[PIC] Is PIC18F2550 internal oscillator enough precise to do a 24h clock?

[Alloy]

Hey
Is PIC18F2550 internal oscillator enough precise to do a 24h clock?
I mean, It will display minutes and hours, and after 23:59 it will flip to 00:00 and so on....
It is supposed to run all time and have minimal components...

 

[andre_luis]

Internal oscillators, even having calibration ( which I'm unaware if this uC has this feature ) they inherently have a very high variance when compared to the ones made with crystals, since they are based on RC networks implemented in silicon, and thus change their characteristics with the temperature. If you have some way to permanently read precise externals signals ( e.g 50/60Hz from mains ) you can atleast implement a correction factor to adjust the clock counter.

 

[Alloy]

Ook, but from a practical point of view, how much deviation will I get?
1 minute per 24 hour?
More, less?

How much using a 12MHz crystal will help?

 

[KlausST]

Hi,

I bet the answer is in the datasheet.
But I'm no PIC user.

Either you find the value in % or in ppm. Or in a chart.

The same is with the xtal.

Klaus

 

[FvM]

All accuracy numbers are in the datasheet why don't you read it?

If internal oscillator error is in the +/-1 % band (can be more), your clock if off by +/- 15 minutes per day. This should answer the question.

50 ppm crystal error translates to 4 sec/day. Still not acceptable for a long running real time clock.