[time-nuts] CPU clock jitter

M. Simon msimon6808 at yahoo.com
Wed Jan 9 11:57:06 UTC 2013


Message: 1
I was thinking more along the lines of large jitter in the CPU clock. A clock shortened by 1/8th cycle could play havoc with CPU operation. Of course you could do the experiment to see what happens. Or you could divide by 10 and only worry about the rest of the problems. 

Simon

Date: Tue, 08 Jan 2013 08:27:18 -0800
From: Ed Breya <eb at telight.com>
To: time-nuts at febo.com
Subject: Re: [time-nuts] (no subject)
Message-ID: <50EC48E6.5020006 at telight.com>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

Yes, that's true, Simon, but remember the initial goals of simplicity 
and long term phase coherence, while jitter doesn't matter so much. The 
longer term average frequency ratio should be right on, while comparing 
at any particular cycles it would be awful.

Ed



------------------------------

Message: 2
Date: Tue, 08 Jan 2013 20:56:32 +0100
From: Magnus Danielson <magnus at rubidium.dyndns.org>
To: time-nuts at febo.com
Subject: Re: [time-nuts] (no subject)
Message-ID: <50EC79F0.3020005 at rubidium.dyndns.org>
Content-Type: text/plain; charset=ISO-8859-1; format=flowed

On 01/08/2013 09:08 AM, M. Simon wrote:
>
 To reduce cycle to cycle jitter I think a divide by 10 is in order (/5,
 /2). And if you can arrange it the MCU should be a little slow.

The cycle jitter doesn't change in time-amplitude by division, it's just 
the jitter relative to the period which changes with the division.

Cheers,
Magnus



 



Engineering is the art of making what you want from what you can get at a profit.


More information about the time-nuts mailing list