How to terminate the function that runs every n seconds

[Grant Edwards]

On 2015-01-14, Mark Lawrence <breamoreboy at yahoo.co.uk> wrote:
> On 14/01/2015 17:37, Grant Edwards wrote:
>> On 2015-01-14, Mark Lawrence <breamoreboy at yahoo.co.uk> wrote:
>>
>>> Reminds me of working on Telematics S200/300/4000/5000 telecomms kit in
>>> the early 90s where the timers were mains based, so a one hour timer
>>> would go off at about one hour, 30 seconds.
>>
>> I don't understand.  Power line frequencies are _very_ accurate and
>> have been relied upon for timekeeping since the 1930s.  We're talking
>> a few hundred PPM over a 24 hour period.  A 30 second error over a one
>> hour period seems _really_ high.
>
> http://www2.nationalgrid.com/uk/services/balancing-services/frequency-response/ 
> "National Grid has a licence obligation to control frequency within the 
> limits specified in the 'Electricity Supply Regulations', i.e. ±1% of 
> nominal system frequency (50.00Hz) save in abnormal or exceptional 
> circumstances.".  I wouldn't describe ±1% as very accurate and certainly 
> not within a few hundred ppm.

Sorry, I should have guessed from the use of the word "mains" that you
were in the UK -- which seems to have much laxer power-line frequency
regulation than the US.

> I'm assuming that this kind of limitation applies around the world,
> so could you be getting confused with some other more accurate
> frequency control?

Power line frequency control in the US is much tighter than the UK.
According to real-world data I've seen, powerline-based timings are
typically accurate to a few seconds per month.  In the US, they
implement freqency corrections every hour to keep long term,
cumulative time errors under a certain limit (the limit ranges from 2
to 10 seconds depending on region).

-- 
Grant Edwards               grant.b.edwards        Yow! I love ROCK 'N ROLL!
                                  at               I memorized the all WORDS
                              gmail.com            to "WIPE-OUT" in 1965!!