[time-nuts] Setting up PRS10 with M12+T GPS

[Magnus Danielson]

Björn,

On 05/03/2012 12:42 AM, bg at lysator.liu.se wrote:
>> Yes, all made clear.
>> Trigger on Ch.4, I see. You have written "green: 10MHz PRS10B" but it must
>> be A, otherwise it seems that the PPS and the 10MHz of the PRS10 are not
>> phase aligned. Moreover, the PRS10B seems to track the PRS10A with "some
>> delay" as the 10MHz from A spreads on more than 5nS (useful, the infinite
>> persistence, isn't it? I use it a lot)
>>
>>
>
> I had
>
>      locked Tbolt 1PPS going into PPS_IN of PRS10A
>      PPS Out of PRS10A going into PPS_IN of PRS10B.
>
> I watched/measured
>      1) PPS out from Tbolt on ch 1  (yellow)
>      2) 10MHz out from PRS10A on ch 2 (green)   // Corrected
>      3) 10MHz out from PRS10B on ch 3 (blue)    // Corrected
>      4) 1PPS out from PRS10 B on ch 4 (red)
>
> I should do some proper measurements with TIC, instead of just 'watching'
> on the scope. Yes the 1pps and 10MHz is ofcause stable within the same
> PRS10. But should not the 10MHz zero crossing be aligned with the 1PPS?

No. See my other posting. As far as I can see, the PRS-10 does not 
bother to create relative phase alignment between 10 MHz and PPS. If you 
want that, you will have to play tricks with the time-offsets (TO) with 
the hints from the time-tags (TT) and then some separate calibration.

It seems like there is still some time-offset calibration to be done.

> Cables were not perfectly matched, but should be within 2-3ns of each
> other.
>
> Yes, persistance is very nice!

Indeed. For stuff like this, I like having a scope in parallel with the 
signals so I see what happens. The counter gives the quality measures, 
but the scope gives the overall picture and you get such a more direct 
response of behaviour as you see the PPS move about.

Cheers,
Magnus