[time-nuts] WWVB PPS clock pulse ?

Lenny Story lenny at Codematic.com
Wed Mar 30 02:30:29 UTC 2011


All,

Ok... so i'll just make it a rising edge pulse of ~500ms.  That should be
enough time for a linux box to detect it.

Thanks,
-Lenny


On Mon, Mar 28, 2011 at 6:54 PM, Rex <rexa at sonic.net> wrote:

> On 3/28/2011 9:46 AM, Lenny Story wrote:
>
>> Greetings,
>>
>> Is there an accepted form for the PPS clock pulses that are produced
>> by the various time equipment in use today ?
>>
>>
>>
> Short answer -- No.
>
> Examples:
> HP GPSDOs like 3801A and 3816A have a positive-going square pulse that is
> 10's of microseconds long. From the 3801 manual, "1 PPS Output
> Characteristics
> Jitter of leading edge: < 200 nanoseconds between pulses.
> Accumulated time error: < 7 microseconds per day unlocked, for
> 24 hours, after one day of stabilization and 2 days of locked operation
> with a fixed antenna location.
> Waveform: Pulse width 10 to 50 microseconds.
> Time Accuracy: < 1 microsecond (Locked to GPS)
> Connector: DB­25 (J3)"
>
> I just looked at the PPS from my 3816A and it is about 3V into 50 ohms and
> a bit over 20 uS long.
>
> The T-bolt manual says 10 uS wide and can be configured either positive or
> negative pulse.
>
> Looking at the M12 receiver manual, looks like a 6-7 mS pulse.
>
>
>
>


More information about the time-nuts mailing list