I've got my spark and fuel tuning very close for part throttle and cruise at many different combinations, now I've locked my BLM's to 128 to begin finer tuning of the fuel using the Integrator.
My question is, say the integrator is 140, how should I use that number to derive the new VE table value for that load/rpm.
Would I still use the 140/128 = 1.09 * (VE value) method to come up with the new value in the table?? This seems like it would adjust way too much since the BLM numbers aren't that far off of 128 to begin with.
Also, the Int value seems to flucuate very quickly, is there a "rule of thumb" as to how long the Int value is constant before using it or should I get a wide band O2 to control it before I go any further?

Nope, the math for the BL correction is different then from the Int.

I perfer to leave the BL active and tune that way.

A WB is only going to help you in areas that are non C/L. In C/L the ecm is designed to toggle over 14.7, to average 14.7.

14.7 isn't a targeted AFR.

So what AFR should we be shooting for? I was told 12.5/1 for a supercharged car and about 13.5/1 for a N/A car. Also, what value do you use for the int value? I know 128 is the value for the corse fuel correction, what is the "magic number" for the int value?

Well for some reason mine likes to stay in cell#4 alot, I've seen other posts where others do the same thing and there doesn't seem to be a solution, so if I can tune my INT only and get all the rpm/load combination pretty close to 128, then it should not be a factor. My fuel cell trim does move around alot and I've adjusted the boundaries so that I use more of them more often, but at almost any rpm and load, it will sometimes be in #4 and I'm not decellerating ???
*
Anyway, my other thought for the formula for the correction factor using the INT then would be:
1) (Current INT value - 128) / 5 (because very 5 changes in int corelates to 1 change in BLM.)
2) (Result from #1 + 128) / 128 (the correction factor)
3) (Result from #2) * (currect value in VE table from rpm/load)
4) Replace value in the table with result from #3.
*
Would this be right?

Not really.

After MUCH testing and experimentation I have found this rather simple formula to work reasonably well for calculating the New VE value using the INT (and BLM locked).

New_VE = Old_VE x (1+((INT - 128)x.0034))

This is on my SD TPI System. You may find that you have to vary the ".0034" factor, but you'll probably find that the factor will be somewhere in the .0030 to .0039 range.

This works reasonably well provided you are no more than "20" from 128. (i.e. INT > 108 and < 148). If the INT is further away from 128 than 20, then you will probably have to do 2 calculations.

Once you find yourself "126 to 130" on the INT in most places, just "unlock" the BLM, reset the computer, and you'll see your BLMs close to 128 all over with only minor differences to the INT.

Edit: Formula changed to correct error.

Thanks a million Glenn.
Do you have any rule of thumb of how long the INT value must be steady before you use it for a calculation since it does swing very quickly alot of the times not staying on the same value for very long?

That's the "transitional" where the ECM is changing from one cell to another. The best thing is to get a LOT of data and then use the average INT value for a particular RPM & MAP range.

I would "sort" on an RPM range of +/-25 RPM at the lower VE table and +/- 75 RPM at the Extended VE table. I would use MAP value +/- 1 for lower MAP(load) readings and +/- 3 for higher MAP (load) readings.

The more data, the less affect the "anomolies" have.

By sorting on MAT temp, that was how I noticed my temperature affected my INT values and I had to recalibrate the MAT to get consistent readings over all operating temperature ranges.

Someone asked about the "math" so I figure I would do an example for those not following the formula.

Assume that you have a current VE value of 66 at 2,000 RPM @ 60 Kpa and you are getting an INT value of 138 (with the BLMs locked). This is how the forumla would work.

New_VE = 66 x (1+((138-128)x.0034)) =

New_VE = 66 x (1+ (10 x .0034)) =

New_VE = 66 x (1+.034) =

New_VE = 66 x 1.034

New_VE = 68.2

So you'd change the VE value at 2,000 RPM 60 Kpa from 66 to 68.2

Edit: Calc and example changed to reflect error found in calc.

Wait a minute, maybe I don't understand something about the INT value, but an INT value below 128 indicates a slight rich condition and the fuel should be cut right? But with the formula above it would actually increase the VE value giving it more fuel
In your example with the 118 value, the VE increased from 66 to 68.2 thus giving more fuel and would drive the INT down.
*
Maybe you meant (INT - 128) instead of (128 - INT) ??

Yes. It should be:

New_VE = Old+VE x (1 + ((INT - 128) x .0034))

Sorry, but I did all these formulas a long time ago and I'm just going off memory.

That's cool, thanks, my wife thinks I'm losing my mind anyway doing this eprom stuff and driving around all the time with my laptop hooked up I was beginning tho think maybe she's right But only after 10 chips now, I've got this thing of a truck running pretty darn good since a few weeks ago I knew nothing about it!
Thanks for all of your time and hardwork on this board Glenn, it is much appreciated!

uhhhh??? could you show an example? I'm a bit confused

sure....
Here's an example straight from my table:
With the BLM cells locked at 128 via the eprom min/max BLM cell boundary parms.
I have an INT reading of 140 at 1100rpms and 50Kpa MAP reading.
The current VE value in the eprom for that 1100rpm/50KPa combination is 50.8
*
The 140 INT value tells me it needs just a touch more fuel.
Using the formula above:
New VE = Current VE * (1+((INT-128) * 0.0034))
New VE = 50.8 * (1+((140 - 128) * 0.0034)
New VE = 50.8 * (1+(12 * 0.0034))
New VE = 50.8 * (1+ 0.0408)
New VE = 50.8 * 1.0408
New VE = 52.9
*
So replace the 50.8 value with 52.9 at the 1100rpm/50KPA location in the VE table.
*
Again, lock the BLMs at 128 and use this method only after you get the BLMs pretty close to 128 all the time.

Thank You! I was not sure exactly to be looking for, now I know! Thank you both!:hail:

Glenn,

I have found that 1 click on Tunercat takes my INT up or down by 4. Anyone else have any experience? And is there anyone to get TC to increment by less than .4?

If you want to increment a value by less than .4, double click on the cell and it will pull up an entry box that you can enter any value you want, but you will notice that after saving the file and reopening it, no matter how the value was changed, either by using the "increment" function or keying it in, it may be slightly different because when it is stored in hex base 16, the base 10 value will get rounded up or down usually by +/-.1 and the TC display rounds to the nearest .1.

2Quick,

Yes I have tried that but TC seems to want to make sure the change is always no less than .4VE. I think you can fake it out by multiplying and then change the VE number. It seems to lose track after 1 change.

Dominic, I can't say I've noticed a pattern either way, sometimes I use the increment (+) sign or decrement and it will still change the value after saving and returning, this is with the $8D ECM definition. You have to remember that each 1/10th value must be converted to the nearest 1/16th value for the hex equivalent, but even sometimes whole numbers will not store that way either, so there may be something else going on as well.

I discussed this a long time ago. You have 255 possible combinations in Machine Code to represent 0-100 VE. 100/255 = 0.4 (more or less). Thus the smallest increment you can change a VE value is 0.4. Occasionally it's 0.3, because 100/255 is not EXACTLY 0.4, but that is the "step increment".

That is why sometimes you will find yourself averaging "127" in a particular RPM/Load range and the next "step increment" will cause you to average "129" at the same RPM/Load. That's when you have it "close enough".

PS: I have found that if you have to chose between 127 and 129, go with the 127. I have found engines run a bitt smoother when they run a "dash" richer. I was running "128/128" perfect at idle, but it wasn't as smooth as when I was running "124-126". Just some practical observations.

Glenn,

Thanks for the response. And yes I agree with the comment regarding an "extra pinch of salt" to the mix. My experience has been the same.

Ahh, that's it because its stored in 8 bits, the way I was thinking would have been at least 24 bits, duh, what a way to conserve space, thanks for clearing that up Glenn.

Remember, the Motorola processor used in these ECMs are all 8 bit processors (which some minor 16 arithimetic instructions). It is "early 1980s technology" more or less.

I don't think GM started to use 16 Bit processors until the LT1s with the 8051. RBob and Grumpy would be able to tell you when GM first started to use 16 bit processors.

I get changing the INT table, but I've got a question for everyone. How exactly do you go about "locking" the BLM table. Is it just a matter of correcting and correcting until it finally rolls everything up to 128? It seems as if no matter how much I correct my VE table off of BLM corrections that it constantly crawls off again. I've disabled EGR functions, but as I drive the car around more and more the BLM's seem to go off simply later than before, but they are still going off...

Any help is appreciated,
Mike

If you are using Tunercat, just go to the ECM Constants Table and set the min/max for BLM's at 128. If you are using a Hex Editor, set 002B9 and 002BA at hex of 80.

Great! THANKS!

-=-Mike

Also, I assume that the changes to the VE table are after you add in both tables values correct? so I would add up table 1 & 2 and then plug that into the equation, work everything through and then adjust table 1 by the amount changed?

thanks,
Mike

Mike,

I think you'll mainly want to work with the values in Table1. Originally I was told that the relationship between Table 1 and 2 were additive, but I recently found out that is not the case. I tried to change my cal by "swapping" over part of the value in Table 1 over to Table 2 but it didn't work well. I have also made changes to Table 2 that would indicate it has a "smaller" effect on the resulting BLM's than if I made the same numeric change to a cell in Table 1. Perhaps Trax, Glenn or Bruce could elaborate.

I think ya'll may be a bit confused, you must work with EACH value in the table individually by applying the equation to each value/cell in the table depending on what the INT value is currently for that RPM/MAP combination. THere should be no "summing" of any values from table1 and 2 together, and no applying a specific "ratio" to the entire table.
Also, Tabvle 1 and Table 2 are used for different RPMs, Table1 is for 0-1600rpms while Table 2 is used for 1600-5600 rpms, at least for the $8D ECM, the only values that should be common for both tables is the VE values at the shared 1600rpm points.
*
This is a very time consuming process as you must go thru the equation for each rpm/map combination that you have recorded with a scan tool using the INT values for each that are given on the scan tool recording, you will not hit all combinations but should at least get around 50-75 different combinations driving under all different and varying conditions.

2QUick,

The Fuel tables on a 7747 overlap up to 3200 rpm.

Whoops, sorry about that, thought we were talking about the $8D still. Have no idea about the 7747.

V8 Astro,

I tried the 0x5E2 fix to no avail. My problem is that I am in Open Loop Idle when I first start up. The rpm's drop almost immediately. I have the kickdown delay set at 5sec. with kickdown steps being 25. So according to RRob, I will stay in open loop until my IAC steps in the IAC x Temp table drop or are equal to that number. However, what I would like to happen is for idle to be elevated until 70C. One other thing is that I do have MAX IAC steps set to 60. I have no other problems with IAC operation other than this cold start one. And at that the motor starts easily when cold. It just idles too low at those cooler engine temps to properly combust the A/F mixture. Once in CL and at 75C, it idles fine.