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

[Tom Van Baak]

Erik,

 > Does anyone know if it is possible with the U6200A to measure every 
period of a PPS?

Yes, use time interval mode instead of period or frequency mode. TI mode 
is a 2-channel measurement so one channel is your GPS/1PPS and the other 
channel is your Rb/1PPS. The resulting readings (phase difference) are 
directly compatible with Stable32 / TimeLab.

 > But why can't you measure the 1 s period of a PPS?

You can. The period of a 1PPS is simply 1 plus the delta between 
successive pairs of TI readings.

If that's too succinct, consider two consecutive TI readings. Say TI[1] 
= GPS[1] - Rb[1] and TI[2] = GPS[2] - Rb[2]. GPS period = GPS[2] - 
GPS[1] which is (TI[2] + Rb[2]) - (TI[1] + Rb[1]) and since if your Rb 
is accurate Rb[2] - Rb[1] is 1 so GPS period is simply 1 + TI[2] - 
TI[1]. It's intuitive once you play with the data, on paper, in Excel, 
or TimeLab.

 > If the counter would be able to do gapless measurement

Well, yes, *if*. But check the U6200A manual. I don't see gapless period 
measurements mentioned as a feature. It's not something often done. If 
you want precise low rate pulse edge measurements you use time interval, 
or timestamping instead. The U6200A, and the 53132A from which it is 
cloned, are not timestamping counters. But both do time interval, so use 
that. That's what almost all of us use.

Note that -- at 1/10th the price -- the TAPR/TICC can do all of this 
because it is designed at its core to be a timestamping counter. It can 
be used for frequency and period, but it does that based on the raw 
timestamp data. The picPET and ProTIC are the same way.

 > so I was surprised the U76200A could not do that

Period measurements usually start with one leading edge and count to the 
next leading edge. Then there's the process of getting the binary data, 
formatting to decimal, and communication. By the time the cycle resumes, 
the 2nd pulse is already gone so the next period measurement starts with 
the 3rd pulse ending with the 4th pulse. Thus the measurement rate is 
half the 1PPS rate. Most period counters work this way. This issue is 
not obvious with higher frequencies but very obvious with 1PPS. Using 
auto-trigger makes it even worse.

All this is why time interval, or especially timestamping, counters are 
a better way to go. You get gap-less / zero dead-time measurements for free.

/tvb