I managed to get a pretty good spreadsheet built up for fixing VE tables. I am running an 8746 $61 ECM and this has worked well for me. I got the idea from Craig's VEPhD and kinda adapted to the lower-resolution VE tables for the 8746. I am pretty sure it would work for 7747 stuff too. I will upload it to Craig's file sharing site for you guys to check out. I'll post a Word doc for instructions too. It's pretty much cut and paste, but there is a bit of observation to it because you have to look for zeroes in the correction table.
It'll be up in an hour or so. I am still working on the Jeep engine.

 

I just uploaded the file to Craig's site. The name of the file is VEFixer. It is super easy to use and has really saved me lots of time in the tuning of my Jeep engine. I didn't put any documentation up, but I will in a few minutes.
EDIT: I put the documentation up too. It is in a text file, not a Word doc.

 

Cool spread sheet, but I'm not exactly understanding how the INT is being used here. Doesn't it just point to what direction the blm is moving? I have one made up for the 747 that just uses the blm data, but for some reason it always seems to overshoot the correction, so I had to "install" a gain **** to adjust how much it corrects. Work like a charm now, two passes will pretty lock you in +- 2 blm's.

 

That sounds pretty cool, BMonte. Care to upload a copy so I can compare the two? Mine uses the INT just about the same way as VEPhD does. I think the convention is that if INT<128, the BLM is going down...if INT>128, the BLM is going up. Essentially, the only difference between mine and VEPhD is the difference in the table resolutions. It still has a few quirks too like filling in NEW VE table cells with zeroes if the BLM or INT in a corresponding cell is a zero. It also likes the number 0.00006 for some reason that I have yet to figure out. But it has worked and cut a LOT of time out of my tuning. I wish there was one for timing tables to tune timing VS knock counts. If anyone knows of any spreadsheets like that, please let me know. I'll probably have the Jeep on the road this weekend for some extended highway runs and I'll probably be needing the capability to quickly tune my spark tables because I have Reserve duty this weekend.
Seriously, upload a copy of your spreadsheet to Craig's site so I can compare them. Mine takes a few tries to lock you in, but it works. My idle is rock steady with BLMs 124 to 128. I'll email a copy of the logs and the BLM/INT tables if anyone is interested.

 

I have used the ALDLTOBIN program for a while.
It does the converting, rewrites the table and then writes you a new bin file when it's done.
You can select the wide or narrow table to use for error correction.

 

Where do you find this ALDLTOBIN program? Will it run on Win ME?

 

it should run on ME, it is dos based, i beleive Jon wrote it perl.

i don't know where a place to download it is, if you give me your e-mail address, i can send it out to you

 

download the program here

ALDLTOBIN

i posted this software to my public server. I found this version of the software doing a simple search on yahoo.com for 'aldltobin'

 

I think Dewey is right. I think Jon wrote it.
I put the program on a 3.5 disk. then after I log, I copy the log file file to the 3.5 and change the name to just blm.txt.
Then I put a copy of the bin I am working with on the disk and rename it something short like 1.bin.
So when you are ready to go you just open the dos window, change to the a: drive and type.

aldltobin 1.bin blm.txt 1227747 W

when you hit return it will load the program and spit out a new bin file called
new_1.bin
from there you can rename it any thing you want.

BTW
I normally just keep my working bins on the 3.5 so I can go back and forth the the house PC.
Then I back up all bins on there.

HTH

Thanks again for the program Jon!!!

 

Ive been using the aldl to bin program for a while now too, its pretty sweet

 

Is that different from the standard Win ALDL program?

 

we all know what WinALDL does. . .datalogging.

WinALDL can output logs of the BLM & INT tables which we then give to ALDLTOBIN which outputs a new binary image based on these values and specific computations.

The what data to use is either the Narrow average 10 or Wide average 10. N or W. It's up to you which you want to use. If you've got lots of Narrow data then use that obviously, else use the Wide.

The ecm can be either the 8746 or 7747 (common ecm's used with winaldl) but all similar ecm's are supported if you know the starting location of the main VE table. Just type in the binary location (ex. 4C2)

 

I have been using the Wide Average 10 BLM and INT tables and it seems to be working really well. I may give this ALDLTOBIN thing a try tonight and see how I like it.
FYI...if you didn't realize this, you can open the text files that WINALDL creates with MS Excel. That is exactly how my VE Fixer spreadsheet works. Then it's all cut and paste from there.

 

Along the same lines you can use excel's graphing capability to graph any number of variables. . . i.e. Coolant Temp vs IAC counts vs whatever and come up with a ton of different ways to visualize the data.

 

I normally use the wide (W) for about the first 2 log runs. By then they are real close and I start working on timing. When I make a nice long run and have plenty of narrow data, then I use narrow (N).

I also watch the number of hits each cell takes. Just to run the rpms up and down to hit a cell once won't help to much with the fine tuning. I spend a lot of time getting to see all the neighbors projects while I'm running around logging data.

If you have real time tuning it's even better!

 

Are you referring to the AutoPROM or the PROMinator? I happen to have the AutoPROM and it is a wonderful piece of equipment. Real-time tuning is da shiznit!

 

I'm still old school. I have the romulator.
Only J@@P owners can afford that new stuff

 

That's because we don't have to spend as much $$ to keep our J@@ps running.

 

I'll try to upload it tonight. I had the same problem as you with the 0's so I ended up using a logic function to return 1 for any o'rs it found. Kinda took a while to figure out. My big problem was that the order of the data was backwards(rpm and map wise) from what your VE curves are so I had to manually enter every formula for each cell. It works quite well so hopefully some of you can give it a try.

 

Yeah that was a program I wrote a while ago. That was like my first programming project I had ever written for a PC. It does a pretty good job considering the slow stock baud rate of the c3 ecms.
One thing that I noticed you can do is treat the INT as a smaller influence. I believe Craig's VEPhDsoftware does; New VE = Old VE * (BLM*INT)/(128^2) . I'm not sure if this would overshoot more or less because I don't have time to think about it . Might give that a try and see where it gets you.

 

That's exactly what I did. I used the BL and INT together and adapted the whole thing to the TBI tables. I tried to adapt Craig's sheet, but I couldn't get the tables to function.
I have an easy fix for the backwards tables from WinALDL. I can just go in and adjust the formulas to account for the reversed data goving the appearance that it actually can read the backwards table and generate the correction factors properly. It will still work, but it will look a little funny.
BMonteSS, what logic function did you use? Care to elaborate a little?

 

Heres what I used in the "Correction Factor" Table:

=((IF(B3=0,128,B3))*((IF(B15=0,128,B15))))/128^2

I also had to manually adjust the "New VE1%" Table manually to account for the MAP & RPM being on different axis' in WinALDL and TunerPro RT.

Hope this helps.
R

 

I've got an Excel spreadsheet with brief instructions and easy to follow but i was hoping someone could help it so it doesnt "Overshoot" the correction like BMmonteSS stated. I've kind of been having the same problem any help or obvious formulas would be greatly appreciated.

 

I'm using JPrevost's New VE = Old VE * (BLM*INT)/(128^2) formula from Jeepguy553's VEFixer spreadsheet with a logic function to treat the 0's as 128.
I also added a sheet to drop .5* of spark from each cell that has 5 or more knock counts in it. I've just started tuning so if the pro's see anything weird in my formulas please let me know. I'm attaching the modified VEFixer that I'm using.

 

Thanks WS6, I've been kind of overshooting it because i've just been using this formula I thought up. New VE = Old VE * (BLM/128) Might get it zeroed in within +,- 2 BLM's now with only a few passes.

 

Fwiw here's mine, it's very similar to WS6's. Basically I took the original and cleaned it up a bit, rearranged the axes to match Tunerpro, and added a VE1+VE2 summary to plot the combined VE table. Just be sure to update the VE2 table on the "VE Tables" sheet (VE1 is filled automatically using the New VE1 table).

NOTE: I use one macro to reset the ALDL tables (Excel will alert you of this when you open it). Sometimes I don't copy the entire BLM/INT tables depending on how many samples I got in each of the cells, so the macro resets both tables to zeros to clear the previous data set.

Finally be sure to use Edit/Paste Special/Values when pasting your BLM/INT data or you'll lose the lean/rich color formatting (if you care).

Enjoy!

 

Wow! How did you get the background and Rich/Lean formatting in there?
I'll have to take a deeper look at this over the weekend. Would you mind if I *Borrow* some of your formatting ideas?

R

 

Do whatever you want with it

Background is set in Format/Sheet/Background. Rich/lean colors are set using conditional formats (Format/Conditional Formatting).

 

I'm not understanding something; how do you keep from "overshooting"? I've been using new VE= old VE * (BLM/128) and I seem to be chasing 128, so I guess I'm overshooting?

I don't understand how to avoid doing this. As you can tell, I'm a noob.

I just downloaded the excel files, and I'm really diggin them. Is the correction factor a given? Meaning, does it ever change, or is this just the way it always is?

 

Have you tried one of the files that also use the INT data?
They use this formula: New VE = Old VE * (BLM*INT)/(128^2)
The correction factor will change with each set of new data you use to calculate the new VE tables.

I'm trying to find formulas to help tune spark once I've got my BLMs all straightened out. Something like add .5* or 1* to the spark table then data log to compare the old VE table to the new and keep the spark if VE went up and put it back if VE goes down, stays the same, or no data was gathered for that cell and drop .5* if there's knock.

R

 

Sounds like you have your work cut out for you. I haven't even attempted the SA table yet.

I realized after I posted that the correction factor would change itself after entering new values in the tables. I'm kind of new at the whole excel formula thing too; what is the ^2 for?

 

^2 means Squared This site might be helpful.
http://phoenix.phys.clemson.edu/tuto...rithmetic.html

Another common operator that you will use quite often during your physics laboratory course is the exponential operator. This is used, for instance, to square a number or take the cube root of another one. You use the caret (^) to perform exponential operations in Excel. Also note that the use of parentheses is often imperative when using the exponential operator. Below are some examples using the this common operator.

Mathematical
Expression Excel
Expression Result
3^2=9
3^(-2)=0.11111
8^(1/3)=2
100^(3/2)=1000

 

Oh, I get it now. That make's perfect sense. I understand why it would be a more accurate formula than the one I was using. Thanks.

 

Did anyone here realise that the correction factors are given by WinALDL in the BLM and INT.txt files? After fiddling with all of these formulas the correction factor was already there if you scrolled down.

Now I read somewhere that BLM was weighted as 5 times more important than INT. I'm thinking ((BLM Correction*5) + INT Correction)/6 for the correction factor to multiply against the VE1% cells. I'm trying to find the posts where I saw someone mention that 1 BLM = 5 INT. I'm hoping that this will get us a more accurate correction as the INT will bias the Correction factor but not as much as the BLM does. I'll link to the posts when I find them.

R

 

The correction factors are there of course but sometimes the correction factors given by WINALDL are completely off. There has to be a way to put in assumed averages and get a correction factor off of those numbers. If the correction factors WINALDL spat out were reliable, wouldn't you think we would see and use them? LOL don't answer that, stupider and more obvious things have been overlooked. But anyways, it's nice to be able to manually edit some of the values if they are way off or there is not enough information there for you.

 

I've been screwing with a table of my own, so please correct me if I'm wrong. The total VE is VE1+VE2, right? None of these work by any calculations that figure VE2 into the equation. Here's what I mean, but forget about the fact that INT is not in the equation and there's a bunch of equal signs:

BLM/128= X*current VE=New VE
115/128=.8984*25.5=22.01

What I'm thinking:
BLM/128=X*(current VE+VE2)=(VE-VE2)=New VE
115/128=.8984*(25.5+39.06)=(58.00-39.06)=18.94

See the difference? I'm looking at my data logs and it seems that this might get me closer to 128, but I have yet to prove it.




I have a question that's been rattling around. If the total VE is VE1+VE2, and VE2 is 39.06 any place there's a cell in VE1, when the total of the two exceeds 100%, are the injectors static?

I noticed there are some places that this is occuring for me. What should the total VE be less than? 90%?

 

The total VE should be less then 85% for good measure. Theres a post started by BMmonteSS that covers this quite well. If you're getting into the 100% VE then you need to up fuel pressure and lower your BPC so your VE doesnt exceed 85%. WIthout the efficiency exceeding this your car can add extra fuel due to WOT, and everything else. An injector is only static if it's on 100% of the time. Go to the TBI board and the sticky entitled Come in for a Free Tune will tell you all you need to know.

 

I have a question that's been rattling around. If the total VE is VE1+VE2, and VE2 is 39.06 any place there's a cell in VE1, when the total of the two exceeds 100%, are the injectors static?

The injectors may or may not be static with 100% VE. The injector duty cycle (DC) is not the same as the VE%. To find out if the injectors are static need to check the injector PW against the available time. The available time being RPM and # cylinders dependent.

You are correct that both VE tables should be taken into account for the VE update. To prevent overshoot a reduction factor can be used. IOW, if the BLM / 128 is 0.8984, then reduce that by half before applying it to the VE%.

RBob.