You don't specify a delay, you send one timed command to make the pin go
HIGH, and another timed command to make it go LOW. And the two
command-times should be "your delay" apart.

This may help:
https://www.youtube.com/watch?v=ppD06ZETnek&t=75s&pp=ygUPZ3Jjb24gdXNycCBncGlv

--M

On Sat, Feb 22, 2025 at 12:12 PM <kavin...@atindriya.co.in> wrote:

> Hi,
>    How to give the delay in uhd_usrp_set_command_time(uhd_usrp_handle
> h,int64_t full_secs,double frac_secs,size_t mboard) function?
>
>
>
>
>
>
>
>
>
> On 2025-02-22 00:10, Marcus D. Leech wrote:
> > On 21/02/2025 23:56, kavin...@atindriya.co.in wrote:
> >> Hi,
> >>   Can I use usleep(1) or any other technique is there for GPIO .
> >>
> > You might be able to use timed commands:
> >
> > https://files.ettus.com/manual/page_timedcmds.html#timedcmds_gen_cmds
> >
> > But since usleep() cannot guarantee any kind of maximum period that
> > your process will go to sleep, it is an unreliable
> >   method when precision and repeatable timing of ANY kind of "doing
> > stuff with external hardware" is required.  Even if
> >   usleep() at microsecond-scale intervals *WAS* reliable and
> > repeatable, there's no guarantee about all the *other* things
> >   that are required to launch a command over a latency-not-guaranteed
> > bus to the external hardware, including kernel-layer
> >   USB drivers, etc, etc.
> >
> >
> >>
> >> On 2025-02-21 23:37, Marcus D. Leech wrote:
> >>> On 21/02/2025 23:22, kavin...@atindriya.co.in wrote:
> >>>> Hi,
> >>>>  Thanks for answering.
> >>>>  Can you explain me clearly?
> >>>>
> >>>> Thanks.
> >>> I'd suggest, at a minimum, looking at the MAN page for usleep().
> >>>
> >>>
> >>>>
> >>>>
> >>>>
> >>>> On 2025-02-21 23:11, Marcus D. Leech wrote:
> >>>>> On 21/02/2025 23:05, kavin...@atindriya.co.in wrote:
> >>>>>> Hi,
> >>>>>>    In USRP B205mini-i I am able to access the GPIO Pin. I have to
> >>>>>> create a Pulse in the GPIO pin. I want to make high the gpio pin
> >>>>>> for 1microsecond and low for 999microseconds. How to achieve this
> >>>>>> timings through coding. Can you give me the solution for this?
> >>>>>>
> >>>>>>    In coding after making high I am giving usleep(1) function but
> >>>>>> when I am check in the oscilloscope it is showing 64microseconds.
> >>>>>> How?
> >>>>>>
> >>>>>>
> >>>>> Because an application-land program executing on an ordinary
> >>>>> general-purpose operating system is unlikely to be able to
> >>>>>   achieve repeatable, 1usec-scale timing.  This has nothing to do,
> >>>>> per se, with UHD or USRPs, but rather, to understanding
> >>>>>   how applications execute in a general-purpose operating system
> >>>>> that isn't ruthlessly optimized for "hard" real-time tasks.
> >>>>>
> >>>>>
> >>>>> _______________________________________________
> >>>>> USRP-users mailing list -- usrp-users@lists.ettus.com
> >>>>> To unsubscribe send an email to usrp-users-le...@lists.ettus.com
> >>> _______________________________________________
> >>> USRP-users mailing list -- usrp-users@lists.ettus.com
> >>> To unsubscribe send an email to usrp-users-le...@lists.ettus.com
> > _______________________________________________
> > USRP-users mailing list -- usrp-users@lists.ettus.com
> > To unsubscribe send an email to usrp-users-le...@lists.ettus.com
> _______________________________________________
> USRP-users mailing list -- usrp-users@lists.ettus.com
> To unsubscribe send an email to usrp-users-le...@lists.ettus.com
>
_______________________________________________
USRP-users mailing list -- usrp-users@lists.ettus.com
To unsubscribe send an email to usrp-users-le...@lists.ettus.com

Reply via email to