ddnspider

iTrader: (26)

I've been doing ALOT of reading and learning as I want to start tuning. I've got the Greg Banish DVD's, book, been on efilive.com, etc. From what I gather, on a rather modified car, you would do the following:

a) obviously get car idling properly, etc.
b) perform VE table updates as required under steady load/throttle/etc, setting a/f to ~14.7 throughout the entire table and correct variance in VE cells between commanded and actual a/f if using a wideband.
c) perform MAF updates as required under steady load/throttle/etc, setting a/f to ~14.7 throughout the entire table and correct variance in MAF Freq. cells between commanded and actual a/f if using a wideband.
d) adjust PE to desired a/f under WOT. You MAF/VE tuning was mainly for idle/part throttle/driveability, your desired WOT a/f is done in the PE table.
e) adjust spark as necessary to gain power and avoid knock

I'm mainly wanting to confirm d)

joecar

Hi Bro,

A few comments:

- VE and MAF are used thru-out the operating range; the PCM uses them interchangeably under various conditions to compute the cylinder airmass;

- only after the PCM has computed the cylinder airmass will it use the desired AFR to compute fuelmass required to produce that AFR (and not just at WOT)... you see where I'm going here...?

- you might want to roughly get idle going, then do VE/MAF, and then re-visit idle;

- while tuning you want to make sure your AFR never goes lean while under load;

- stock spark is usually ok; to adjust spark safely it is best to rent dyno time;

- do you have EFIlive...? If you do, and you're boosted, you might want to run an EFILive COS since this provides boost VE/spark tables; how much boost and how big is your engine...? If you're able to exceed the PCM's MAF limits, then you can do various things like run MAF-less or scale the IFR.

- if do you have EFILive, do you have V1 or V2, and which wideband do you have...?

- we have a method that will get both your VE and MAF tables correct in a single log session;

- with different alcohol/gasoline composition, you might want to do your tuning effort using EQR or Lambda instead of AFR (both in commanded fuel and wideband measurement);

- there's more stuff for you to read (and ask about) on the EFILive forum (such as estimation of dynamic air temperature).

ddnspider

iTrader: (26)

Thanks for the response Joe. Here's where it gets sticky for me though:

98 car thats boosted. Without it getting ugly and scaling Map/VE/spark, what are the options? From what I've come across in Banish's book, what makes the most sense in this particular case is to perform AutoVE/AutoMAF with lambda=1 for basically the entire VE table and most of the MAF range. Then use PE to actually set my desired a/f based off RPM/MAP since as soon as I'm nearing boost, the VE table will jump straight to 105kPa.

My particular car is a 98 M6 with a 383 and a T76 turbo kit. Not idea for getting my feet wet in tuning, but I'm enjoying learning. Using EFI Live V1.

Here's my other question. For ANY car, if you're following the AutoVE/AutoMAF tutorials with the car set in open loop, you are usually doing low rpm 1st and monitoring the % error between commanded lambda=1 and actual lambda. But what about for high load/rpm, do you do the same % error based on what you want your actual lambda to be. For example, if I want my actual lambda to be .9 for high rpm/load, do I perform the same monitoring/tweaks to the VE/MAF tables until lambda=.9? If so, then what is the point of PE? I guess I'm getting hung up on the fact of being able to reach my desired lambda via the VE/MAF tables OR PE.

joecar

AutoVE/AutoMAF: you set the open loop fuel air table (OLFA) to the desired Lambda (i.e. Lambda = 1.0 for part throttle; Lambda = 0.87 for WOT; Lambda = 0.90 for almost WOT), and you can leave PE disabled during tuning; you log like this and then apply the wideband correction (BEN) to VE/MAF; for example, PCM commands Lambda 0.87, wideband reports Lambda 0.9, the correction (called BEN) to VE/MAF is 0.9/0.87 (i.e. measured was 0.9 which is leaner than desired 0.87, so VE/MAF must be increased by the ratio of Lambda's).

VE and MAF set airmass;

OLFA and PE set desired Lambda;

OLFA is based on MAP kPa and RPM;

PE is based on RPM and is triggered by TP (throttle position);

You can run purely from OLFA, but PE gives you some control of Lambda via TP (MAP usually lags TP by a small amount of time).

You would do AutoVE/AutoMAF with a mix of the following driving:
- light throttle,
- part throttle,
- WOT,
- WOT with some load (e.g going up a hill, dragging the brakes for a short time).

[ Oh, you have a 1998 LS1, sorry there are no EFILive COS's for this, people usually swap to a 0411 PCM if they want COS ]

joecar

Do you know what your peak MAF grams/second are (are you maxing out your PCM's MAF upper airflow limit)...?

Various people re-pin their 1998 to accept a 2002 PCM since this allows various other OS's to be run, notably these other OS's have COS upgrades;

without a COS, it is hard to tune for boost since the PCM pegs at the 105 kPa column; the COS's have a boost VE table that extends up from 105 kPa.

IF you're using fuel that contains alcohol, then you want to use EQR (equivalence ratio) and/or Lambda... I can setup your EFILive calc pids for this with your V1.

ddnspider

iTrader: (26)

Ah, okay, maybe it's starting to come together for me now. Correct any mistakes:

The commanded A/F is set in the OLFA table. Technically its F/A, unless you change units to AFR. You can slowly richen A/F in this table to account for idle/cruise, then part throttle, then WOT in the 95-100kPa range. This is strictly MAP vs ECT.

The PE modifier based on RPM is used for WOT exclusively, these values would likely coincide with the values in the OLFA table in the 95-100kPa range, but this is used to set A/F based on RPM at WOT.

So you set what you want the A/F to be in the OLFA table, disable PE, and run the AutoVE/MAF. Here is where you would make the map showing %error of commanded (i.e. values in the OLFA table) versus actual (via LC1 wideband for example). Since the OLFA table is only A/F versus ECT, then this map of MAP vs. RPM you would see the %error of the A/F you commanded for a given MAP regardless of RPM. In other words, your OLFA value of say 13.1 at 75kPa would try to command that for all RPM's in the map that was created. Any error would then mean there is a different amount of airflow at that value so you would change the coinciding cell in the VE or MAF table.

ddnspider

iTrader: (26)

Im running straight 93 octane. My tuner told me Im maxing out my MAF. This is why I assumed that I would focus on the OLFA in the 100kPa and PE vs RPM to set my WOT A/F since I'm maxing the MAF and the VE table will go straight to the 105kPa column anytime the car is in boost. I'm still learning, just trying to understand the process. I realize I'm limited on the 98 PCM.

ddnspider

iTrader: (26)

Joe,

Looks like you may have answered my question in a thread I dug up on EFILive

joecar

Yes, you got it.

The %error would really take the form of a correction multiplier:

correction multiplier = commanded equivratio * wideband lambda

[ noting that commanded equivratio = 1/(commanded lambda) ]