As you and Jim suggested, I have done a test with even simpler software. I have tested with this code:

#include <89LP214.h>
#include <absacc.h>
#include <stdio.h>

#define	OUTP1	P1_6
#define	OUTP2	P1_4
#define	OUTP3	P1_1
#define	OUTP4	P1_0


void main(void)
{
TMOD = 0x11;	//Setup timers - 2 * Mode 1
CLKREG = 0xF0;	//Divider timerclock med 16
P1M0 = 0x00;	//P1 i output mode
P1M1 = 0xff;	//P1 i output mode

while(1)
	{
	TR1 = 1;
	//EA = 1;

	PCON = 1;	//Start idle mode (go to sleep)
	}
}

 

The should do nothing except start the timer, and enter idle mode. Using this code, I get 2.53mA with the ceramic resonator, and 2.41mA with the internal clock. With no �µC installed, the consumption (voltage regulator) is 0.21mA.

The 8MHz/internal clock consumption matches the datasheet just fine, but does the crystal oscillator really "steal" this much power?

Mikkel