A/D code on a PIC

complete

Am I reading this right? You set the go/done bit (adcon0,1) then test to see if it is low. How could it ever be low if you just set it in the previous line? Don't you have to set it after the conversion is read? Bob

complete

Maybe "goto $-2" should be "goto $-1"

complete

Am I reading this right? You set the go/done bit (adcon0,1) then test to see if it is low. How could it ever be low if you just set it in the previous line? Don't you have to preset it up where you start conversion right after the ADC is started, third line in ADCREAD before the delays? Bob

Actually, no. The '2455 is word-addressable. $-2 takes you back to the prev instruction. That threw me too when I went from PIC16's to 18's. :)

My reading of the datasheet is that the micro sets the go/done bit low when it is finished, although the text is a bit ambiguous. From the datasheet:

"When the A/D conversion is complete, the result is loaded into the ADRESH:ADRESL register pair, the GO/DONE bit (ADCON0 register) is cleared..."

One thing I should mention is that my program does run, rather than locking up in the loop we're discussing. Obviously it's getting set low at some point, but I'm not seeing any useful readings.

I *so* wish they'd provided an example of A/D code in the datasheet. :(

No appnotes?

see

previous

Yes, but you have to set the go/done bit high to start the process that's what "go"means. You can test the bit as you are doing to see if the process is complete but you should not reset it again until you want the next conversion. You certainly shouldn't set it before you test it even though there is a loop there. BTW if you have implemented a loop to test the bit, you don't need the delays because the test loop will create the necessary delay.

There are plenty of app ntes on the microchip site. You appear to be loading ADCON0 with 0x00 in your read routine which will turn the ad converter off.

You should take a look in the family reference manual for the part in question. It has sample code for devices such as the A/D.

BAJ

They have source for the 18C452, which does not appear to have an ADCON2 register, and uses register bits which are unused on the '2455. But they do use interrupts in a way I'll have to investigate further. :)

Is that not appropriate? I set the ADON bit on the next line. Should ADON be set once and left that way? These are the types of things I'm still not clear on.

I see that now, you labeled it start conversion, typical aquisition is around 13uS you seem to only have 5uS. I cant see what your doing with the result.

I'll remember that. I haven't yet graduated from the 8-bit 16's