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 potentiometer (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 KL25Z can only run one firmware program at a time, so if you install the Pinscape firmware on your KL25Z, it will replace and erase your existing VirtuaPin proprietary firmware. If you do this, the only way to restore your VirtuaPin firmware is to physically ship the KL25Z back to VirtuaPin and ask them to re-flash it. They don't allow you to do this at home, and they don't even allow you to back up your firmware, since they want to protect their proprietary software from copying. For all of these reasons, if you want to run the Pinscape software, I strongly recommend that you buy a "blank" retail KL25Z to use with Pinscape. They only cost about $15 and are available at several online retailers, including Amazon, Mouser, and eBay. The blank retail boards don't come with any proprietary firmware pre-installed, so installing Pinscape won't delete anything that you paid extra for.

With those warnings in mind, if you're absolutely sure that you don't mind permanently erasing your VirtuaPin firmware, it is at least possible to use Pinscape as a replacement for the VirtuaPin firmware. Pinscape uses the same button wiring conventions as the VirtuaPin setup, so you can keep your buttons (although you'll have to update the GPIO pin mappings in the Config Tool to match your physical wiring). As of the June, 2021 firmware, the Vishay VCNL4010 plunger sensor that comes with the VirtuaPin v3 plunger kit is supported, so you can also keep your plunger, if you have that chip. (You should check to be sure that's the sensor chip you have before committing to this route, if keeping the plunger sensor is important to you. The older VirtuaPin plunger kits came with different IR sensors that the Pinscape software doesn't handle.)



File content as of revision 13:72dda449c3c0:

/* Copyright 2014 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.

// Pinscape Controller
// "Pinscape" is the name of my custom-built virtual pinball cabinet.  I wrote this
// software to perform a number of tasks that I needed for my cabinet.  It runs on a
// Freescale KL25Z microcontroller, which is a small and inexpensive device that
// attaches to the host PC via USB and can interface with numerous types of external
// hardware.
// I designed the software and hardware in this project especially for Pinscape, but 
// it uses standard interfaces in Windows and Visual Pinball, so it should be
// readily usable in anyone else's VP-based cabinet.  I've tried to document the
// hardware in enough detail for anyone else to duplicate the entire project, and
// the full software is open source.
// The device appears to the host computer as a USB joystick.  This works with the
// standard Windows joystick device drivers, so there's no need to install any
// software on the PC - Windows should recognize it as a joystick when you plug
// it in and shouldn't ask you to install anything.  If you bring up the control
// panel for USB Game Controllers, this device will appear as "Pinscape Controller".
// *Don't* do any calibration with the Windows control panel or third-part 
// calibration tools.  The device calibrates itself automatically for the
// accelerometer data, and has its own special calibration procedure for the
// plunger (see below).
// The controller provides the following functions.  It should be possible to use
// any subet of the features without using all of them.  External hardware for any
// particular function can simply be omitted if that feature isn't needed.
//  - Nudge sensing via the KL25Z's on-board accelerometer.  Nudge accelerations are
//    processed into a physics model of a rolling ball, and changes to the ball's
//    motion are sent to the host computer via the joystick interface.  This is designed
//    especially to work with Visuall Pinball's nudge handling to produce realistic 
//    on-screen results in VP.  By doing some physics modeling right on the device, 
//    rather than sending raw accelerometer data to VP, we can produce better results
//    using our awareness of the real physical parameters of a pinball cabinet.
//    VP's nudge handling has to be more generic, so it can't make the same sorts
//    of assumptions that we can about the dynamics of a real cabinet.
//    The nudge data reports are compatible with the built-in Windows USB joystick 
//    drivers and with VP's own joystick input scheme, so the nudge sensing is almost 
//    plug-and-play.  There are no Windiows drivers to install, and the only VP work 
//    needed is to customize a few global preference settings.
//  - Plunger position sensing via an attached TAOS TSL 1410R CCD linear array sensor.  
//    The sensor must be wired to a particular set of I/O ports on the KL25Z, and must 
//    be positioned adjacent to the plunger with proper lighting.  The physical and
//    electronic installation details are desribed in the project documentation.  We read 
//    the CCD to determine how far back the plunger is pulled, and report this to Visual 
//    Pinball via the joystick interface.  As with the nudge data, this is all nearly
//    plug-and-play, in that it works with the default Windows USB drivers and works 
//    with the existing VP handling for analog plunger input.  A few VP settings are
//    needed to tell VP to allow the plunger.
//    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 might change slightly 
//    when you put everything back together.)  To calibrate, you have to attach a
//    momentary switch (e.g., a push-button switch) between one of the KL25Z ground
//    pins (e.g., jumper J9 pin 12) and PTE29 (J10 pin 9).  Press and hold the
//    button for about two seconds - the LED on the KL25Z wlil flash blue while
//    you hold the button, and will turn solid blue when you've held it down long
//    enough to enter calibration mode.  This mode will last about 15 seconds.
//    Simply pull the plunger all the way back, hold it for a few moments, and
//    gradually return it to the starting position.  *Don't* release it - we want
//    to measure the maximum retracted position and the rest position, but NOT
//    the maximum forward position when the outer barrel spring is compressed.
//    After about 15 seconds, the device will save the new calibration settings
//    to its flash memory, and the LED will return to the regular "heartbeat" 
//    flashes.  If this is the first time you calibrated, you should observe the
//    color of the flashes change from yellow/green to blue/green to indicate
//    that the plunger has been calibrated.
//    Note that while Visual Pinball itself has good native support for analog 
//    plungers, most of the VP tables in circulation don't implement the necessary
//    scripting features to make this work properly.  Therefore, you'll have to do
//    a little scripting work for each table you download to add the required code
//    to that individual table.  The work has to be customized for each table, so
//    I haven't been able to automate this process, but I have tried to reduce it
//    to a relatively simple recipe that I've documented separately.
//  - In addition to the CCD sensor, a button should be attached (also described in 
//    the project documentation) to activate calibration mode for the plunger.  When 
//    calibration mode is activated, the software reads the plunger position for about 
//    10 seconds when to note the limits of travel, and uses these limits to ensure
//    accurate reports to VP that properly report the actual position of the physical
//    plunger.  The calibration is stored in non-volatile memory on the KL25Z, so it's
//    only necessary to calibrate once - the calibration will survive power cycling
//    and reboots of the PC.  It's only necessary to recalibrate if the CCD sensor or
//    the plunger are removed and reinstalled, since the relative alignment of the
//    parts could cahnge slightly when reinstalling.
//  - 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.  See the buttonMap[] array
//    below for the available GPIO ports and their assigned joystick button numbers.
//    If you're not using a GPIO port, you can just leave it unconnected - the digital
//    inputs have built-in pull-up resistors, so an unconnected port is the same as
//    an open switch (an "off" state for 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!).
// The on-board LED on the KL25Z flashes to indicate the current device status:
//    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 red/green = the LedWiz unti number has been changed, so a reset
//        is needed.  You can simply unplug the device and plug it back in,
//        or presss and hold the reset button on the device for a few seconds.
//    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,
//        so you should use the config command to disable the CCD software 
//        features if you won't be attaching a CCD.
//    alternating blue/green = everything's working
// Software configuration: you can change option settings by sending special
// USB commands from the PC.  I've provided a Windows program for this purpose;
// refer to the documentation for details.  For reference, here's the format
// of the USB command for option changes:
//    length of report = 8 bytes
//    byte 0 = 65 (0x41)
//    byte 1 = 1 (0x01)
//    byte 2 = new LedWiz unit number, 0x01 to 0x0f
//    byte 3 = feature enable bit mask:
//             0x01 = enable CCD (default = on)
// Plunger calibration mode: the host can activate plunger calibration mode
// by sending this packet.  This has the same effect as pressing and holding
// the plunger calibration button for two seconds, to allow activating this
// mode without attaching a physical button.
//    length = 8 bytes
//    byte 0 = 65 (0x41)
//    byte 1 = 2 (0x02)
// Exposure reports: the host can request a report of the full set of pixel
// values for the next frame by sending this special packet:
//    length = 8 bytes
//    byte 0 = 65 (0x41)
//    byte 1 = 3 (0x03)
// We'll respond with a series of special reports giving the exposure status.
// Each report has the following structure:
//    bytes 0:1 = 11-bit index, with high 5 bits set to 10000.  For 
//                example, 0x04 0x80 indicates index 4.  This is the 
//                starting pixel number in the report.  The first report 
//                will be 0x00 0x80 to indicate pixel #0.  
//    bytes 2:3 = 16-bit unsigned int brightness level of pixel at index
//    bytes 4:5 = brightness of pixel at index+1
//    etc for the rest of the packet
// This still has the form of a joystick packet at the USB level, but
// can be differentiated by the host via the status bits.  It would have
// been cleaner to use a different Report ID at the USB level, but this
// would have necessitated a different container structure in the report
// descriptor, which would have broken LedWiz compatibility.  Given that
// constraint, we have to re-use the joystick report type, making for
// this somewhat kludgey approach.
#include "mbed.h"
#include "math.h"
#include "USBJoystick.h"
#include "MMA8451Q.h"
#include "tsl1410r.h"
#include "FreescaleIAP.h"
#include "crc32.h"

// ---------------------------------------------------------------------------
// Configuration details

// Our USB device vendor ID, product ID, and version.  
// We use the vendor ID for the LedWiz, so that the PC-side software can
// identify us as capable of performing LedWiz commands.  The LedWiz uses
// a product ID value from 0xF0 to 0xFF; the last four bits identify the
// unit number (e.g., product ID 0xF7 means unit #7).  This allows multiple
// LedWiz units to be installed in a single PC; the software on the PC side
// uses the unit number to route commands to the devices attached to each
// unit.  On the real LedWiz, the unit number must be set in the firmware
// at the factory; it's not configurable by the end user.  Most LedWiz's
// ship with the unit number set to 0, but the vendor will set different
// unit numbers if requested at the time of purchase.  So if you have a
// single LedWiz already installed in your cabinet, and you didn't ask for
// a non-default unit number, your existing LedWiz will be unit 0.
// We use unit #7 by default.  There doesn't seem to be a requirement that
// unit numbers be contiguous (DirectOutput Framework and other software
// seem happy to have units 0 and 7 installed, without 1-6 existing).
// Marking this unit as #7 should work for almost everybody out of the box;
// the most common case seems to be to have a single LedWiz installed, and
// it's probably extremely rare to more than two.
// Note that the USB_PRODUCT_ID value set here omits the unit number.  We
// take the unit number from the saved configuration.  We provide a
// configuration command that can be sent via the USB connection to change
// the unit number, so that users can select the unit number without having
// to install a different version of the software.  We'll combine the base
// product ID here with the unit number to get the actual product ID that
// we send to the USB controller.
const uint16_t USB_VENDOR_ID = 0xFAFA;
const uint16_t USB_PRODUCT_ID = 0x00F0;
const uint16_t USB_VERSION_NO = 0x0006;
const uint8_t DEFAULT_LEDWIZ_UNIT_NUMBER = 0x07;

// Number of pixels we read from the sensor on each frame.  This can be
// less than the physical pixel count if desired; we'll read every nth
// piexl if so.  E.g., with a 1280-pixel physical sensor, if npix is 320,
// we'll read every 4th pixel.  It takes time to read each pixel, so the
// fewer pixels we read, the higher the refresh rate we can achieve.
// It's therefore better not to read more pixels than we have to.
// VP seems to have an internal resolution in the 8-bit range, so there's
// no apparent benefit to reading more than 128-256 pixels when using VP.
// Empirically, 160 pixels seems about right.  The overall travel of a
// standard pinball plunger is about 3", so 160 pixels gives us resolution
// of about 1/50".  This seems to take full advantage of VP's modeling
// ability, and is probably also more precise than a human player's
// perception of the plunger position.
const int npix = 160;

// On-board RGB LED elements - we use these for diagnostic displays.
DigitalOut ledR(LED1), ledG(LED2), ledB(LED3);

// calibration button - switch input and LED output
DigitalIn calBtn(PTE29);
DigitalOut calBtnLed(PTE23);

// Joystick button input pin assignments.  You can wire up to
// 32 GPIO ports to buttons (equipped with momentary switches).
// Connect each switch between the desired GPIO port and ground
// (J9 pin 12 or 14).  When the button is pressed, we'll tell the
// host PC that the corresponding joystick button is pressed.  We
// debounce the keystrokes in software, so you can simply wire
// directly to pushbuttons with no additional external hardware.
// Note that we assign 24 buttons by default, even though the USB
// joystick interface can handle up to 32 buttons.  VP itself only
// allows mapping of up to 24 buttons in the preferences dialog 
// (although it can recognize 32 buttons internally).  If you want 
// more buttons, you can reassign pins that are assigned by default
// as LedWiz outputs.  To reassign a pin, find the pin you wish to
// reassign in the LedWizPortMap array below, and change the pin name 
// there to NC (for Not Connected).  You can then change one of the
// "NC" entries below to the reallocated pin name.  The limit is 32
// buttons total.
// Note: PTD1 (pin J2-12) should NOT be assigned as a button input,
// as this pin is physically connected on the KL25Z to the on-board
// indicator LED's blue segment.  This precludes any other use of
// the pin.
PinName buttonMap[] = {
    PTC2,      // J10 pin 10, joystick button 1
    PTB3,      // J10 pin 8,  joystick button 2
    PTB2,      // J10 pin 6,  joystick button 3
    PTB1,      // J10 pin 4,  joystick button 4
    PTE30,     // J10 pin 11, joystick button 5
    PTE22,     // J10 pin 5,  joystick button 6
    PTE5,      // J9 pin 15,  joystick button 7
    PTE4,      // J9 pin 13,  joystick button 8
    PTE3,      // J9 pin 11,  joystick button 9
    PTE2,      // J9 pin 9,   joystick button 10
    PTB11,     // J9 pin 7,   joystick button 11
    PTB10,     // J9 pin 5,   joystick button 12
    PTB9,      // J9 pin 3,   joystick button 13
    PTB8,      // J9 pin 1,   joystick button 14
    PTC12,     // J2 pin 1,   joystick button 15
    PTC13,     // J2 pin 3,   joystick button 16
    PTC16,     // J2 pin 5,   joystick button 17
    PTC17,     // J2 pin 7,   joystick button 18
    PTA16,     // J2 pin 9,   joystick button 19
    PTA17,     // J2 pin 11,  joystick button 20
    PTE31,     // J2 pin 13,  joystick button 21
    PTD6,      // J2 pin 17,  joystick button 22
    PTD7,      // J2 pin 19,  joystick button 23
    PTE1,      // J2 pin 20,  joystick button 24

    NC,        // not used,   joystick button 25
    NC,        // not used,   joystick button 26
    NC,        // not used,   joystick button 27
    NC,        // not used,   joystick button 28
    NC,        // not used,   joystick button 29
    NC,        // not used,   joystick button 30
    NC,        // not used,   joystick button 31
    NC         // not used,   joystick button 32

// LED-Wiz emulation output pin assignments.  
// The LED-Wiz protocol allows setting individual intensity levels
// on all outputs, with 48 levels of intensity.  This can be used
// to control lamp brightness and motor speeds, among other things.
// Unfortunately, the KL25Z only has 10 PWM channels, so while we 
// can support the full complement of 32 outputs, we can only provide 
// PWM dimming/speed control on 10 of them.  The remaining outputs 
// can only be switched fully on and fully off - we can't support
// dimming on these, so they'll ignore any intensity level setting 
// requested by the host.  Use these for devices that don't have any
// use for intensity settings anyway, such as contactors and knockers.
// Ports with pins assigned as "NC" are not connected.  That is,
// there's no physical pin for that LedWiz port number.  You can
// send LedWiz commands to turn NC ports on and off, but doing so
// will have no effect.  The reason we leave some ports unassigned
// is that we don't have enough physical GPIO pins to fill out the
// full LedWiz complement of 32 ports.  Many pins are already taken
// for other purposes, such as button inputs or the plunger CCD
// interface.
// The mapping between physical output pins on the KL25Z and the
// assigned LED-Wiz port numbers is essentially arbitrary - you can
// customize this by changing the entries in the array below if you
// wish to rearrange the pins for any reason.  Be aware that some
// of the physical outputs are already used for other purposes
// (e.g., some of the GPIO pins on header J10 are used for the
// CCD sensor - but you can of course reassign those as well by
// changing the corresponding declarations elsewhere in this module).
// The assignments we make here have two main objectives: first,
// to group the outputs on headers J1 and J2 (to facilitate neater
// wiring by keeping the output pins together physically), and
// second, to make the physical pin layout match the LED-Wiz port
// numbering order to the extent possible.  There's one big wrench
// in the works, though, which is the limited number and discontiguous
// placement of the KL25Z PWM-capable output pins.  This prevents
// us from doing the most obvious sequential ordering of the pins,
// so we end up with the outputs arranged into several blocks.
// Hopefully this isn't too confusing; for more detailed rationale,
// read on...
// With the LED-Wiz, the host software configuration usually 
// assumes that each RGB LED is hooked up to three consecutive ports
// (for the red, green, and blue components, which need to be 
// physically wired to separate outputs to allow each color to be 
// controlled independently).  To facilitate this, we arrange the 
// PWM-enabled outputs so that they're grouped together in the 
// port numbering scheme.  Unfortunately, these outputs aren't
// together in a single group in the physical pin layout, so to
// group them logically in the LED-Wiz port numbering scheme, we
// have to break up the overall numbering scheme into several blocks.
// So our port numbering goes sequentially down each column of
// header pins, but there are several break points where we have
// to interrupt the obvious sequence to keep the PWM pins grouped
// logically.
// In the list below, "pin J1-2" refers to pin 2 on header J1 on
// the KL25Z, using the standard pin numbering in the KL25Z 
// documentation - this is the physical pin that the port controls.
// "LW port 1" means LED-Wiz port 1 - this is the LED-Wiz port
// number that you use on the PC side (in the DirectOutput config
// file, for example) to address the port.  PWM-capable ports are
// marked as such - we group the PWM-capable ports into the first
// 10 LED-Wiz port numbers.
// If you wish to reallocate a pin in the array below to some other
// use, such as a button input port, simply change the pin name in
// the entry to NC (for Not Connected).  This will disable the given
// logical LedWiz port number and free up the physical pin.
// If you wish to reallocate a pin currently assigned to the button
// input array, simply change the entry for the pin in the buttonMap[]
// array above to NC (for "not connected"), and plug the pin name into
// a slot of your choice in the array below.
// Note: PTD1 (pin J2-12) should NOT be assigned as an LedWiz output,
// as this pin is physically connected on the KL25Z to the on-board
// indicator LED's blue segment.  This precludes any other use of
// the pin.
struct {
    PinName pin;
    bool isPWM;
} ledWizPortMap[32] = {
    { PTA1, true },      // pin J1-2,  LW port 1  (PWM capable - TPM 2.0 = channel 9)
    { PTA2, true },      // pin J1-4,  LW port 2  (PWM capable - TPM 2.1 = channel 10)
    { PTD4, true },      // pin J1-6,  LW port 3  (PWM capable - TPM 0.4 = channel 5)
    { PTA12, true },     // pin J1-8,  LW port 4  (PWM capable - TPM 1.0 = channel 7)
    { PTA4, true },      // pin J1-10, LW port 5  (PWM capable - TPM 0.1 = channel 2)
    { PTA5, true },      // pin J1-12, LW port 6  (PWM capable - TPM 0.2 = channel 3)
    { PTA13, true },     // pin J2-2,  LW port 7  (PWM capable - TPM 1.1 = channel 13)
    { PTD5, true },      // pin J2-4,  LW port 8  (PWM capable - TPM 0.5 = channel 6)
    { PTD0, true },      // pin J2-6,  LW port 9  (PWM capable - TPM 0.0 = channel 1)
    { PTD3, true },      // pin J2-10, LW port 10 (PWM capable - TPM 0.3 = channel 4)
    { PTD2, false },     // pin J2-8,  LW port 11
    { PTC8, false },     // pin J1-14, LW port 12
    { PTC9, false },     // pin J1-16, LW port 13
    { PTC7, false },     // pin J1-1,  LW port 14
    { PTC0, false },     // pin J1-3,  LW port 15
    { PTC3, false },     // pin J1-5,  LW port 16
    { PTC4, false },     // pin J1-7,  LW port 17
    { PTC5, false },     // pin J1-9,  LW port 18
    { PTC6, false },     // pin J1-11, LW port 19
    { PTC10, false },    // pin J1-13, LW port 20
    { PTC11, false },    // pin J1-15, LW port 21
    { PTE0, false },     // pin J2-18, LW port 22
    { NC, false },       // Not used,  LW port 23
    { NC, false },       // Not used,  LW port 24
    { NC, false },       // Not used,  LW port 25
    { NC, false },       // Not used,  LW port 26
    { NC, false },       // Not used,  LW port 27
    { NC, false },       // Not used,  LW port 28
    { NC, false },       // Not used,  LW port 29
    { NC, false },       // Not used,  LW port 30
    { NC, false },       // Not used,  LW port 31
    { NC, false }        // Not used,  LW port 32

// 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

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

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

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

// ---------------------------------------------------------------------------
// LedWiz emulation

static int pbaIdx = 0;

// LedWiz output pin interface.  We create a cover class to virtualize
// digital vs PWM outputs and give them a common interface.  The KL25Z
// unfortunately doesn't have enough hardware PWM channels to support 
// PWM on all 32 LedWiz outputs, so we provide as many PWM channels as
// we can (10), and fill out the rest of the outputs with plain digital
// outs.
class LwOut
    virtual void set(float val) = 0;
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;
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;
class LwUnusedOut: public LwOut
    LwUnusedOut() { }
    virtual void set(float val) { }

// output pin array
static LwOut *lwPin[32];

// initialize the output pin array
void initLwOut()
    for (int i = 0 ; i < countof(lwPin) ; ++i)
        PinName p = (i < countof(ledWizPortMap) ? ledWizPortMap[i].pin : NC);
        if (p == NC)
            lwPin[i] = new LwUnusedOut();
        else if (ledWizPortMap[i].isPWM)
            lwPin[i] = new LwPwmOut(p);
            lwPin[i] = new LwDigOut(p);

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

// profile (brightness/blink) state for each LedWiz output
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

static float wizState(int idx)
    if (wizOn[idx]) 
        // on - map profile brightness state to PWM level
        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
            return val/48.0;
        else if (val == 49)
            // 49 is undefined in the LedWiz documentation.  Even so, DOF2
            // *does* set outputs to 49 in some cases where it intends for
            // them to be fully on.  This is a DOF2 bug, but the real LedWiz 
            // treats 49 as fully on, so it's a harmless bug when used with 
            // real LedWiz units.  For the sake of bug-for-bug compatibility, 
            // we must do the same thing.
            return 1.0;
        else if (val >= 129 && val <= 132)
            // Values of 129-132 select different flashing modes.  We don't
            // support any of these.  Instead, simply treat them as fully on.  
            // Note that DOF doesn't ever use modes 129-132, as it implements 
            // all flashing modes itself on the host side, so this limitation 
            // won't have any effect on DOF users.  You can observe it using 
            // LedBlinky, though.
            return 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).
            // NB: the 49 and 129-132 cases are broken out above for the sake
            // of documentation.  We end up using 1.0 as the return value for
            // everything outside of the defined 0-48 range, so we could collapse
            // this whole thing to a single 'else' branch, but I wanted to call 
            // out the specific reasons for handling the settings above as we do.
            return 1.0;
        // off - show at 0 intensity
        return 0.0;

static void updateWizOuts()
    for (int i = 0 ; i < 32 ; ++i)

// ---------------------------------------------------------------------------
// Button input

// button input map array
DigitalIn *buttonDigIn[32];

// timer for button reports
static Timer buttonTimer;

// initialize the button inputs
void initButtons()
    // create the digital inputs
    for (int i = 0 ; i < countof(buttonDigIn) ; ++i)
        if (i < countof(buttonMap) && buttonMap[i] != NC)
            buttonDigIn[i] = new DigitalIn(buttonMap[i]);
            buttonDigIn[i] = 0;
    // start the button timer

// read the raw button input state
uint32_t readButtonsRaw()
    // start with all buttons off
    uint32_t buttons = 0;
    // scan the button list
    uint32_t bit = 1;
    for (int i = 0 ; i < countof(buttonDigIn) ; ++i, bit <<= 1)
        if (buttonDigIn[i] != 0 && !buttonDigIn[i]->read())
            buttons |= bit;
    // return the button list
    return buttons;

// Read buttons with debouncing.  We keep a circular buffer
// of recent input readings.  We'll AND together the status of
// each button over the past 50ms.  A button that has been on
// continuously for 50ms will be reported as ON.  All others
// will be reported as OFF.
uint32_t readButtonsDebounced()
    struct reading {
        int dt;           // time since previous reading
        uint32_t b;       // button state at this reading
    static reading readings[8];  // circular buffer of readings
    static int ri = 0;    // reading buffer index (next write position)
    // get the write pointer
    reading *r = &readings[ri];

    // figure the time since the last reading, and read the raw button state
    r->dt = buttonTimer.read_ms();
    uint32_t b = r->b = readButtonsRaw();
    // start timing the next interval
    // AND together readings over 50ms
    int ms = 0;
    for (int i = 1 ; i < countof(readings) && ms < 50 ; ++i)
        // find the next prior reading, wrapping in the circular buffer
        int j = ri - i;
        if (j < 0) 
            j = countof(readings) - 1;
        reading *rj = &readings[j];
        // AND the buttons for this reading
        b &= rj->b;
        // count the time
        ms += rj->dt;
    // advance the write position for next time
    ri += 1;
    if (ri >= countof(readings)) 
        ri = 0;
    // return the debounced result
    return b;

// ---------------------------------------------------------------------------
// Non-volatile memory (NVM)

// Structure defining our NVM storage layout.  We store a small
// amount of persistent data in flash memory to retain calibration
// data when powered off.
struct NVM
    // checksum - we use this to determine if the flash record
    // has been properly initialized
    uint32_t checksum;

    // signature value
    static const uint32_t SIGNATURE = 0x4D4A522A;
    static const uint16_t VERSION = 0x0003;
    // Is the data structure valid?  We test the signature and 
    // checksum to determine if we've been properly stored.
    int valid() const
        return (d.sig == SIGNATURE 
                && d.vsn == VERSION
                && == sizeof(NVM)
                && checksum == CRC32(&d, sizeof(d)));
    // save to non-volatile memory
    void save(FreescaleIAP &iap, int addr)
        // update the checksum and structure size
        checksum = CRC32(&d, sizeof(d)); = sizeof(NVM);
        // erase the sector

        // save the data
        iap.program_flash(addr, this, sizeof(*this));
    // reset calibration data for calibration mode
    void resetPlunger()
        // set extremes for the calibration data
        d.plungerMax = 0;
        d.plungerZero = npix;
        d.plungerMin = npix;
    // stored data (excluding the checksum)
        // Signature, structure version, and structure size - further verification 
        // that we have valid initialized data.  The size is a simple proxy for a
        // structure version, as the most common type of change to the structure as
        // the software evolves will be the addition of new elements.  We also
        // provide an explicit version number that we can update manually if we
        // make any changes that don't affect the structure size but would affect
        // compatibility with a saved record (e.g., swapping two existing elements).
        uint32_t sig;
        uint16_t vsn;
        int sz;
        // has the plunger been manually calibrated?
        int plungerCal;
        // plunger calibration min and max
        int plungerMin;
        int plungerZero;
        int plungerMax;
        // is the CCD enabled?
        int ccdEnabled;
        // LedWiz unit number
        uint8_t ledWizUnitNo;
    } d;

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

class MyUSBJoystick: public USBJoystick
    MyUSBJoystick(uint16_t vendor_id, uint16_t product_id, uint16_t product_release) 
        : USBJoystick(vendor_id, product_id, product_release, true)
        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_; 

// ---------------------------------------------------------------------------
// Some simple math service routines

inline float square(float x) { return x*x; }
inline float round(float x) { return x > 0 ? floor(x + 0.5) : ceil(x - 0.5); }

// ---------------------------------------------------------------------------
// 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).

// 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 MMA8451!.  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 conidtion.
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;

// ---------------------------------------------------------------------------
// 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;
    // initialize the LedWiz ports
    // initialize the button input ports
    // we don't need a reset yet
    bool needReset = false;
    // clear the I2C bus for the accelerometer
    // set up a flash memory controller
    FreescaleIAP iap;
    // use the last sector of flash for our non-volatile memory structure
    int flash_addr = (iap.flash_size() - SECTOR_SIZE);
    NVM *flash = (NVM *)flash_addr;
    NVM cfg;
    // check for valid flash
    bool flash_valid = flash->valid();
    // if the flash is valid, load it; otherwise initialize to defaults
    if (flash_valid) {
        memcpy(&cfg, flash, sizeof(cfg));
        printf("Flash restored: plunger cal=%d, min=%d, zero=%d, max=%d\r\n", 
            cfg.d.plungerCal, cfg.d.plungerMin, cfg.d.plungerZero, cfg.d.plungerMax);
    else {
        printf("Factory reset\r\n");
        cfg.d.sig = cfg.SIGNATURE;
        cfg.d.vsn = cfg.VERSION;
        cfg.d.plungerCal = 0;
        cfg.d.plungerZero = 0;
        cfg.d.plungerMin = 0;
        cfg.d.plungerMax = npix;
        cfg.d.ledWizUnitNo = DEFAULT_LEDWIZ_UNIT_NUMBER;
        cfg.d.ccdEnabled = true;
    // Create the joystick USB client.  Note that we use the LedWiz unit
    // number from the saved configuration.
    MyUSBJoystick js(
        USB_PRODUCT_ID | cfg.d.ledWizUnitNo,

    // plunger calibration button debounce timer
    Timer calBtnTimer;
    int calBtnLit = 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;
    // 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);
    // create the CCD array object
    TSL1410R ccd(PTE20, PTE21, PTB0);
    // last accelerometer report, in mouse coordinates
    int x = 0, y = 0, z = 0;
    // previous two plunger readings, for "debouncing" the results (z0 is
    // the most recent, z1 is the one before that)
    int z0 = 0, z1 = 0, z2 = 0;
    // Firing in progress: we set this when we detect the start of rapid 
    // plunger movement from a retracted position towards the rest position.
    // The actual plunger spring return speed seems to be too slow for VP, 
    // so when we detect the start of this motion, we immediately tell VP
    // to return the plunger to rest, then we monitor the real plunger 
    // until it atcually stops.
    int firing = 0;

    // start the first CCD integration cycle
    // 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:
    //    0x01  -> plunger sensor enabled
    uint16_t statusFlags = (cfg.d.ccdEnabled ? 0x01 : 0x00);
    // flag: send a pixel dump after the next read
    bool reportPix = false;

    // we're all set up - now just loop, processing sensor reports and 
    // host requests
    for (;;)
        // Look for an incoming report.  Continue processing input as
        // long as there's anything pending - this ensures that we
        // handle input in as timely a fashion as possible by deferring
        // output tasks as long as there's input to process.
        HID_REPORT report;
        while (js.readNB(&report))
            // all Led-Wiz reports are 8 bytes exactly
            if (report.length == 8)
                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 (0-7)
                    //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 < 32 ; ++i, bit <<= 1)
                        if (bit == 0x100) {
                            bit = 1;
                        wizOn[i] = ((data[ri] & bit) != 0);
                    // update the physical outputs
                    // 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 = Set Configuration:
                        //     data[2] = LedWiz unit number (0x00 to 0x0f)
                        //     data[3] = feature enable bit mask:
                        //               0x01 = enable CCD
                        // we'll need a reset if the LedWiz unit number is changing
                        uint8_t newUnitNo = data[2] & 0x0f;
                        needReset |= (newUnitNo != cfg.d.ledWizUnitNo);
                        // set the configuration parameters from the message
                        cfg.d.ledWizUnitNo = newUnitNo;
                        cfg.d.ccdEnabled = data[3] & 0x01;
                        // update the status flags
                        statusFlags = (statusFlags & ~0x01) | (data[3] & 0x01);
                        // if the ccd is no longer enabled, use 0 for z reports
                        if (!cfg.d.ccdEnabled)
                            z = 0;
                        // save the configuration
              , flash_addr);
                    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;
                    // LWZ-PBA - full state dump; each byte is one output
                    // in the current bank.  pbaIdx keeps track of the bank;
                    // this is incremented implicitly by each PBA message.
                    //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                    
                    if (pbaIdx == 24)
                        pbaIdx = 0;
                        pbaIdx += 8;
        // check for plunger calibration
        if (!calBtn)
            // 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;
            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
                cfg.d.plungerCal = 1;
      , flash_addr);
                // the flash state is now valid
                flash_valid = true;
            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) {
                calBtnLed = 1;
                ledR = 1;
                ledG = 1;
                ledB = 0;
            else {
                calBtnLed = 0;
                ledR = 1;
                ledG = 1;
                ledB = 1;
        // read the plunger sensor, if it's enabled
        uint16_t pix[npix];
        if (cfg.d.ccdEnabled)
            // start with the previous reading, in case we don't have a
            // clear result on this frame
            int znew = z;

            // read the array
  , npix);
            // get the average brightness at each end of the sensor
            long avg1 = (long(pix[0]) + long(pix[1]) + long(pix[2]) + long(pix[3]) + long(pix[4]))/5;
            long avg2 = (long(pix[npix-1]) + long(pix[npix-2]) + long(pix[npix-3]) + long(pix[npix-4]) + long(pix[npix-5]))/5;
            // figure the midpoint in the brightness; multiply by 3 so that we can
            // compare sums of three pixels at a time to smooth out noise
            long midpt = (avg1 + avg2)/2 * 3;
            // Work from the bright end to the dark end.  VP interprets the
            // Z axis value as the amount the plunger is pulled: zero is the
            // rest position, and the axis maximum is fully pulled.  So we 
            // essentially want to report how much of the sensor is lit,
            // since this increases as the plunger is pulled back.
            int si = 1, di = 1;
            if (avg1 < avg2)
                si = npix - 2, di = -1;
            // If the bright end and dark end don't differ by enough, skip this
            // reading entirely - we must have an overexposed or underexposed frame.
            // Otherwise proceed with the scan.
            if (labs(avg1 - avg2) > 0x1000)
                uint16_t *pixp = pix + si;           
                for (int n = 1 ; n < npix - 1 ; ++n, pixp += di)
                    // if we've crossed the midpoint, report this position
                    if (long(pixp[-1]) + long(pixp[0]) + long(pixp[1]) < midpt)
                        // note the new position
                        int pos = n;
                        // Calibrate, or apply calibration, depending on the mode.
                        // In either case, normalize to our range.  VP appears to
                        // ignore negative Z axis values.
                        if (calBtnState == 3)
                            // calibrating - note if we're expanding the calibration envelope
                            if (pos < cfg.d.plungerMin)
                                cfg.d.plungerMin = pos;
                            if (pos < cfg.d.plungerZero)
                                cfg.d.plungerZero = pos;
                            if (pos > cfg.d.plungerMax)
                                cfg.d.plungerMax = pos;
                            // normalize to the full physical range while calibrating
                            znew = int(round(float(pos)/npix * JOYMAX));
                            // Running normally - normalize to the calibration range.  Note
                            // that values below the zero point are allowed - the zero point
                            // represents the park position, where the plunger sits when at
                            // rest, but a mechanical plunger has a smmall amount of travel
                            // in the "push" direction.  We represent forward travel with
                            // negative z values.
                            if (pos > cfg.d.plungerMax)
                                pos = cfg.d.plungerMax;
                            znew = int(round(float(pos - cfg.d.plungerZero)
                                / (cfg.d.plungerMax - cfg.d.plungerZero + 1) * JOYMAX));
                        // done

            // Determine if the plunger is being fired - i.e., if the player
            // has just released the plunger from a retracted position.
            // We treat firing as an event.  That is, we tell VP when the
            // plunger is fired, and then stop sending data until the firing
            // is complete, allowing VP to carry out the firing motion using
            // its internal model plunger rather than trying to track the
            // intermediate positions of the mechanical plunger throughout
            // the firing motion.  This is essential because the firing
            // motion is too fast for us to track - in the time it takes us
            // to read one frame, the plunger can make it all the way to the
            // zero position and bounce back halfway.  Fortunately, the range
            // of motions for the plunger is limited, so if we see any rapid
            // change of position toward the rest position, it's reasonably
            // safe to interpret it as a firing event.  
            // This isn't foolproof.  The user can trick us by doing a 
            // controlled rapid forward push but stopping short of the rest 
            // position.  We'll misinterpret that as a firing event.  But 
            // that's not a natural motion that a user would make with a
            // plunger, so it's probably an acceptable false positive.
            // Possible future enhancement: we could add a second physical
            // sensor that detects when the plunger reaches the zero position
            // and asserts an interrupt.  In the interrupt handler, set a
            // flag indicating the zero position signal.  On each scan of
            // the CCD, also check that flag; if it's set, enter firing
            // event mode just as we do now.  The key here is that the
            // secondary sensor would have to be something much faster
            // than our CCD scan - it would have to react on, say, the
            // millisecond time scale.  A simple mechanical switch or a
            // proximity sensor could work.  This would let us detect
            // with certainty when the plunger physically fires, eliminating
            // the case where the use can fool us with motion that's fast
            // enough to look like a release but doesn't actually reach the
            // starting position.
            // To detremine when a firing even occurs, we watch for rapid
            // motion from a retracted position towards the rest position -
            // that is, large position changes in the negative direction over
            // a couple of consecutive readings.  When we see a rapid move
            // toward zero, we set our internal 'firing' flag, immediately
            // report to VP that the plunger has returned to the zero 
            // position, and then suspend reports until the mechanical
            // readings indicate that the plunger has come to rest (indicated
            // by several readings in a row at roughly the same position).
            // Tolerance for firing is 1/3 of the current pull distance, or
            // about 1/2", whichever is greater.  Making this value too small
            // makes for too many false positives.  Empirically, 1/4" is too
            // twitchy, so set a floor at about 1/2".  But we can be less
            // sensitive the further back the plunger is pulled, since even
            // a long pull will snap back quickly.  Note that JOYMAX always
            // corresponds to about 3", no matter how many pixels we're
            // reading, since the physical sensor is about 3" long; so we
            // factor out the pixel count calculate (approximate) physical
            // distances based on the normalized axis range.
            // Firing pattern: when firing, don't simply report a solid 0,
            // but instead report a series of pseudo-bouces.  This looks
            // more realistic, beacause the real plunger is also bouncing
            // around during this time.  To get maximum firing power in
            // the simulation, though, our pseudo-bounces are tiny cmopared
            // to the real thing.
            const int restTol = JOYMAX/24;
            int fireTol = z/3 > JOYMAX/6 ? z/3 : JOYMAX/6;
            static const int firePattern[] = { 
                -JOYMAX/12, -JOYMAX/12, -JOYMAX/12, 
            if (firing != 0)
                // Firing in progress - we've already told VP to send its
                // model plunger all the way back to the rest position, so
                // send no further reports until the mechanical plunger
                // actually comes to rest somewhere.
                if (abs(z0 - z2) < restTol && abs(znew - z2) < restTol)
                    // the plunger is back at rest - firing is done
                    firing = 0;
                    // resume normal reporting
                    z = z2;
                else if (firing < countof(firePattern))
                    // firing - report the next position in the pseudo-bounce 
                    // pattern
                    z = firePattern[firing++];
                    // firing, out of pseudo-bounce items - just report the
                    // rest position
                    z = 0;
            else if (z0 < z2 && z1 < z2 && znew < z2
                     && (z0 < z2 - fireTol 
                         || z1 < z2 - fireTol
                         || znew < z2 - fireTol))
                // Big jumps toward rest position in last two readings - 
                // firing has begun.  Report an immediate return to the
                // rest position, and send no further reports until the
                // physical plunger has come to rest.  This effectively
                // detaches VP's model plunger from the real world for
                // the duration of the spring return, letting VP evolve
                // its model without trying to synchronize with the
                // mechanical version.  The release motion is too fast
                // for that to work well; we can't take samples quickly
                // enough to get prcise velocity or acceleration
                // readings.  It's better to let VP figure the speed
                // and acceleration through modeling.  Plus, that lets
                // each virtual table set the desired parameters for its
                // virtual plunger, rather than imposing the actual
                // mechanical charateristics of the physical plunger on
                // every table.
                firing = 1;
                // report the first firing pattern position
                z = firePattern[0];
                // everything normal; report the 3rd recent position on
                // tape delay
                z = z2;
            // shift in the new reading
            z2 = z1;
            z1 = z0;
            z0 = znew;
            // plunger disabled - pause 10ms to throttle updates to a
            // reasonable pace

        // 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;
        // update the buttons
        uint32_t buttons = readButtonsDebounced();
        // Send the status report.  Note that the nominal x and y axes
        // are reversed - this makes it more intuitive to set up in VP.
        // If we mount the Freesale card flat on the floor of the cabinet
        // with the USB connectors facing the front of the cabinet, this
        // arrangement of our nominal axes aligns with VP's standard
        // setting, so that we can configure VP with X Axis = X on the
        // joystick and Y Axis = Y on the joystick.
        js.update(y, x, z, buttons, statusFlags);
        // If we're in pixel dump mode, report all pixel exposure values
        if (reportPix)
            // we have satisfied this request
            reportPix = false;
            // send reports for all pixels
            int idx = 0;
            while (idx < npix)
                js.updateExposure(idx, npix, pix);
            // The pixel dump requires many USB reports, since each report
            // can only send a few pixel values.  An integration cycle has
            // been running all this time, since each read starts a new
            // cycle.  Our timing is longer than usual on this round, so
            // the integration won't be comparable to a normal cycle.  Throw
            // this one away by doing a read now, and throwing it away - that 
            // will get the timing of the *next* cycle roughly back to normal.
  , npix);
        if (x != 0 || y != 0)
            printf("%d,%d\r\n", x, y);

        // provide a visual status indication on the on-board LED
        if (calBtnState < 2 && hbTimer.read_ms() > 1000) 
            if (js.isSuspended() || !js.isConnected())
                // 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 (needReset)
                // connected, need to reset due to changes in config parameters -
                // flash red/green
                hb = !hb;
                ledR = (hb ? 0 : 1);
                ledG = (hb ? 1 : 0);
                ledB = 0;
            else if (cfg.d.ccdEnabled && !cfg.d.plungerCal)
                // 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