Hi Martin,

According to the Datasheet there is a tolerance of 150 ns in the
Timings. I'll guess a resolution of 300 ns would be fine too.
I could use inline assembler for precise time wasting but I'd prefer
the usage of a timer implementation. 

I'd like to see this driver in a pull request some day ;)

Am Dienstag, den 16.02.2016, 18:17 +0100 schrieb Martin:
> Hi Simon,
> 
> I'm not the timer expert, but to reliably get a 100 ns resolution
> sounds 
> rather tough an a system that runs with 170MHz [1].
> I think there's not too much air between xtimer and the bare silicon.
> You can have a look into the timer configuration [2] and the timer 
> interrupts [3] of your board.
> 
> But I guess this is not what you want.
> 
> Best regards,
> Martin
> 
> [1] https://github.com/RIOT-OS/RIOT/wiki/Board%3A-STM32F4discovery
> [2] 
> https://github.com/RIOT-OS/RIOT/blob/master/boards/stm32f4discovery/i
> nclude/periph_conf.h#L53
> [3] 
> https://github.com/RIOT-OS/RIOT/blob/master/cpu/stm32f4/periph/timer.
> c#L288
> 
> Am 02/16/2016 um 05:17 PM schrieb Simon Brummer:
> > 
> > Hi Everybody,
> > 
> > I'll try to write a driver for a programmable RGB-Ledstripe. To
> > programm the timing for the Ledstripe I need a timerresolution of
> > 100
> > ns. As far as I can see, xtimer features only a 1 µs resolution. Is
> > there a way to adjust xtimers resolution or is there another way to
> > use
> > timers in a generic way?
> > 
> > My Project runs on a stm32f4 discovery board.
> > 
> > 
> > Simon
> > _______________________________________________
> > users mailing list
> > [email protected]
> > https://lists.riot-os.org/mailman/listinfo/users
> _______________________________________________
> users mailing list
> [email protected]
> https://lists.riot-os.org/mailman/listinfo/users
_______________________________________________
users mailing list
[email protected]
https://lists.riot-os.org/mailman/listinfo/users

Reply via email to