AVR-Chat

0.001%!

Where do you buy your capacitors - Planet Ultron?   Given their very nature
and the limitations on measurement you'd find it difficult to match better
than about 1%.

-----Original Message-----
From: David VanHorn [mailto:dvanhorn@cedar.net] 
Sent: March 14, 2004 6:46 AM
To: AVR-Chat@yahoogroups.com; AVR-Chat@yahoogroups.com
Subject: Re: [AVR-Chat] Oscillator Lock?!?!


At 08:38 AM 3/14/2004 -0500, John Johnson wrote:

>Not to split hairs here, but the two caps will never be equal. 

I think you're very much splitting hairs. 

Look at the crystal manufacturer's data sheets, and tell me where they show
unequal values in any normal application. 

Or are you making the case, that like two resistors, they may be 0.001% off
no matter how carefully they match them?

At 08:52 AM 3/15/2004 -0700, Cobb, Quentin wrote:

>0.001%!
>
>Where do you buy your capacitors - Planet Ultron?   Given their very nature
>and the limitations on measurement you'd find it difficult to match better
>than about 1%.

I was stretching the point to make a point. 

Theoretically speaking, no two small caps will ever be exactly the same value.
However, for any practical definition of the word, the two oscillator caps are intended to be the same value, and if by some miracle they are EXACTLY matched, they will still work just fine.

Just make sure CKOPT is set to 1.

I'm seeing this on an M128, but it may pertain to others as well. 

I have the part in sleep, with the CKDIV set to 128, and everything I can, shut down.
When I see a wakeup condition, I would like to jump to reset, but so far, that seems not to work.. However, if I jump to a jmp here loop, and let the WDT do the reset, then it does work. 

On Jmp Reset, I can't really define what's wrong, since I can't Jtag it in that condition, but it seems that some part of the chip doesn't get actually reset without a pulse on the reset line, or a WDT.

Anyone else seen this, or have evidence to the contrary?

Umm...I think you should delete this and repost as a NEW topic. :)

--- In AVR-Chat@yahoogroups.com, David VanHorn <dvanhorn@c...> wrote:
> 
> I'm seeing this on an M128, but it may pertain to others as well. 
> 
> I have the part in sleep, with the CKDIV set to 128, and everything
I can, shut down.
> When I see a wakeup condition, I would like to jump to reset, but so
far, that seems not to work.. However, if I jump to a jmp here loop,
and let the WDT do the reset, then it does work. 
> 
> On Jmp Reset, I can't really define what's wrong, since I can't Jtag
it in that condition, but it seems that some part of the chip doesn't
get actually reset without a pulse on the reset line, or a WDT.
> 
> Anyone else seen this, or have evidence to the contrary?