputs. If you opt for the basic core setup, you'll have enough outputs for all of the toys in a really well-equipped cabinet. If your ambitions go beyond merely well-equipped and run to the ridiculously extravagant, just add an extra board or two. The modular design also means that you can add to the system over time.

Expansion Board project page

Update notes

If you have a Pinscape V1 setup already installed, you should be able to switch to the new version pretty seamlessly. There are just a couple of things to be aware of.

First, the "configuration" procedure is completely different in the new version. Way better and way easier, but it's not what you're used to from V1. In V1, you had to edit the project source code and compile your own custom version of the program. No more! With V2, you simply install the standard, pre-compiled .bin file, and select options using the Pinscape Config Tool on Windows.

Second, if you're using the TSL1410R optical sensor for your plunger, there's a chance you'll need to boost your light source's brightness a little bit. The "shutter speed" is faster in this version, which means that it doesn't spend as much time collecting light per frame as before. The software actually does "auto exposure" adaptation on every frame, so the increased shutter speed really shouldn't bother it, but it does require a certain minimum level of contrast, which requires a certain minimal level of lighting. Check the plunger viewer in the setup tool if you have any problems; if the image looks totally dark, try increasing the light level to see if that helps.

New Features

V2 has numerous new features. Here are some of the highlights...

Dynamic configuration: as explained above, configuration is now handled through the Config Tool on Windows. It's no longer necessary to edit the source code or compile your own modified binary.

Improved plunger sensing: the software now reads the TSL1410R optical sensor about 15x faster than it did before. This allows reading the sensor at full resolution (400dpi), about 400 times per second. The faster frame rate makes a big difference in how accurately we can read the plunger position during the fast motion of a release, which allows for more precise position sensing and faster response. The differences aren't dramatic, since the sensing was already pretty good even with the slower V1 scan rate, but you might notice a little better precision in tricky skill shots.

Keyboard keys: button inputs can now be mapped to keyboard keys. The joystick button option is still available as well, of course. Keyboard keys have the advantage of being closer to universal for PC pinball software: some pinball software can be set up to take joystick input, but nearly all PC pinball emulators can take keyboard input, and nearly all of them use the same key mappings.

Local shift button: one physical button can be designed as the local shift button. This works like a Shift button on a keyboard, but with cabinet buttons. It allows each physical button on the cabinet to have two PC keys assigned, one normal and one shifted. Hold down the local shift button, then press another key, and the other key's shifted key mapping is sent to the PC. The shift button can have a regular key mapping of its own as well, so it can do double duty. The shift feature lets you access more functions without cluttering your cabinet with extra buttons. It's especially nice for less frequently used functions like adjusting the volume or activating night mode.

Night mode: the output controller has a new "night mode" option, which lets you turn off all of your noisy devices with a single button, switch, or PC command. You can designate individual ports as noisy or not. Night mode only disables the noisemakers, so you still get the benefit of your flashers, button lights, and other quiet devices. This lets you play late into the night without disturbing your housemates or neighbors.

Gamma correction: you can designate individual output ports for gamma correction. This adjusts the intensity level of an output to make it match the way the human eye perceives brightness, so that fades and color mixes look more natural in lighting devices. You can apply this to individual ports, so that it only affects ports that actually have lights of some kind attached.

IR Remote Control: the controller software can transmit and/or receive IR remote control commands if you attach appropriate parts (an IR LED to send, an IR sensor chip to receive). This can be used to turn on your TV(s) when the system powers on, if they don't turn on automatically, and for any other functions you can think of requiring IR send/receive capabilities. You can assign IR commands to cabinet buttons, so that pressing a button on your cabinet sends a remote control command from the attached IR LED, and you can have the controller generate virtual key presses on your PC in response to received IR commands. If you have the IR sensor attached, the system can use it to learn commands from your existing remotes.

Yet more USB fixes: I've been gradually finding and fixing USB bugs in the mbed library for months now. This version has all of the fixes of the last couple of releases, of course, plus some new ones. It also has a new "last resort" feature, since there always seems to be "just one more" USB bug. The last resort is that you can tell the device to automatically reboot itself if it loses the USB connection and can't restore it within a given time limit.

More Downloads

  • Custom VP builds: I created modified versions of Visual Pinball 9.9 and Physmod5 that you might want to use in combination with this controller. The modified versions have special handling for plunger calibration specific to the Pinscape Controller, as well as some enhancements to the nudge physics. If you're not using the plunger, you might still want it for the nudge improvements. The modified version also works with any other input controller, so you can get the enhanced nudging effects even if you're using a different plunger/nudge kit. The big change in the modified versions is a "filter" for accelerometer input that's designed to make the response to cabinet nudges more realistic. It also makes the response more subdued than in the standard VP, so it's not to everyone's taste. The downloads include both the updated executables and the source code changes, in case you want to merge the changes into your own custom version(s).

    Note! These features are now standard in the official VP releases, so you don't need my custom builds if you're using 9.9.1 or later and/or VP 10. I don't think there's any reason to use my versions instead of the latest official ones, and in fact I'd encourage you to use the official releases since they're more up to date, but I'm leaving my builds available just in case. In the official versions, look for the checkbox "Enable Nudge Filter" in the Keys preferences dialog. My custom versions don't include that checkbox; they just enable the filter unconditionally.
  • Output circuit shopping list: This is a saved shopping cart at with the parts needed to build one copy of the high-power output circuit for the LedWiz emulator feature, for use with the standalone KL25Z (that is, without the expansion boards). The quantities in the cart are for one output channel, so if you want N outputs, simply multiply the quantities by the N, with one exception: you only need one ULN2803 transistor array chip for each eight output circuits. If you're using the expansion boards, you won't need any of this, since the boards provide their own high-power outputs.
  • Cary Owens' optical sensor housing: A 3D-printable design for a housing/mounting bracket for the optical plunger sensor, designed by Cary Owens. This makes it easy to mount the sensor.
  • Lemming77's potentiometer mounting bracket and shooter rod connecter: Sketchup designs for 3D-printable parts for mounting a slide potentiometer as the plunger sensor. These were designed for a particular slide potentiometer that used to be available from an seller but is no longer listed. You can probably use this design as a starting point for other similar devices; just check the dimensions before committing the design to plastic.

Copyright and License

The Pinscape firmware is copyright 2014, 2021 by Michael J Roberts. It's released under an MIT open-source license. See License.

Warning to VirtuaPin Kit Owners

This software isn't designed as a replacement for the VirtuaPin plunger kit's firmware. If you bought the VirtuaPin kit, I recommend that you don't install this software. The VirtuaPin kit uses the same KL25Z microcontroller that Pinscape uses, but the rest of its hardware is different and incompatible. In particular, the Pinscape firmware doesn't include support for the IR proximity sensor used in the VirtuaPin plunger kit, so you won't be able to use your plunger device with the Pinscape firmware. In addition, the VirtuaPin setup uses a different set of GPIO pins for the button inputs from the Pinscape defaults, so if you do install the Pinscape firmware, you'll have to go into the Config Tool and reassign all of the buttons to match the VirtuaPin wiring.



File content as of revision 37:ed52738445fc:

/* Copyright 2014, 2015 M J Roberts, MIT License
* Permission is hereby granted, free of charge, to any person obtaining a copy of this software
* and associated documentation files (the "Software"), to deal in the Software without
* restriction, including without limitation the rights to use, copy, modify, merge, publish,
* distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
* The above copyright notice and this permission notice shall be included in all copies or
* substantial portions of the Software.

// The Pinscape Controller
// A comprehensive input/output controller for virtual pinball machines
// This project implements an I/O controller designed for use in custom-built virtual
// pinball cabinets.  It can handle nearly all of the functions involved in connecting 
// pinball simulation software on a Windows PC with devices in the cabinet, including
// input devices such as buttons and sensors, and output devices that generate visual
// or mechanical feedback during play, like lights, solenoids, and shaker motors.
// You can use one, some, or all of the functions, in any combination.  You can select
// options and configure the controller using a setup tool that runs on Windows.
// The main functions are:
//  - Nudge sensing, via the KL25Z's on-board accelerometer.  Nudging the cabinet
//    causes small accelerations that the accelerometer can detect; these are sent to
//    Visual Pinball (or other pinball emulator software) on the PC via the joystick
//    interface, using the X and Y axes.  VP and most other PC pinball emulators have 
//    native handling for this type of nudge input, so all you have to do is set some 
//    preferences in VP to let it know that an accelerometer is attached.
//  - Plunger position sensing, via a number of sensor options.  To use this feature,
//    you need to choose a sensor and set it up, connect the sensor electrically to 
//    the KL25Z, and configure the Pinscape software on the KL25Z to let it know how 
//    the sensor is hooked up.  The Pinscape software monitors the sensor and sends
//    readings to Visual Pinball via the joystick Z axis.  VP and other PC software
//    has native support for this type of input as well; as with the nudge setup,
//    you just have to set some options in VP to activate the plunger.
//    The Pinscape software supports optical sensors (the TAOS TSL1410R and TSL1412R 
//    linear sensor arrays) as well as slide potentiometers.  The specific equipment
//    that's supported, along with physical mounting and wiring details, can be found
//    in the Build Guide.
//    Note that while VP has its own built-in support for plunger devices like this
//    one, many existing VP tables will ignore it, because they use custom scripting 
//    that's only designed for keyboard plunger input.  The Build Guide has advice on
//    adjusting tables to add plunger support when necessary.
//    For best results, the plunger sensor should be calibrated.  The calibration
//    is stored in non-volatile memory on board the KL25Z, so it's only necessary
//    to do the calibration once, when you first install everything.  (You might
//    also want to re-calibrate if you physically remove and reinstall the CCD 
//    sensor or the mechanical plunger, since their alignment shift change slightly 
//    when you put everything back together.)  You can optionally install a
//    dedicated momentary switch or pushbutton to activate the calibration mode;
//    this is describe in the project documentation.  If you don't want to bother
//    with the extra button, you can also trigger calibration using the Windows 
//    setup software, which you can find on the Pinscape project page.
//    The calibration procedure is described in the project documentation.  Briefly,
//    when you trigger calibration mode, the software will scan the CCD for about
//    15 seconds, during which you should simply pull the physical plunger back
//    all the way, hold it for a moment, and then slowly return it to the rest
//    position.  (DON'T just release it from the retracted position, since that
//    let it shoot forward too far.  We want to measure the range from the park
//    position to the fully retracted position only.)
//  - Button input wiring.  24 of the KL25Z's GPIO ports are mapped as digital inputs
//    for buttons and switches.  The software reports these as joystick buttons when
//    it sends reports to the PC.  These can be used to wire physical pinball-style
//    buttons in the cabinet (e.g., flipper buttons, the Start button) and miscellaneous 
//    switches (such as a tilt bob) to the PC.  Visual Pinball can use joystick buttons
//    for input - you just have to assign a VP function to each button using VP's
//    keyboard options dialog.  To wire a button physically, connect one terminal of
//    the button switch to the KL25Z ground, and connect the other terminal to the
//    the GPIO port you wish to assign to the button.
//  - LedWiz emulation.  The KL25Z can appear to the PC as an LedWiz device, and will
//    accept and process LedWiz commands from the host.  The software can turn digital
//    output ports on and off, and can set varying PWM intensitiy levels on a subset
//    of ports.  (The KL25Z can only provide 6 PWM ports.  Intensity level settings on
//    other ports is ignored, so non-PWM ports can only be used for simple on/off
//    devices such as contactors and solenoids.)  The KL25Z can only supply 4mA on its
//    output ports, so external hardware is required to take advantage of the LedWiz
//    emulation.  Many different hardware designs are possible, but there's a simple
//    reference design in the documentation that uses a Darlington array IC to
//    increase the output from each port to 500mA (the same level as the LedWiz),
//    plus an extended design that adds an optocoupler and MOSFET to provide very
//    high power handling, up to about 45A or 150W, with voltages up to 100V.
//    That will handle just about any DC device directly (wtihout relays or other
//    amplifiers), and switches fast enough to support PWM devices.
//    The device can report any desired LedWiz unit number to the host, which makes
//    it possible to use the LedWiz emulation on a machine that also has one or more
//    actual LedWiz devices intalled.  The LedWiz design allows for up to 16 units
//    to be installed in one machine - each one is invidually addressable by its
//    distinct unit number.
//    The LedWiz emulation features are of course optional.  There's no need to 
//    build any of the external port hardware (or attach anything to the output 
//    ports at all) if the LedWiz features aren't needed.  Most people won't have
//    any use for the LedWiz features.  I built them mostly as a learning exercise,
//    but with a slight practical need for a handful of extra ports (I'm using the
//    cutting-edge 10-contactor setup, so my real LedWiz is full!).
//  - Enhanced LedWiz emulation with TLC5940 PWM controller chips.  You can attach
//    external PWM controller chips for controlling device outputs, instead of using
//    the limited LedWiz emulation through the on-board GPIO ports as described above. 
//    The software can control a set of daisy-chained TLC5940 chips, which provide
//    16 PWM outputs per chip.  Two of these chips give you the full complement
//    of 32 output ports of an actual LedWiz, and four give you 64 ports, which
//    should be plenty for nearly any virtual pinball project.  A private, extended
//    version of the LedWiz protocol lets the host control the extra outputs, up to
//    128 outputs per KL25Z (8 TLC5940s).  To take advantage of the extra outputs
//    on the PC side, you need software that knows about the protocol extensions,
//    which means you need the latest version of DirectOutput Framework (DOF).  VP
//    uses DOF for its output, so VP will be able to use the added ports without any
//    extra work on your part.  Older software (e.g., Future Pinball) that doesn't
//    use DOF will still be able to use the LedWiz-compatible protocol, so it'll be
//    able to control your first 32 ports (numbered 1-32 in the LedWiz scheme), but
//    older software won't be able to address higher-numbered ports.  That shouldn't
//    be a problem because older software wouldn't know what to do with the extra
//    devices anyway - FP, for example, is limited to a pre-defined set of outputs.
//    As long as you put the most common devices on the first 32 outputs, and use
//    higher numbered ports for the less common devices that older software can't
//    use anyway, you'll get maximum functionality out of software new and old.
// STATUS LIGHTS:  The on-board LED on the KL25Z flashes to indicate the current 
// device status.  The flash patterns are:
//    two short red flashes = the device is powered but hasn't successfully
//        connected to the host via USB (either it's not physically connected
//        to the USB port, or there was a problem with the software handshake
//        with the USB device driver on the computer)
//    short red flash = the host computer is in sleep/suspend mode
//    long yellow/green = everything's working, but the plunger hasn't
//        been calibrated; follow the calibration procedure described above.
//        This flash mode won't appear if the CCD has been disabled.  Note
//        that the device can't tell whether a CCD is physically attached;
//        if you don't have a CCD attached, you can set the appropriate option 
//        in config.h or use the  Windows config tool to disable the CCD 
//        software features.
//    alternating blue/green = everything's working, and the plunger has
//        been calibrated
// USB PROTOCOL:  please refer to USBProtocol.h for details on the USB
// message protocol.

#include "mbed.h"
#include "math.h"
#include "USBJoystick.h"
#include "MMA8451Q.h"
#include "tsl1410r.h"
#include "FreescaleIAP.h"
#include "crc32.h"
#include "TLC5940.h"
#include "74HC595.h"
#include "nvm.h"
#include "plunger.h"
#include "ccdSensor.h"
#include "potSensor.h"
#include "nullSensor.h"

#include "config.h"

// ---------------------------------------------------------------------------
// utilities

// number of elements in an array
#define countof(x) (sizeof(x)/sizeof((x)[0]))

// floating point square of a number
inline float square(float x) { return x*x; }

// floating point rounding
inline float round(float x) { return x > 0 ? floor(x + 0.5) : ceil(x - 0.5); }

// --------------------------------------------------------------------------
// USB product version number
const uint16_t USB_VERSION_NO = 0x0008;

// --------------------------------------------------------------------------
// Joystick axis report range - we report from -JOYMAX to +JOYMAX
#define JOYMAX 4096

// ---------------------------------------------------------------------------
// On-board RGB LED elements - we use these for diagnostic displays.
// Note that LED3 (the blue segment) is hard-wired on the KL25Z to PTD1,
// so PTD1 shouldn't be used for any other purpose (e.g., as a keyboard
// input or a device output).  (This is kind of unfortunate in that it's 
// one of only two ports exposed on the jumper pins that can be muxed to 
// SPI0 SCLK.  This effectively limits us to PTC5 if we want to use the 
// SPI capability.)
DigitalOut ledR(LED1), ledG(LED2), ledB(LED3);

// ---------------------------------------------------------------------------
// Wire protocol value translations.  These translate byte values from
// the USB protocol to local native format.

// unsigned 16-bit integer 
inline uint16_t wireUI16(const uint8_t *b)
    return b[0] | ((uint16_t)b[1] << 8);

inline int16_t wireI16(const uint8_t *b)
    return (int16_t)wireUI16(b);

inline uint32_t wireUI32(const uint8_t *b)
    return b[0] | ((uint32_t)b[1] << 8) | ((uint32_t)b[2] << 16) | ((uint32_t)b[3] << 24);

inline int32_t wireI32(const uint8_t *b)
    return (int32_t)wireUI32(b);

inline PinName wirePinName(int c)
    static const PinName p[] =  {
        NC,    PTA1,  PTA2,  PTA4,  PTA5,  PTA12, PTA13, PTA16, PTA17, PTB0,   // 0-9
        PTB1,  PTB2,  PTB3,  PTB8,  PTB9,  PTB10, PTB11, PTC0,  PTC1,  PTC2,   // 10-19
        PTC3,  PTC4,  PTC5,  PTC6,  PTC7,  PTC8,  PTC9,  PTC10, PTC11, PTC12,  // 20-29
        PTC13, PTC16, PTC17, PTD0,  PTD1,  PTD2,  PTD3,  PTD4,  PTD5,  PTD6,   // 30-39
        PTD7,  PTE0,  PTE1,  PTE2,  PTE3,  PTE4,  PTE5,  PTE20, PTE21, PTE22,  // 40-49 
        PTE23, PTE29, PTE30, PTE31                                             // 50-53
    return (c < countof(p) ? p[c] : NC);

// ---------------------------------------------------------------------------
// LedWiz emulation, and enhanced TLC5940 output controller
// There are two modes for this feature.  The default mode uses the on-board
// GPIO ports to implement device outputs - each LedWiz software port is
// connected to a physical GPIO pin on the KL25Z.  The KL25Z only has 10
// PWM channels, so in this mode only 10 LedWiz ports will be dimmable; the
// rest are strictly on/off.  The KL25Z also has a limited number of GPIO
// ports overall - not enough for the full complement of 32 LedWiz ports
// and 24 VP joystick inputs, so it's necessary to trade one against the
// other if both features are to be used.
// The alternative, enhanced mode uses external TLC5940 PWM controller
// chips to control device outputs.  In this mode, each LedWiz software
// port is mapped to an output on one of the external TLC5940 chips.
// Two 5940s is enough for the full set of 32 LedWiz ports, and we can
// support even more chips for even more outputs (although doing so requires
// breaking LedWiz compatibility, since the LedWiz USB protocol is hardwired
// for 32 outputs).  Every port in this mode has full PWM support.

// Current starting output index for "PBA" messages from the PC (using
// the LedWiz USB protocol).  Each PBA message implicitly uses the
// current index as the starting point for the ports referenced in
// the message, and increases it (by 8) for the next call.
static int pbaIdx = 0;

// Generic LedWiz output port interface.  We create a cover class to 
// virtualize digital vs PWM outputs, and on-board KL25Z GPIO vs external 
// TLC5940 outputs, and give them all a common interface.  
class LwOut
    // Set the output intensity.  'val' is 0.0 for fully off, 1.0 for
    // fully on, and fractional values for intermediate intensities.
    virtual void set(float val) = 0;

// LwOut class for virtual ports.  This type of port is visible to
// the host software, but isn't connected to any physical output.
// This can be used for special software-only ports like the ZB
// Launch Ball output, or simply for placeholders in the LedWiz port
// numbering.
class LwVirtualOut: public LwOut
    LwVirtualOut() { }
    virtual void set(float val) { }

// Active Low out.  For any output marked as active low, we layer this
// on top of the physical pin interface.  This simply inverts the value of
// the output value, so that 1.0 means fully off and 0.0 means fully on.
class LwInvertedOut: public LwOut
    LwInvertedOut(LwOut *o) : out(o) { }
    virtual void set(float val) { out->set(1.0 - val); }
    LwOut *out;

// The TLC5940 interface object.  We'll set this up with the port 
// assignments set in config.h.
TLC5940 *tlc5940 = 0;
void init_tlc5940(Config &cfg)
    if (cfg.tlc5940.nchips != 0)
        tlc5940 = new TLC5940(cfg.tlc5940.sclk, cfg.tlc5940.sin, cfg.tlc5940.gsclk,
            cfg.tlc5940.blank, cfg.tlc5940.xlat, cfg.tlc5940.nchips);

// LwOut class for TLC5940 outputs.  These are fully PWM capable.
// The 'idx' value in the constructor is the output index in the
// daisy-chained TLC5940 array.  0 is output #0 on the first chip,
// 1 is #1 on the first chip, 15 is #15 on the first chip, 16 is
// #0 on the second chip, 32 is #0 on the third chip, etc.
class Lw5940Out: public LwOut
    Lw5940Out(int idx) : idx(idx) { prv = -1; }
    virtual void set(float val)
        if (val != prv)
           tlc5940->set(idx, (int)((prv = val) * 4095));
    int idx;
    float prv;

// 74HC595 interface object.  Set this up with the port assignments in
// config.h.
HC595 *hc595 = 0;

// initialize the 74HC595 interface
void init_hc595(Config &cfg)
    if (cfg.hc595.nchips != 0)
        hc595 = new HC595(cfg.hc595.nchips, cfg.hc595.sin, cfg.hc595.sclk, cfg.hc595.latch, cfg.hc595.ena);

// LwOut class for 74HC595 outputs.  These are simple digial outs.
// The 'idx' value in the constructor is the output index in the
// daisy-chained 74HC595 array.  0 is output #0 on the first chip,
// 1 is #1 on the first chip, 7 is #7 on the first chip, 8 is
// #0 on the second chip, etc.
class Lw595Out: public LwOut
    Lw595Out(int idx) : idx(idx) { prv = -1; }
    virtual void set(float val)
        if (val != prv)
           hc595->set(idx, (prv = val) == 0.0 ? 0 : 1);
    int idx;
    float prv;

// Default LedWiz mode - using on-board GPIO ports.  In this mode, we
// assign a KL25Z GPIO port to each LedWiz output.  We have to use a
// mix of PWM-capable and Digital-Only ports in this configuration, 
// since the KL25Z hardware only has 10 PWM channels, which isn't
// enough to fill out the full complement of 32 LedWiz outputs.

// LwOut class for a PWM-capable GPIO port
class LwPwmOut: public LwOut
    LwPwmOut(PinName pin) : p(pin) { prv = -1; }
    virtual void set(float val) 
        if (val != prv)
            p.write(prv = val); 
    PwmOut p;
    float prv;

// LwOut class for a Digital-Only (Non-PWM) GPIO port
class LwDigOut: public LwOut
    LwDigOut(PinName pin) : p(pin) { prv = -1; }
    virtual void set(float val) 
         if (val != prv)
            p.write((prv = val) == 0.0 ? 0 : 1); 
    DigitalOut p;
    float prv;

// Array of output physical pin assignments.  This array is indexed
// by LedWiz logical port number - lwPin[n] is the maping for LedWiz
// port n (0-based).  
// Each pin is handled by an interface object for the physical output 
// type for the port, as set in the configuration.  The interface 
// objects handle the specifics of addressing the different hardware
// types (GPIO PWM ports, GPIO digital ports, TLC5940 ports, and
// 74HC595 ports).
static int numOutputs;
static LwOut **lwPin;

// Number of LedWiz emulation outputs.  This is the number of ports
// accessible through the standard (non-extended) LedWiz protocol
// messages.  The protocol has a fixed set of 32 outputs, but we
// might have fewer actual outputs.  This is therefore set to the
// lower of 32 or the actual number of outputs.
static int numLwOutputs;

// Current absolute brightness level for an output.  This is a float
// value from 0.0 for fully off to 1.0 for fully on.  This is the final
// derived value for the port.  For outputs set by LedWiz messages, 
// this is derived from the LedWiz state, and is updated on each pulse 
// timer interrupt for lights in flashing states.  For outputs set by 
// extended protocol messages, this is simply the brightness last set.
static float *outLevel;

// initialize the output pin array
void initLwOut(Config &cfg)
    // Count the outputs.  The first disabled output determines the
    // total number of ports.
    numOutputs = MAX_OUT_PORTS;
    int i;
    for (i = 0 ; i < MAX_OUT_PORTS ; ++i)
        if (cfg.outPort[i].typ == PortTypeDisabled)
            numOutputs = i;
    // the real LedWiz protocol can access at most 32 ports, or the
    // actual number of outputs, whichever is lower
    numLwOutputs = (numOutputs < 32 ? numOutputs : 32);
    // allocate the pin array
    lwPin = new LwOut*[numOutputs];    
    // Allocate the current brightness array.
    outLevel = new float[numOutputs < 32 ? 32 : numOutputs];
    // create the pin interface object for each port
    for (i = 0 ; i < numOutputs ; ++i)
        // get this item's values
        int typ = cfg.outPort[i].typ;
        int pin = cfg.outPort[i].pin;
        int flags = cfg.outPort[i].flags;
        int activeLow = flags & PortFlagActiveLow;

        // create the pin interface object according to the port type        
        switch (typ)
        case PortTypeGPIOPWM:
            // PWM GPIO port
            lwPin[i] = new LwPwmOut(wirePinName(pin));
        case PortTypeGPIODig:
            // Digital GPIO port
            lwPin[i] = new LwDigOut(wirePinName(pin));
        case PortTypeTLC5940:
            // TLC5940 port
            lwPin[i] = new Lw5940Out(pin);
        case PortType74HC595:
            // 74HC595 port
            lwPin[i] = new Lw595Out(pin);

        case PortTypeVirtual:
            // virtual or unknown
            lwPin[i] = new LwVirtualOut();
        // if it's Active Low, layer an inverter
        if (activeLow)
            lwPin[i] = new LwInvertedOut(lwPin[i]);

        // turn it off initially      

// LedWiz output states.
// The LedWiz protocol has two separate control axes for each output.
// One axis is its on/off state; the other is its "profile" state, which
// is either a fixed brightness or a blinking pattern for the light.
// The two axes are independent.
// Note that the LedWiz protocol can only address 32 outputs, so the
// wizOn and wizVal arrays have fixed sizes of 32 elements no matter
// how many physical outputs we're using.

// on/off state for each LedWiz output
static uint8_t wizOn[32];

// Profile (brightness/blink) state for each LedWiz output.  If the
// output was last updated through an LedWiz protocol message, it
// will have one of these values:
//   0-48 = fixed brightness 0% to 100%
//   129 = ramp up / ramp down
//   130 = flash on / off
//   131 = on / ramp down
//   132 = ramp up / on
// Special value 255:  If the output was updated through the 
// extended protocol, we'll set the wizVal entry to 255, which has 
// no meaning in the LedWiz protocol.  This tells us that the value 
// in outLevel[] was set directly from the extended protocol, so it 
// shouldn't be derived from wizVal[].
static uint8_t wizVal[32] = {
    48, 48, 48, 48, 48, 48, 48, 48,
    48, 48, 48, 48, 48, 48, 48, 48,
    48, 48, 48, 48, 48, 48, 48, 48,
    48, 48, 48, 48, 48, 48, 48, 48

// LedWiz flash speed.  This is a value from 1 to 7 giving the pulse
// rate for lights in blinking states.
static uint8_t wizSpeed = 2;

// Current LedWiz flash cycle counter.
static uint8_t wizFlashCounter = 0;

// Get the current brightness level for an LedWiz output.
static float wizState(int idx)
    // if the output was last set with an extended protocol message,
    // use the value set there, ignoring the output's LedWiz state
    if (wizVal[idx] == 255)
        return outLevel[idx];
    // if it's off, show at zero intensity
    if (!wizOn[idx])
        return 0;

    // check the state
    uint8_t val = wizVal[idx];
    if (val <= 48)
        // PWM brightness/intensity level.  Rescale from the LedWiz
        // 0..48 integer range to our internal PwmOut 0..1 float range.
        // Note that on the actual LedWiz, level 48 is actually about
        // 98% on - contrary to the LedWiz documentation, level 49 is 
        // the true 100% level.  (In the documentation, level 49 is
        // simply not a valid setting.)  Even so, we treat level 48 as
        // 100% on to match the documentation.  This won't be perfectly
        // ocmpatible with the actual LedWiz, but it makes for such a
        // small difference in brightness (if the output device is an
        // LED, say) that no one should notice.  It seems better to
        // err in this direction, because while the difference in
        // brightness when attached to an LED won't be noticeable, the
        // difference in duty cycle when attached to something like a
        // contactor *can* be noticeable - anything less than 100%
        // can cause a contactor or relay to chatter.  There's almost
        // never a situation where you'd want values other than 0% and
        // 100% for a contactor or relay, so treating level 48 as 100%
        // makes us work properly with software that's expecting the
        // documented LedWiz behavior and therefore uses level 48 to
        // turn a contactor or relay fully on.
        return val/48.0;
    else if (val == 49)
        // 49 is undefined in the LedWiz documentation, but actually
        // means 100% on.  The documentation says that levels 1-48 are
        // the full PWM range, but empirically it appears that the real
        // range implemented in the firmware is 1-49.  Some software on
        // the PC side (notably DOF) is aware of this and uses level 49
        // to mean "100% on".  To ensure compatibility with existing 
        // PC-side software, we need to recognize level 49.
        return 1.0;
    else if (val == 129)
        //   129 = ramp up / ramp down
        return wizFlashCounter < 128 
            ? wizFlashCounter/128.0 
            : (256 - wizFlashCounter)/128.0;
    else if (val == 130)
        //   130 = flash on / off
        return wizFlashCounter < 128 ? 1.0 : 0.0;
    else if (val == 131)
        //   131 = on / ramp down
        return wizFlashCounter < 128 ? 1.0 : (255 - wizFlashCounter)/128.0;
    else if (val == 132)
        //   132 = ramp up / on
        return wizFlashCounter < 128 ? wizFlashCounter/128.0 : 1.0;
        // Other values are undefined in the LedWiz documentation.  Hosts
        // *should* never send undefined values, since whatever behavior an
        // LedWiz unit exhibits in response is accidental and could change
        // in a future version.  We'll treat all undefined values as equivalent 
        // to 48 (fully on).
        return 1.0;

// LedWiz flash timer pulse.  This fires periodically to update 
// LedWiz flashing outputs.  At the slowest pulse speed set via
// the SBA command, each waveform cycle has 256 steps, so we
// choose the pulse time base so that the slowest cycle completes
// in 2 seconds.  This seems to roughly match the real LedWiz
// behavior.  We run the pulse timer at the same rate regardless
// of the pulse speed; at higher pulse speeds, we simply use
// larger steps through the cycle on each interrupt.  Running
// every 1/127 of a second = 8ms seems to be a pretty light load.
Timeout wizPulseTimer;
#define WIZ_PULSE_TIME_BASE  (1.0/127.0)
static void wizPulse()
    // increase the counter by the speed increment, and wrap at 256
    wizFlashCounter += wizSpeed;
    wizFlashCounter &= 0xff;
    // if we have any flashing lights, update them
    int ena = false;
    for (int i = 0 ; i < numLwOutputs ; ++i)
        if (wizOn[i])
            uint8_t s = wizVal[i];
            if (s >= 129 && s <= 132)
                ena = true;

    // Set up the next timer pulse only if we found anything flashing.
    // To minimize overhead from this feature, we only enable the interrupt
    // when we need it.  This eliminates any performance penalty to other
    // features when the host software doesn't care about the flashing 
    // modes.  For example, DOF never uses these modes, so there's no 
    // need for them when running Visual Pinball.
    if (ena)
        wizPulseTimer.attach(wizPulse, WIZ_PULSE_TIME_BASE);

// Update the physical outputs connected to the LedWiz ports.  This is 
// called after any update from an LedWiz protocol message.
static void updateWizOuts()
    // update each output
    int pulse = false;
    for (int i = 0 ; i < numLwOutputs ; ++i)
        pulse |= (wizVal[i] >= 129 && wizVal[i] <= 132);
    // if any outputs are set to flashing mode, and the pulse timer
    // isn't running, turn it on
    if (pulse)
        wizPulseTimer.attach(wizPulse, WIZ_PULSE_TIME_BASE);
    // flush changes to 74HC595 chips, if attached
    if (hc595 != 0)
// ---------------------------------------------------------------------------
// Button input

// button state
struct ButtonState
    ButtonState() : di(NULL), pressed(0), t(0), js(0), keymod(0), keycode(0) { }
    // DigitalIn for the button
    DigitalIn *di;

    // current on/off state
    int pressed;
    // Sticky time remaining for current state.  When a
    // state transition occurs, we set this to a debounce
    // period.  Future state transitions will be ignored
    // until the debounce time elapses.
    float t;
    // joystick button mask for the button, if mapped as a joystick button
    uint32_t js;
    // keyboard modifier bits and scan code for the button, if mapped as a keyboard key
    uint8_t keymod;
    uint8_t keycode;
    // media control key code
    uint8_t mediakey;

} buttonState[MAX_BUTTONS];

// timer for button reports
static Timer buttonTimer;

// initialize the button inputs
void initButtons(Config &cfg, bool &kbKeys)
    // presume we'll find no keyboard keys
    kbKeys = false;
    // create the digital inputs
    ButtonState *bs = buttonState;
    for (int i = 0 ; i < MAX_BUTTONS ; ++i, ++bs)
        PinName pin = wirePinName(cfg.button[i].pin);
        if (pin != NC)
            // set up the GPIO input pin for this button
            bs->di = new DigitalIn(pin);
            // note if it's a keyboard key of some kind (including media keys)
            uint8_t val = cfg.button[i].val;
            switch (cfg.button[i].typ)
            case BtnTypeJoystick:
                // joystick button - get the button bit mask
                bs->js = 1 << val;
            case BtnTypeKey:
                // regular keyboard key - note the scan code
                bs->keycode = val;
                kbKeys = true;
            case BtnTypeModKey:
                // keyboard mod key - note the modifier mask
                bs->keymod = val;
                kbKeys = true;
            case BtnTypeMedia:
                // media key - note the code
                bs->mediakey = val;
                kbKeys = true;
    // start the button timer

// Button data
uint32_t jsButtons = 0;

// Keyboard state
    bool changed;       // flag: changed since last report sent
    int nkeys;          // number of active keys in the list
    uint8_t data[8];    // key state, in USB report format: byte 0 is the modifier key mask,
                        // byte 1 is reserved, and bytes 2-7 are the currently pressed key codes
} kbState = { false, 0, { 0, 0, 0, 0, 0, 0, 0, 0 } };

// Media key state
    bool changed;       // flag: changed since last report sent
    uint8_t data;       // key state byte for USB reports
} mediaState = { false, 0 };

// read the button input state; returns true if there are any button
// state changes to report, false if not
bool readButtons(Config &cfg)
    // no changes detected yet
    bool changes = false;
    // start with an empty list of USB key codes
    uint8_t modkeys = 0;
    uint8_t keys[7] = { 0, 0, 0, 0, 0, 0, 0 };
    int nkeys = 0;
    // clear the joystick buttons
    uint32_t newjs = 0;
    // start with no media keys pressed
    uint8_t mediakeys = 0;

    // figure the time elapsed since the last scan
    float dt =;
    // reset the time for the next scan
    // scan the button list
    ButtonState *bs = buttonState;
    for (int i = 0 ; i < MAX_BUTTONS ; ++i, ++bs)
        // read this button
        if (bs->di != 0)
            // deduct the elapsed time since the last update
            // from the button's remaining sticky time
            bs->t -= dt;
            if (bs->t < 0)
                bs->t = 0;
            // If the sticky time has elapsed, note the new physical
            // state of the button.  If we still have sticky time
            // remaining, ignore the physical state; the last state
            // change persists until the sticky time elapses so that
            // we smooth out any "bounce" (electrical transients that
            // occur when the switch contact is opened or closed).
            if (bs->t == 0)
                // get the new physical state
                int pressed = !bs->di->read();
                // update the button's logical state if this is a change
                if (pressed != bs->pressed)
                    // store the new state
                    bs->pressed = pressed;
                    // start a new sticky period for debouncing this
                    // state change
                    bs->t = 0.075;

            // if it's pressed, add it to the appropriate key state list
            if (bs->pressed)
                // OR in the joystick button bit, mod key bits, and media key bits
                newjs |= bs->js;
                modkeys |= bs->keymod;
                mediakeys |= bs->mediakey;
                // if it has a keyboard key, add the scan code to the active list
                if (bs->keycode != 0 && nkeys < 7)
                    keys[nkeys++] = bs->keycode;

    // check for joystick button changes
    if (jsButtons != newjs)
        changes = true;
        jsButtons = newjs;
    // Check for changes to the keyboard keys
    if ([0] != modkeys
        || kbState.nkeys != nkeys
        || memcmp(keys, &[2], 6) != 0)
        // we have changes - set the change flag and store the new key data
        kbState.changed = true;
        changes = true;[0] = modkeys;
        if (nkeys <= 6) {
            // 6 or fewer simultaneous keys - report the key codes
            kbState.nkeys = nkeys;
            memcpy(&[2], keys, 6);
        else {
            // more than 6 simultaneous keys - report rollover (all '1' key codes)
            kbState.nkeys = 6;
            memset(&[2], 1, 6);
    // Check for changes to media keys
    if ( != mediakeys)
        mediaState.changed = true; = mediakeys;
        changes = true;
    // return the change indicator
    return changes;

// ---------------------------------------------------------------------------
// Customization joystick subbclass

class MyUSBJoystick: public USBJoystick
    MyUSBJoystick(uint16_t vendor_id, uint16_t product_id, uint16_t product_release,
        bool waitForConnect, bool enableJoystick, bool useKB) 
        : USBJoystick(vendor_id, product_id, product_release, waitForConnect, enableJoystick, useKB)
        suspended_ = false;
    // are we connected?
    int isConnected()  { return configured(); }
    // Are we in suspend mode?
    int isSuspended() const { return suspended_; }
    virtual void suspendStateChanged(unsigned int suspended)
        { suspended_ = suspended; }

    // are we suspended?
    int suspended_; 

// ---------------------------------------------------------------------------
// Accelerometer (MMA8451Q)

// The MMA8451Q is the KL25Z's on-board 3-axis accelerometer.
// This is a custom wrapper for the library code to interface to the
// MMA8451Q.  This class encapsulates an interrupt handler and 
// automatic calibration.
// We install an interrupt handler on the accelerometer "data ready" 
// interrupt to ensure that we fetch each sample immediately when it
// becomes available.  The accelerometer data rate is fiarly high
// (800 Hz), so it's not practical to keep up with it by polling.
// Using an interrupt handler lets us respond quickly and read
// every sample.
// We automatically calibrate the accelerometer so that it's not
// necessary to get it exactly level when installing it, and so
// that it's also not necessary to calibrate it manually.  There's
// lots of experience that tells us that manual calibration is a
// terrible solution, mostly because cabinets tend to shift slightly
// during use, requiring frequent recalibration.  Instead, we
// calibrate automatically.  We continuously monitor the acceleration
// data, watching for periods of constant (or nearly constant) values.
// Any time it appears that the machine has been at rest for a while
// (about 5 seconds), we'll average the readings during that rest
// period and use the result as the level rest position.  This is
// is ongoing, so we'll quickly find the center point again if the 
// machine is moved during play (by an especially aggressive bout
// of nudging, say).

// I2C address of the accelerometer (this is a constant of the KL25Z)
const int MMA8451_I2C_ADDRESS = (0x1d<<1);

// SCL and SDA pins for the accelerometer (constant for the KL25Z)
#define MMA8451_SCL_PIN   PTE25
#define MMA8451_SDA_PIN   PTE24

// Digital in pin to use for the accelerometer interrupt.  For the KL25Z,
// this can be either PTA14 or PTA15, since those are the pins physically
// wired on this board to the MMA8451 interrupt controller.
#define MMA8451_INT_PIN   PTA15

// accelerometer input history item, for gathering calibration data
struct AccHist
    AccHist() { x = y = d = 0.0; xtot = ytot = 0.0; cnt = 0; }
    void set(float x, float y, AccHist *prv)
        // save the raw position
        this->x = x;
        this->y = y;
        this->d = distance(prv);
    // reading for this entry
    float x, y;
    // distance from previous entry
    float d;
    // total and count of samples averaged over this period
    float xtot, ytot;
    int cnt;

    void clearAvg() { xtot = ytot = 0.0; cnt = 0; }    
    void addAvg(float x, float y) { xtot += x; ytot += y; ++cnt; }
    float xAvg() const { return xtot/cnt; }
    float yAvg() const { return ytot/cnt; }
    float distance(AccHist *p)
        { return sqrt(square(p->x - x) + square(p->y - y)); }

// accelerometer wrapper class
class Accel
    Accel(PinName sda, PinName scl, int i2cAddr, PinName irqPin)
        : mma_(sda, scl, i2cAddr), intIn_(irqPin)
        // remember the interrupt pin assignment
        irqPin_ = irqPin;

        // reset and initialize
    void reset()
        // clear the center point
        cx_ = cy_ = 0.0;
        // start the calibration timer
        iAccPrv_ = nAccPrv_ = 0;
        // reset and initialize the MMA8451Q
        // set the initial integrated velocity reading to zero
        vx_ = vy_ = 0;
        // set up our accelerometer interrupt handling
        intIn_.rise(this, &Accel::isr);
        mma_.setInterruptMode(irqPin_ == PTA14 ? 1 : 2);
        // read the current registers to clear the data ready flag
        mma_.getAccXYZ(ax_, ay_, az_);

        // start our timers
    void get(int &x, int &y) 
         // disable interrupts while manipulating the shared data
         // read the shared data and store locally for calculations
         float ax = ax_, ay = ay_;
         float vx = vx_, vy = vy_;
         // reset the velocity sum for the next run
         vx_ = vy_ = 0;

         // get the time since the last get() sample
         float dt = tGet_.read_us()/1.0e6;
         // done manipulating the shared data
         // adjust the readings for the integration time
         vx /= dt;
         vy /= dt;
         // add this sample to the current calibration interval's running total
         AccHist *p = accPrv_ + iAccPrv_;
         p->addAvg(ax, ay);

         // check for auto-centering every so often
         if (tCenter_.read_ms() > 1000)
             // add the latest raw sample to the history list
             AccHist *prv = p;
             iAccPrv_ = (iAccPrv_ + 1) % maxAccPrv;
             p = accPrv_ + iAccPrv_;
             p->set(ax, ay, prv);

             // if we have a full complement, check for stability
             if (nAccPrv_ >= maxAccPrv)
                 // check if we've been stable for all recent samples
                 static const float accTol = .01;
                 AccHist *p0 = accPrv_;
                 if (p0[0].d < accTol
                     && p0[1].d < accTol
                     && p0[2].d < accTol
                     && p0[3].d < accTol
                     && p0[4].d < accTol)
                     // Figure the new calibration point as the average of
                     // the samples over the rest period
                     cx_ = (p0[0].xAvg() + p0[1].xAvg() + p0[2].xAvg() + p0[3].xAvg() + p0[4].xAvg())/5.0;
                     cy_ = (p0[0].yAvg() + p0[1].yAvg() + p0[2].yAvg() + p0[3].yAvg() + p0[4].yAvg())/5.0;
                // not enough samples yet; just up the count
             // clear the new item's running totals
             // reset the timer
         // report our integrated velocity reading in x,y
         x = rawToReport(vx);
         y = rawToReport(vy);
         if (x != 0 || y != 0)        
             printf("%f %f %d %d %f\r\n", vx, vy, x, y, dt);
    // adjust a raw acceleration figure to a usb report value
    int rawToReport(float v)
        // scale to the joystick report range and round to integer
        int i = int(round(v*JOYMAX));
        // if it's near the center, scale it roughly as 20*(i/20)^2,
        // to suppress noise near the rest position
        static const int filter[] = { 
            -18, -16, -14, -13, -11, -10, -8, -7, -6, -5, -4, -3, -2, -2, -1, -1, 0, 0, 0, 0,
            0, 0, 0, 0, 1, 1, 2, 2, 3, 4, 5, 6, 7, 8, 10, 11, 13, 14, 16, 18
        return (i > 20 || i < -20 ? i : filter[i+20]);

    // interrupt handler
    void isr()
        // Read the axes.  Note that we have to read all three axes
        // (even though we only really use x and y) in order to clear
        // the "data ready" status bit in the accelerometer.  The
        // interrupt only occurs when the "ready" bit transitions from
        // off to on, so we have to make sure it's off.
        float x, y, z;
        mma_.getAccXYZ(x, y, z);
        // calculate the time since the last interrupt
        float dt = tInt_.read_us()/1.0e6;

        // integrate the time slice from the previous reading to this reading
        vx_ += (x + ax_ - 2*cx_)*dt/2;
        vy_ += (y + ay_ - 2*cy_)*dt/2;
        // store the updates
        ax_ = x;
        ay_ = y;
        az_ = z;
    // underlying accelerometer object
    MMA8451Q mma_;
    // last raw acceleration readings
    float ax_, ay_, az_;
    // integrated velocity reading since last get()
    float vx_, vy_;
    // timer for measuring time between get() samples
    Timer tGet_;
    // timer for measuring time between interrupts
    Timer tInt_;

    // Calibration reference point for accelerometer.  This is the
    // average reading on the accelerometer when in the neutral position
    // at rest.
    float cx_, cy_;

    // timer for atuo-centering
    Timer tCenter_;

    // Auto-centering history.  This is a separate history list that
    // records results spaced out sparesely over time, so that we can
    // watch for long-lasting periods of rest.  When we observe nearly
    // no motion for an extended period (on the order of 5 seconds), we
    // take this to mean that the cabinet is at rest in its neutral 
    // position, so we take this as the calibration zero point for the
    // accelerometer.  We update this history continuously, which allows
    // us to continuously re-calibrate the accelerometer.  This ensures
    // that we'll automatically adjust to any actual changes in the
    // cabinet's orientation (e.g., if it gets moved slightly by an
    // especially strong nudge) as well as any systematic drift in the
    // accelerometer measurement bias (e.g., from temperature changes).
    int iAccPrv_, nAccPrv_;
    static const int maxAccPrv = 5;
    AccHist accPrv_[maxAccPrv];
    // interurupt pin name
    PinName irqPin_;
    // interrupt router
    InterruptIn intIn_;

// ---------------------------------------------------------------------------
// Clear the I2C bus for the MMA8451Q.  This seems necessary some of the time
// for reasons that aren't clear to me.  Doing a hard power cycle has the same
// effect, but when we do a soft reset, the hardware sometimes seems to leave
// the MMA's SDA line stuck low.  Forcing a series of 9 clock pulses through
// the SCL line is supposed to clear this condition.  I'm not convinced this
// actually works with the way this component is wired on the KL25Z, but it
// seems harmless, so we'll do it on reset in case it does some good.  What
// we really seem to need is a way to power cycle the MMA8451Q if it ever 
// gets stuck, but this is simply not possible in software on the KL25Z. 
// If the accelerometer does get stuck, and a software reboot doesn't reset
// it, the only workaround is to manually power cycle the whole KL25Z by 
// unplugging both of its USB connections.
void clear_i2c()
    // assume a general-purpose output pin to the I2C clock
    DigitalOut scl(MMA8451_SCL_PIN);
    DigitalIn sda(MMA8451_SDA_PIN);
    // clock the SCL 9 times
    for (int i = 0 ; i < 9 ; ++i)
        scl = 1;
        scl = 0;
// ---------------------------------------------------------------------------
// Simple binary (on/off) input debouncer.  Requires an input to be stable 
// for a given interval before allowing an update.
class Debouncer
    Debouncer(bool initVal, float tmin)
        this->stable = this->prv = initVal;
        this->tmin = tmin;
    // Get the current stable value
    bool val() const { return stable; }

    // Apply a new sample.  This tells us the new raw reading from the
    // input device.
    void sampleIn(bool val)
        // If the new raw reading is different from the previous
        // raw reading, we've detected an edge - start the clock
        // on the sample reader.
        if (val != prv)
            // we have an edge - reset the sample clock
            // this is now the previous raw sample for nxt time
            prv = val;
        else if (val != stable)
            // The new raw sample is the same as the last raw sample,
            // and different from the stable value.  This means that
            // the sample value has been the same for the time currently
            // indicated by our timer.  If enough time has elapsed to
            // consider the value stable, apply the new value.
            if ( > tmin)
                stable = val;
    // current stable value
    bool stable;

    // last raw sample value
    bool prv;
    // elapsed time since last raw input change
    Timer t;
    // Minimum time interval for stability, in seconds.  Input readings 
    // must be stable for this long before the stable value is updated.
    float tmin;

// ---------------------------------------------------------------------------
// Turn off all outputs and restore everything to the default LedWiz
// state.  This sets outputs #1-32 to LedWiz profile value 48 (full
// brightness) and switch state Off, sets all extended outputs (#33
// and above) to zero brightness, and sets the LedWiz flash rate to 2.
// This effectively restores the power-on conditions.
void allOutputsOff()
    // reset all LedWiz outputs to OFF/48
    for (int i = 0 ; i < numLwOutputs ; ++i)
        outLevel[i] = 0;
        wizOn[i] = 0;
        wizVal[i] = 48;
    // reset all extended outputs (ports >32) to full off (brightness 0)
    for (int i = 32 ; i < numOutputs ; ++i)
        outLevel[i] = 0;
    // restore default LedWiz flash rate
    wizSpeed = 2;
    // flush changes to hc595, if applicable
    if (hc595 != 0)

// ---------------------------------------------------------------------------
// TV ON timer.  If this feature is enabled, we toggle a TV power switch
// relay (connected to a GPIO pin) to turn on the cab's TV monitors shortly
// after the system is powered.  This is useful for TVs that don't remember
// their power state and don't turn back on automatically after being
// unplugged and plugged in again.  This feature requires external
// circuitry, which is built in to the expansion board and can also be
// built separately - see the Build Guide for the circuit plan.
// Theory of operation: to use this feature, the cabinet must have a 
// secondary PC-style power supply (PSU2) for the feedback devices, and
// this secondary supply must be plugged in to the same power strip or 
// switched outlet that controls power to the TVs.  This lets us use PSU2
// as a proxy for the TV power state - when PSU2 is on, the TV outlet is 
// powered, and when PSU2 is off, the TV outlet is off.  We use a little 
// latch circuit powered by PSU2 to monitor the status.  The latch has a 
// current state, ON or OFF, that we can read via a GPIO input pin, and 
// we can set the state to ON by pulsing a separate GPIO output pin.  As 
// long as PSU2 is powered off, the latch stays in the OFF state, even if 
// we try to set it by pulsing the SET pin.  When PSU2 is turned on after 
// being off, the latch starts receiving power but stays in the OFF state, 
// since this is the initial condition when the power first comes on.  So 
// if our latch state pin is reading OFF, we know that PSU2 is either off 
// now or *was* off some time since we last checked.  We use a timer to 
// check the state periodically.  Each time we see the state is OFF, we 
// try pulsing the SET pin.  If the state still reads as OFF, we know 
// that PSU2 is currently off; if the state changes to ON, though, we 
// know that PSU2 has gone from OFF to ON some time between now and the 
// previous check.  When we see this condition, we start a countdown
// timer, and pulse the TV switch relay when the countdown ends.
// This scheme might seem a little convoluted, but it neatly handles
// all of the different cases that can occur:
// - Most cabinets systems are set up with "soft" PC power switches, 
//   so that the PC goes into "Soft Off" mode (ACPI state S5, in Windows
//   parlance) when the user turns off the cabinet.  In this state, the
//   motherboard supplies power to USB devices, so the KL25Z continues
//   running without interruption.  The latch system lets us monitor
//   the power state even when we're never rebooted, since the latch
//   will turn off when PSU2 is off regardless of what the KL25Z is doing.
// - Some cabinet builders might prefer to use "hard" power switches,
//   cutting all power to the cabinet, including the PC motherboard (and
//   thus the KL25Z) every time the machine is turned off.  This also
//   applies to the "soft" switch case above when the cabinet is unplugged,
//   a power outage occurs, etc.  In these cases, the KL25Z will do a cold
//   boot when the PC is turned on.  We don't know whether the KL25Z
//   will power up before or after PSU2, so it's not good enough to 
//   observe the *current* state of PSU2 when we first check - if PSU2
//   were to come on first, checking the current state alone would fool
//   us into thinking that no action is required, because we would never
//   have known that PSU2 was ever off.  The latch handles this case by
//   letting us see that PSU2 *was* off before we checked.
// - If the KL25Z is rebooted while the main system is running, or the 
//   KL25Z is unplugged and plugged back in, we will correctly leave the 
//   TVs as they are.  The latch state is independent of the KL25Z's 
//   power or software state, so it's won't affect the latch state when
//   the KL25Z is unplugged or rebooted; when we boot, we'll see that 
//   the latch is already on and that we don't have to turn on the TVs.
//   This is important because TV ON buttons are usually on/off toggles,
//   so we don't want to push the button on a TV that's already on.

// Current PSU2 state:
//   1 -> default: latch was on at last check, or we haven't checked yet
//   2 -> latch was off at last check, SET pulsed high
//   3 -> SET pulsed low, ready to check status
//   4 -> TV timer countdown in progress
//   5 -> TV relay on
int psu2_state = 1;

// PSU2 power sensing circuit connections
DigitalIn *psu2_status_sense;
DigitalOut *psu2_status_set;

// TV ON switch relay control
DigitalOut *tv_relay;

// Timer interrupt
Ticker tv_ticker;
float tv_delay_time;
void TVTimerInt()
    // time since last state change
    static Timer tv_timer;

    // Check our internal state
    switch (psu2_state)
    case 1:
        // Default state.  This means that the latch was on last
        // time we checked or that this is the first check.  In
        // either case, if the latch is off, switch to state 2 and
        // try pulsing the latch.  Next time we check, if the latch
        // stuck, it means that PSU2 is now on after being off.
        if (!psu2_status_sense->read())
            // switch to OFF state
            psu2_state = 2;
            // try setting the latch
    case 2:
        // PSU2 was off last time we checked, and we tried setting
        // the latch.  Drop the SET signal and go to CHECK state.
        psu2_state = 3;
    case 3:
        // CHECK state: we pulsed SET, and we're now ready to see
        // if that stuck.  If the latch is now on, PSU2 has transitioned
        // from OFF to ON, so start the TV countdown.  If the latch is
        // off, our SET command didn't stick, so PSU2 is still off.
        if (psu2_status_sense->read())
            // The latch stuck, so PSU2 has transitioned from OFF
            // to ON.  Start the TV countdown timer.
            psu2_state = 4;
            // The latch didn't stick, so PSU2 was still off at
            // our last check.  Try pulsing it again in case PSU2
            // was turned on since the last check.
            psu2_state = 2;
    case 4:
        // TV timer countdown in progress.  If we've reached the
        // delay time, pulse the relay.
        if ( >= tv_delay_time)
            // turn on the relay for one timer interval
            psu2_state = 5;
    case 5:
        // TV timer relay on.  We pulse this for one interval, so
        // it's now time to turn it off and return to the default state.
        psu2_state = 1;

// Start the TV ON checker.  If the status sense circuit is enabled in
// the configuration, we'll set up the pin connections and start the
// interrupt handler that periodically checks the status.  Does nothing
// if any of the pins are configured as NC.
void startTVTimer(Config &cfg)
    // only start the timer if the status sense circuit pins are configured
    if (cfg.TVON.statusPin != NC && cfg.TVON.latchPin != NC && cfg.TVON.relayPin != NC)
        psu2_status_sense = new DigitalIn(cfg.TVON.statusPin);
        psu2_status_set = new DigitalOut(cfg.TVON.latchPin);
        tv_relay = new DigitalOut(cfg.TVON.relayPin);
        tv_delay_time = cfg.TVON.delayTime;
        // Set up our time routine to run every 1/4 second.  
        tv_ticker.attach(&TVTimerInt, 0.25);

// ---------------------------------------------------------------------------
// In-memory configuration data structure.  This is the live version in RAM
// that we use to determine how things are set up.
// When we save the configuration settings, we copy this structure to
// non-volatile flash memory.  At startup, we check the flash location where
// we might have saved settings on a previous run, and it's valid, we copy 
// the flash data to this structure.  Firmware updates wipe the flash
// memory area, so you have to use the PC config tool to send the settings
// again each time the firmware is updated.
NVM nvm;

// For convenience, a macro for the Config part of the NVM structure
#define cfg (nvm.d.c)

// flash memory controller interface
FreescaleIAP iap;

// figure the flash address as a pointer along with the number of sectors
// required to store the structure
NVM *configFlashAddr(int &addr, int &numSectors)
    // figure how many flash sectors we span, rounding up to whole sectors
    numSectors = (sizeof(NVM) + SECTOR_SIZE - 1)/SECTOR_SIZE;

    // figure the address - this is the highest flash address where the
    // structure will fit with the start aligned on a sector boundary
    addr = iap.flash_size() - (numSectors * SECTOR_SIZE);
    // return the address as a pointer
    return (NVM *)addr;

// figure the flash address as a pointer
NVM *configFlashAddr()
    int addr, numSectors;
    return configFlashAddr(addr, numSectors);

// Load the config from flash
void loadConfigFromFlash()
    // We want to use the KL25Z's on-board flash to store our configuration
    // data persistently, so that we can restore it across power cycles.
    // Unfortunatly, the mbed platform doesn't explicitly support this.
    // mbed treats the on-board flash as a raw storage device for linker
    // output, and assumes that the linker output is the only thing
    // stored there.  There's no file system and no allowance for shared
    // use for other purposes.  Fortunately, the linker ues the space in
    // the obvious way, storing the entire linked program in a contiguous
    // block starting at the lowest flash address.  This means that the
    // rest of flash - from the end of the linked program to the highest
    // flash address - is all unused free space.  Writing our data there
    // won't conflict with anything else.  Since the linker doesn't give
    // us any programmatic access to the total linker output size, it's
    // safest to just store our config data at the very end of the flash
    // region (i.e., the highest address).  As long as it's smaller than
    // the free space, it won't collide with the linker area.
    // Figure how many sectors we need for our structure
    NVM *flash = configFlashAddr();
    // if the flash is valid, load it; otherwise initialize to defaults
    if (flash->valid()) 
        // flash is valid - load it into the RAM copy of the structure
        memcpy(&nvm, flash, sizeof(NVM));
        // flash is invalid - load factory settings nito RAM structure

void saveConfigToFlash()
    int addr, sectors;
    configFlashAddr(addr, sectors);, addr);

// ---------------------------------------------------------------------------
// Plunger Sensor

// the plunger sensor interface object
PlungerSensor *plungerSensor = 0;

// Create the plunger sensor based on the current configuration.  If 
// there's already a sensor object, we'll delete it.
void createPlunger()
    // delete any existing sensor object
    if (plungerSensor != 0)
        delete plungerSensor;
    // create the new sensor object according to the type
    switch (cfg.plunger.sensorType)
    case PlungerType_TSL1410RS:
        // pins are: SI, CLOCK, AO
        plungerSensor = new PlungerSensorTSL1410R(cfg.plunger.sensorPin[0], cfg.plunger.sensorPin[1], cfg.plunger.sensorPin[2], NC);
    case PlungerType_TSL1410RP:
        // pins are: SI, CLOCK, AO1, AO2
        plungerSensor = new PlungerSensorTSL1410R(cfg.plunger.sensorPin[0], cfg.plunger.sensorPin[1], cfg.plunger.sensorPin[2], cfg.plunger.sensorPin[3]);
    case PlungerType_TSL1412RS:
        // pins are: SI, CLOCK, AO1, AO2
        plungerSensor = new PlungerSensorTSL1412R(cfg.plunger.sensorPin[0], cfg.plunger.sensorPin[1], cfg.plunger.sensorPin[2], NC);
    case PlungerType_TSL1412RP:
        // pins are: SI, CLOCK, AO1, AO2
        plungerSensor = new PlungerSensorTSL1412R(cfg.plunger.sensorPin[0], cfg.plunger.sensorPin[1], cfg.plunger.sensorPin[2], cfg.plunger.sensorPin[3]);
    case PlungerType_Pot:
        // pins are: AO
        plungerSensor = new PlungerSensorPot(cfg.plunger.sensorPin[0]);
    case PlungerType_None:
        plungerSensor = new PlungerSensorNull();

// ---------------------------------------------------------------------------
// Reboot - resets the microcontroller
void reboot(USBJoystick &js)
    // disconnect from USB
    // wait a few seconds to make sure the host notices the disconnect
    // reset the device
    while (true) { }

// ---------------------------------------------------------------------------
// Translate joystick readings from raw values to reported values, based
// on the orientation of the controller card in the cabinet.
void accelRotate(int &x, int &y)
    int tmp;
    switch (cfg.orientation)
    case OrientationFront:
        tmp = x;
        x = y;
        y = tmp;
    case OrientationLeft:
        x = -x;
    case OrientationRight:
        y = -y;
    case OrientationRear:
        tmp = -x;
        x = -y;
        y = tmp;

// ---------------------------------------------------------------------------
// Device status.  We report this on each update so that the host config
// tool can detect our current settings.  This is a bit mask consisting
// of these bits:
//    0x0001  -> plunger sensor enabled
//    0x8000  -> RESERVED - must always be zero
// Note that the high bit (0x8000) must always be 0, since we use that
// to distinguish special request reply packets.
uint16_t statusFlags;
// flag: send a pixel dump after the next read
bool reportPix = false;

// ---------------------------------------------------------------------------
// Calibration button state:
//  0 = not pushed
//  1 = pushed, not yet debounced
//  2 = pushed, debounced, waiting for hold time
//  3 = pushed, hold time completed - in calibration mode
int calBtnState = 0;

// calibration button debounce timer
Timer calBtnTimer;

// calibration button light state
int calBtnLit = false;

// ---------------------------------------------------------------------------
// Handle a configuration variable update.  'data' is the USB message we
// received from the host.
void configVarMsg(uint8_t *data)
    switch (data[1])
    case 1:
        // USB identification (Vendor ID, Product ID)
        cfg.usbVendorID = wireUI16(data+2);
        cfg.usbProductID = wireUI16(data+4);
    case 2:
        // Pinscape Controller unit number - note that data[2] contains
        // the nominal unit number, 1-16
        if (data[2] >= 1 && data[2] <= 16)
            cfg.psUnitNo = data[2];
    case 3:
        // Enable/disable joystick
        cfg.joystickEnabled = data[2];
    case 4:
        // Accelerometer orientation
        cfg.orientation = data[2];

    case 5:
        // Plunger sensor type
        cfg.plunger.sensorType = data[2];
    case 6:
        // Set plunger pin assignments
        cfg.plunger.sensorPin[0] = wirePinName(data[2]);
        cfg.plunger.sensorPin[1] = wirePinName(data[3]);
        cfg.plunger.sensorPin[2] = wirePinName(data[4]);
        cfg.plunger.sensorPin[3] = wirePinName(data[5]);
    case 7:
        // Plunger calibration button and indicator light pin assignments = wirePinName(data[2]); = wirePinName(data[3]);
    case 8:
        // ZB Launch Ball setup
        cfg.plunger.zbLaunchBall.port = (int)(unsigned char)data[2];
        cfg.plunger.zbLaunchBall.btn = (int)(unsigned char)data[3];
        cfg.plunger.zbLaunchBall.pushDistance = (float)wireUI16(data+4) / 1000.0;
    case 9:
        // TV ON setup
        cfg.TVON.statusPin = wirePinName(data[2]);
        cfg.TVON.latchPin = wirePinName(data[3]);
        cfg.TVON.relayPin = wirePinName(data[4]);
        cfg.TVON.delayTime = (float)wireUI16(data+5) / 100.0;
    case 10:
        // TLC5940NT PWM controller chip setup
        cfg.tlc5940.nchips = (int)(unsigned char)data[2];
        cfg.tlc5940.sin = wirePinName(data[3]);
        cfg.tlc5940.sclk = wirePinName(data[4]);
        cfg.tlc5940.xlat = wirePinName(data[5]);
        cfg.tlc5940.blank = wirePinName(data[6]);
        cfg.tlc5940.gsclk = wirePinName(data[7]);
    case 11:
        // 74HC595 shift register chip setup
        cfg.hc595.nchips = (int)(unsigned char)data[2];
        cfg.hc595.sin = wirePinName(data[3]);
        cfg.hc595.sclk = wirePinName(data[4]);
        cfg.hc595.latch = wirePinName(data[5]);
        cfg.hc595.ena = wirePinName(data[6]);
    case 12:
        // button setup
            // get the button number
            int idx = data[2];
            // if it's in range, set the button data
            if (idx > 0 && idx <= MAX_BUTTONS)
                // adjust to an array index
                // set the values
                cfg.button[idx].pin = data[3];
                cfg.button[idx].typ = data[4];
                cfg.button[idx].val = data[5];
    case 13:
        // LedWiz output port setup
            // get the port number
            int idx = data[2];
            // if it's in range, set the port data
            if (idx > 0 && idx <= MAX_OUT_PORTS)
                // adjust to an array index
                // set the values
                cfg.outPort[idx].typ = data[3];
                cfg.outPort[idx].pin = data[4];
                cfg.outPort[idx].flags = data[5];

// ---------------------------------------------------------------------------
// Handle an input report from the USB host.  Input reports use our extended
// LedWiz protocol.
void handleInputMsg(HID_REPORT &report, USBJoystick &js, int &z)
    // all Led-Wiz reports are exactly 8 bytes
    if (report.length == 8)
        // LedWiz commands come in two varieties:  SBA and PBA.  An
        // SBA is marked by the first byte having value 64 (0x40).  In
        // the real LedWiz protocol, any other value in the first byte
        // means it's a PBA message.  However, *valid* PBA messages
        // always have a first byte (and in fact all 8 bytes) in the
        // range 0-49 or 129-132.  Anything else is invalid.  We take
        // advantage of this to implement private protocol extensions.
        // So our full protocol is as follows:
        // first byte =
        //   0-48     -> LWZ-PBA
        //   64       -> LWZ SBA 
        //   65       -> private control message; second byte specifies subtype
        //   129-132  -> LWZ-PBA
        //   200-228  -> extended bank brightness set for outputs N to N+6, where
        //               N is (first byte - 200)*7
        //   other    -> reserved for future use
        uint8_t *data =;
        if (data[0] == 64) 
            // LWZ-SBA - first four bytes are bit-packed on/off flags
            // for the outputs; 5th byte is the pulse speed (1-7)
            //printf("LWZ-SBA %02x %02x %02x %02x ; %02x\r\n",
            //       data[1], data[2], data[3], data[4], data[5]);

            // update all on/off states
            for (int i = 0, bit = 1, ri = 1 ; i < numLwOutputs ; ++i, bit <<= 1)
                // figure the on/off state bit for this output
                if (bit == 0x100) {
                    bit = 1;
                // set the on/off state
                wizOn[i] = ((data[ri] & bit) != 0);
                // If the wizVal setting is 255, it means that this
                // output was last set to a brightness value with the
                // extended protocol.  Return it to LedWiz control by
                // rescaling the brightness setting to the LedWiz range
                // and updating wizVal with the result.  If it's any
                // other value, it was previously set by a PBA message,
                // so simply retain the last setting - in the normal
                // LedWiz protocol, the "profile" (brightness) and on/off
                // states are independent, so an SBA just turns an output
                // on or off but retains its last brightness level.
                if (wizVal[i] == 255)
                    wizVal[i] = (uint8_t)round(outLevel[i]*48);
            // set the flash speed - enforce the value range 1-7
            wizSpeed = data[5];
            if (wizSpeed < 1)
                wizSpeed = 1;
            else if (wizSpeed > 7)
                wizSpeed = 7;

            // update the physical outputs
            if (hc595 != 0)
            // reset the PBA counter
            pbaIdx = 0;
        else if (data[0] == 65)
            // Private control message.  This isn't an LedWiz message - it's
            // an extension for this device.  65 is an invalid PBA setting,
            // and isn't used for any other LedWiz message, so we appropriate
            // it for our own private use.  The first byte specifies the 
            // message type.
            if (data[1] == 1)
                // 1 = Old Set Configuration:
                //     data[2] = LedWiz unit number (0x00 to 0x0f)
                //     data[3] = feature enable bit mask:
                //               0x01 = enable plunger sensor

                // get the new LedWiz unit number - this is 0-15, whereas we
                // we save the *nominal* unit number 1-16 in the config                
                uint8_t newUnitNo = (data[2] & 0x0f) + 1;

                // we'll need a reset if the LedWiz unit number is changing
                bool needReset = (newUnitNo != cfg.psUnitNo);
                // set the configuration parameters from the message
                cfg.psUnitNo = newUnitNo;
                cfg.plunger.enabled = data[3] & 0x01;
                // update the status flags
                statusFlags = (statusFlags & ~0x01) | (data[3] & 0x01);
                // if the plunger is no longer enabled, use 0 for z reports
                if (!cfg.plunger.enabled)
                    z = 0;
                // save the configuration
                // reboot if necessary
                if (needReset)
            else if (data[1] == 2)
                // 2 = Calibrate plunger
                // (No parameters)
                // enter calibration mode
                calBtnState = 3;
            else if (data[1] == 3)
                // 3 = pixel dump
                // (No parameters)
                reportPix = true;
                // show purple until we finish sending the report
                ledR = 0;
                ledB = 0;
                ledG = 1;
            else if (data[1] == 4)
                // 4 = hardware configuration query
                // (No parameters)
                    cfg.psUnitNo - 1,   // report 0-15 range for unit number (we store 1-16 internally)
            else if (data[1] == 5)
                // 5 = all outputs off, reset to LedWiz defaults
            else if (data[1] == 6)
                // 6 = Save configuration to flash.
                // Reboot the microcontroller.  Nearly all config changes
                // require a reset, and a reset only takes a few seconds, 
                // so we don't bother tracking whether or not a reboot is
                // really needed.
        else if (data[0] == 66)
            // Extended protocol - Set configuration variable.
            // The second byte of the message is the ID of the variable
            // to update, and the remaining bytes give the new value,
            // in a variable-dependent format.
        else if (data[0] >= 200 && data[0] <= 228)
            // Extended protocol - Extended output port brightness update.  
            // data[0]-200 gives us the bank of 7 outputs we're setting:
            // 200 is outputs 0-6, 201 is outputs 7-13, 202 is 14-20, etc.
            // The remaining bytes are brightness levels, 0-255, for the
            // seven outputs in the selected bank.  The LedWiz flashing 
            // modes aren't accessible in this message type; we can only 
            // set a fixed brightness, but in exchange we get 8-bit 
            // resolution rather than the paltry 0-48 scale that the real
            // LedWiz uses.  There's no separate on/off status for outputs
            // adjusted with this message type, either, as there would be
            // for a PBA message - setting a non-zero value immediately
            // turns the output, overriding the last SBA setting.
            // For outputs 0-31, this overrides any previous PBA/SBA
            // settings for the port.  Any subsequent PBA/SBA message will
            // in turn override the setting made here.  It's simple - the
            // most recent message of either type takes precedence.  For
            // outputs above the LedWiz range, PBA/SBA messages can't
            // address those ports anyway.
            int i0 = (data[0] - 200)*7;
            int i1 = i0 + 7 < numOutputs ? i0 + 7 : numOutputs; 
            for (int i = i0 ; i < i1 ; ++i)
                // set the brightness level for the output
                float b = data[i-i0+1]/255.0;
                outLevel[i] = b;
                // if it's in the basic LedWiz output set, set the LedWiz
                // profile value to 255, which means "use outLevel"
                if (i < 32) 
                    wizVal[i] = 255;
                // set the output
            // update 74HC595 outputs, if attached
            if (hc595 != 0)
            // Everything else is LWZ-PBA.  This is a full "profile"
            // dump from the host for one bank of 8 outputs.  Each
            // byte sets one output in the current bank.  The current
            // bank is implied; the bank starts at 0 and is reset to 0
            // by any LWZ-SBA message, and is incremented to the next
            // bank by each LWZ-PBA message.  Our variable pbaIdx keeps
            // track of our notion of the current bank.  There's no direct
            // way for the host to select the bank; it just has to count
            // on us staying in sync.  In practice, the host will always
            // send a full set of 4 PBA messages in a row to set all 32
            // outputs.
            // Note that a PBA implicitly overrides our extended profile
            // messages (message prefix 200-219), because this sets the
            // wizVal[] entry for each output, and that takes precedence
            // over the extended protocol settings.
            //printf("LWZ-PBA[%d] %02x %02x %02x %02x %02x %02x %02x %02x\r\n",
            //       pbaIdx, data[0], data[1], data[2], data[3], data[4], data[5], data[6], data[7]);

            // Update all output profile settings
            for (int i = 0 ; i < 8 ; ++i)
                wizVal[pbaIdx + i] = data[i];

            // Update the physical LED state if this is the last bank.
            // Note that hosts always send a full set of four PBA
            // messages, so there's no need to do a physical update
            // until we've received the last bank's PBA message.
            if (pbaIdx == 24)
                if (hc595 != 0)
                pbaIdx = 0;
                pbaIdx += 8;

// ---------------------------------------------------------------------------
// Main program loop.  This is invoked on startup and runs forever.  Our
// main work is to read our devices (the accelerometer and the CCD), process
// the readings into nudge and plunger position data, and send the results
// to the host computer via the USB joystick interface.  We also monitor
// the USB connection for incoming LedWiz commands and process those into
// port outputs.
int main(void)
    // turn off our on-board indicator LED
    ledR = 1;
    ledG = 1;
    ledB = 1;
    // clear the I2C bus for the accelerometer
    // load the saved configuration
    // start the TV timer, if applicable
    // we're not connected/awake yet
    bool connected = false;
    time_t connectChangeTime = time(0);

    // create the plunger sensor interface

    // set up the TLC5940 interface and start the TLC5940 clock, if applicable

    // enable the 74HC595 chips, if present
    // initialize the LedWiz ports
    // start the TLC5940 clock
    if (tlc5940 != 0)
    // initialize the button input ports
    bool kbKeys = false;
    initButtons(cfg, kbKeys);

    // Create the joystick USB client.  Note that we use the LedWiz unit
    // number from the saved configuration.
    MyUSBJoystick js(cfg.usbVendorID, cfg.usbProductID, USB_VERSION_NO, true, cfg.joystickEnabled, kbKeys);
    // last report timer - we use this to throttle reports, since VP
    // doesn't want to hear from us more than about every 10ms
    Timer reportTimer;
    // set the initial status flags
    statusFlags = (cfg.plunger.enabled ? 0x01 : 0x00);

    // initialize the calibration buttons, if present
    DigitalIn *calBtn = ( == NC ? 0 : new DigitalIn(;
    DigitalOut *calBtnLed = ( == NC ? 0 : new DigitalOut(;

    // initialize the calibration button 
    calBtnState = 0;
    // set up a timer for our heartbeat indicator
    Timer hbTimer;
    int hb = 0;
    uint16_t hbcnt = 0;
    // set a timer for accelerometer auto-centering
    Timer acTimer;
    // create the accelerometer object
    Accel accel(MMA8451_SCL_PIN, MMA8451_SDA_PIN, MMA8451_I2C_ADDRESS, MMA8451_INT_PIN);
    // last accelerometer report, in joystick units (we report the nudge
    // acceleration via the joystick x & y axes, per the VP convention)
    int x = 0, y = 0;
    // last plunger report position, in 'npix' normalized pixel units
    int pos = 0;
    // last plunger report, in joystick units (we report the plunger as the
    // "z" axis of the joystick, per the VP convention)
    int z = 0;
    // most recent prior plunger readings, for tracking release events(z0 is
    // reading just before the last one we reported, z1 is the one before that, 
    // z2 the next before that)
    int z0 = 0, z1 = 0, z2 = 0;
    // Simulated "bounce" position when firing.  We model the bounce off of
    // the barrel spring when the plunger is released as proportional to the
    // distance it was retracted just before being released.
    int zBounce = 0;
    // Simulated Launch Ball button state.  If a "ZB Launch Ball" port is
    // defined for our LedWiz port mapping, any time that port is turned ON,
    // we'll simulate pushing the Launch Ball button if the player pulls 
    // back and releases the plunger, or simply pushes on the plunger from
    // the rest position.  This allows the plunger to be used in lieu of a
    // physical Launch Ball button for tables that don't have plungers.
    // States:
    //   0 = default
    //   1 = cocked (plunger has been pulled back about 1" from state 0)
    //   2 = uncocked (plunger is pulled back less than 1" from state 1)
    //   3 = launching, plunger is forward beyond park position
    //   4 = launching, plunger is behind park position
    //   5 = pressed and holding (plunger has been pressed forward beyond 
    //       the park position from state 0)
    int lbState = 0;
    // button bit for ZB launch ball button
    const uint32_t lbButtonBit = (1 << (cfg.plunger.zbLaunchBall.btn - 1));
    // Time since last lbState transition.  Some of the states are time-
    // sensitive.  In the "uncocked" state, we'll return to state 0 if
    // we remain in this state for more than a few milliseconds, since
    // it indicates that the plunger is being slowly returned to rest
    // rather than released.  In the "launching" state, we need to release 
    // the Launch Ball button after a moment, and we need to wait for 
    // the plunger to come to rest before returning to state 0.
    Timer lbTimer;
    // Launch Ball simulated push timer.  We start this when we simulate
    // the button push, and turn off the simulated button when enough time
    // has elapsed.
    Timer lbBtnTimer;
    // Simulated button states.  This is a vector of button states
    // for the simulated buttons.  We combine this with the physical
    // button states on each USB joystick report, so we will report
    // a button as pressed if either the physical button is being pressed
    // or we're simulating a press on the button.  This is used for the
    // simulated Launch Ball button.
    uint32_t simButtons = 0;
    // Firing in progress: we set this when we detect the start of rapid 
    // plunger movement from a retracted position towards the rest position.
    // When we detect a firing event, we send VP a series of synthetic
    // reports simulating the idealized plunger motion.  The actual physical
    // motion is much too fast to report to VP; in the time between two USB
    // reports, the plunger can shoot all the way forward, rebound off of
    // the barrel spring, bounce back part way, and bounce forward again,
    // or even do all of this more than once.  This means that sampling the 
    // physical motion at the USB report rate would create a misleading 
    // picture of the plunger motion, since our samples would catch the 
    // plunger at random points in this oscillating motion.  From the 
    // user's perspective, the physical action that occurred is simply that 
    // the plunger was released from a particular distance, so it's this 
    // high-level event that we want to convey to VP.  To do this, we
    // synthesize a series of reports to convey an idealized version of
    // the release motion that's perfectly synchronized to the VP reports.  
    // Essentially we pretend that our USB position samples are exactly 
    // aligned in time with (1) the point of retraction just before the 
    // user released the plunger, (2) the point of maximum forward motion 
    // just after the user released the plunger (the point of maximum 
    // compression as the plunger bounces off of the barrel spring), and 
    // (3) the plunger coming to rest at the park position.  This series
    // of reports is synthetic in the sense that it's not what we actually
    // see on the CCD at the times of these reports - the true plunger
    // position is oscillating at high speed during this period.  But at
    // the same time it conveys a more faithful picture of the true physical
    // motion to VP, and allows VP to reproduce the true physical motion 
    // more faithfully in its simulation model, by correcting for the
    // relatively low sampling rate in the communication path between the
    // real plunger and VP's model plunger.
    // If 'firing' is non-zero, it's the index of our current report in
    // the synthetic firing report series.
    int firing = 0;

    // start the first CCD integration cycle
    // we're all set up - now just loop, processing sensor reports and 
    // host requests
    for (;;)
        // Look for an incoming report.  Process a few input reports in
        // a row, but stop after a few so that a barrage of inputs won't
        // starve our output event processing.  Also, pause briefly between
        // reads; allowing reads to occur back-to-back seems to occasionally 
        // stall the USB pipeline (for reasons unknown; I'd fix the underlying 
        // problem if I knew what it was).
        HID_REPORT report;
        for (int rr = 0 ; rr < 4 && js.readNB(&report) ; ++rr, wait_ms(1))
            handleInputMsg(report, js, z);
        // check for plunger calibration
        if (calBtn != 0 && !calBtn->read())
            // check the state
            switch (calBtnState)
            case 0: 
                // button not yet pushed - start debouncing
                calBtnState = 1;
            case 1:
                // pushed, not yet debounced - if the debounce time has
                // passed, start the hold period
                if (calBtnTimer.read_ms() > 50)
                    calBtnState = 2;
            case 2:
                // in the hold period - if the button has been held down
                // for the entire hold period, move to calibration mode
                if (calBtnTimer.read_ms() > 2050)
                    // enter calibration mode
                    calBtnState = 3;
                    // reset the plunger calibration limits
            case 3:
                // Already in calibration mode - pushing the button here
                // doesn't change the current state, but we won't leave this
                // state as long as it's held down.  So nothing changes here.
            // Button released.  If we're in calibration mode, and
            // the calibration time has elapsed, end the calibration
            // and save the results to flash.
            // Otherwise, return to the base state without saving anything.
            // If the button is released before we make it to calibration
            // mode, it simply cancels the attempt.
            if (calBtnState == 3 && calBtnTimer.read_ms() > 15000)
                // exit calibration mode
                calBtnState = 0;
                // save the updated configuration
       = 1;
            else if (calBtnState != 3)
                // didn't make it to calibration mode - cancel the operation
                calBtnState = 0;
        // light/flash the calibration button light, if applicable
        int newCalBtnLit = calBtnLit;
        switch (calBtnState)
        case 2:
            // in the hold period - flash the light
            newCalBtnLit = ((calBtnTimer.read_ms()/250) & 1);
        case 3:
            // calibration mode - show steady on
            newCalBtnLit = true;
            // not calibrating/holding - show steady off
            newCalBtnLit = false;
        // light or flash the external calibration button LED, and 
        // do the same with the on-board blue LED
        if (calBtnLit != newCalBtnLit)
            calBtnLit = newCalBtnLit;
            if (calBtnLit) {
                if (calBtnLed != 0)
                ledR = 1;
                ledG = 1;
                ledB = 0;
            else {
                if (calBtnLed != 0)
                ledR = 1;
                ledG = 1;
                ledB = 1;
        // If the plunger is enabled, and we're not already in a firing event,
        // and the last plunger reading had the plunger pulled back at least
        // a bit, watch for plunger release events until it's time for our next
        // USB report.
        if (!firing && cfg.plunger.enabled && z >= JOYMAX/6)
            // monitor the plunger until it's time for our next report
            while (reportTimer.read_ms() < 15)
                // do a fast low-res scan; if it's at or past the zero point,
                // start a firing event
                int pos0;
                if (plungerSensor->lowResScan(pos0) && pos0 <=
                    firing = 1;

        // read the plunger sensor, if it's enabled
        if (cfg.plunger.enabled)
            // start with the previous reading, in case we don't have a
            // clear result on this frame
            int znew = z;
            if (plungerSensor->highResScan(pos))
                // We got a new reading.  If we're in calibration mode, use it
                // to figure the new calibration, otherwise adjust the new reading
                // for the established calibration.
                if (calBtnState == 3)
                    // Calibration mode.  If this reading is outside of the current
                    // calibration bounds, expand the bounds.
                    if (pos <
               = pos;
                    if (pos <
               = pos;
                    if (pos >
               = pos;
                    // normalize to the full physical range while calibrating
                    znew = int(round(float(pos)/plungerSensor->npix * JOYMAX));
                    // Not in calibration mode, so normalize the new reading to the 
                    // established calibration range.  
                    // Note that negative values are allowed.  Zero represents the
                    // "park" position, where the plunger sits when at rest.  A mechanical 
                    // plunger has a small amount of travel in the "push" direction,
                    // since the barrel spring can be compressed slightly.  Negative
                    // values represent travel in the push direction.
                    if (pos >
                        pos =;
                    znew = int(round(float(pos -
                        / ( - + 1) * JOYMAX));

            // If we're not already in a firing event, check to see if the
            // new position is forward of the last report.  If it is, a firing
            // event might have started during the high-res scan.  This might
            // seem unlikely given that the scan only takes about 5ms, but that
            // 5ms represents about 25-30% of our total time between reports,
            // there's about a 1 in 4 chance that a release starts during a
            // scan.  
            if (!firing && z0 > 0 && znew < z0)
                // The plunger has moved forward since the previous report.
                // Watch it for a few more ms to see if we can get a stable
                // new position.
                int pos0;
                if (plungerSensor->lowResScan(pos0))
                    int pos1 = pos0;
                    Timer tw;
                    while (tw.read_ms() < 6)
                        // read the new position
                        int pos2;
                        if (plungerSensor->lowResScan(pos2))
                            // If it's stable over consecutive readings, stop looping.
                            // (Count it as stable if the position is within about 1/8".
                            // pos1 and pos2 are reported in pixels, so they range from
                            // 0 to npix.  The overall travel of a standard plunger is
                            // about 3.2", so we have (npix/3.2) pixels per inch, hence
                            // 1/8" is (npix/3.2)*(1/8) pixels.)
                            if (abs(pos2 - pos1) < int(plungerSensor->npix/(3.2*8)))
                            // If we've crossed the rest position, and we've moved by
                            // a minimum distance from where we starting this loop, begin
                            // a firing event.  (We require a minimum distance to prevent
                            // spurious firing from random analog noise in the readings
                            // when the plunger is actually just sitting still at the 
                            // rest position.  If it's at rest, it's normal to see small
                            // random fluctuations in the analog reading +/- 1% or so
                            // from the 0 point, especially with a sensor like a
                            // potentionemeter that reports the position as a single 
                            // analog voltage.)  Note that we compare the latest reading
                            // to the first reading of the loop - we don't require the
                            // threshold motion over consecutive readings, but any time
                            // over the stability wait loop.
                            if (pos1 <
                                && abs(pos2 - pos0) > int(plungerSensor->npix/(3.2*8)))
                                firing = 1;
                            // the new reading is now the prior reading
                            pos1 = pos2;
            // Check for a simulated Launch Ball button press, if enabled
            if (cfg.plunger.zbLaunchBall.port != 0)
                const int cockThreshold = JOYMAX/3;
                const int pushThreshold = int(-JOYMAX/3 * cfg.plunger.zbLaunchBall.pushDistance);
                int newState = lbState;
                switch (lbState)
                case 0:
                    // Base state.  If the plunger is pulled back by an inch
                    // or more, go to "cocked" state.  If the plunger is pushed
                    // forward by 1/4" or more, go to "pressed" state.
                    if (znew >= cockThreshold)
                        newState = 1;
                    else if (znew <= pushThreshold)
                        newState = 5;
                case 1:
                    // Cocked state.  If a firing event is now in progress,
                    // go to "launch" state.  Otherwise, if the plunger is less
                    // than 1" retracted, go to "uncocked" state - the player
                    // might be slowly returning the plunger to rest so as not
                    // to trigger a launch.
                    if (firing || znew <= 0)
                        newState = 3;
                    else if (znew < cockThreshold)
                        newState = 2;
                case 2:
                    // Uncocked state.  If the plunger is more than an inch
                    // retracted, return to cocked state.  If we've been in
                    // the uncocked state for more than half a second, return
                    // to the base state.  This allows the user to return the
                    // plunger to rest without triggering a launch, by moving
                    // it at manual speed to the rest position rather than
                    // releasing it.
                    if (znew >= cockThreshold)
                        newState = 1;
                    else if (lbTimer.read_ms() > 500)
                        newState = 0;
                case 3:
                    // Launch state.  If the plunger is no longer pushed
                    // forward, switch to launch rest state.
                    if (znew >= 0)
                        newState = 4;
                case 4:
                    // Launch rest state.  If the plunger is pushed forward
                    // again, switch back to launch state.  If not, and we've
                    // been in this state for at least 200ms, return to the
                    // default state.
                    if (znew <= pushThreshold)
                        newState = 3;
                    else if (lbTimer.read_ms() > 200)
                        newState = 0;                    
                case 5:
                    // Press-and-Hold state.  If the plunger is no longer pushed
                    // forward, AND it's been at least 50ms since we generated
                    // the simulated Launch Ball button press, return to the base 
                    // state.  The minimum time is to ensure that VP has a chance
                    // to see the button press and to avoid transient key bounce
                    // effects when the plunger position is right on the threshold.
                    if (znew > pushThreshold && lbTimer.read_ms() > 50)
                        newState = 0;
                // change states if desired
                if (newState != lbState)
                    // If we're entering Launch state OR we're entering the
                    // Press-and-Hold state, AND the ZB Launch Ball LedWiz signal 
                    // is turned on, simulate a Launch Ball button press.
                    if (((newState == 3 && lbState != 4) || newState == 5)
                        && wizOn[cfg.plunger.zbLaunchBall.port-1])
                        simButtons |= lbButtonBit;
                    // if we're switching to state 0, release the button
                    if (newState == 0)
                        simButtons &= ~(1 << (cfg.plunger.zbLaunchBall.btn - 1));
                    // switch to the new state
                    lbState = newState;
                    // start timing in the new state
                // If the Launch Ball button press is in effect, but the
                // ZB Launch Ball LedWiz signal is no longer turned on, turn
                // off the button.
                // If we're in one of the Launch states (state #3 or #4),
                // and the button has been on for long enough, turn it off.
                // The Launch mode is triggered by a pull-and-release gesture.
                // From the user's perspective, this is just a single gesture
                // that should trigger just one momentary press on the Launch
                // Ball button.  Physically, though, the plunger usually
                // bounces back and forth for 500ms or so before coming to
                // rest after this gesture.  That's what the whole state
                // #3-#4 business is all about - we stay in this pair of
                // states until the plunger comes to rest.  As long as we're
                // in these states, we won't send duplicate button presses.
                // But we also don't want the one button press to continue 
                // the whole time, so we'll time it out now.
                // (This could be written as one big 'if' condition, but
                // I'm breaking it out verbosely like this to make it easier
                // for human readers such as myself to comprehend the logic.)
                if ((simButtons & lbButtonBit) != 0)
                    int turnOff = false;
                    // turn it off if the ZB Launch Ball signal is off
                    if (!wizOn[cfg.plunger.zbLaunchBall.port-1])
                        turnOff = true;
                    // also turn it off if we're in state 3 or 4 ("Launch"),
                    // and the button has been on long enough
                    if ((lbState == 3 || lbState == 4) && lbBtnTimer.read_ms() > 250)
                        turnOff = true;
                    // if we decided to turn off the button, do so
                    if (turnOff)
                        simButtons &= ~lbButtonBit;
            // If a firing event is in progress, generate synthetic reports to 
            // describe an idealized version of the plunger motion to VP rather 
            // than reporting the actual physical plunger position.
            // We use the synthetic reports during a release event because the
            // physical plunger motion when released is too fast for VP to track.
            // VP only syncs its internal physics model with the outside world 
            // about every 10ms.  In that amount of time, the plunger moves
            // fast enough when released that it can shoot all the way forward,
            // bounce off of the barrel spring, and rebound part of the way
            // back.  The result is the classic analog-to-digital problem of
            // sample aliasing.  If we happen to time our sample during the
            // release motion so that we catch the plunger at the peak of a
            // bounce, the digital signal incorrectly looks like the plunger
            // is moving slowly forward - VP thinks we went from fully
            // retracted to half retracted in the sample interval, whereas
            // we actually traveled all the way forward and half way back,
            // so the speed VP infers is about 1/3 of the actual speed.
            // To correct this, we take advantage of our ability to sample 
            // the CCD image several times in the course of a VP report.  If
            // we catch the plunger near the origin after we've seen it
            // retracted, we go into Release Event mode.  During this mode,
            // we stop reporting the true physical plunger position, and
            // instead report an idealized pattern: we report the plunger
            // immediately shooting forward to a position in front of the
            // park position that's in proportion to how far back the plunger
            // was just before the release, and we then report it stationary
            // at the park position.  We continue to report the stationary
            // park position until the actual physical plunger motion has
            // stabilized on a new position.  We then exit Release Event
            // mode and return to reporting the true physical position.
            if (firing)
                // Firing in progress.  Keep reporting the park position
                // until the physical plunger position comes to rest.
                const int restTol = JOYMAX/24;
                if (firing == 1)
                    // For the first couple of frames, show the plunger shooting
                    // forward past the zero point, to simulate the momentum carrying
                    // it forward to bounce off of the barrel spring.  Show the 
                    // bounce as proportional to the distance it was retracted
                    // in the prior report.
                    z = zBounce = -z0/6;
                else if (firing == 2)
                    // second frame - keep the bounce a little longer
                    z = zBounce;
                else if (firing > 4
                    && abs(znew - z0) < restTol
                    && abs(znew - z1) < restTol 
                    && abs(znew - z2) < restTol)
                    // The physical plunger has come to rest.  Exit firing
                    // mode and resume reporting the actual position.
                    firing = false;
                    z = znew;
                    // until the physical plunger comes to rest, simply 
                    // report the park position
                    z = 0;
                // not in firing mode - report the true physical position
                z = znew;

            // shift the new reading into the recent history buffer
            z2 = z1;
            z1 = z0;
            z0 = znew;

        // update the buttons
        bool buttonsChanged = readButtons(cfg);
        // send a keyboard report if we have new data to report
        if (kbState.changed)
            kbState.changed = false;

        // send the media control report, if applicable
        if (mediaState.changed)
            mediaState.changed = false;

        // If it's been long enough since our last USB status report,
        // send the new report.  We throttle the report rate because
        // it can overwhelm the PC side if we report too frequently.
        // VP only wants to sync with the real world in 10ms intervals,
        // so reporting more frequently creates I/O overhead without 
        // doing anything to improve the simulation.
        if (cfg.joystickEnabled && reportTimer.read_ms() > 10)
            // read the accelerometer
            int xa, ya;
            accel.get(xa, ya);
            // confine the results to our joystick axis range
            if (xa < -JOYMAX) xa = -JOYMAX;
            if (xa > JOYMAX) xa = JOYMAX;
            if (ya < -JOYMAX) ya = -JOYMAX;
            if (ya > JOYMAX) ya = JOYMAX;
            // store the updated accelerometer coordinates
            x = xa;
            y = ya;
            // Report the current plunger position UNLESS the ZB Launch Ball 
            // signal is on, in which case just report a constant 0 value.  
            // ZB Launch Ball turns off the plunger position because it
            // tells us that the table has a Launch Ball button instead of
            // a traditional plunger.
            int zrep = (cfg.plunger.zbLaunchBall.port != 0 && wizOn[cfg.plunger.zbLaunchBall.port-1] ? 0 : z);
            // rotate X and Y according to the device orientation in the cabinet
            accelRotate(x, y);

            // send the joystick report
            js.update(x, y, zrep, jsButtons | simButtons, statusFlags);
            // we've just started a new report interval, so reset the timer

        // If we're in pixel dump mode, report all pixel exposure values
        if (reportPix)
            // send the report            

            // we have satisfied this request
            reportPix = false;
        // If joystick reports are turned off, send a generic status report
        // periodically for the sake of the Windows config tool.
        if (!cfg.joystickEnabled && reportTimer.read_ms() > 200)

        if (x != 0 || y != 0)
            printf("%d,%d\r\n", x, y);

        // check for connection status changes
        int newConnected = js.isConnected() && !js.isSuspended();
        if (newConnected != connected)
            // give it a few seconds to stabilize
            time_t tc = time(0);
            if (tc - connectChangeTime > 3)
                // note the new status
                connected = newConnected;
                connectChangeTime = tc;
                // if we're no longer connected, turn off all outputs
                if (!connected)

        // provide a visual status indication on the on-board LED
        if (calBtnState < 2 && hbTimer.read_ms() > 1000) 
            if (!newConnected)
                // suspended - turn off the LED
                ledR = 1;
                ledG = 1;
                ledB = 1;

                // show a status flash every so often                
                if (hbcnt % 3 == 0)
                    // disconnected = red/red flash; suspended = red
                    for (int n = js.isConnected() ? 1 : 2 ; n > 0 ; --n)
                        ledR = 0;
                        ledR = 1;
            else if (cfg.plunger.enabled && !
                // connected, plunger calibration needed - flash yellow/green
                hb = !hb;
                ledR = (hb ? 0 : 1);
                ledG = 0;
                ledB = 1;
                // connected - flash blue/green
                hb = !hb;
                ledR = 1;
                ledG = (hb ? 0 : 1);
                ledB = (hb ? 1 : 0);
            // reset the heartbeat timer