Regarding the AC switch: I believe subarus with older ECUTek use the AC switch and/or the defrost switch for resetting trims or other input functions. So one of those may be a likely option for the evo. I'll see if I can do some logging later today.

Would be fine with me, but I don't use the boost drop on my E85 maps . I'm not sure what people's typical setup is for this... do you just use it as information (IE oh, its not just a flashy light I really am hitting bad knock) or do you use it as some sort of control (IE: Oh crap, lots of detonation... let me cool down the cylinder temps with some fuel and low boost)?

 

 

On gasolene dropping boost really helps alleviate knock. On E85 its pretty darn hard to knock so thats where the AFR safety comes in handy, to drop boost if you go lean.

 

well i just think 1 timer would be better.

so you can still enable/disable boost drop on HIGH knock, but there is 1 timer for both flashing and boost drop...

maybe just a little less confusing?

 

For boost drop on knock I have it set to 500ms (half a second). The reason I have it set to 500 ms is because the knock sum in and of itself is based off a decay timer. When knock drops below my threshold it is because knock has subsided for quite some time already and the timer is just finally decaying to below my threshold level. Knock is something that can "just happen" for a short period of time and the very next pull it may be fine.

For AFR safety I have my timer set for 10,000ms or 10 seconds. The reason for this choice is because if I have an AFR safety trigger it is due to a fueling problem. IE: my fuel filter is clogged, my fuel pump o-ring has torn or is leaking, I blew a FPR line, etc... All of these things are a serious event and won't correct themselves. Thus the triggering of an AFR safety event means that I, as a driver, REALLY want to know and a 10 second timer no only protects me from each successive gear I may pull but it gives me a 10 second warning that something is very very wrong.

 

yup.

but the afr safety boost drop timer will STILL be independant of the knock boost drop timer.

i was talking about merging the "knock boost drop timer" and the "high knock cel fast flash timer"

 

Oh, thats fine then if it saves you programming headache.

 

Ok, AC Stuff... I came, I logged, I made a post. Let me know if you want me to post any logs.

I checked the bit-wise AC relay/switch/power steering and they correspond in some way but were flipping up and down constantly. Either the code is implemented wrong in evoscan or they are not correctly defined. (I'm guessing #1, they did seem semi-distinct, but probably not interpreted correctly).

As far as MUT 9A it is definitely directly related. Some things I noticed:

*first and most importantly for our purposes so far* On putting the car to the 'ON' position without starting it will register 8 for 13-16 seconds then flip to 0, pushing the AC on button does nothing to change this while up or down.

With the car running and no AC it registers 192. It will stay at 192 if you turn the cool switch all the way to Max AC (on USDM Evos this is still manual and should engage the AC according the manual, but doesn't seem to do so (either via timing/load changes or according the MUT 9A).

Pushing the switch immediately changes the value to 193 while the car is idling, and pushing it again to turn off immediately changes back to 192.

With low load (~50 load) acceleration all the way up to 7k and no AC on the value flips from 192 to 128 at exactly 3000 RPM and then flips back to 192 when you let off.

AC on and low load acceleration (~70 ->50 load) results in a flip from 193 to 129 at 3k, then 129 ->128 just after 4500 RPM (probably happened at 4500RPM but the logger was not fast enough to catch it). It then went to 192 when I let off at 7k (assuming it has some decay where it keeps the compressor off).

With high load acceleration it flipped from 193 to 129 as soon as I got on it (wasn't logging TPS but load did not rise until 1 or 2 samples later) and 2875 RPM. After letoff it flipped back to 193.
----These are a little confusing I would think this one and the last one would be backwards, but maybe it does not decay to 128 until you have been accelerating for a certain amount of time.

It didn't seem to flip in 1st gear on normal light accelleration with the AC off, but 2nd flipped at 2875 again from 192->128 with really light 1st through 3rd accelleration / normal driving.


One thing I forgot to test was setting the air location to defrost which should also engage the AC. I also forgot to test what it registered at highway cruising speeds.

------

In the end, it does not appear that checking MUT 9A would be valuable for switching, but that does not mean the ECU does not know what this switch is doing....

 

so bit0 of MUT9A is when the A/C ECU wants the compressor turned on right?

what if you:

LOG MUT9A
start car
turn on A/C
adjust controlls so it will cycle on/off a lot - ie low fan speed, medium temp
keep watching 9A.
wait for compressor to turn off and then see what 9A is.

I think you will find that when the compressor is ON 9A bit0 is ON, otherwise 9A bit0 is OFF..

 

Remember the compressor turns on/off as it sees fit determined by the air temperature in your HVAC. You'll see 1 or 2 air temp sensors in your airbox if you tear it apart, so it can be completely normal to have AC on but the compressor not pumping. I thing "max ac" means that the compressor is on all the time though..........

 

yes that's precisely what I have been trying to say

HVAC ask's for the compressor to turn on, ECU has override control.

therefore the ECU never really sees the A/C switch ON/OFF status ...

 

OK I have merged the timers as per discussion before....

Should have this ready for testing today

 

I don't use the AC very much at all (pretty much only when my girlfriend finally slaps me) but I don't ever notice it turning off by itself when I have it on (load changes, etc). I would think that would be really hard on the AC clutch. It might divert the air somewhere but that seems dumb too. If you can give me something more specific that is likely to turn it off I can test it (IE what control combination?). If you think setting the temp to very warm and running the AC for a few will do the trick I can try that, but I would think if there are sensors in the airbox that would work backwards. I'm guessing there is a "sweet spot".

 

Ah junk. Well, do we know if the defrost button is wired in? It didn't seem to flip 9A at all, but I checked with my friend and he confirmed that was a ECU integrated function on is WRX with ECUTek.

 

nah i dont think the demister is either

regarding A/C on/off, the compressor doesn't stay on all the time. if you just let it idle you can hear it tuning off/on every few mins... once it reaches its "desired" temp in the cabin... its climate control yeh