On Fri, Jul 8, 2016 at 8:12 AM, 'Oliver Seitz' via jallib <
[email protected]> wrote:
>
> Matt wrote:
>
> Using a prescaler reduces accuracy
> but would let you count at a higher frequency. Using a 8bit
> timer instead of a 16 bit timer is more accurate.
>
> Hi Matt!
>
> The plan is to use a 12f1572. According to the datasheet, both TMR1 can
> count up to 16,666MHz without or 33,333MHz with internal prescaler. To
> count the max. prescaled frequency, core clock must be 32MHz with
> prescaler=8.
>
> TMR0 can handle shorter pulses, but the period is like for TMR1 defined as
> "Greater of: 20 or (Tcy + 40)/prescale value". Min Tcy is 200ns, so the
> second term can never go below (240/8=) 30. This may be a datasheet error,
> but the way it is written it won't count faster than 33,333MHz.
>
> If you're counting, say, 10000000Hz or 10MHz, a prescaler value of 8
> introduces a possible error of -7Hz, because there may be counts left in
> the prescaler which cannot be read. The internal oscillator is to be used
> as reference with +/-2% of accuracy, or +/-200000Hz. Even when using a
> crystal with +/-15ppm, the reference introduces a possible error of
> +/-150Hz. Therefore, when it comes to higher frequency, I'd say the
> prescaler error can be accepted. When the frequency appears to be low, the
> prescaler can be switched off.
>
Actually the error will go up to 3% at least. I'm using it to read a
moisture sensor I've designed which will work outside. Still at 3% the
error is insignificant because the frequency value will be compared with
the programmed A2D value... So I will cut the trail anyway.
Oliver, thanks for your comments, it helps.
Vasile
>
> I can't imagine however why using an 8-bit counter should be more accurate.
>
> Greets,
> Kiste
>
> --
> You received this message because you are subscribed to the Google Groups
> "jallib" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to [email protected].
> To post to this group, send email to [email protected].
> Visit this group at https://groups.google.com/group/jallib.
> For more options, visit https://groups.google.com/d/optout.
>
--
You received this message because you are subscribed to the Google Groups
"jallib" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To post to this group, send email to [email protected].
Visit this group at https://groups.google.com/group/jallib.
For more options, visit https://groups.google.com/d/optout.
