From: Jonathan Kirwan
Subject: Re: Querry, Kick Starting Crystal Oscilator.
References: <3DCA905A.59FD928B@mmm.com.DELETETHIS> <3DD26B65.firstname.lastname@example.org> <email@example.com>
X-Newsreader: Forte Agent 1.92/32.572
NNTP-Posting-Date: Thu, 14 Nov 2002 01:53:08 GMT
Organization: AT&T Broadband
Date: Thu, 14 Nov 2002 01:53:08 GMT
On Wed, 13 Nov 2002 16:10:55 GMT, Spehro Pefhany
>On Wed, 13 Nov 2002 15:09:50 GMT, the renowned Fred Bloggs
>>You should forget about playing games with the 32kHz crystal and go with
>>a 4.096MHz crystal driving a /128 prescalar into the processor. Since it
>> has been demonstrated that Q roughly sets the start-up time, and if
>>you notice that this is actually a number of cycles, then increasing the
>>frequency will decrease start-up time proportionately. A 4.096MHz should
>>be in full oscillation in 20ms.
>Okay so far.
>>The increased power dissipation will be
>>negligible compared to the processor.
>Many modern CMOS processors use tens of uA or less at 32kHz clock.
>Some only one or two uA. For example, this 16-bit micro:
>> If you survey some typical
>>oscillators such as at http://www.ecsxtal.com/surfosc.htm , then you
>>will see that the ~4MHz oscillators have a start-up time in the 4-30ms
>>range. The 4.096MHz crystal is readily available in several styles.
>Those packaged oscillators draw a huge amount of current, over 10mA
>for some.. you could run thousands of CPUs with that much juice. ;-)
It's not the least bit difficult to find packaged oscillators at
25mA at 10MHz. Been there. Darned things can be, by far, the
hottest thing on the board!
>If current consumption is not a factor, then this solution is the
>obvious one, but then why not just run the whole micro at 4MHz?
>Can you make a 4MHz oscillator that runs reliably on just a few uA?
That's exactly the issue. Who uses a 32MHz oscillator who isn't
concerned about power consumption? I'm still pondering that