comparator AC1 problem with ATtiny1616

[wolfgang6444]

I am having serious problems using comparator AC1 on attiny1616.
Basic-Idea:
use AC1.AINP2(PB0) as positive input and AC1.AINN0(PA5) as negative input and observe the comparator output on AC1.OUT(PB3).

Here is the extraction of my code:
void setup() {
VREF.CTRLA=VREF_ADC0REFSEL_4V34_gc | VREF_DAC0REFSEL_4V34_gc; //references doe ADC0 and DAC0
VREF.CTRLC=VREF_DAC1REFSEL_2V5_gc;
VREF.CTRLB|=VREF_DAC1REFEN_bm | VREF_ADC0REFEN_bm; //reference for AC1
AC1.MUXCTRLA=AC_MUXPOS_PIN2_gc | AC_MUXNEG_PIN0_gc;
AC1.CTRLA=AC_OUTEN_bp | AC_INTMODE_NEGEDGE_gc | AC_HYSMODE_50mV_gc |AC_ENABLE_bm;
AC1.INTCTRL=AC_ENABLE_bm; //enable interrupt for AC1
PORTB.DIR &= ~0x3; //disable digital output-driver on PB3
PORTB.DIRSET &= ~0x3;

PB3(RXD) is constantly low.
Any idea on what I am missing? Do I need to configure PB3 to see the output?

[hmeijdam]

Hard to tell since you don't show complete code. I see you enabling an interrupt, but there is no ISR.

[wolfgang6444]

Well the entire project is a bit more involved: This is a loco-controller for some maerklin MM2-system. It includes PWM motor-control with BEMF measurement. In the long run I want to use the comparator to get an early indication for power drop in order to potentially save data to NVM before I brown-out. So I want to use the interrupt - not the output. That's why the interrupt is configured - the output is for debugging only. In the long run I also want to use the reference rather than AIN0.
Maybe I should create a simpler testcase first ...
Anyway, I have attached the code as is. Maybe something is obvious to somebody ...
loco_decoder_MM2.txt

[hmeijdam]

Posting a minimal sketch that compiles and demonstrates your problem is always a smart idea. That way people can load that easily in their development environment and play/experiment with your findings.

I quickly looked in your sketch and I don't see you clearing the interrupt flag in your ISR(18), so I would not be surprised that ISR keeps firing. You do clear the flag in the TCB0 ISR.

[wolfgang6444]

O.K.
attached is about as simple as I can make it. No interrupt, no Vref, no nothing ...
I am moving PA6 and PA7 and I am expecting to see Pa5 to move => Nothing
I am missing something fundamental here ...

comp_test.txt

[jvasileff]

Your code has:

AC0.CTRLA |= AC_OUTEN_bp | AC_ENABLE_bm;

but I believe it should be:

AC0.CTRLA |= AC_OUTEN_bm | AC_ENABLE_bm;

(Note _bp vs _bm)

FYI, you'll have better luck having people look at your code snippets if you post them as part of your comment rather than as an attachment, and also be sure to format the code correctly. To learn how, see the link "Markdown is supported" under the text area, and then search "Quoting code".

[wolfgang6444]

Great, that does the trick! I did not really care for the difference between _bm, _bp and _gc, but it becomes obvious if you look at what the directive expands to.
Many thanks!