10bits adc using 16f877

[sixty9sandals]

i don't know what you expect, but it display something, it's just reverted. Let's say when the value=0, all i/o=1... but it's working.

anyways, here's a shorter method

        DEFINE ADC_BITS 10
        DEFINE ADC_CLOCK 3
        DEFINE ADC_SAMPLEUS 50
        
        MEASURED_VALUE VAR word
        TRISA = %00000001
        TRISB = %00000000
        TRISD.0 = %0
        TRISD.1 = %0
        ADCON0 = %11000101
        ADCON1 = %10000000


START:
        ADCIN 0 , MEASURED_VALUE
        PORTB= MEASURED_VALUE.LOWBYTE
        PORTD.0= MEASURED_VALUE.8
        PORTD.1= MEASURED_VALUE.9
        GOTO START

hi..
i've try out ur code..when i supplied 1.5v as an input then the leds displayed 0100010010 instead of 100110011...why is this happen? any suggestion to solve it?

thanks a lot

[mister_e]

274 instead of 304... it's not as this far. Everything is related on acurracy of the VREF. IF it's REALLY 5.000000000V then you should have
(1.5/5)*1024=307

If my maths are good after 20 hours awake, ((307-274)/1024)*100=3.22 % of error. so your supply line should be around... 5-(3.22%)=4.9844 Volt

If i use a serial communication here and if i inject 1.499 volt, VDD=5.014, i should read... (1.499/5.016)*1024=306 right? i have 305 wich is real darn close.

If your ADC input is a little bit noisy, it may change the results, same thing with the VDD line. If you respect the minimum acquisition time and the maximum input impedance... you will have the nice results.

If you need accuracy, you may decide to use a precision voltage source.

mmm, maybe 1023 instead of 1024....

[sixty9sandals]

thanks a lot for ur help even after 20hours awake..its realy help me..possibly its caused by the noise..im gonna check it again..thanks a lot

[sixty9sandals]

i've check it..but the regulator input is 4.8V..is there posibbly coz by my wrong ADCON1 register?..

thank you

[mister_e]

Regulator INput or OUTput?

if it's output, the results make sense. use the previous calc i did.

If it's INput... and if it's a simple 7805... the output signal will be pretty bad

[Sean_Goddard]

If you have a noise problem, try setting the settling time on the ADC to a few microseconds, then initiating a conversion. At the same time, make the PIC SLEEP for around the same length of time as the conversion. This should SIGNIFICANTLY reduce the noise across the PIC silicon substrate and give you a lot better accuracy.