A/C isn't turning off when the button is turned off. My battery was dead this morning which is very unusual, it hasn't been weak at all, just out of the blue dead; not 100% mind you but enough that lights were dim and it wouldn't even try to turn over. 20min on the charger and fired right up with no fuss. BUT then I noticed it was like the A/C was on, and trying to press the button to turn it on/off did nothing, the light didn't come on, wouldn't come on and wouldn't turn it off. Soooo got to thinking, well most ECU's that at this point in car design controlled the A/C compressor as well, tend to command the electric clutch on the A/C compressor to disengage at high rpm and throttle.... soooo I stood on the gas causing it to kick down a couple gears and went up to near redline, let off and... few min later no more freezing cold air coming out of the vents, which was after I pulled over and tried just turning the car off and starting it back up.

I had thought perhaps the low voltage on trying to start it initially caused something to freak out as electronics tend to hate low voltage. It was fine the rest of the way to work, started fine at the end of the day to go home, BUT again the A/C was on and it wasn't commanded to be, the light on the button did work now, and eventually the A/C did go back off. But I'm not sure what's going on. Only thought I have is to pull the negative battery terminal and wait 20min or so, re-attach and see if that does the trick if it continues. Is there a relay that could be failing?

 

As soon as you set the air to get on the windshield, A/C will automatically turn on. You can even hear the click when you move the control knob...

This is to help defogging the windshield.

 

Wasn’t on defrost/defog, was working fine today though.

 

The A/C request signal comes from a chain of switches (HVAC blower on, A/C button pressed OR one of the windshield outlet positions selected, A/C line pressure okay, compressor temperature okay) applying ground to pin A21 of the MICU (BCM). That request signal gets turned into a CAN packet, which gets routed through the gauge cluster (which acts as a communications hub) to the ECM/PCM. The ECM controls the radiator fans and compressor clutch through relays in the "auxiliary under-hood relay box" (box with 3 relays in it, front left). The ECM also watches the evaporator temperature (the bit that cools the cabin air) via a dedicated sensor and cycles the compressor off and on to prevent ice formation.

 

Problem seems to no longer be present, so I’m assuming it was caused by the low voltage attempt to start. You mentioned the gauge cluster is a CAN hub of sorts so, I’m guessing that might have been the point of origin for the problem. Been working normally the last couple days though so I’m just chalking it up to low voltage electronic anger.

 

You could check for DTCs.

Checking ground connections wouldn't hurt.