Clock and Dual Termometer

[fratello]

Thanks for atention ! All three programs are - without read_temperature - verry accurate !
I try, in all three cases, to put the "read_temp" code in other places... the results are inadequates !

[Acetronics2]

@ first you can try your program with a 9 bits resolution, to have the quickest sensor conversion time ...

THEN you will know where is something wrong ...

Alain

[malc-c]

Not checked the data sheet so can't confirm if this is an option, but for the sake of a couple of ££ why not use a DS1307 RTC chip ??

[cncmachineguy]

I try, in all three cases, to put the "read_temp" code in other places... the results are inadequates !

Still waiting to hear what the results are. Is it slow, fast, sometimes right, sometimes wrong? Is it close or all over the place?

[fratello]

The clock run ok without read_temp, in all cases and run faster ( strange for me ) if use this subroutine.
I try now with 9 bits resolution of DS18B20, wait for result !
Thanks for support !

[fratello]

If DS18B20 it's in 9 bit mode, the clock run faster (!) with ~ 2 sec/hour (5 sec in 3 hours).

[ardhuru]

Thats about .055%; you ARE using a crystal, not a resonator, right?

Anand

[fratello]

Yes, I use a crystal. But, without code for DS18B20, the clock run correct (~ 1 sec/day).

[HenrikOlsson]

Hi,
If I read this correctly you have an interrupt frequency of 61.035Hz (prescaler 1:64, 256*64=16384, 1/0.016384=61.035Hz) when you ideally should have 61Hz - that's 0.057% off and as Anand says 2 seconds over one hour is 0.055%.

I currently have no idea why it starts running correct when you remove the DS1820 code - are you sure that's the case?