COS 3 Lean Cruise

[vbhero]

First up, details of car etc

2002 VX SS Commodore A4 with 3.73 diff gears
215 / 223 .604 .610 112LSA Cam
COS3 (01290003) in OLSD.

Being in Aus, we are lucky that we can legit use the Lean Cruise function, so given that it works well in stockers, I have decided to go ahead and use it rather than play with B3647.
I copied the B3639 values from a stock HSV tune, choosing one that has 3.73 diff gears standard so it would be as close as possible to my cruise values. But modified the enablers.

Initially it seemed to work, but then watching the WBO2, I noted that the values seemed off via my dodgy mental EQ to Lambda conversion. When I reviewed the log, yep I was right, the commanded EQR was off when compared to what the B3639 indicates it should be. I for the life of me, cannot see why that is the case, it seems to work then not.
Any ideas?
When you see the log, you'll see another oddity I need to chase in my install, WBO2 #2 has random spikes. I usually filter that data out for obvious reasons, its odd as it only shows when logging, I never see these spikes on my DLG-1 display.

Thanks
Pete

20151118 COS3 Current.ctz
20151117 Lean Cruise Fault.efi

[5.7ute]

One thing that happens is that lean cruise can enable while enriched. For instance, if you enrich the higher load areas in B3647 and these cells are being referenced when LC enables. This commanded EQ ratio will be counted as the base for all the LC calculations. This will stay active in this point until LC deactivates either by speed or PE coming in.
The only work around I had for this issue with the COS is to either control the enrichment and lean cruise with B3647. Or let PE enable easier to prevent lean cruise enabling under load.
I do not know if this is what you are seeing, as I don't have Efilive on this computer to check your files. But it is something I found a few years ago when playing with lean cruise on mine.

[vbhero]

One thing that happens is that lean cruise can enable while enriched. For instance, if you enrich the higher load areas in B3647 and these cells are being referenced when LC enables. This commanded EQ ratio will be counted as the base for all the LC calculations. This will stay active in this point until LC deactivates either by speed or PE coming in.
The only work around I had for this issue with the COS is to either control the enrichment and lean cruise with B3647. Or let PE enable easier to prevent lean cruise enabling under load.
I do not know if this is what you are seeing, as I don't have Efilive on this computer to check your files. But it is something I found a few years ago when playing with lean cruise on mine.

Thanks mate. I have read that elsewhere and you just reminded me to check. But my B3647 is set to stoich in the applicable cells so rules that out.
I have before this used B3647 with reliable lean out at the appropriate times. Just wasn't a fan of transient lean out under part throttle when poking around town.
That plus, LC is there as a function so I thought why not do it that way. But now I am wondering if there is a bug in the OS, or is there something wrong with mine.

[joecar]

Is B4001 IFR table correct...?

[vbhero]

Is B4001 IFR table correct...?

I am running stock OEM Injectors and haven't touched that table. I have just compared it to my stock tune and it is the same

[joecar]

My understanding is that Holden implemented a returnless fuel system, much like GM implemented on the Carmaro/Firebird/Corvette...

in this system, the FPR is located in the fuel pump/sender module inside the fuel tank, and is referenced to barometric pressure (i.e. un-referenced wrt to manifold);

the FPR maintains constant pressure at all times while the fuel pump is running (regardless of any variable-speed mechanism on the fuel pump, if that is present);

this means that rail pressure will be held constant throughout the engine's operating range, and this pressure is applied at the top of each injector...

constant rail pressure requires the PCM to account for manifold pressure/vacuum at the bottom of each injector...

this is done by sloping the IFR table using the squareroot pressure relationship, which looks like an-almost straight line sloping up as vacuum increases.


If you look at your IFR table, it is curvy... this implies that at various places along the curve the IFR values in the IFR table will be incorrect.


Lean cruise, by virtue of dropping out of CL, requires that the IFR table be correct in order to command the exact fueling ratio specified in table B3639 (and any other fueling table that runs in OL, i.e. where there is no CL to correct fueling to stoich).

[joecar]

Yes, the upshot of all this is that Holden implemented the IFR table incorrectly.

[vbhero]

My problem isn't a commanded vs actual. It's a commanded is wrong if that makes sense
If you look at the log and compare it with B3639, it's not commanding the correct EQ at times. When you look towards the end where I am doing 100km/h, the commanded EQ does not match what the tune is set to.
For some of it, it's not even going into lean cruise yet the tables tell me I should be at ~0.9355 EQ. You can then see where I slowed to shake it up a bit then after I accelerated it eventually leans to 0.96 commanded but according to B3639 it should have been commanding ~0.85 EQ. B3647 is commanding 1.00 lambda so all should be sweet.
I cannot see any other table that should have been affecting it.
I have been watching it more over the past few days, and at times it seems spot on, others its not. I cannot pinpoint a common condition that dictates when it isn't working.
Hoping the brains trust is able to help

[joecar]

Ah, ok, I see (sorry, I missed the point)... hmmm, sounds like some other tables are getting involved, but we don't see them.

[vbhero]

Ah, ok, I see (sorry, I missed the point)... hmmm, sounds like some other tables are getting involved, but we don't see them.

Well it's reassuring to know that its not just me that cannot see them hahaha. Nothing worse than not seeing the cause of a problem, posting it up only to have someone point out the obvious lol

I have just changed the LC settings, including the enablers etc to 100% stock to see if there is some weird interaction happening as a result. The enablers I have set are quite aggressive, so maybe that has something to do with it.
It's curious that from what I have searched, no one else seems to have reported similar symptoms. I wonder if that is because not many are using LC or if there is something wrong with my flash of COS3.