now simulating front O2 signal using WB signal
Thread Starter
EvoM Guru
iTrader: (50)
Joined: Mar 2006
Posts: 9,675
Likes: 132
From: Tri-Cities, WA // Portland, OR
I'm one of those idiots who did/does run closed loop under WOT conditions. Actually if you build it correctly its a safety feature. I set up my MAFT Pro to target a certain AFR above a certian TPS and load value. I then set window limits (-10% and +10% for instance) so that the MAFT Pro can only correct so far. It has saved my *** in the past when I had a fuel supply issue in that it added fuel when I was loosing fuel pressure. If somebody is paranoid of fuel being taken out then just make your limits 0% and +10% so that it only adds fuel so as to save the engine.
The bad part is that you would need to set up some sort of gain feature so that the correction isn't hunting and/or overshooting/undershooting. Also an old laggy wideband sensor would cause issues in that it wouldn't be able to correct enough to provide good closed loop operation. Some controllers don't even though an error when a wideband sensor is to slow. The LC1 lets me know when its had it but you can see it in the laggy sensor response for a few months before it throws a code.
It would just be a good idea to be able to enable or disable such a feature. I typically only run closed loop AFR when I'm going for glory passes in the fall. The AFR drifts a few tenths from pass to pass and gear to gear. Closed loop really helps you get every last hp out of the car. Yes I tune my car that tight.
I think that for the moment, I am going to just differentiate between idle and cruise by having a crossover rpm and an AFR setting for below and for above that crossover value. Much less code to write and will execute more quickly. If someone can come up with a good reason to actually use a closed loop table, I'll give it more consideration.
I can think of a few different reasons to vary closed loop AFR with rpm (hell even afr vs load). Many cars with cams would idle much better if you could run them richer than stoich. So 1250rpms and below you'd be running 13.8:1 AFR. You would then slowly blend the closed loop AFR to be say 15.5:1 AFR at 2000 or even 2500 rpm's. If you quickly go from 13.8:1 to 15.5:1 you will feel a sharp decline in power and or stutter.
I agree with Dan L. Its like the equivalent of older late 70's and early 80's cars running lean burn carbs that transitioned very quickly to a very lean mix off of a fairly rich idle and transition circuit. They tended to stumble or fall flat on their face with no power. You need to transition carefully. Maybe limit timing to about 15-20 degrees in that area too?
His idea about limits to get around a failing sensor may be a very valid one. But what happens when it fails ? Also do Zeitronix ones fail in a similar manner? Never had one but Im sure many guys/gals are using one on this forum.
mr. fred I do think this idea is really really great. There are many issues and limits that need thinking about to implement it properly as VW or BMW have found out. IIRC Innovate is trying to get car manufacturers interested in doing a WB control of the AFR for complete closed loop control. You might ask Klaus Almendinger if he has insight gained from his experience. He may even be able to help out. It would be a working application ported to a stock type OBD2 system which may be something he needs to show to the manufacturers. Maybe even qualify for assistance from the govt as a gas saving emmissions reducing device. :-) Worth 2 tries !!!
Milburn
His idea about limits to get around a failing sensor may be a very valid one. But what happens when it fails ? Also do Zeitronix ones fail in a similar manner? Never had one but Im sure many guys/gals are using one on this forum.
mr. fred I do think this idea is really really great. There are many issues and limits that need thinking about to implement it properly as VW or BMW have found out. IIRC Innovate is trying to get car manufacturers interested in doing a WB control of the AFR for complete closed loop control. You might ask Klaus Almendinger if he has insight gained from his experience. He may even be able to help out. It would be a working application ported to a stock type OBD2 system which may be something he needs to show to the manufacturers. Maybe even qualify for assistance from the govt as a gas saving emmissions reducing device. :-) Worth 2 tries !!!
Milburn
A failing WB sensor is a valid concern. However what stops people from tuning open loop maps on a bad sensor right now? I'm on my third sensor in my DSM now due to leaded fuel usage. As long as you keep an eye on it (like everybody should now from time to time on their open loop tunes) its not an issue.
Thread Starter
EvoM Guru
iTrader: (50)
Joined: Mar 2006
Posts: 9,675
Likes: 132
From: Tri-Cities, WA // Portland, OR
Another thought. (Only somewhat related, but not really worth starting a new thread for.) I wonder what it would take to rescale the trims to provide a wider range? Think automatically accommodation of E-85, or other mixtures, via trim adjustments. That, plus table-driven AFR targets, and you've got a very good implementation of flex-fuel. 
(Spoken as a recent E-85 convert, who is getting a little tired of trying to make sure he can track down a station any time he takes a road trip.
)

(Spoken as a recent E-85 convert, who is getting a little tired of trying to make sure he can track down a station any time he takes a road trip.
)
Another thought. (Only somewhat related, but not really worth starting a new thread for.) I wonder what it would take to rescale the trims to provide a wider range? Think automatically accommodation of E-85, or other mixtures, via trim adjustments. That, plus table-driven AFR targets, and you've got a very good implementation of flex-fuel. 
(Spoken as a recent E-85 convert, who is getting a little tired of trying to make sure he can track down a station any time he takes a road trip.
)

(Spoken as a recent E-85 convert, who is getting a little tired of trying to make sure he can track down a station any time he takes a road trip.
)






