[time-nuts] Re: Picotest U6200A problem with measuring PPS period very second.

[Bob kb8tq]

Hi

*Could* it be perfect in some ideal device? Sure it could. In any real world
device I’ve ever used (and I think I’ve used just about everything on the market
…..) , it’s been far from perfect. 

As you get out to even a modest tau ( say 1K sec) you are looking at something
in the 10,000 to 100,000 samples range. There is a lot of room for all sorts of
issues to creep in. 

Rounding is only one of many ways things can go tilt. It is the most obvious
and generally the easiest to understand. 

Since there is a quick / simple / (near) zero cost solution ( look at a delta  to a second
pps stream) crossing more sources of trouble off the list is normally a pretty easy
decision. 

Bob

> On Feb 27, 2022, at 1:50 PM, Erik Kaashoek <erik at kaashoek.com> wrote:
> 
> Bob
> Would this rounding error be reduced if the phase is measured in a gapless
> way, e.g. against a continuos running phase counter that is captured at the
> moment of the PPS and the period is calculated as the delta between current
> and previous capture. Any rounding in this delta calculation can cause
> problems so it should be done in double precission I guess.
> Will run some tests if the current period calculations are phase stable.
> 
> 
> On Sun, Feb 27, 2022, 18:19 Bob kb8tq <kb8tq at n1k.org> wrote:
> 
>> Hi
>> 
>> If your goal is an ADEV measurement, you will need to tag the
>> edges against a second “stable” PPS signal. A sequence of period
>> measures can get you into issues related to round off problems.
>> 
>> Bob
>> 
>>> On Feb 27, 2022, at 11:27 AM, Erik Kaashoek <erik at kaashoek.com> wrote:
>>> 
>>> But why can't you measure the 1 s period of a PPS?
>>> If the counter would be able to do gapless measurement and you set the
>> gate to 0.9s it should be able to measure the period of a PPS, at least my
>> own build counter behaves that way so I was surprised the U76200A could not
>> do that.
>>> 
>>> On 27-2-2022 14:25, Azelio Boriani wrote:
>>>> Without a reference? If the same PPS acts as start and as stop then
>>>> one (I suppose) rising edge is start and the next rising edge is stop,
>>>> so a measurement every second is not possible. Maybe a dedicated
>>>> hardware with two channels where one edge is the start for a channel
>>>> and stop for the other is necessary.
>>>> 
>>>> On Sun, Feb 27, 2022 at 12:08 PM Erik Kaashoek <erik at kaashoek.com>
>> wrote:
>>>>> I'm trying to measure the phase of a 1 second PPS from a cheap GPS
>>>>> module using a Picotest U6200A
>>>>> Set up in time/period mode, input 1. Trigger level is set to 1V, DC
>>>>> coupling, all seems to work well but there is only one measurement
>> every
>>>>> two seconds which either misses output the in between periods of
>>>>> averages over two periods.
>>>>> Does anyone know if it is possible with the U6200A to measure every
>>>>> period of a PPS?
>>>>> Erik.
>>>>> _______________________________________________
>>>>> time-nuts mailing list -- time-nuts at lists.febo.com -- To unsubscribe
>> send an email to time-nuts-leave at lists.febo.com
>>>>> To unsubscribe, go to and follow the instructions there.
>>>> _______________________________________________
>>>> time-nuts mailing list -- time-nuts at lists.febo.com -- To unsubscribe
>> send an email to time-nuts-leave at lists.febo.com
>>>> To unsubscribe, go to and follow the instructions there.
>>> _______________________________________________
>>> time-nuts mailing list -- time-nuts at lists.febo.com -- To unsubscribe
>> send an email to time-nuts-leave at lists.febo.com
>>> To unsubscribe, go to and follow the instructions there.
>> _______________________________________________
>> time-nuts mailing list -- time-nuts at lists.febo.com -- To unsubscribe send
>> an email to time-nuts-leave at lists.febo.com
>> To unsubscribe, go to and follow the instructions there.
> _______________________________________________
> time-nuts mailing list -- time-nuts at lists.febo.com -- To unsubscribe send an email to time-nuts-leave at lists.febo.com
> To unsubscribe, go to and follow the instructions there.