68SS98Z28

I have a 1998 Z with the A4. I just starting programming with HPTuners and have the SLP lid, smooth bellows and bottom breather performance pack.

I was looking at the stock VE table and there seems to be four zones if you look at the pressure lines: at 15 and 20, there is a dramatic double hump and the 20 line overlaps the 25 and 30 lines above it. pressure lines 25-55 have only the hint of a first hump and there is no overlap in any of the lines in this domain. Lines 60-85 differ in that there is an inversion or dip rather than the first hump. There is also no overlap in this region. Lines 90-105 return to the first hump but they don't rise as fast as the 60-85 group so there is overlap as they fall below the earlier group.

I downloaded a program from Horist's page of a 98z with the A4. The poster states that it is a hypertech program smoothed once. In this file, there is no overlap of any of the pressure lines. From 15-55, the double hump is presserved with no inversion. From 60-105 there is only a hint of the first hump, and there is no falling off at the highest pressures. This program essentially only has 1 zone in that there is no overlap of any lines in either dimension and the "form" of the lines is carried throughout the whole map, with only minor shape change.

Anyone have any thoughts on:
1. Why would the 20 line overlap the lines above on stock file?
2. Why does the 60-85 group on the stock program invert the first hump?
3. Why do the highest pressure lines on the stock program fall off at high RPMs?

The downloaded program seems to make more sense--anyone have any comment on the effects of this VE table on performance, or which program is more appropriate if programming for power/torque? How about with a lid vs. without?

jimmyblue

iTrader: (11)

You will run across ugly-to-the-eye data in the tune file
and in my opinion it's a desperate cry for help, like a Goth
chick with 7 face piercings.

The VE map is granular enough, and the motor a regular
enough machine, that the response surface for VE should
be smooth, continuous and well behaved (aside from any
discontinuous inputs, like the EGR opening, which mine
has none).

As a rule when I see something that looks like a zit, I
pop it. That is, smooth it in to whatever an eyeball
interpolation makes look "right". Not necessarily using
the smoothing functions for very local kookiness, though
I will do that just for esthetics at the end. I don't want
to change the "normal" points, just whack-the-mole.

As far as the high pressure lines falling off, -if- it were a
real effect I would expect it to relate to the poorer flow
of more dense air in the intale / head runners, and more
exhaust stack-up out the back. Though why it would be
strongly "nonlinear" with MAP, I can't imagine.

Steve Bryant

iTrader: (1)

I think that one of the keys to all tuning is the use of a wide band AFR. The most important thing in setting up the VE table (in my opinion) is to disable the mass air flow, closed loop and power enrichment temporarily and go to speed density mode. Then retard the timing in the high octane spark table by like 5 degrees across the board so you don't have to worry about knock retard issues while you are working with mixture. Then drive and log data (AFR/MAP/RPM) so that you can find out whether you are rich or lean regarding stoichiometric (14.68:1). You want to try to record this data in steady state conditions (only mild acceleration or deceleration at most). Once you've adjusted your VE values in the various cells until you can log in the range of about 14.5:1 - 14.9:1, you have your VE table pretty well squared away. You might want to apply a small amount of smoothing to the whole table.

You can work with the VE table and get it in fair shape using long term fuel trend data (LTFT). However it is not terribly accurate and takes longer. You cannot set up anything regarding power enrichment without a WBAFR. Also, in determining what to do about KR issues once you start working with that, you cannot know if you should retard the time unless you know where you stand on mixture.

Steve

68SS98Z28

Steve, your comment on the timing retard is interesting. I am logging in SD mode with the MAF disabled by setting the failure frequency to 0. I don't have wide band but do have time so i am trying to dial in the VE with LTFTs. So far that appears to be working. I am finding it difficult to hit many cells without some abrupt throttle changes so while i haven't done 100% throttle i have done about 75% and some rapid changes.

I also copied the high octane table to the low octane table as many here have suggested for the logging. I show NR at flow rates above 0.36. Some logging sessions i see average of 1-2 degrees others i see average NR up to 4 degrees. Maximum NR i see as high as 6-8 degrees.

So--i always burn premium, should i be tuning this way? Once i've tuned the VE tables should i keep the high octane table on the low octane table? Should i be concerned that i have some NR while loggin the LTFTs in SD mode? Should i be concerned if i see NR once i re-enable the MAF?

Thanks in advance for your thoughts.

Paul

Steve Bryant

iTrader: (1)

Paul,
Retard the timing 5 to 7 degrees to do this logging on the high octane table then make the low octane table match the high octane table. That way, the spark isn't retarded by the knock retard value. In fact, you shouldn't have hardly any KR while you are adjusting AFR if you retard the time like this. Also, your car will run like a dog, but that's OK. You don't need a lot of acceleration/torque anyway when you are just trying to focus on getting steady state AFR right for the various cells in the VE table. I think that it helps a lot to retard the timing because it takes care of your engine and it allows you to focus on one aspect of tuning at a time.

Once you have SD AFR under control, I would re-enable the MAF and then work on Spark Advance. You cannot correctly set up PE without a WBAFR. When you finish with spark advance, set up the cells of the low octane spark table to 5 or 6 degrees below the high octane table.

Start saving $25/week until you can buy a basic WBAFR with an analog output so you can log AFR. If you are going to learn how to tune and not jeopardize your engine, you need a wideband.

Steve

dewmanshu

iTrader: (4)

That's just funny there.


WB definitely makes the complicated things in tuning less complicated...and safer.

PurplePiss

I disagree with reducing the timing to VE tune. Timing will affect the burn, which will affect how much air is allowed to enter the cylinder. If you set it at 20 degrees and set your VE, it will be lean when you start running 26-30 degrees of timing (assuming the timing is not too far advanced that your losing power).

Also, what is everyone's opinion on setting the Stoich to 13 for VE tuning? I would think that this would also skew it when you go back to 14.7. I just leave it at 14.7 and leave PE on. You can filter it out by AFR or by TPS.

bshell

iTrader: (24)

PP - if you are using a wb to tune the VE / MAF tables, just set PE to where you want it and tune the VE table to 14.7. No need to use the filter function if you are tuning with the wb as you are using AFR error (doesn't matter what the commanded AFR is). I normally set my initial PE table a little on the conservative side (rich) until I get the higher kPa rows dialed in on the VE table and then bring PE up to my target. I also set total timing to something sane (25*-26* on my car) before starting on the VE table. Once I get close to finishing the VE table, I'll start upping the advance a little at a time.

Paul - 98s are prone to showing false knock. Make sure the KR is due to timing advance and not false before making huge cuts in your Low Octane table. Mixture will also play a part in producing a knock event if it is off enough.

I also tried tuning my car using fuel trims when I first bought hpt. Spent 5 months tweeking and re-tweeking the VE table using LTFT + STFT.

I upgraded to the EIO and bought an LC-1 in March and haven't looked back since. Tuning with a wb is so much easier and quicker, plus you can tune the ENTIRE VE table instead of just part throttle.

I am in Honolulu (Pearl Harbor area) also and can help you out if you need it.

PurplePiss

How do you start "bumping timing" Do you do it across the entire table? Do you just do .60-.80? How do you decide if you should be running 12.9 with 22* or 12.5 with 28* of timing.

I started this process and found that I could run decent timing in 1st and 2nd, but when it came back up through the rpm range in third I would get knock due to cyclinder heat. Then you have to penalize 1st and 2nd by reducing timing. We almost need a timing map per gear, and a PE table per gear. LOL.

bshell

iTrader: (24)

Timing / PE multiplier vs gear would be nice for the A4s, but M6s don't differentiate between each gear (its all "in gear") For now, I limit my WOT tuning to 3rd, 4th, and 5th (I have 4.10s). I used 5th for low rpm and 3rd for high rpm with 4th filling in the middle. I can get more cell counts that way. Although its been harder to do since our track closed for good!

I generally bump timing in the high load cells. For me, that's .68 on up to .80 g/cyl. The rest of my table is copied from the high octane table (I'm in full-time OLSD). You can use jimmyblue's method here: https://ls1tech.com/forums/pcm-diagnostics-tuning/521022-partial-throttle-tuning.html for part throttle timing (see post #8).

I also start my PE table with values from 12.4ish to 12.7 - 12.8ish (richer at peak torque, leaner at peak hp). After dialing VE in, I now run 12.6 to 13.1 with 26* to 28* of advance (26 around peak tq, up to 28 at peak hp).

SSpdDmon

iTrader: (10)

That was a fun one to edit....all your words Jimmy! You trying to tell us something?