Hi all,
Have created a map on V3 firmware 3.023.
I’ve transferred over all the information I could from my V2 map, including VE, injector data, fueling > general parameters.
Ignoring cold and warmup fueling, just looking at fueling once at operating temperature - the short term trim is having to add between 10 and 20% extra fuel to maintain lambda on target. I haven’t driven the car in load yet but at 3000rpm high idle the short term trim is adding pretty much the same as at idle (1000rpm).
This leads me to think that there is something scaling the fuel dose calculation.
My first thought was that I’d picked the wrong option here as it differs from V2. I assume V2 would have just acted as if it was in MAP REFERENCED. Regardless, with the fuel pressure sensor option selected or not, the effect seems negligible.
Second thought was the new ethanol correction table vs the fuel scale factor in V2. My fuel scale factor in v2 was 159.5%, so for the correction table in V3 I set E100 to 59.5 and interpolated back to 0 at E0. This reduced the short term trim slightly (was previously pinned at max 120%)
I notice there is a VE airflow correction, but I wouldn’t have thought the VE should need to change.
Has anyone experienced anything similar?
Happy to give any settings changes a go if that’ll help with the development.
I’ve attached a log on v3 idling to warm, the v3 map and also my v2 map.
I’ve checked both projects on the simulator and the pw on both is the same.
What I can see that there is no vvti set in V3 so the cam doesnt move (the same on the log).
Maybe that is the reason ?
Here is a log on my V2 map running from cold up to operating temp. idle to warm v2.zip (648.0 KB)
Once warm (but before the fan kicks on) the short term correction hovers around 103-104%. Much less correction than the ~115% on the V3 map. Same VE, same injector data, same fuel in the tank.
Didn’t get a chance to swap back over to V3 and play with the VVTi. Tomorrow job.
At about 11 minutes into the v2 log and 11:30 into the v3 log the RPM (approximately), MAP, IAT and CTC are all the same. There is a slight difference in injector PW. Approx 2.0ms for v2 and approx 2.1ms for v3. Not sure how significant that is…
EDIT: didn’t spot it before but the ignition timing is fairly different (approx. 6 deg) due to the idle ignition strategy, that might be part of the inj PW difference.
Have there been changes to the LSU4.9 wideband accuracy in V3? Do you think that some of the difference in short term trims could be attributed to that?
I can’t imagine it would be anything close to a 10% discrepancy though…
Waiting to get others thoughts on this as I’m not sure what else I could try short of adjusting the tune to suit - feel like that shouldn’t be required.
Honestly i think its VVT related. I was chasing a miss at idle after rebuilding my engine and was blaming it on the cams. Turned out i had to “zero” out the camshafts again after the rebuild. it was off by 4 degrees but had massive difference in idle performance and quality (in fact i had to adjust my ve table by 5-10%).
But i checked the maps and found the cam Offset is different so i assume you checked that already?
The WBO sensor support is written from scratch and has better precision than V2. There is option to calibrate WBO circuit. You need to disconnect LSU 4.9 connector and from menu Tools select Calibrate WBO sensor. Please write here what was the calibration constant. It increase the precision around lambda 1 (from 0.95 to 1.1). As we have lambda around 1 it can play the role. You can try compare results at lambda target 0.9
The PW difference is about 10% if we consider injectors opening time. It would be easier to compare the results under the load.
Yeah I was playing with the offset in V3 to make sure the cam position was 0 when no vvt was commanded (idle). Had to move it quite far as you can see in the screenshots.
Like Jadzwin says I think it’s going to be easier to work out what’s going on with a bit of load. Will take it for a drive tomorrow arvo.
Took the car for a decent drive tonight. Good news is the car drove really well considering how much of the map is a first guess. Pretty impressed with the closed loop boost control with default PID settings.
Not so good news is that it’s definitely running a fair bit leaner than V2 in all conditions but primarily in idle and cruise.
I grabbed this log off the EDL-1 and noticed also that the auto tune function gives incorrect proposed VE numbers. I haven’t tried it with a log saved directly onto the laptop rather than from the EDL.
I had noticed your post previous to this and went and tried the Autotune function and noticed the same results with the Laptop connected vs EDL1. The proposed VE numbers are all very low compared to the current VE Tables…
