Crystal osc running but PIC not using it

[p15218]

I'm trying to use the PBP Count function with a bit more stability than the 16F690's internal RC oscillator provides.

My problem is that I have the crystal oscillator running with the following edit to 16f690.inc file:
; __config _INTRC_OSC_NOCLKOUT & _WDT_ON & _MCLRE_ON & _CP_OFF ------- original code
__config _HS_OSC & _WDT_ON & _MCLRE_ON & _CP_OFF --------my edit

BUT it matters not if I remove or short the crystal, the code still runs!
I've checked the 4 MHz crystal's pins with a scope & counter and it is active at about 3999990 Hz, close enough to 4 MHz for me.
MCLR is pulled high with a 4.7k Ohm resistor.

What else do I have to do to enable the crystal oscillator internal to the PIC ?
thanks
Bob

[Charles Linquis]

Your problem lies with the line

__config _INTRC_OSC_NOCLKOUT

The "_INTC_OSC" part means to use the internal oscillator.

You should change this to "_XT_OSC" instead

[p15218]

Your problem lies with the line

__config _INTRC_OSC_NOCLKOUT

The "_INTC_OSC" part means to use the internal oscillator.

You should change this to "_XT_OSC" instead


Well thanks but I did...notice the ";" commenting the "_INTC_OSC" line out.
Changing to "_XT_OSC" has activity at the crystal as well as _HS_OSC did with the same result that pulling the crystal makes no difference, the code runs no matter.

[Darrel Taylor]

Make sure your programmer is updating the CONFIG settings of the chip.
Usually there's an option for that.

It may still be using the CONFIG settings from a previous program.
<br>

[Archangel]

Hi Bob,
This particular PIC has special Oscillator, options and Has FSCM, which stands for Fail Safe Clock Monitor - datasheet page 49 of 294 and as I understand it, it automaticly deploys an internal clock when oscillator failure occurs, pretty darn nifty feature unless you are doing what You want to . Easiest way is use a cheaper PIC.

[mister_e]

Easiest way is use a cheaper PIC.

And never learn... not for me.

[Archangel]

And never learn... not for me.

Me no understand . . . was I wrong?

[mister_e]

may bee hi deed not reed care foolee dee hole tred ;)

'The Easiest solution and use another PIC' reply always drive me nuts when the user don't need to and have a nice 'learning opportunitie' on hand... i know i may have badly interpret it anyways... Need to sleep... sorry :(

Good news, you are right! Assuming the Fail-Safe clock monitor AND the int-ext-switchover mode are both enable, at POR, the internal OSC run @ 4MHz.... hence why it always work without noticeable bugs.

[p15218]

Ah, thanks to Joe S. for the info. I have verified this by using a 4.91 Mhz crystal, far enough out to show errors in frequency counting.
Pulling the crystal had the PIC instantly switch to it's internal RC source. THANKS!
Bob

[Archangel]

may bee hi deed not reed care foolee dee hole tred ;)

'The Easiest solution and use another PIC' reply always drive me nuts when the user don't need to and have a nice 'learning opportunitie' on hand... .

ICE e e e e gotit EWE SCAERT MEE . . thought I missed something in de datasheet, possible as it was about 4 am here when I wrote that.:) HE HE You don't miss very many MASTER_E, I humble grasshopper, wish, only to learn . . .