I need help to get my closed loop to run as nice as my open loop!

[schpenxel]

It always makes me go huh when people know my name on here, lol

[Paz]

I've been playing with filters and reading up, I'm trying to filter out decel and also about a 350ms delay when AFR drops from normal into PE in my AFR error graph.

What i have come up with so far (which seems to work in my mind) is- [6310]-15 or [50121.slope(350)]=0, so [FTC]-decel or (AFR Cmd.slope(350ms)=false, does it sound like it should work, from my scan it is removing all decel and transients from what i can tell?

It always makes me go huh when people know my name on here, lol

Please take it as a complement mate, guys such as yourself, Ed, Greg, etc just to name a few are the saving grace around here to help us learners

[07GTS]

if u use that filter and also look at your scan have it so u can view the graph and chart v time, then set the graph to only show viewed data then go to a place u have decel or a change in commanded and just move the cursor over the spot and see if the graph still populates or not then u can see what its removing or not and u can then adjust the delay time ect to suit

[Paz]

Seems to be working GTS, tested it as you suggested and also when i transfer data into my VE now i don't seem to get the huge spikes as i was previously, it's flowing a lot better.

[07GTS]

sounds good, howd the lowering the min PW go any better..?

[Paz]

Am seeing it go down below 0.7 a fair bit since dropping it, I'm fairly lean in decel now but could also be me adjusting the VE correctly since i wasn't filtering it out before

[07GTS]

yea ive spent along time tuning mine and testing the effects of alot of the tables and u can get yourself in a circle quite often, but its all good learning and if something odd happens u can be quicker to find the area of issue and adjust to help ect

[Paz]

So how close does one worry to get their VE table (as in % error)?

I've got most areas now from about -2% to about +2.3% in AFR error, is it worth trying to narrow it down even more?

[07GTS]

that should be close enough u will find the last few percent change throughout the year so u will only chase your tail, personally i try to have only negative error -2 -3 so i know if anything its rich not lean but thats just me

[schpenxel]

+/- 2% is plenty close, IMO anyways

[Paz]

Awesome, I can now start on trying to get my IAT offset range settings sorted out.

I do have a few places higher up in the rpm/boost ranges that still need fine tuning but will wait till i can get it on a dyno for that.

Is there a good read somewhere for setting up my power/boost enrich commanded AFRs?

I get how it works and have got it coming into line pretty much where i want it but what commanded AFR should i be shooting for at what PSI etc?

Currently it is set up safe, I've used settings that Greg (Ghuggins) has recommended on the forum here but not sure if it is actually too rich, or is this something that needs to be done on a dyno to show actual gains (like spark)?

[07GTS]

sometimes u can feel the difference when changing AFR on the street and even EOIT i could feel a difference especially when i got it wrong, what boost u running and just on 98 fuel ?

[Paz]

At the moment i'm at about 8psi max GTS, and yep 98 (don't have E85 available out here).

[07GTS]

think i was about 12-12.2 afr on 98 with about 16 deg timing up top and bout 10psi back then, ooh u would love E85 once u go E u wont go back i just get it in barrels at the moment so its the same every time also thinking bout going flex just encase

[Jason W]

Have you tried changing your End Of Injection Timing settings? I ended up maxing out the Boundary to 8, then have the normal and makeup tables around 7.9 and 7.5 respectively in the 80-105 degree ECT fields.
I struggled for so long to get idle right and had to run in open loop at around 13AFR just to get enough fuel into the cylinders after the overlap sucked most of it out! After making this change it's a completely different car. Mine always seemed to be ok in open loop when cold but then when switching to closed loop then it started to run really bad as the O2 sensors couldn't read accurately due to the raw fuel in the exhaust being my best guess...
Oh, I am pretty sure my cam is 112LSA for reference.

[Paz]

Yep Jason I've adjusted my EOI settings, I've got a rather large cam (well for a boosted application lol) and have set it to hopefully a pretty close figure (near as my skill level will allow anyway).

My biggest issue (i think) is the single turbo and the O2 sensor location (both in same pipe) trying to run in CL, I will try again soon once i have my OL worked out (sorting a lean issue now due to IATs and Inj tip temp tables).