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

/media/uploads/mjr/pinscape_no_background_small_L7Miwr6.jpg

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 vpforums.org. Also visit my Pinscape Resources page for more about this project and other virtual pinball projects I'm working on.

Downloads

  • 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.

Documentation

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 mouser.com 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 Aliexpress.com 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.)

Revision:
5:a70c0bce770d
Parent:
4:02c7cd7b2183
Child:
6:cc35eb643e8f
--- a/main.cpp	Thu Jul 24 05:50:36 2014 +0000
+++ b/main.cpp	Sun Jul 27 18:24:51 2014 +0000
@@ -1,3 +1,109 @@
+/* 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.
+*
+* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING
+* BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+* NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
+* DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN 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 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.
+//
+//    Unfortunately, analog plungers are not well supported by individual tables,
+//    so some work is required for each table to give it proper support.  I've tried
+//    to reduce this to a recipe and document it in the project documentation.
+//
+//  - 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.
+//
+//  - 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!).
+
+
 #include "mbed.h"
 #include "USBJoystick.h"
 #include "MMA8451Q.h"
@@ -5,25 +111,35 @@
 #include "FreescaleIAP.h"
 #include "crc32.h"
 
-// customization of the joystick class to expose connect/suspend status
-class MyUSBJoystick: public USBJoystick
-{
-public:
-    MyUSBJoystick(uint16_t vendor_id, uint16_t product_id, uint16_t product_release) 
-        : USBJoystick(vendor_id, product_id, product_release, false)
-    {
-        suspended_ = false;
-    }
-    
-    int isConnected() { return configured(); }
-    int isSuspended() const { return suspended_; }
-    
-protected:
-    virtual void suspendStateChanged(unsigned int suspended)
-        { suspended_ = suspended; }
+
+// ---------------------------------------------------------------------------
+//
+// Configuration details
+//
 
-    int suspended_; 
-};
+// 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.
+const uint16_t USB_VENDOR_ID = 0xFAFA;
+const uint16_t USB_PRODUCT_ID = 0x00F7;
+const uint16_t USB_VERSION_NO = 0x0004;
 
 // On-board RGB LED elements - we use these for diagnostic displays.
 DigitalOut ledR(LED1), ledG(LED2), ledB(LED3);
@@ -32,6 +148,24 @@
 DigitalIn calBtn(PTE29);
 DigitalOut calBtnLed(PTE23);
 
+// 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
+
+
+// ---------------------------------------------------------------------------
+//
+// LedWiz emulation
+//
+
 static int pbaIdx = 0;
 
 // on/off state for each LedWiz output
@@ -70,22 +204,14 @@
     ledB = wizState(2);
 }
 
-struct AccPrv
-{
-    AccPrv() : x(0), y(0) { }
-    float x;
-    float y;
-    
-    double dist(AccPrv &b)
-    {
-        float dx = x - b.x, dy = y - b.y;
-        return sqrt(dx*dx + dy*dy);
-    }
-};
+// ---------------------------------------------------------------------------
+//
+// Non-volatile memory (NVM)
+//
 
-// Non-volatile memory structure.  We store persistent a small
+// Structure defining our NVM storage layout.  We store a small
 // amount of persistent data in flash memory to retain calibration
-// data between sessions.
+// data when powered off.
 struct NVM
 {
     // checksum - we use this to determine if the flash record
@@ -113,33 +239,211 @@
     } d;
 };
 
-// Accelerometer handler
-const int MMA8451_I2C_ADDRESS = (0x1d<<1);
+
+// ---------------------------------------------------------------------------
+//
+// Customization joystick subbclass
+//
+
+class MyUSBJoystick: public USBJoystick
+{
+public:
+    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_; }
+    
+protected:
+    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 some
+// special data processing to produce more realistic results in
+// Visual Pinball.
+//
+// We install an interrupt handler on the accelerometer "data ready" 
+// interrupt in order to ensure that we fetch each sample immediately
+// when it becomes available.  Since our main program loop is busy
+// reading the CCD virtually all of the time, it wouldn't be practical
+// to keep up with the accelerometer data stream by polling.
+//
+// Visual Pinball is nominally designed to accept raw accelerometer
+// data as nudge input, but in practice, this doesn't produce
+// very realistic results.  VP simply applies accelerations from a
+// physical accelerometer directly to its modeled ball(s), but the
+// data stream coming from a real accelerometer isn't as clean as
+// an idealized physics simulation.  The problem seems to be that the
+// accelerometer samples capture instantaneous accelerations, not
+// integrated acceleration over time.  In other words, adding samples 
+// over time doesn't accurately reflect the actual net acceleration
+// experienced.  The longer the sampling period, the greater the
+// divergence between the sum of a series of samples and the actual
+// net acceleration.  The effect in VP is to leave the ball with
+// an unrealistically high residual velocity over the course of a
+// nudge event.
+//
+// This is where our custom data processing comes into play.  Rather
+// than sending raw accelerometer samples, we apply the samples to
+// our own virtual model ball.  What we send VP is the accelerations
+// experienced by the ball in our model, not the actual accelerations
+// we read from the MMA8451Q.  Now, that might seem like an unnecessary
+// middleman, because VP is just going to apply the accelerations to
+// its own model ball.  But it's a useful middleman: what we can do
+// in our model that VP can't do in its model is take into account
+// our special knowledge of the physical cabinet configuration.  VP
+// has to work generically with any sort of nudge input device, but
+// we can make assumptions about what kind of physical environment
+// we're operating in.
+//
+// The key assumption we make about our physical environment is that
+// accelerations from nudges should net out to zero over intervals on
+// the order of a couple of seconds.  Nudging a pinball cabinet makes
+// the cabinet accelerate briefly in the nudge direction, then rebound,
+// then re-rebound, and so on until the swaying motion damps out and
+// the table returns roughly to rest.  The table doesn't actually go
+// anywhere in these transactions, so the net acceleration experienced
+// is zero by the time the motion has damped out.  The damping time
+// depends on the degree of force of the nudge, but is a second or
+// two in most cases.
+//
+// We can't just assume that all motion and/or acceleration must stop 
+// in a second or two, though.  For one thing, the player can nudge
+// the table repeatedly for long periods.  (Doing this too aggressivly
+// will trigger a tilt, so there are limits, but a skillful player
+// can keep nudging a table almost continuously without tilting it.)
+// For another, a player could actually pick up one end of the table
+// for an extended period, applying a continuous acceleration the
+// whole time.
+//
+// The strategy we use to cope with these possibilities is to model a
+// ball, rather like VP does, but with damping that scales with the
+// current speed.  We'll choose a damping function that will bring
+// the ball to rest from any reasonable speed within a second or two
+// if there are no ongoing accelerations.  The damping function must
+// also be weak enough that new accelerations dominate - that is,
+// the damping function must not be so strong that it cancels out
+// ongoing physical acceleration input, such as when the player
+// lifts one end of the table and holds it up for a while.
+//
+// What we report to VP is the acceleration experienced by our model
+// ball between samples.  Our model ball starts at rest, and our damping
+// function ensures that when it's in motion, it will return to rest in
+// a short time in the absence of further physical accelerations.  The
+// sum or our reports to VP from a rest state to a subsequent rest state
+// will thus necessarily equal exactly zero.  This will ensure that we 
+// don't leave VP's model ball with any residual velocity after an 
+// isolated nudge.
+//
+// We do one more bit of data processing: automatic calibration.  When
+// we observe the accelerometer input staying constant (within a noise
+// window) for a few seconds continously, we'll assume that the cabinet
+// is at rest.  It's safe to assume that the accelerometer isn't
+// installed in such a way that it's perfectly level, so at the
+// cabinet's neutral rest position, we can expect to read non-zero
+// accelerations on the x and y axes from the component along that
+// axis of the Earth's gravity.  By watching for constant acceleration
+// values over time, we can infer the reseting position of the device
+// and take that as our zero point.  By doing this continuously, we
+// don't have to assume that the machine is perfectly motionless when
+// initially powered on - we'll organically find the zero point as soon
+// as the machine is undisturbed for a few moments.  We'll also deal
+// gracefully with situations where the machine is jolted so much in
+// the course of play that its position is changed slightly.  The result
+// should be to make the zeroing process reliable and completely 
+// transparent to the user.
+//
+
+// point structure
+struct FPoint
+{
+    float x, y;
+    
+    FPoint() { }
+    FPoint(float x, float y) { this->x = x; this->y = y; }
+    
+    void set(float x, float y) { this->x = x; this->y = y; }
+    void zero() { this->x = this->y = 0; }
+    
+    FPoint &operator=(FPoint &pt) { this->x = pt.x; this->y = pt.y; return *this; }
+    FPoint &operator-=(FPoint &pt) { this->x -= pt.x; this->y -= pt.y; return *this; }
+    FPoint &operator+=(FPoint &pt) { this->x += pt.x; this->y += pt.y; return *this; }
+    FPoint &operator*=(float f) { this->x *= f; this->y *= f; return *this; }
+    FPoint &operator/=(float f) { this->x /= f; this->y /= f; return *this; }
+    float magnitude() const { return sqrt(x*x + y*y); }
+    
+    float distance(FPoint &b)
+    {
+        float dx = x - b.x;
+        float dy = y - b.y;
+        return sqrt(dx*dx + dy*dy);
+    }
+};
+
+
+// accelerometer wrapper class
 class Accel
 {
 public:
     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
+        reset();
+    }
+    
+    void reset()
+    {
+        // assume initially that the device is perfectly level
+        center_.zero();
+        tCenter_.start();
+        iAccPrv_ = nAccPrv_ = 0;
+
+        // reset and initialize the MMA8451Q
+        mma_.init();
+        
         // set the initial ball velocity to zero
-        vx_ = vy_ = 0;
+        v_.zero();
         
         // set the initial raw acceleration reading to zero
-        xRaw_ = yRaw_ = 0;
+        araw_.zero();
+        vsum_.zero();
 
         // enable the interrupt
-        mma_.setInterruptMode(irqPin == PTA14 ? 1 : 2);
+        mma_.setInterruptMode(irqPin_ == PTA14 ? 1 : 2);
         
         // set up the interrupt handler
         intIn_.rise(this, &Accel::isr);
         
         // read the current registers to clear the data ready flag
         float z;
-        mma_.getAccXYZ(xRaw_, yRaw_, z);
+        mma_.getAccXYZ(araw_.x, araw_.y, z);
 
         // start our timers
         tGet_.start();
         tInt_.start();
+        tRest_.start();
     }
     
     void get(float &x, float &y, float &rx, float &ry) 
@@ -148,11 +452,11 @@
          __disable_irq();
          
          // read the shared data and store locally for calculations
-         float vx = vx_, vy = vy_, xRaw = xRaw_, yRaw = yRaw_;
+         FPoint vsum = vsum_, araw = araw_;
+         
+         // reset the velocity sum
+         vsum_.zero();
 
-         // reset the velocity
-         vx_ = vy_ = 0;
-         
          // get the time since the last get() sample
          float dt = tGet_.read_us()/1.0e6;
          tGet_.reset();
@@ -160,16 +464,178 @@
          // done manipulating the shared data
          __enable_irq();
          
-         // calculate the acceleration since the last get(): a = dv/dt
-         x = vx/dt;
-         y = vy/dt;         
+         // check for auto-centering every so often
+         if (tCenter_.read_ms() > 1000)
+         {
+             // add the latest raw sample to the history list
+             accPrv_[iAccPrv_] = araw_;
+             
+             // commit the history entry
+             iAccPrv_ = (iAccPrv_ + 1) % maxAccPrv;
+
+             // 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 = .005;
+                 if (accPrv_[0].distance(accPrv_[1]) < accTol
+                     && accPrv_[0].distance(accPrv_[2]) < accTol
+                     && accPrv_[0].distance(accPrv_[3]) < accTol
+                     && accPrv_[0].distance(accPrv_[4]) < accTol)
+                 {
+                     // figure the new center as the average of these samples
+                     center_.set(
+                        (accPrv_[0].x + accPrv_[1].x + accPrv_[2].x + accPrv_[3].x + accPrv_[4].x)/5.0,
+                        (accPrv_[0].y + accPrv_[1].y + accPrv_[2].y + accPrv_[3].y + accPrv_[4].y)/5.0);
+                 }
+             }
+             else
+             {
+                // not enough samples yet; just up the count
+                ++nAccPrv_;
+             }
+            
+             // reset the timer
+             tCenter_.reset();
+         }
+
+         // Calculate the velocity vector for the model ball.  Start
+         // with the accumulated velocity from the accelerations since
+         // the last reading.
+         FPoint dv = vsum;
+
+         // remember the previous velocity of the model ball
+         FPoint vprv = v_;
+         
+         // If we have residual motion, check for damping.
+         //
+         // The dmaping we model here isn't friction - we leave that sort of
+         // detail to the pinball simulator on the PC.  Instead, our form of
+         // damping is just an attempt to compensate for measurement errors
+         // from the accelerometer.  During a nudge event, we should see a
+         // series of accelerations back and forth, as the table sways in
+         // response to the push, rebounds from the sway, rebounds from the
+         // rebound, etc.  We know that in reality, the table itself doesn't
+         // actually go anywhere - it just sways, and when the swaying stops,
+         // it ends up where it started.  If we use the accelerometer input
+         // to do dead reckoning on the location of the table, we know that
+         // it has to end up where it started.  This means that the series of
+         // position changes over the course of the event should cancel out -
+         // the displacements should add up to zero.  
          
-         // return the raw accelerometer data in rx,ry
-         rx = xRaw;
-         ry = yRaw;
+          to model friction and other forces
+         // on the ball.  Instead, the damping we apply is to compensate for
+         // measurement errors in the accelerometer.  During a nudge event,
+         // a real pinball cabinet typically ends up at the same place it
+         // started - it sways in response to the nudge, but the swaying
+         // quickly damps out and leaves the table unmoved.  You don't
+         // typically apply enough force to actually pick up the cabinet
+         // and move it, or slide it across the floor - and doing so would
+         // trigger a tilt, in which case the ball goes out of play and we
+         // don't really have to worry about how realistically it behaves
+         // in response to the acceleration.
+         if (vprv.magnitude() != 0)
+         {
+             // The model ball is moving.  If the current motion has been
+             // going on for long enough, apply damping.  We wait a short
+             // time before we apply damping to allow small continuous
+             // accelerations (from tiling the table) to get the ball
+             // rolling.
+             if (tRest_.read_ms() > 100)
+             {
+             }
+         }
+         else
+         {
+             // the model ball is at rest; if the instantaneous acceleration
+             // is also near zero, reset the rest timer
+             if (dv.magnitude() < 0.025)
+                 tRest_.reset();
+         }
+         
+         // If the current velocity change is near zero, damp the ball's
+         // velocity.  The idea is that the total series of accelerations 
+         // from a nudge should net to zero, since a nudge doesn't
+         // actually move the table anywhere.  
+         // 
+         // Ideally, this wouldn't be necessary, because the raw
+         // accelerometer readings should organically add up to zero over
+         // the course of a nudge.  In practice, the accelerometer isn't
+         // perfect; it can only sample so fast, so it can't capture every
+         // instantaneous change; and each reading has some small measurement
+         // error, which becomes significant when many readings are added
+         // together.  The damping is an attempt to reconcile the imperfect
+         // measurements with what how expect the real physical system to
+         // behave - we know what the outcome of an event should be, so we
+         // adjust our measurements to get the expected outcome.
+         //
+         // If the ball's velocity is large at this point, assume that this
+         // wasn't a nudge event at all, but a sustained inclination - as
+         // though the player picked up one end of the table and held it
+         // up for a while, to accelerate the ball down the sloped table.
+         // In this case just reset the velocity to zero without doing
+         // any damping, so that we don't pass through any deceleration
+         // to the pinball simulation.  In this case we want to leave it
+         // to the pinball simulation to do its own modeling of friction
+         // or bouncing to decelerate the ball.  Our correction is only
+         // realistic for brief events that naturally net out to neutral
+         // accelerations.
+         if (dv.magnitude() < .025)
+         {
+            // check the ball's speed
+            if (v_.magnitude() < .25)
+            {
+                // apply the damping
+                FPoint damp(damping(v_.x), damping(v_.y));
+                dv -= damp;
+                ledB = 0;
+            }
+            else
+            {
+                // the ball is going too fast - simply reset it
+                v_ = dv;
+                vprv = dv;
+                ledB = 1;
+            }
+         }
+         else
+             ledB = 1;
+         
+         // apply the velocity change for this interval
+         v_ += dv;
+         
+         // return the acceleration since the last update (change in velocity
+         // over time) in x,y
+         dv /= dt;
+         x = (v_.x - vprv.x) / dt;
+         y = (v_.y - vprv.y) / dt;
+         
+         // report the calibrated instantaneous acceleration in rx,ry
+         rx = araw.x - center_.x;
+         ry = araw.y - center_.y;
      }    
     
 private:
+    // velocity damping function
+    float damping(float v)
+    {
+        // scale to -2048..2048 range, and get the absolute value
+        float a = fabs(v*2048.0);
+        
+        // damp out small velocities immediately
+        if (a < 20)
+            return v;
+        
+        // calculate the cube root of the scaled value
+        float r = exp(log(a)/3.0);
+        
+        // rescale
+        r /= 2048.0;
+        
+        // apply the sign and return the result
+        return (v < 0 ? -r : r);
+    }
+
     // interrupt handler
     void isr()
     {
@@ -178,39 +644,101 @@
         // 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 z;
-        mma_.getAccXYZ(xRaw_, yRaw_, z);
+        float x, y, z;
+        mma_.getAccXYZ(x, y, z);
+
+        // store the raw results
+        araw_.set(x, y);
+        zraw_ = z;
         
         // calculate the time since the last interrupt
         float dt = tInt_.read_us()/1.0e6;
         tInt_.reset();
         
-        // Accelerate the model ball: v = a*dt.  Assume that the raw
-        // data from the accelerometer reflects the average physical
-        // acceleration over the interval since the last sample.
-        vx_ += xRaw_ * dt;
-        vy_ += yRaw_ * dt;
+        // Add the velocity to the running total.  First, calibrate the
+        // raw acceleration to our centerpoint, then multiply by the time
+        // since the last sample to get the velocity resulting from
+        // applying this acceleration for the sample time.
+        FPoint rdt((x - center_.x)*dt, (y - center_.y)*dt);
+        vsum_ += rdt;
     }
     
-    // current modeled ball velocity
-    float vx_, vy_;
-    
-    // last raw axis readings
-    float xRaw_, yRaw_;
-    
     // underlying accelerometer object
     MMA8451Q mma_;
     
-    // interrupt router
-    InterruptIn intIn_;
+    // last raw acceleration readings
+    FPoint araw_;
+    float zraw_;
+    
+    // total velocity change since the last get() sample
+    FPoint vsum_;
+    
+    // current modeled ball velocity
+    FPoint v_;
     
     // timer for measuring time between get() samples
     Timer tGet_;
     
     // timer for measuring time between interrupts
     Timer tInt_;
+    
+    // time since last rest
+    Timer tRest_;
+
+    // calibrated center point - this is the position where we observe
+    // constant input for a few seconds, telling us the orientation of
+    // the accelerometer device when at rest
+    FPoint center_;
+
+    // timer for atuo-centering
+    Timer tCenter_;
+    
+    // recent accelerometer readings, for auto centering
+    int iAccPrv_, nAccPrv_;
+    static const int maxAccPrv = 5;
+    FPoint 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;
+        wait_us(20);
+        scl = 0;
+        wait_us(20);
+    }
+}
+
+// ---------------------------------------------------------------------------
+//
+// 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
@@ -218,6 +746,12 @@
     ledG = 1;
     ledB = 1;
     
+    // clear the I2C bus for the accelerometer
+    clear_i2c();
+    
+    // Create the joystick USB client
+    MyUSBJoystick js(USB_VENDOR_ID, USB_PRODUCT_ID, USB_VERSION_NO);
+
     // set up a flash memory controller
     FreescaleIAP iap;
     
@@ -234,11 +768,17 @@
     // 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.  VP doesn't seem to have very high
-    // resolution internally for the plunger, so it's probably not necessary
-    // to use the full resolution of the sensor - about 160 pixels seems
-    // perfectly adequate.  We can read the sensor faster (and thus provide
-    // a higher refresh rate) if we read fewer pixels in each frame.
+    // 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;
 
     // if the flash is valid, load it; otherwise initialize to defaults
@@ -271,34 +811,22 @@
     // set up a timer for our heartbeat indicator
     Timer hbTimer;
     hbTimer.start();
-    int t0Hb = hbTimer.read_ms();
     int hb = 0;
+    uint16_t hbcnt = 0;
     
     // set a timer for accelerometer auto-centering
     Timer acTimer;
     acTimer.start();
-    int t0ac = acTimer.read_ms();
     
-    // Create the joystick USB client
-    MyUSBJoystick js(0xFAFA, 0x00F7, 0x0003);
-
     // create the accelerometer object
-    Accel accel(PTE25, PTE24, MMA8451_I2C_ADDRESS, PTA15);
+    Accel accel(MMA8451_SCL_PIN, MMA8451_SDA_PIN, MMA8451_I2C_ADDRESS, MMA8451_INT_PIN);
     
     // create the CCD array object
     TSL1410R ccd(PTE20, PTE21, PTB0);
     
-    // recent accelerometer readings, for auto centering
-    int iAccPrv = 0, nAccPrv = 0;
-    const int maxAccPrv = 5;
-    AccPrv accPrv[maxAccPrv];
-
     // last accelerometer report, in mouse coordinates
     int x = 127, y = 127, z = 0;
 
-    // raw accelerator centerpoint, on the unit interval (-1.0 .. +1.0)
-    float xCenter = 0.0, yCenter = 0.0;    
-    
     // start the first CCD integration cycle
     ccd.clear();
 
@@ -542,116 +1070,55 @@
         float xa, ya, rxa, rya;
         accel.get(xa, ya, rxa, rya);
         
-        // check for auto-centering every so often
-        if (acTimer.read_ms() - t0ac > 1000) 
-        {
-            // add the sample to the history list
-            accPrv[iAccPrv].x = xa;
-            accPrv[iAccPrv].y = ya;
-            
-            // store the slot
-            iAccPrv += 1;
-            iAccPrv %= maxAccPrv;
-            nAccPrv += 1;
-            
-            // If we have a full complement, check for stability.  The
-            // raw accelerometer input is in the rnage -4096 to 4096, but
-            // the class cover normalizes to a unit interval (-1.0 .. +1.0).
-            const float accTol = .005;
-            if (nAccPrv >= maxAccPrv
-                && accPrv[0].dist(accPrv[1]) < accTol
-                && accPrv[0].dist(accPrv[2]) < accTol
-                && accPrv[0].dist(accPrv[3]) < accTol
-                && accPrv[0].dist(accPrv[4]) < accTol)
-            {
-                // figure the new center
-                xCenter = (accPrv[0].x + accPrv[1].x + accPrv[2].x + accPrv[3].x + accPrv[4].x)/5.0;
-                yCenter = (accPrv[0].y + accPrv[1].y + accPrv[2].y + accPrv[3].y + accPrv[4].y)/5.0;
-            }
-            
-            // reset the auto-center timer
-            acTimer.reset();
-            t0ac = acTimer.read_ms();
-        }
-        
-        // adjust for our auto centering
-        xa -= xCenter;
-        ya -= yCenter;
-        
-        // confine to the unit interval
+        // confine the accelerometer results to the unit interval
         if (xa < -1.0) xa = -1.0;
         if (xa > 1.0) xa = 1.0;
         if (ya < -1.0) ya = -1.0;
         if (ya > 1.0) ya = 1.0;
 
-        // figure the new mouse report data
-        int xnew = (int)(127 * xa);
-        int ynew = (int)(127 * ya);
+        // scale to our -127..127 reporting range
+        int xnew = int(127 * xa);
+        int ynew = int(127 * ya);
 
         // store the updated joystick coordinates
         x = xnew;
         y = ynew;
         z = znew;
         
-        // if we're in USB suspend or disconnect mode, spin
-        if (js.isSuspended() || !js.isConnected())
-        {
-            // go dark (turn off the indicator LEDs)
-            ledG = 1;
-            ledB = 1;
-            ledR = 1;
-            
-            // wait until we're connected and come out of suspend mode
-            for (uint32_t n = 0 ; js.isSuspended() || !js.isConnected() ; ++n)
-            {
-                // spin for a bit
-                wait(1);
-                
-                // if we're suspended, do a brief red flash; otherwise do a long red flash
-                if (js.isSuspended())
-                {
-                    // suspended - flash briefly ever few seconds
-                    if (n % 3 == 0)
-                    {
-                        ledR = 0;
-                        wait(0.05);
-                        ledR = 1;
-                    }
-                }
-                else
-                {
-                    // running, not connected - flash red
-                    ledR = !ledR;
-                }
-            }
-        }
-
         // Send the status report.  It doesn't really matter what
         // coordinate system we use, since Visual Pinball has config
         // options for rotations and axis reversals, but reversing y
         // at the device level seems to produce the most intuitive 
         // results for the Windows joystick control panel view, which
         // is an easy way to check that the device is working.
-        js.update(x, -y, z, int(rxa*127), int(rya*127), 0);
+        //
+        // $$$ button updates are for diagnostics, so we can see that the
+        // device is sending data properly if the accelerometer gets stuck
+        js.update(x, -y, z, int(rxa*127), int(rya*127), hb ? 0x5500 : 0xAA00);
         
         // show a heartbeat flash in blue every so often if not in 
         // calibration mode
-        if (calBtnState < 2 && hbTimer.read_ms() - t0Hb > 1000) 
+        if (calBtnState < 2 && hbTimer.read_ms() > 1000) 
         {
-            if (js.isSuspended())
+            if (js.isSuspended() || !js.isConnected())
             {
-                // suspended - turn off the LEDs entirely
+                // suspended - turn off the LED
                 ledR = 1;
                 ledG = 1;
                 ledB = 1;
-            }
-            else if (!js.isConnected())
-            {
-                // not connected - flash red
-                hb = !hb;
-                ledR = (hb ? 0 : 1);
-                ledG = 1;
-                ledB = 1;
+
+                // show a status flash every so often                
+                if (hbcnt % 3 == 0)
+                {
+                    // disconnected = red flash; suspended = red-red
+                    for (int n = js.isConnected() ? 1 : 2 ; n > 0 ; --n)
+                    {
+                        ledR = 0;
+                        wait(0.05);
+                        ledR = 1;
+                        wait(0.25);
+                    }
+                }
             }
             else if (flash_valid)
             {
@@ -665,14 +1132,14 @@
             {
                 // connected, factory reset - flash yellow/green
                 hb = !hb;
-                ledR = (hb ? 0 : 1);
-                ledG = 0;
+                //ledR = (hb ? 0 : 1);
+                //ledG = 0;
                 ledB = 1;
             }
             
             // reset the heartbeat timer
             hbTimer.reset();
-            t0Hb = hbTimer.read_ms();
+            ++hbcnt;
         }
     }
 }