My BLM's are pretty much on the LEAN side with values of 145 + acorss the board with the exception of some that are at 130 or relatively close to 128.

I think the reason for runing lean might be due to the headers.

Today I took the car for a drive and took 30mins of data logs. I sorted out the data RPM,KPA,BLM and had a great nice big list of values.

I started adjusting the whole VE table and after about 4 hrs I pretty much had all the cells complete except for KPA's above 70 and high RPM ranges. RPM ranges 400-600 I pretty much left alone cause I never get into those anyway.

So I have all my new VE values ready to get burned and am keen to see how it runs. I used Traxions method of calculating the new VE taking rpms within 25 either way of the desired rpm 1175-1225 for the 1200RPM Cells.

Reading through Traxions article again I notice he states:

looking at my modified VE values there are parts where there is not the increase like Traxion says.

My VE value at 1300 @ 55kPa = 83.4
&
My VE value at 1300 @ 60kPa = 73.78

I did take averages of BLM's around those readings and this is what I got for my new values.

This is one example of this happeing, I have a lot more places where going higher in the kPa and same RPM has lower VE values then at lower kPa and same RPM.

According to Traxion this should not be happening.

So what is the verdict on this? Do they have to have this trend or is it not that important??

Thanks,
Michael

PS. I have my new VE values all on spread sheet in comparison to the Stock values using the spreadsheet program that someone here created. If someone would not mind just having a look what I am talking about I could email it.

 

Hey Michael,

As a general rule of thumb, the VE of an engine will definitely increase until the motor hits peak torque. Peak Torque usually always matches the peak VE.

One possible reason that you are experiencing this decrease is that you are on the edge of a BLM cell (hysteresis of 100rpms) and could be operating in 2 different Block Learn Cells.

Tim

 

What do you recommend me doing then Tim?

Think I should just program as it is now and see how it runs and then adjust again or lower the numbers so they do increase like they should?

 

I always go by 'Whatever the car asks for - give it'. If the car is asking for a curve like you have now ... and you don't have anything suggesting that you should really strive towards the norm ... then leave it that way If the car is happy like that then leave it like that. Once you get everything nailed down then drive around for a long time to allow the computer to really learn those settings ... then rescan and see if you see anything different.

Also check your timing in those areas to make sure there are no huge jumps between those map readings at that rpm.

Tim

 

Thanks Tim for the replies you have answered the question pretty well and I will give it a go and then rescan and make adjustments accordingly.

Michael

 

Just thought I'd share my experiences tweaking the VE tables on my more or less stock 91 L98..

My first several burns were only with specific VE cell corrections based on my ALDL data. This seemed to work pretty well in terms of getting the BLM's to around 128, but drivability problems started to creep in (odd hesitations and surges) and I also had odd gaps in the VE table like you describe (a higher kPa value having a lower VE for the same RPM range).

Then one day I decided to take a different approach. I looked at the ALDL data only in spots where I have a VERY good reading of a specific VE cell (ie, in a specific rpm/map range for at least several seconds, enough for the blm/int to stabilize to a specific value) and used these as "known good" readings, and updated by VE tables accordingly. I then visually "smoothed" the values inbetween these "known good" readings, and simply throwing out values that didn't make sense (such as the aforementioned higher kpa/lower VE spots). Burned chip, drove car. wow! drivability problems went away. And of course I logged a bunch of new data.

now that the VE tables were "smooth", there were no huge jumps between VE cells and the ALDL data was much more usable than before (instead of BLM's constantly fluctuating). I still had corrections to make, but this time around i didn't need to smooth the results nearly as much. I did this a few times iteratively and am now very happy with the results. Towards the end, I really didn't have to smooth anything at all, because the VE table was so close that i wasn't getting any weird results from the aldl data stream anymore. The car was now going through the VE table smoothly and transitioning between cells smoothly instead of with big jumps in BLM.

my theory on why it solved the drivabvility problems is as follows: there are lots of VE cells that the car never spends any real time in, so you simply can't tune them by monitoring the ALDL data stream, as you might not even get a data frame while the car is in that VE cell. but the VE cell still affects the fuel curve for a moment, and VE cells that are way off could still affect drivability (too much or too little fuel for a brief moment) yet never be in that cell long enough to report it in the data stream. My biggest problem was an acceleration stumble - i have a 2400 stall converter, so if i'm pulling away from a stop a little quickly I can go straight from 600rpm to 2000rpm in a split second. That's a lot of VE cells to cover! When I smoothed everything out, the stumble/hesitation completely went away, drivability overal improved and the BLM's were easier to monitor and correct for.

 

They sure as hell makes sens 91! I have also taken a simialr approach like that since my VE values were all over the place.

I did it like you said and after one burn I was in the area of 118 - 135 blms all around. After my second burn I am around 122-133 and slowly getting there

I think getting 128 blm's constantly is impossible. Too many factors get in the way.

 

Yea, this is exactly the approach that I suggested in my VE tuning post (sticky post at the top) ...

"After adjusting VE points in the lower table with this new information I use TunerCat’s graphing functionality and view how the lower VE curve looks. The points that have been adjusted will be obvious. I use these points as a guide for adjusting the other points in the curve to achieve a nice uniform even flowing curve. This may or may not be correct. Later readings from Diacom will show how ‘off’ my guestimates are. "

Tim

 

Tim,

Since the Main Fuel Table on a 7747 doesn't exceed 3200rpm, and based on dyno runs my peak TQ is at 3800, then I should see my VE% continuously increase and be highest at 3200/100kPa. Correct?
Its interesting that you can review the MAP voltage values and see that happening. MAP voltage is usually highest at or around the TQ peak and then it'll start to tail off. I like to see how quickly it tails off to indicate if the HP has increased or not. The slower the MAP voltage tails off, the further up the rpm scale HP peak has moved.
I've also experienced some "funky"parts of the Main Fuel Map. There seems to be a "knee" at the 1600-1800rpm level which is just before the motor starts to come "on cam" at 2000rpm.