Re: [ntp:questions] [Is deze mail veilig?] [Is this e-mail safe?] [Cet e-mail est-il sans danger?] Re: Shared PPS source/Multiple PPS sources
On 07/02/15 10:29, Rob wrote: I presume you meant this followup to the multi PPS sources to a single system and then it is not true either, of course our systems have at least 4 cores and they can service multiple interrupts at the same time. On a single-core system I'd invert one of the PPS signals and then appropriately fudge the offset so that the two interrupts do not compete with each other. ___ questions mailing list questions@lists.ntp.org http://lists.ntp.org/listinfo/questions
Re: [ntp:questions] [Is deze mail veilig?] [Is this e-mail safe?] [Cet e-mail est-il sans danger?] Re: Shared PPS source/Multiple PPS sources
On 2015-02-07, Jan Ceuleers jan.ceule...@computer.org wrote: On 07/02/15 10:29, Rob wrote: I presume you meant this followup to the multi PPS sources to a single system and then it is not true either, of course our systems have at least 4 cores and they can service multiple interrupts at the same time. On a single-core system I'd invert one of the PPS signals and then appropriately fudge the offset so that the two interrupts do not compete with each other. Except I would not trust the gps to make the length of the pulse EXACTLY 1ms to the nanosecond say. Ie, the pulse length could vary by the 10usec. But I have not tested this and it may well depend on the manufacturer. The pulse length is usually there to make sure that the interrupt hardware sees the pulse, and the manufacturer of the gps may not decide it is worth making the pulse length exact. ___ questions mailing list questions@lists.ntp.org http://lists.ntp.org/listinfo/questions
Re: [ntp:questions] [Is deze mail veilig?] [Is this e-mail safe?] [Cet e-mail est-il sans danger?] Re: Shared PPS source/Multiple PPS sources
On 07/02/15 19:10, William Unruh wrote: Except I would not trust the gps to make the length of the pulse EXACTLY 1ms to the nanosecond say. Ie, the pulse length could vary by the 10usec. But I have not tested this and it may well depend on the manufacturer. The pulse length is usually there to make sure that the interrupt hardware sees the pulse, and the manufacturer of the gps may not decide it is worth making the pulse length exact. That is a fair point. The result would be high jitter. Still worth a go though, since an inverter is cheap (and one might be needed anyway for level conversion purposes). If the device does not provide a programmable offset, that is (as Mike Cook suggested). ___ questions mailing list questions@lists.ntp.org http://lists.ntp.org/listinfo/questions