nuttx/nuttx-mirror
1
2
Fork 0
mirror of https://github.com/apache/nuttx.git synced 2025-01-13 07:28:38 +08:00
Code Issues Projects Releases Packages Wiki Activity Actions

drivers/leds/ws2812: optimize reset bytes

Browse source 
It's not necessary to send a reset pulse after sending the RGB data
since the first thing that a new transmission does is sending the long
reset pulse. I have left 1 reset byte after the RGB data to keep some
padding given that the bits inside the byte are shifted by 1 bit although
I think it wouldn't be necessary either. But just in case.

This reduces the memory usage (you can easily save 60 bytes) and increases
the maximum refresh rate.
This commit is contained in:
Diego Herranz 2020-12-09 09:36:15 +00:00 committed by Xiang Xiao
parent 866bb35d2d
commit d9d649105d
1 changed files with 3 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

6
drivers/leds/ws2812.c
View file

@ -78,17 +78,17 @@
#define WS2812_RW_PIXEL_SIZE 4
/* Transmit buffer looks like:
* [<----reset bytes---->|<-RGBn->...<-RGB0->|<----reset bytes---->]
* [<----N reset bytes---->|<-RGBn->...<-RGB0->|<----1 reset byte---->]
*
* It is important that this is shipped as close to one chunk as possible
* in order to meet timing requirements and to keep MOSI from going high
* between transactions. Some chips will leave MOSI at the state of the
* MSB of the last byte for this reason it is recommended to shift the
* bits that represents the zero or one waveform so that the MSB is 0.
* The reset clocks will pad the shortened low at the end.
* The reset byte after the RGB data will pad the shortened low at the end.
*/
#define TXBUFF_SIZE(n) (WS2812_RST_CYCLES * 2 + n * WS2812_BYTES_PER_LED)
#define TXBUFF_SIZE(n) (WS2812_RST_CYCLES + n * WS2812_BYTES_PER_LED + 1)
/****************************************************************************
* Private Types