EGO correction sometimes stucks at 100.

EGO correction sometimes stucks at 100.#41638

By Yoosyn - Mon Mar 23, 2020 12:01 am

- Mon Mar 23, 2020 12:01 am #41638

As someone started topic on facebook.
Sometimes during DFCO when ego correction shuts off at 100 it refuses to continue work when conditions are met again. It is very rare. Sometimes it happens once per 100km, sometimes once per 20km. Other time I can cross whole country. It seems it can be triggered by DFCO or WOT (EGO active below TPS threshold). When I see it happen and want corrections back i press reset button

I can provide any data. Here is one log and tune. I can find more logs

2020-02-21_18-16-16_ego.msl

(372.26 KiB) Downloaded 245 times

CurrentTune.msq

(56.62 KiB) Downloaded 245 times

note: I know speeduino is not designed for dailydriving and EGO correction is not most important feature. Maybe it is small bug and it will be easy to fix. I would be very happy

. Im using my car as daily and sometimes I'm doing longer trips across europe with it. It would be nice to have dependable EGO correction.

Re: EGO correction sometimes stucks at 100.#41645

By Ursidae - Mon Mar 23, 2020 6:53 am

- Mon Mar 23, 2020 6:53 am #41645

I posted the issue on facebook. At work atm but I'll provide a log and my tune later in this thread.

Re: EGO correction sometimes stucks at 100.#41655

By Ursidae - Mon Mar 23, 2020 7:54 pm

- Mon Mar 23, 2020 7:54 pm #41655

I might've found a mechanism for this to occur. Check corrections.ino rows 463-467. First there's a check of ego condition and then the ignition event no is checked.

This may happen from what I can see:
Ignition event no is an uint16 with max value 65535. Assume that the next update point is close to max value and when this point approaches egocorr is disabled by any of the conditions. Then the ignition count may cycle over to 0 but the update point is still a high number and egocorr will be disabled until ignition count reaches the update point again.

Re: EGO correction sometimes stucks at 100.#41656

By Ursidae - Mon Mar 23, 2020 8:01 pm

- Mon Mar 23, 2020 8:01 pm #41656

As far as I understand Ignitioncount will cycle over after about 16min for a 4 cyl engine at 2000rpm average. 65535/2 ign per rev/2000rpm

Re: EGO correction sometimes stucks at 100.#41678

By Yoosyn - Tue Mar 24, 2020 9:46 pm

- Tue Mar 24, 2020 9:46 pm #41678

Ursidae wrote: ↑Mon Mar 23, 2020 7:54 pm I might've found a mechanism for this to occur. Check corrections.ino rows 463-467. First there's a check of ego condition and then the ignition event no is checked.

This may happen from what I can see:
Ignition event no is an uint16 with max value 65535. Assume that the next update point is close to max value and when this point approaches egocorr is disabled by any of the conditions. Then the ignition count may cycle over to 0 but the update point is still a high number and egocorr will be disabled until ignition count reaches the update point again.

This is on par with impression I have! When going back home from my friend after 15km there is city speed limit and some downhill where DFCO kicks in for a longer time. I often had a issue with correction stuck after this part of road