Boost VE Tuning Tips

[ferocity02]

What tips do you guys have for tuning the boost VE table? I've found it difficult to get much data in the boost VE table. After eliminating cells with less than 20 frames and applying filters, I usually only have a few cells left of usable data in the boost VE table.

I've been copying the 105 kPa column from the vacuum VE table into the boost VE table, and then sloping it 15-20% up to 285 kPa. But my most recent boost VE values are lower than those in the 105 kPa column. So my method is probably not the best.

Any tips?

[cyipher]

Ive been running into the same thing with hitting cells, With the help of picnic_george we have been basicly hand forming the boost ve and main ve table based upon the low cell hits. On my setup I transision into boost so fast that its hard to hit cells more then once unless in full boost. I think without a loading dyno this is the only way.

[andysc3]

I use 10 frames of data per cell. I get more hits and its still fairly accurate. Then smooth the table as needed

[ferocity02]

I use 10 frames of data per cell. I get more hits and its still fairly accurate. Then smooth the table as needed

How important is smoothing the table? Obviously removing huge dips or spikes is important. But if the data you're getting is causing a bumpy VE table, then why change it? That's what the data says it should be. If you smooth it out then fueling will be off.

Will bumpiness cause the driveability to be noticeably worse?

[andysc3]

I have mine smoothed out at I don't have any drivability issues. And afr is still within +/- .1 from my commanded afr

[joecar]

What tips do you guys have for tuning the boost VE table? I've found it difficult to get much data in the boost VE table. After eliminating cells with less than 20 frames and applying filters, I usually only have a few cells left of usable data in the boost VE table.

I've been copying the 105 kPa column from the vacuum VE table into the boost VE table, and then sloping it 15-20% up to 285 kPa. But my most recent boost VE values are lower than those in the 105 kPa column. So my method is probably not the best.

Any tips?Post log and tune files...

What wideband...?

FlashScan V1 or V2...?

[ferocity02]

Here's a log, can't post the tune at this time since I'm working with my tuner on it. AEM wideband and V1.

During tuning, should I have the IAT vs. VE table be all 1's, or does it not matter? At 6psi my IAT's don't really go up at all. They are pretty steady at ~10F above ambient.

[ferocity02]

I keep having an issue where one log will tell me to add or remove ~10% in some cells, so I'll correct the VE table, and then the next log will tell me the opposite. This is mostly an issue in decel and idle. Other areas are +/- 2% between logs.

For example, this morning it told me to knock 12% off the 3200 RPM 20 kpA cell, so I did, and then this afternoon it told me to add 8% back... The difference in average IAT between the two logs was about 12F.

[andysc3]

What injectors are you using?

[minytrker]

The more data you can get the better you can build your ve table. I personally use alot of filters and use different cell counts depending on what Im doing. I filter out anything less than 50 cell count on some parts of the ve. As far as smoothing a ve, the smoother the better obviously but the the way the pcm interpolates everything it smooths out big spikes anyways. Usually with enough data and if you dont have another table off that is fighting you the ve will end up smooth on its own. If you keep going rich/lean, lean/rich usually another table or modifier is off.

[ferocity02]

What injectors are you using?

FIC 72# at 4bar


The more data you can get the better you can build your ve table. I personally use alot of filters and use different cell counts depending on what Im doing. I filter out anything less than 50 cell count on some parts of the ve. As far as smoothing a ve, the smoother the better obviously but the the way the pcm interpolates everything it smooths out big spikes anyways. Usually with enough data and if you dont have another table off that is fighting you the ve will end up smooth on its own. If you keep going rich/lean, lean/rich usually another table or modifier is off.

Should I make the VE vs. IAT multiplier table A0014 all 1's for initial tuning? Right now it's a constant slope line increasing with IAT. But I feel like 12F difference in IAT isn't enough to throw off a few cells this much between logs. I'm doing OLSD as well, so the computer isn't correcting anywhere. The decel and idle cells I hit quite a bit, usually 200+ for decel and 2000+ for idle.

I think the issue I am having was how I'm extrapolating for the 400RPM row and 15kPa column. Since I idle at 700RPM the 400 row will have an effect. What I've been doing is copying the 800 row to the 400 row and subtracting ~7% so it slopes down nicely. What I think I should do for now on is leave the 400 row alone so the corrections aren't over-correcting. Same story for the 15kPa column which I can't hit unless I'm above 3200RPM.

[joecar]

A0008 is easier to deal with (you have more control over it) than A0014.

Start out with A0014 set to all 1's and get the VE table corrected initially.

[ferocity02]

A0008 is easier to deal with (you have more control over it) than A0014.

Start out with A0014 set to all 1's and get the VE table corrected initially.

I have A0008 setup already to richen it up at ECT under 104F, but it's 1.0 above 104F so it's not affecting the data.

So I think one issue I was having is that I log in the morning on the way to work, and in the afternoon on the way home. The difference in ambient temps between the two is 10-15F, which causes a similar difference in IAT's. So if I make A0014 all 1's then there is no compensation for changes in IAT's between logs. So my morning log will be rich and my afternoon log will be lean. Also there is also a good amount of variation in IAT in each log itself depending on if I'm in traffic or if I'm on the highway, in boost, or vacuum, etc.

So I set A0014 to all 1's today and logged the change in BEN's with IAT at idle and found that the curve is pretty darn close to what I had in there, ans also close to other peoples' curves. I made the reference temp (VE multiplier of 1.0) 86F since that is closest to my normal IAT in my area, adjusted my VE tables accordingly, and will log again tomorrow. I'll leave the A0014 curve in there so it can account for the change in IAT's while tuning. I know this isn't ideal, but I think it's the best I can do for tuning on the street. Anyone see any major issues with leaving A0014 active?

I also think I need to start using a percentage of the correction factor rather than the entire thing so I'm not overshooting.

[joecar]

You have the answer.

Post a pic of your A0014.

[ferocity02]

Here it is