Tekkotsu: LedEngine Class Reference

This implements a collection of LED special effects so that the code can be reused various places as feedback to to the user.

The 'c' prefix on a function (cset(), cflash(), ccycle()) means it will call clear() before doing its own setting. This is a quick way to replace whatever is currently displayed with whatever you are about to display.

A flash() will invert and override the current setting, so that it will "reset" after the flash. Flashes change mid-range values to extremes to make the flash visible (ie not just (1-current)) Normal invert() will do simple inverses (just (1-current)), and doesn't change back automatically.

Cycling ("pulsing") and single-value setting are mutually exclusive; one will cut off the other

getSetting() returns value of last set(); getValue() returns what's actually being returned to Motion at the moment

There's some nice functions for using the LEDs to display numbers. This is handy for when you want to be free of the terminal.

The ERS-220 and ERS-7 have enough LEDs that they just use a "count the lights" style of display instead of this pseudo-arabic display. (look close to see that green bar LED at the top of the 210, which doesn't show up well in the camera image for some reason. )

The ERS-7 also has a "mode" which controls how certain face LEDs are interpreted. By setting ERS7Info::LEDABModeOffset / ERS7Info::LEDABModeMask, you can switch between the modes -- a value of 0 is mode "A", and a value of 1 is mode "B".

Definition at line 69 of file LedEngine.h.


Public Types
	onedigit
	can display a number -9 thru 9, using the current robot model's numMask. For negative numbers, blinks the top bar - fast if it's supposed to be on, slow if it would otherwise be off
	twodigit
	can display a number -99 thru 99, using setOneOfTwo(). For negative numbers, sets the top bar to 1 (off otherwise).
	major
	shows overall value
	minor
	shows value within major tick
	none
	if you want to leave blank
enum	numStyle_t { onedigit, twodigit }
	Use these to specify a style for displaying numbers using displayNumber(). More...
enum	percentStyle_t { major, minor, none }
	Use these to specify a style for displaying a percentage value [0-1] using displayPercent(). More...
Public Member Functions
	LedEngine ()
	constructor - don't forget to call if you inherit
virtual	~LedEngine ()
	destructor
int	updateLEDs (const MotionCommand *caller, LEDBitMask_t mask=AllLEDMask)
	call this from a MotionCommand's updateOutputs() - makes calls to MotionManager to update LED values
int	updateLEDs (OutputCmd cmds[NumLEDs])
	call this from a MotionCommand's updateOutputs() - performs the calculations to update LEDs' values
int	updateLEDFrames (OutputCmd cmds[NumLEDs][NumFrames])
	call this from a MotionCommand's updateOutputs() - performs the calculations to update LEDs' values
void	recalcFlashEnd ()
	recalculates nextFlashEnd so we can tell when a flash has completed
int	isDirty ()
	returns true if there are changes since the last updateLEDs()
void	invert (LEDBitMask_t leds)
	sets the leds specified by leds to the inverse of their current value
void	cset (LEDBitMask_t leds, float value)
	sets the leds specified by leds to value, clears all the rest
void	set (LEDBitMask_t leds, float value)
	sets the leds specified by leds to value
void	cflash (LEDBitMask_t leds, float value, unsigned int ms)
	sets the leds specified by leds to value for ms milliseconds, then sets back. Clears ~leds
void	flash (LEDBitMask_t leds, float value, unsigned int ms)
	sets the leds specified by leds to value for ms milliseconds, then sets back.
void	flash (LEDBitMask_t leds, unsigned int ms)
	sets the leds specified by leds to either a much higher or much lower value for ms milliseconds, then sets back.
void	ccycle (LEDBitMask_t leds, unsigned int period, float amp, float offset=0, int phase=0)
	causes the leds specified by leds to cycle between low and high, clears others. See cycle() for parameter documentation.
void	cycle (LEDBitMask_t leds, unsigned int period, float amp, float offset=0, int phase=0)
	causes the leds specified by leds to cycle between low and high; values calculated for cycle will be clipped to [0,1] for more sensible blending of square wave approximations (high amplitude sine wave)
void	clear ()
	sets all leds to 0.
void	extendFlash (unsigned int ms)
	adds ms to all flash times (may resurrect previously completed flashes)
float	getSetting (LEDOffset_t led_id)
	returns the current setting of the LED specified by led_id (the value you passed in set())
float	getValue (LEDOffset_t led_id, unsigned int planahead=0)
	returns the current value of the LED specified by led_id (the value being expressed - may change if cycling for instance)
void	displayNumber (int x, numStyle_t style)
	Allows convenient display of numerical information to the LEDs on the face.
void	displayPercent (float x, percentStyle_t left_style, percentStyle_t right_style)
	Allows convenient display of percentage information to the LEDs on the face.
Static Public Attributes
static const LEDBitMask_t	ERS210numMasks [11]
	holds a series of bit masks for the onedigit style of numerical display (-9:9 and '.')
static const LEDBitMask_t	ERS220numMasks [11]
	bit masks for the ondigit style of numberical display - just count the LEDs on the head
static const LEDBitMask_t	ERS7numMasks [11]
Protected Member Functions
float	calcValue (unsigned int i, unsigned int t)
	Calculates the current value of led i for current time t.
void	setOneOfTwo (unsigned int x, unsigned int low, unsigned int mid, unsigned int high)
	used by displayNumber() to determine settings of LEDs when using numStyle_t::twodigit
void	setColumn (float x, unsigned int low, unsigned int mid, unsigned int high, unsigned int top)
	used by displayPercent() to determine settings of LEDs
Static Protected Member Functions
static float	calcInvert (float value)
	Performs the 'invert' calculation based on current value (returns 1-value).
static float	calcFlash (float value)
	Performs the 'flash' calculation based on current value (uses calcInvert() if value upper or lower third, 0 or 1 otherwise).
static float	calcCycle (unsigned int period, float amp, float offset, unsigned int t)
	Performs the 'cycle' calculation based on desired period, amplituted, amplitude offset, and time since start. See cycle().
Protected Attributes
LEDInfo	infos [NumLEDs]
	the information regarding each of the LEDs
bool	dirty
	true if changes since last updateLEDs: either at least one LED is cycling or a flash has begun/ended
unsigned int	numCycling
	the number of LEDs currently cycling (if non-zero, always dirty)
unsigned int	nextFlashEnd
	the soonest time a flash will end (and we'll become dirty)
Classes
struct	LEDInfo
	Holds all the information needed by each of the LEDs. More...

Member Enumeration Documentation

enum LedEngine::numStyle_t

Use these to specify a style for displaying numbers using displayNumber().

Enumerator:

onedigit	can display a number -9 thru 9, using the current robot model's numMask. For negative numbers, blinks the top bar - fast if it's supposed to be on, slow if it would otherwise be off
twodigit	can display a number -99 thru 99, using setOneOfTwo(). For negative numbers, sets the top bar to 1 (off otherwise).

Definition at line 129 of file LedEngine.h.

enum LedEngine::percentStyle_t

Use these to specify a style for displaying a percentage value [0-1] using displayPercent().

Enumerator:

major	shows overall value
minor	shows value within major tick
none	if you want to leave blank

Definition at line 134 of file LedEngine.h.

Member Function Documentation

void LedEngine::cycle	(	LEDBitMask_t	leds,
		unsigned int	period,
		float	amp,
		float	offset = `0`,
		int	phase = `0`
	)

causes the leds specified by leds to cycle between low and high; values calculated for cycle will be clipped to [0,1] for more sensible blending of square wave approximations (high amplitude sine wave)

Parameters:

	leds	the bitmask of leds to apply this to
	period	the period of the cycle (milliseconds), includes an on and off
	amp	the amplitude of the cycle - note that this is clipped at 0 and 1.
	offset	the vertical offset of the cycle - simply shifts the baseline of the cycle up or down
	phase	the phase within the cycle to start at (specify in milliseconds)

When this function is called, the starting time is stored as current time + phase.

The equation used is

$\cos(\frac{2\pi(t-starttime)}{period})*(\frac{-amp}{2})+.5+offset$

The idea is that with a amplitude=1 and offset=0, it will start at 0, ramp up to 1, and then ramp down again. The arguments to this function will let you control all parameters of the cycle.

You can get a blink-on/off instead of cycle on/off by using a very large amplitude.

Definition at line 253 of file LedEngine.cc.

Referenced by ccycle(), and EmergencyStopMC::EmergencyStopMC().

void LedEngine::displayNumber	(	int	x,
		numStyle_t	style
	)

Allows convenient display of numerical information to the LEDs on the face.

If overflow occurs, the face LEDs are set to flash on and off 3 every 333 milliseconds

Definition at line 289 of file LedEngine.cc.

void LedEngine::displayPercent	(	float	x,
		percentStyle_t	left_style,
		percentStyle_t	right_style
	)

Allows convenient display of percentage information to the LEDs on the face.

Besides allowing a two-digit display, the 'edge' bar for each type is blinked to denote how full it is. So you can get up to a two-digit, base 5 display, with an extra digit of estimated value.

If overflow (>1) occurs, sets everything to .75.
If underflow (<0) occurs, sets everything to .25.

The left and right columns are combined with an OR operation. (they overlap on the top bar) Left and right designations are dog centric!

Definition at line 356 of file LedEngine.cc.

int LedEngine::updateLEDFrames ( OutputCmd cmds[NumLEDs][NumFrames] )

call this from a MotionCommand's updateOutputs() - performs the calculations to update LEDs' values

Parameters:

cmds

on input, used for weight values - on return, holds the resulting OutputCmd's

Definition at line 166 of file LedEngine.cc.

Referenced by LedMC::updateOutputs().

int LedEngine::updateLEDs ( OutputCmd cmds[NumLEDs] )

call this from a MotionCommand's updateOutputs() - performs the calculations to update LEDs' values

Parameters:

cmds

on input, used for weight values - on return, holds the resulting OutputCmd's

Definition at line 156 of file LedEngine.cc.

int LedEngine::updateLEDs	(	const MotionCommand *	caller,
		LEDBitMask_t	mask = `AllLEDMask`
	)

call this from a MotionCommand's updateOutputs() - makes calls to MotionManager to update LED values

Parameters:

	caller	pass the "parent" motioncommand's address here (usually will pass 'this')
	mask	a bitmask of which leds to update (uses weight of 1)

Definition at line 144 of file LedEngine.cc.

Member Data Documentation

const LEDBitMask_t LedEngine::ERS210numMasks [static]

Initial value:

 {
  ERS210Info::BotRLEDMask|ERS210Info::BotLLEDMask|ERS210Info::TopBrLEDMask, 
  
  ERS210Info::BotLLEDMask|ERS210Info::MidLLEDMask|ERS210Info::TopLLEDMask,  
  
  ERS210Info::BotRLEDMask|ERS210Info::BotLLEDMask|ERS210Info::TopLLEDMask|ERS210Info::TopBrLEDMask, 
  
  ERS210Info::BotRLEDMask|ERS210Info::BotLLEDMask|ERS210Info::MidRLEDMask|ERS210Info::TopLLEDMask
  |ERS210Info::TopBrLEDMask, 
  
  ERS210Info::BotLLEDMask|ERS210Info::MidLLEDMask|ERS210Info::TopRLEDMask|ERS210Info::TopLLEDMask,  

  ERS210Info::BotRLEDMask|ERS210Info::BotLLEDMask|ERS210Info::TopRLEDMask|ERS210Info::TopBrLEDMask, 

  ERS210Info::BotRLEDMask|ERS210Info::BotLLEDMask|ERS210Info::MidRLEDMask|ERS210Info::MidLLEDMask
  |ERS210Info::TopRLEDMask|ERS210Info::TopBrLEDMask, 

  ERS210Info::BotLLEDMask|ERS210Info::MidLLEDMask|ERS210Info::TopLLEDMask|ERS210Info::TopBrLEDMask,  

  ERS210Info::BotRLEDMask|ERS210Info::BotLLEDMask|ERS210Info::MidRLEDMask|ERS210Info::MidLLEDMask
  |ERS210Info::TopRLEDMask|ERS210Info::TopLLEDMask|ERS210Info::TopBrLEDMask, 

  ERS210Info::BotLLEDMask|ERS210Info::MidLLEDMask|ERS210Info::TopRLEDMask|ERS210Info::TopLLEDMask
  |ERS210Info::TopBrLEDMask,  

  ERS210Info::BotLLEDMask 
}

holds a series of bit masks for the onedigit style of numerical display (-9:9 and '.')

the hope is that these actually resemble the shapes of the numbers so people can recognize them more easily - without converting base 2 in their heads.

Definition at line 125 of file LedEngine.h.

Referenced by displayNumber().

const LEDBitMask_t LedEngine::ERS220numMasks [static]

bit masks for the ondigit style of numberical display - just count the LEDs on the head

This is "Count the dots" style

Definition at line 126 of file LedEngine.h.

Referenced by displayNumber().

const LEDBitMask_t LedEngine::ERS7numMasks [static]

writes a number on the display

Definition at line 127 of file LedEngine.h.

Referenced by displayNumber().