LED patterns - Revision history

90.180.214.199: /* Low Level */

2014-01-07T08:58:06Z

Low Level

← Older revision		Revision as of 08:58, 7 January 2014
Line 257:		Line 257:

	See [[N900_Hardware_LED]] for hardware programming details.		See [[N900_Hardware_LED]] for hardware programming details.

	** engine1_leds is in the form "0000rgb00" (9 digits, "1" or "0"), where rgb control which leds should participate. The 6 "0" in the above pattern correspond to the 6 kbd white leds controlled by mce via engine3, so readout of engine3_leds gives 111100011 ;-)
	** engine1_load accepts a string as specified above (e.g. 9d804000427f0d7f7f007f0042000000). Though LP5523 has an absolute storage of 96 steps a 16bits, which can dynamically partitioned between the three engines, the driver limits this to a fixed 16 words a 16bit per engine

	~~After that, when the state is changed to "run", the program will be run by the controller.~~

	In order for the leds to work, the led_current of the appropriate leds need to be set to non-zero values. This is accomplished by writing to files leds:lp5523:[bgr]/'''led_current''' under the same directory. '''Don't mess with''' the default settings here that come with driver init, they are suitable and changing to higher values may and for certain values '''will kill your LEDS'''!.

	== References ==		== References ==

90.180.214.199: /* Low Level */

2014-01-07T08:56:45Z

Low Level

← Older revision		Revision as of 08:56, 7 January 2014
Line 250:		Line 250:

	=== Low Level ===		=== Low Level ===
	'''~~this section should get merged with~~ [[N900_Hardware_LED]] !'''		'''see also [[N900_Hardware_LED]] !'''

	~~The N8x0 had a LP5521<ref name="lp5521-spec">http://www.national.com/pf/LP/LP5521.html</ref>~~ ~~<ref name="lp5521-driver">http://mxr.maemo.org/fremantle/source/kernel/drivers/i2c/chips/lp5521.c</ref> programmable three channel controller LED Driver which handles the leds.~~
	The N900 has a LP5523 <ref name="lp5523-spec">http://www.national.com/pf/LP/LP5523.html</ref> <ref name="leds-lp5523-driver">http://mxr.maemo.org/fremantle/source/kernel/drivers/leds/leds-lp5523.c</ref>, which has 3 engines like LP5521 but can drive 9 LEDs via a multiplexer. Actually the multiplexer is between the engine (CPU) and the 9 PWM blocks for the 9 LED outputs - each of those 9 PWM blocks has its own accessable registers for current and brightness, and they are exposed to sysfs. This chip also has double the RAM size for 'programs', plus 2 per-machine and one global register variable and a true set of conditional branch and arithmetic commands.

	Alas the LP5523 driver doesn't support any of those advanced features of LP5523, and uses one of those engines for a kbd backlight ramp-up/down.<br>		Alas the LP5523 driver doesn't support any of those advanced features of LP5523, and uses one of those engines for a kbd backlight ramp-up/down.<br>

90.180.214.199: /* e080, e100, e200 - Wait for engine trigger */

2014-01-06T13:40:13Z

e080, e100, e200 - Wait for engine trigger

← Older revision		Revision as of 13:40, 6 January 2014
Line 168:		Line 168:

	Although the trigger information for engine 3 is listed we have no control over the keyboard LED, so it has no effect. Attempting to add a pattern for engine 3 after the pattern for engine 2 causes the pattern to no longer function.		Although the trigger information for engine 3 is listed we have no control over the keyboard LED, so it has no effect. Attempting to add a pattern for engine 3 after the pattern for engine 2 causes the pattern to no longer function.

			...and there's more. Engine can actually do sub/add arithmetics, branches, and select which leds it controls.

			Pavel Machek created a simple compiler, it is at https://gitorious.org/tui/tui/source/5b3f5cacf8e208d3ea50d6066e549940d85e55be:maemo/notcc.py , and it can use more features of the chip than described here.

	=== Default LED Patterns for Nokia N900 ===		=== Default LED Patterns for Nokia N900 ===

90.180.214.199: /* Low Level */

2014-01-06T13:37:55Z

Low Level

← Older revision		Revision as of 13:37, 6 January 2014
Line 257:		Line 257:

	** engine1_leds is in the form "0000rgb00" (9 digits, "1" or "0"), where rgb control which leds should participate. The 6 "0" in the above pattern correspond to the 6 kbd white leds controlled by mce via engine3, so readout of engine3_leds gives 111100011 ;-)		** engine1_leds is in the form "0000rgb00" (9 digits, "1" or "0"), where rgb control which leds should participate. The 6 "0" in the above pattern correspond to the 6 kbd white leds controlled by mce via engine3, so readout of engine3_leds gives 111100011 ;-)
	** engine1_load accepts a string as specified above (e.g. 9d804000427f0d7f7f007f0042000000). Though LP5523 has an absolute storage of 96 steps a 16bits, which can dynamically partitioned between the three engines, the driver ~~is retarded and~~ limits this to a fixed 16 words a 16bit per engine ~~(<s>ToDo: check if driver supports >16 instructions / machine</s> no it doesn't :-/)~~		** engine1_load accepts a string as specified above (e.g. 9d804000427f0d7f7f007f0042000000). Though LP5523 has an absolute storage of 96 steps a 16bits, which can dynamically partitioned between the three engines, the driver limits this to a fixed 16 words a 16bit per engine

	After that, when the state is changed to "run", the program will be run by the controller.		After that, when the state is changed to "run", the program will be run by the controller.

90.180.214.199: /* Low Level */

2014-01-06T13:36:57Z

Low Level

← Older revision		Revision as of 13:36, 6 January 2014
Line 251:		Line 251:
	The N900 has a LP5523 <ref name="lp5523-spec">http://www.national.com/pf/LP/LP5523.html</ref> <ref name="leds-lp5523-driver">http://mxr.maemo.org/fremantle/source/kernel/drivers/leds/leds-lp5523.c</ref>, which has 3 engines like LP5521 but can drive 9 LEDs via a multiplexer. Actually the multiplexer is between the engine (CPU) and the 9 PWM blocks for the 9 LED outputs - each of those 9 PWM blocks has its own accessable registers for current and brightness, and they are exposed to sysfs. This chip also has double the RAM size for 'programs', plus 2 per-machine and one global register variable and a true set of conditional branch and arithmetic commands.		The N900 has a LP5523 <ref name="lp5523-spec">http://www.national.com/pf/LP/LP5523.html</ref> <ref name="leds-lp5523-driver">http://mxr.maemo.org/fremantle/source/kernel/drivers/leds/leds-lp5523.c</ref>, which has 3 engines like LP5521 but can drive 9 LEDs via a multiplexer. Actually the multiplexer is between the engine (CPU) and the 9 PWM blocks for the 9 LED outputs - each of those 9 PWM blocks has its own accessable registers for current and brightness, and they are exposed to sysfs. This chip also has double the RAM size for 'programs', plus 2 per-machine and one global register variable and a true set of conditional branch and arithmetic commands.

	Alas the LP5523 driver doesn't support any of those advanced features of LP5523, and ~~even abuses~~ one of those ~~extremely smart~~ engines for a ~~stupid occasional~~ kbd backlight ramp-up/down.<br>		Alas the LP5523 driver doesn't support any of those advanced features of LP5523, and uses one of those engines for a kbd backlight ramp-up/down.<br>
	''This could as well get done in mce by a 25ms timer that writes 40 new brightness values to the master control of kbd LEDs during 1 second. ~~There~~'~~s really no reason why you need to assign 1 of those 3 machines to this occasional short ramp function and break true rainbow 3 color effects on indicator LED. Sound rationale: while kbd bl ramp the~~ '~~'CPU is awake anyway''. No obvious benefit in offloading the task to a LP5532 engine~~ [~~except from saving 50 lines of code in mce~~]~~!)''~~		''This could as well get done in mce by a 25ms timer that writes 40 new brightness values to the master control of kbd LEDs during 1 second.''

			See [[N900_Hardware_LED]] for hardware programming details.

	The controller can be programmed via "/sys/class/i2c-adapter/i2c-2/2-0032" (FIXME: is this stable?). In that directory there are three files corresponding to the three channels: engine1_mode, engine2_mode and engine3_mode. Each file "stores" one of "run", "load", "disabled" (it seems that there is another mode "direct" but it doesn't work):
	* run: The program runs
	* disabled: The channel is disabled - not running
	* load: In that state, there are two more files in that directory:
	** engine1_leds is in the form "0000rgb00" (9 digits, "1" or "0"), where rgb control which leds should participate. The 6 "0" in the above pattern correspond to the 6 kbd white leds controlled by mce via engine3, so readout of engine3_leds gives 111100011 ;-)		** engine1_leds is in the form "0000rgb00" (9 digits, "1" or "0"), where rgb control which leds should participate. The 6 "0" in the above pattern correspond to the 6 kbd white leds controlled by mce via engine3, so readout of engine3_leds gives 111100011 ;-)
	** engine1_load accepts a string as specified above (e.g. 9d804000427f0d7f7f007f0042000000). Though LP5523 has an absolute storage of 96 steps a 16bits, which can dynamically partitioned between the three engines, the driver is retarded and limits this to a fixed 16 words a 16bit per engine (<s>ToDo: check if driver supports >16 instructions / machine</s> no it doesn't :-/)		** engine1_load accepts a string as specified above (e.g. 9d804000427f0d7f7f007f0042000000). Though LP5523 has an absolute storage of 96 steps a 16bits, which can dynamically partitioned between the three engines, the driver is retarded and limits this to a fixed 16 words a 16bit per engine (<s>ToDo: check if driver supports >16 instructions / machine</s> no it doesn't :-/)
Line 263:		Line 261:

	In order for the leds to work, the led_current of the appropriate leds need to be set to non-zero values. This is accomplished by writing to files leds:lp5523:[bgr]/'''led_current''' under the same directory. '''Don't mess with''' the default settings here that come with driver init, they are suitable and changing to higher values may and for certain values '''will kill your LEDS'''!.		In order for the leds to work, the led_current of the appropriate leds need to be set to non-zero values. This is accomplished by writing to files leds:lp5523:[bgr]/'''led_current''' under the same directory. '''Don't mess with''' the default settings here that come with driver init, they are suitable and changing to higher values may and for certain values '''will kill your LEDS'''!.

	~~NOTE: There must be a delay (how much? The specs of LP5521 mentions it) between writes to sysfs files.~~

	== References ==		== References ==

90.180.214.199: /* References */

2014-01-06T13:33:25Z

References

← Older revision		Revision as of 13:33, 6 January 2014
Line 267:		Line 267:

	== References ==		== References ==
	Also see [[N900_Hardware_LED]]! ~~(somebody should have cleaned this duplication rather than ruthlessly copying)~~		Also see [[N900_Hardware_LED]]!

	<references />		<references />

imported_>joerg_rw: /* Low Level */ details on mux

2013-10-16T13:28:55Z

Low Level: details on mux

← Older revision		Revision as of 13:28, 16 October 2013
Line 249:		Line 249:

	The N8x0 had a LP5521<ref name="lp5521-spec">http://www.national.com/pf/LP/LP5521.html</ref> <ref name="lp5521-driver">http://mxr.maemo.org/fremantle/source/kernel/drivers/i2c/chips/lp5521.c</ref> programmable three channel controller LED Driver which handles the leds.		The N8x0 had a LP5521<ref name="lp5521-spec">http://www.national.com/pf/LP/LP5521.html</ref> <ref name="lp5521-driver">http://mxr.maemo.org/fremantle/source/kernel/drivers/i2c/chips/lp5521.c</ref> programmable three channel controller LED Driver which handles the leds.
	~~(actually~~ N900 has a LP5523 <ref name="lp5523-spec">http://www.national.com/pf/LP/LP5523.html</ref> <ref name="leds-lp5523-driver">http://mxr.maemo.org/fremantle/source/kernel/drivers/leds/leds-lp5523.c</ref>, which has 3 engines like LP5521 but can drive 9 LEDs via a multiplexer. This chip also has double the RAM size for 'programs', plus 2 per-machine and one global register variable and a true set of conditional branch and arithmetic commands.		The N900 has a LP5523 <ref name="lp5523-spec">http://www.national.com/pf/LP/LP5523.html</ref> <ref name="leds-lp5523-driver">http://mxr.maemo.org/fremantle/source/kernel/drivers/leds/leds-lp5523.c</ref>, which has 3 engines like LP5521 but can drive 9 LEDs via a multiplexer. Actually the multiplexer is between the engine (CPU) and the 9 PWM blocks for the 9 LED outputs - each of those 9 PWM blocks has its own accessable registers for current and brightness, and they are exposed to sysfs. This chip also has double the RAM size for 'programs', plus 2 per-machine and one global register variable and a true set of conditional branch and arithmetic commands.

	Alas the LP5523 driver doesn't support any of those advanced features of LP5523, and even abuses one of those extremely smart engines for a stupid occasional kbd backlight ramp-up/down.<br>		Alas the LP5523 driver doesn't support any of those advanced features of LP5523, and even abuses one of those extremely smart engines for a stupid occasional kbd backlight ramp-up/down.<br>

imported_>joerg_rw: /* Low Level */ typo

2013-10-16T12:26:17Z

Low Level: typo

← Older revision		Revision as of 12:26, 16 October 2013
Line 252:		Line 252:

	Alas the LP5523 driver doesn't support any of those advanced features of LP5523, and even abuses one of those extremely smart engines for a stupid occasional kbd backlight ramp-up/down.<br>		Alas the LP5523 driver doesn't support any of those advanced features of LP5523, and even abuses one of those extremely smart engines for a stupid occasional kbd backlight ramp-up/down.<br>
	''This could as well get done in mce by a 25ms timer that writes 40 new brightness values to the master control of kbd LEDs during 1 second. There's really no reason why you need to assign 1 of those 3 machines to this occasional short ramp function and break true rainbow 3 color effects on indicator LED. Sound rationale: while kbd bl ramp the ''CPU is awake anyway''. No obvious benefit in offloading the task to a ~~LP553~~ engine [except from saving 50 lines of code in mce]!)''		''This could as well get done in mce by a 25ms timer that writes 40 new brightness values to the master control of kbd LEDs during 1 second. There's really no reason why you need to assign 1 of those 3 machines to this occasional short ramp function and break true rainbow 3 color effects on indicator LED. Sound rationale: while kbd bl ramp the ''CPU is awake anyway''. No obvious benefit in offloading the task to a LP5532 engine [except from saving 50 lines of code in mce]!)''

	The controller can be programmed via "/sys/class/i2c-adapter/i2c-2/2-0032" (FIXME: is this stable?). In that directory there are three files corresponding to the three channels: engine1_mode, engine2_mode and engine3_mode. Each file "stores" one of "run", "load", "disabled" (it seems that there is another mode "direct" but it doesn't work):		The controller can be programmed via "/sys/class/i2c-adapter/i2c-2/2-0032" (FIXME: is this stable?). In that directory there are three files corresponding to the three channels: engine1_mode, engine2_mode and engine3_mode. Each file "stores" one of "run", "load", "disabled" (it seems that there is another mode "direct" but it doesn't work):

imported_>stefanl: /* Low Level */

2011-10-23T16:25:23Z

Low Level

← Older revision		Revision as of 16:25, 23 October 2011
Line 260:		Line 260:
	** engine1_leds is in the form "0000rgb00" (9 digits, "1" or "0"), where rgb control which leds should participate. The 6 "0" in the above pattern correspond to the 6 kbd white leds controlled by mce via engine3, so readout of engine3_leds gives 111100011 ;-)		** engine1_leds is in the form "0000rgb00" (9 digits, "1" or "0"), where rgb control which leds should participate. The 6 "0" in the above pattern correspond to the 6 kbd white leds controlled by mce via engine3, so readout of engine3_leds gives 111100011 ;-)
	** engine1_load accepts a string as specified above (e.g. 9d804000427f0d7f7f007f0042000000). Though LP5523 has an absolute storage of 96 steps a 16bits, which can dynamically partitioned between the three engines, the driver is retarded and limits this to a fixed 16 words a 16bit per engine (<s>ToDo: check if driver supports >16 instructions / machine</s> no it doesn't :-/)		** engine1_load accepts a string as specified above (e.g. 9d804000427f0d7f7f007f0042000000). Though LP5523 has an absolute storage of 96 steps a 16bits, which can dynamically partitioned between the three engines, the driver is retarded and limits this to a fixed 16 words a 16bit per engine (<s>ToDo: check if driver supports >16 instructions / machine</s> no it doesn't :-/)
	After that, when the state is changed to "run", the program will be ~~ran~~ by the controller.		After that, when the state is changed to "run", the program will be run by the controller.

	In order for the leds to work, the led_current of the appropriate leds need to be set to non-zero values. This is accomplished by writing to files leds:lp5523:[bgr]/'''led_current''' under the same directory. '''Don't mess with''' the default settings here that come with driver init, they are suitable and changing to higher values may and for certain values '''will kill your LEDS'''!.		In order for the leds to work, the led_current of the appropriate leds need to be set to non-zero values. This is accomplished by writing to files leds:lp5523:[bgr]/'''led_current''' under the same directory. '''Don't mess with''' the default settings here that come with driver init, they are suitable and changing to higher values may and for certain values '''will kill your LEDS'''!.

imported_>joerg_rw: /* Low Level */ variables A B C

2011-08-06T16:01:11Z

Low Level: variables A B C

← Older revision		Revision as of 16:01, 6 August 2011
Line 249:		Line 249:

	The N8x0 had a LP5521<ref name="lp5521-spec">http://www.national.com/pf/LP/LP5521.html</ref> <ref name="lp5521-driver">http://mxr.maemo.org/fremantle/source/kernel/drivers/i2c/chips/lp5521.c</ref> programmable three channel controller LED Driver which handles the leds.		The N8x0 had a LP5521<ref name="lp5521-spec">http://www.national.com/pf/LP/LP5521.html</ref> <ref name="lp5521-driver">http://mxr.maemo.org/fremantle/source/kernel/drivers/i2c/chips/lp5521.c</ref> programmable three channel controller LED Driver which handles the leds.
	(actually N900 has a LP5523 <ref name="lp5523-spec">http://www.national.com/pf/LP/LP5523.html</ref> <ref name="leds-lp5523-driver">http://mxr.maemo.org/fremantle/source/kernel/drivers/leds/leds-lp5523.c</ref>, which has 3 engines like LP5521 but can drive 9 LEDs via a multiplexer. This chip also has double the RAM size for 'programs', plus a true set of conditional branch and arithmetic commands.		(actually N900 has a LP5523 <ref name="lp5523-spec">http://www.national.com/pf/LP/LP5523.html</ref> <ref name="leds-lp5523-driver">http://mxr.maemo.org/fremantle/source/kernel/drivers/leds/leds-lp5523.c</ref>, which has 3 engines like LP5521 but can drive 9 LEDs via a multiplexer. This chip also has double the RAM size for 'programs', plus 2 per-machine and one global register variable and a true set of conditional branch and arithmetic commands.

	Alas the LP5523 driver doesn't support any of those advanced features of LP5523, and even abuses one of those extremely smart engines for a stupid occasional kbd backlight ramp-up/down.<br>		Alas the LP5523 driver doesn't support any of those advanced features of LP5523, and even abuses one of those extremely smart engines for a stupid occasional kbd backlight ramp-up/down.<br>