AVR-Chat

> If its too fast, you will loose accuracy.

Thanks for the quick responses!  I guess the next question is how
much accuracy do I lose?  For example, what kind of variation could
I expect with an 8 MHz clock and a prescale of 2?  I have the results
left justified and am only interested in 8 bit readings but want a 
high sample rate.  

Craig

At 03:23 PM 9/1/2004, Craig Limber wrote:

>> If its too fast, you will loose accuracy.
>
>Thanks for the quick responses!  I guess the next question is how
>much accuracy do I lose?  For example, what kind of variation could
>I expect with an 8 MHz clock and a prescale of 2?  I have the results
>left justified and am only interested in 8 bit readings but want a 
>high sample rate.  

That's the fun part. 
Atmel gives us no guidance here.
What worries me, is that you might get one result on one chip, and another on another chip, with a different date code. Or, it might be consistent..

--- In AVR-Chat@yahoogroups.com, Craig Limber <ffdude@s...> wrote:

> ... how much accuracy do I lose? 
> ... with an 8 MHz clock and a
> prescale of 2?  I ... am only
> interested in 8 bit readings but
> want a high sample rate.  

I don't have the ATmega128 data sheet with me, but this is fairly 
well spelled out in the ATmega64 data sheet. The Absolute Accuracy 
falls from 1.5 LSB at 200 kHz ADC clock to 3 LSB at 1 MHz (the 
maximum), both figures being typical with no worst case given.

The smallest prescaler you can use with an 8 MHz clock is 8. All bets 
are off if you try to go faster.

What do you mean by "8 bit readings"? Resolution? Absolute accuracy? 
Monotonicity? At 1 MHz clock (77 kHz sample rate) you're close to the 
limit of the built-in ADC.  If the typical figures given hold up, you 
will get 8-bits of absolute accuracy.

Graham.

My hunch, and its JUST that, is that a prescale of 2 with a
16MHz processor clock, would be a major problem. Again, a
hunch, I would not expect much better than a few bits
"accuracy" if that.

The rule I have used, without any substantiation or
validation, is that every doubling of the clock rate above
the maximum specified loses one bit.

Jim


On Wed, 1 Sep 2004 14:23:10 -0600
 Craig Limber <ffdude@shaw.ca> wrote:
> 
> 
> 
> 
> > If its too fast, you will loose accuracy.
> 
> Thanks for the quick responses!  I guess the next
> question is how
> much accuracy do I lose?  For example, what kind of
> variation could
> I expect with an 8 MHz clock and a prescale of 2?  I
> have the results
> left justified and am only interested in 8 bit readings
> but want a 
> high sample rate.  
> 
> Craig
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Yahoo! Groups Sponsor
> 
> 
>   ADVERTISEMENT 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> Yahoo! Groups Links
> 
> To visit your group on the web, go
> to:http://groups.yahoo.com/group/AVR-Chat/ 
> To unsubscribe from this group, send an email
> to:AVR-Chat-unsubscribe@yahoogroups.com 
> Your use of Yahoo! Groups is subject to the Yahoo! Terms
> of Service.
> 
> 
> 
> 
> 
> 
> 
> 
> 

---------------------------------------------------------------
The Think Different Store
http://www.thinkdifferentstore.com/
For All Your Mac Gear
---------------------------------------------------------------

At 11:52 PM 9/1/04 -0500, you wrote:
>That's the fun part. 
>Atmel gives us no guidance here.
>What worries me, is that you might get one result on one chip, and another
on another chip, with a different date code. Or, it might be consistent.. 
>

How's the raging headache going David? You seem to be getting double
posting as well as double vision :-)

OK it's time to switch off the ICE50 and relaaaaaxxxx

Regards

John Samperi

******************************************************
                                Ampertronics Pty. Ltd.
               11 Brokenwood Place Baulkham Hills NSW 2153
                 Tel. (02) 9674-6495       Fax (02) 9674-8745	
                        Email: samperi@ampertronics.com.au
                          Website  http://ampertronics.com.au
* Electronic Design   * Technical Services   * Contract Assembly
******************************************************