[ntp:questions] What prevents continuous time within an operating system ?

MAYER Hans mayer at iiasa.ac.at
Wed Jan 20 19:59:06 UTC 2016



Thanks for your reply.

> The mother board you are using may have a very cheap oscillator ...

Yes, this may be the reason. 
Definitely it's not an instability of the time source. 
It could also be the GPIO module. I have seen this on different systems but always same hardware type. 

> ... that is very temperature sensitive.

I was thinking about this too. I monitored the temperature. See below. 
The last column is the temperature in Celsius times 1000 for the same time frame I posted the images. 
OK it's the CPU temperature and not the room respectively quartz temperature. 
I have never seen  a correlation between time offset and temperature. 
And for my opinion the time offset jumps to quickly that it could be the temperature. 

// Hans


20160111 170001 1452528001 59905 37800
20160111 171001 1452528601 60505 37800
20160111 172001 1452529201 61105 37800
20160111 173001 1452529801 61705 37800
20160111 174001 1452530401 62305 37800
20160111 175001 1452531001 62905 37800
20160111 180001 1452531601 63505 37800
20160111 181001 1452532201 64105 37800
20160111 182001 1452532801 64705 37800
20160111 183001 1452533401 65305 37800
20160111 184001 1452534001 65905 37800
20160111 185001 1452534601 66505 37800
20160111 190001 1452535201 67105 37800
20160111 191001 1452535801 67705 37800
20160111 192001 1452536401 68305 37800
20160111 193001 1452537001 68905 36200
20160111 194001 1452537601 69505 37800
20160111 195001 1452538201 70105 37800
20160111 200001 1452538801 70705 37800



________________________________________
From: questions [questions-bounces+mayer=iiasa.ac.at at lists.ntp.org] On Behalf Of Leon McCalla [leonmccalla at hotmail.com]
Sent: Monday, January 18, 2016 3:05 PM
To: questions at lists.ntp.org
Subject: Re: [ntp:questions] What prevents continuous time within an operating system ?

The mother board ypu are using may hava a very cheap oscillator that is very temperature sensitive. You may find that. As the air temperature in the room changes 4 or 5 degrees, the cpu frequency may change +/- 0.1% or more.

It doesn't seem like a big deal but 0.1% is 3 seconds per hour. NTPd could be working non stop trying to compensate for this.

Leon.

Sent from my android device.



More information about the questions mailing list