[time-nuts] DS3231m 1HZ output stability

[d0ct0r]

Hello,

I am experimenting to compare 1PPS outputs from Dallas/Maxim DS3231m 
chip and 1PPS output from generic GPS module. I attached both signal to 
MCU timers with free run counters and measure the interval between of 
raising edges.

Here is the typical results:

# Delta GPS: 3659, Delta RTC: 3660
# Delta GPS: 3659, Delta RTC: 3659
> # Delta GPS: 3658, Delta RTC: 1984
> # Delta GPS: 3658, Delta RTC: 1984
# Delta GPS: 3659, Delta RTC: 3659
# Delta GPS: 3659, Delta RTC: 3660

In my tests, Delta is the number of MCU "tics" (counter value) and its 
depends of MCU clock settings.
Usually "Delta" pretty much close for RTC and GPS. But sometimes, 
DS3231m shows some big difference.
Meantime, 1PPS from GPS module is much more consistence.

I would like to ask if somebody did measurement for that Dallas DS32XX 
series chips. Is that 1PPS (1HZ square wave) stable there or it has some 
issues ? I am thinking that "burp" in my tests could be because of my 
poor code or may be because of DS32XX doing temperature conversion every 
second, which cause an impact to the 1HZ output. I would be interested 
to hear any information about. From the datasheet of DS3231m , that 1HZ 
from DS32XX should be stable.


Regards,

V.P.