Hi all,

 

IMO, I wouldnt obsess over it. Once your actual and commanded AFRs are within +/- 3-5 percent or so, your good to go. Itll be nearly impossible to get it perfect. Just the humidity in the air alone can throw things off a few percent. I usually worry more about how it drives then anything else. As far as the fueling goes, I follow the above 5 percent rule when calibrating the fueling. Once its within that range, its tuned as far as Im concerned.

 

That, plus the charcoal canister purges will keep ypu from EVER getting it "perfect"....

 

I'm not going to say much about "how close is close enough", but I will say that you want to stay on top of your BLM cells and make sure they don't split in different ways.

For example, if you have one BLM cell at say 122, and one at 134 (about a +/- 5% difference from 128), and then you cross from a BLM cell at 122 to a BLM cell at 134, you're going to notice a burp in the way the engine runs because you just added 10% to your overall fuel as you cross the cell boundary.

that's one of the reasons to lock the blm at least for tuning purposes and getting as close as possible, so that you don't end up with some VE cells needing +5% and others needing -5% causing weird artifacts that are hard to chase to do BLM cells being all over the place.

My opinion is that keeping the total BLM spread as small as possible is just as important as getting close to 128/128 as possible.

EDITED for technical mistake. going from BLM 122 to BLM 134 _adds_ 10% fuel.

 

I have teh same take on fueling. It definatly goes without saying that precision is what matters most when it comes to drivability. Thats the part where you want to get ****. You cant really tell teh difference between 14.5 and 15.5 at part throttle. You can, however, notice if the fueling on one power cycle is 14.5 and the next is 15.5. the engine will run real uneaven. Smoothing the VE table once its within range will help make sure that there are no sudden changes in fueling while the ECM moves from cell to cell.

 

Thanks you for the great reply's,

 

Getting the Prop gains right helps smooth thngs out also.

 

If you go into the the Constants in TunerCat for your 8D TDF, there are two constants called the Max BLM and Min BLM. Typically the Max is set at 160 and the Min is set at 80 (or maybe 90...it's been awhile sinced I checked). Just set those bot the 128 and then you tune strictly on the INT value.

Then, just drive along at various "loads" while data capturing with Diacom (or whatever scan tool you are using) and then review the INT values you are getting at various RPM/MAP values. Then make corresponding changes to the VE tables until you get your INT values as close as possible to 128. Generally if you get within +/- 4 (ie 124-132, you are close enough).

When the INT values is higher than 128, you are lean and need to richen the mixture so you need to add fuel from the corresponding cell in the VE table. And vice versa, when the INT value is less than 128, you are rich and need to lean the mixture, so you decrease the value in the corresponding RPM/MAP cell in the VE Table.

Now go give it a try and play with it. Once you start doing it, it becomes really clear what everyone is talking about.

Of course, if you had a WideBand O2 sensor and Real Time Bin Editor like TunerCat R/T, you can do this "on-the-fly" with someone driving while you make changes...but you miss the fun of learning.

 

Thanks

 

The ECM will be using the INT only to perform corrections and will be using the stock O2, with a neutral INT cooresponding to ~14.7:1. If you have a wideband O2, just set the enable temp for closed loop to the max allowed, set the max, min BLM to 128, and itll be in open loop. You can then tune by the desired and reported AFRs. You should try to get the AFRs to within a few percent of the commanded AFR and make sure rapid fluctuations are less then 1-2% under steady or semi-steady state. The fluctuations will be what you want to really minimize.

 

Question for Dimented:

with EBL I will assume that the Learn function will adjust the VE using input from WB02 only. I was planning on replacing my 5 year old Bosch NB 02 with a heated GM but since I use a permanently installed WB maybe I just run the WB02 only? would that not be a better idea?

Also from my LM1 there is an output "stereo" connector. how does one patch that into the EBL/ECU? currently the ECU is back under dash(took years) so if I drop it where to access the ECU for that input?

 

..

 

Gee, I wonder how many times I suggested a WB

For WOT, ONLY a WB can do that properly. But for part throttle, it's better than nothing. For tuning, WB is the proper tool. Tuning via the stock O2 using the INT as a guide is better than nothing. Your other option is nothing.

And, in response to your prior question - just in case it was not clear about tuning with the INT: the ECM still makes corrections to based on the O2 readings. The only advantage I find to tuning by the INT (with the BLM's locked) is it simplifies the tuning process. What change do you make when you have a BLM = 120 and an INT = 140 or BLM = 137 and INT = 118? That is why I prefer to use the INT by itself.

As you start to make changes to the VE cells, you will start to notice a relationship between the amount of change you make to the cell and how it affects the INT reading (the next time you scan).

Do it long enough, you may even come up with a mathematical algorythm that allows you to calculate the amount of change you need to make to a particular VE value based on the current INT reading. (ie. Current_VE_Value x (128/Current_INT_Value) x "magic number" = NEW VE Value) ... that will get you very close to 128/128 the next time your run.

BUT, that "magic number" is a little different depending on the mods to the engine. So you will have to compute your own "magic number" for your own engine.

But remember this, this ONLY tunes your part-throttle tables. With properly tuned VE Tables, you can make an "intellegent guess" at the amout of extra fuel you will need to add for P/E...but it is only a "guess". I guarantee the first time you put an actual WB to it, you will be stunned who wrong your "intellegent guess" was not "intellegent".

But, do that enough times, and you may compute another "magic number" that will make your "intellegent guess" more "intellegent".

Sorry, I am hurting really bad now and cannot type or expand any more. Need to rest and reduce pain.

 

I dont have EBL, but Id imagine that RBob should have a table in there. Match the table output to what you program into the LM1. I dont know if autolearn will work with the narrow band or not.

For hooking up the wideband, just use a stereo plug and leave a pigtail with adiquate slack within easy reach so you can just plug the unit into the ECM. I beleive the actual input for the WB is on the EBL PCB. Ask RBob to be sure.

 

..

 

If you do get a WB, then consider going to a Real Time Bin Editor too. You would be amazed who quickly this shortens the tuning time. What would take days/weeks to accomplish can be done in a couple of hours. All you need is one person to concentrate on driving while the other makes the changes to the bin while the car is actually running.

This is when a good "dyno hill" comes in handy. A nice long hill with a reasonably steep grade is the best for tuning many of those higher MAP cells and the deceleration helps get the lower MAP cells.

Then you can look at tuning so you can run solely in "open loop". A few people find they can't get their car to run properly in "closed loop", or their car just runs much smoother & better in "open loop". I know a couple of guys that just flat out prefer the way their car runs in "open loop" and just forget about "closed loop". Then they even experiment with the AF ratio they really want to run and some even don't bother with the P/E tables any more and do everything in the main VE Tables....

Once you get the WB & Real Time Bin Editor, then you can truly have fun experimenting and get the perfect tune for YOUR engine and have it run the way YOU want it. Best of all, you can do in a fraction of the time...and time is really your most valuable resource.

 

Something

 

A couple of my buddies are using the LM-1 and quite like it. I have never used the Ostrich so I cannot comment about it, but others seem to like it.

 

i have the ostrich and, while i am still a "rookie" at tuning, i really do like it.

now, from the rookie side, how do you tune open loop? is that your main ve table, still? i am running a '730 with the $88...v6. are the tables basically labled the same? (i haven't really looked at any v8 files cause i'm affraid i'll confuse myself). sorry for somewhat hijacking the thread but it still relates to the topic.

 

AM91

 

cool, cause i thought of something else. how do you make it stay in open loop? raise the temp for closed loop...right? i think i've read that, maybe, but where do i do that?

 

I got tired of main VE tables and went MAF and tuned the MAF tables. Took all of 2 hours to make it darn near perfect. From there, I enabled open loop a/f ratio lookup turning off closed loop in the process. With a wideband hooked up, I show 13-16.5:1 depending on the load. At WOT, my engine likes 12.3-12.4:1 for peak torque and slightly lean it out to 12.6:1 for higher rpms. I dislike running closed loop, I just don't see the point of cycling the a/f ratio, very rich then very lean when the engine runs much smoother staying at a constant a/f ratio.

EDIT- To disable closed loop Raise the temp for closed loop way high and set the BLM/INT limits to 128/128

 

I was hoping someone would chime in with a MAF setup. So, am I correct that in order to tune the MAF tables, you would lock the BLM's at 128, and change the MAF tables while driving at different load conditions/rpm's trying to get INT somewhere near 128 for the different conditions? Or do you tune the MAF tables to try to get AFR in the proper range for different load conditions/rpm's using a WBO2?

Not trying to hyjack this thread and I believe this is still relevant to the original quest for tuning 128/128. But this is really great info for MAF guys as well.

 

You could use the INT to tune, but the only thing youd use is the airflow as your reference. the load and RPMs are irrelevant as the fueling is based off of the airflow and AFR. Its MUCH faster to use a wideband, though. With the wideband, you can get points along the true MAF curve and just refit the curve. It only takes a few runs to dial in a MAF table. It may take longer to do the stock ECMs with multiple tables, but the idea is the same.

 

subscribing to this....good info.

 

oops.

 

Hello again,:

 

You need to try to hold the engine at a nearly constant load and RPM and let the integrator settle out. Most dataloggers have provisions to overlay the BLM/INT data onto a grid of cells like the VE table, or at least show you where you are in the table. I think thats how TP handles it. Once you have the steady state integrator and the MAP/RPM that it was at, you go to that same point in your VE table and adjust that cell based on the INT correction. For example, if the engine was running around 2000 RPM at 30 kPa and I had an INT of 120, I need to go to the VE cell at 2000 RPM and 30 KPa and reduce the VE a little bit as its too much and causing the motor to run rich.

This is where it gets confusing. In alot of computers, the INTegrator term actually represents injector pulsewidth. The whole closed loop O2 routine is pulsewidth based, meaning that the corrections are not multipliers like the BLM, but rather some small ammount of pulsewidth that allows the routine to have short term control over the fueling to make fast corrections as well as make the O2 cycle. This is very non-linear, and is difficult to use by itself for tuning. Youd have to know what your current injector pulsewidth is along with the INT to see what the percentage of fuel correction is. Maybe Grim knows for sure how its handled in the SD TPI ECMs. If its just a plain multiplier, then the INT will work for tuning. If its a pulsewidth, then its only useful as a rough guide for what the fueling is doing over the short term.

 

OK, thank you for the reply,

 

8D adds or subtracts the CL corr (which includes the INT) to the PW.
Some of this is up in the 730 tuning section.
Just not all in one thread IIRC.

 

Hello, all again

 

Assuming the above info by Z69, the formula would be something like (current inj. pulsewidth + (INT-128))/current injector pulsewidth in order to get the percentage of fuel correction. Stick to using the BLMs for now. The int is useful for seeing which way the fueling is going, but too non-linear for most tuning.

 

I've kinda sat back from the whole 128/128 thing and rethought everything.

We all know the nbo2 is only gonna get you so far, I decided closed loop is not needed. Think about it, every carb prior to 1983 didn't have closed loop, and they all ran fine. People didn't tune with 02 sensors either.

Putting all that aside, i've started tuning open loop, a static tune that feels best, rather than looks best on a log. Sure the WBo2 is great for pinpointing a set AFR WOT on the dyno, but honestly nobody knows exactly what AFR the car will make post power at, so it's better to keep trying until it stops building power than just tuning to a set AFR.

Thats my opinion. I'm done with closed loop. I can get a much better steady tune in open loop, without surging or bucking or any of the crazy stuff.

-- Joe

 

The main point of the wideband is to get feedback on the calibration to see if its correct or not. An engine will run anywhere from 11:1 - 16:1, so its hard to see if your VE tables, MAF table, or whatever is correct just by feel alone. It can get you close, but with a wideband, you can dial the tables in so you can be sure your getting the AFR you want. With a properly tuned setup, the actual and desired AFRs can be within just a few percent of eachother.

 

Well my feeling is the wideband is just gonna tell you where it is you made the most power. You can't dial in to a specific AFR and expect results. The AFR your motor runs best at could be anything. All the wideband will do is tell you what it is once you've found the sweet spot.

I found my wideband completly useless for anything other than what I just described. I sold it, and don't miss it.

Sure it will tell you if you're stupid rich or stupid lean. But so will the plugs.

-- Joe

 

If the engine is running properly, it tells you your actual AFR. I tuned by reading the plugs and doing things caveman style and it works to a good extent, but its only roughed in that way. With the wideband and a mere 15 minutes, I got it nearly bang on, and was immediatly rewarded with a smoother, more efficient running engine. the wideband cant be used when things are way off, but once things are close, its a good sanity check and also useful for seeing whats going on during fast transients. The WB is another important puzzlepiece in tuning. Without it, thats one less piece of information that you have at your disposal. It cant be taken all by itself as gospel, but with all the other information, it helps form a complete picture of whats going on.

There are some things that I never wouldve fixed or diagnosed without it. Its another tool that every tuner should have in their toolbox.

 

I hope this is not considered hijacking this thread I have question about how to make changes needed to the VE table from the int values.

(ie. Current_VE_Value x (128/Current_INT_Value) x "magic number" = NEW VE Value) ... that will get you very close to 128/128 the next time your run.

How do you figure out the "magic number"?

Thank you

 

How many of you tune in open loop and USE the
wideband

 

Hi Joe. I found when tuning on dynos that 12.8-13.1 typically resulted in the max power. Below 12.8 to 12.4 tend to result in a small decrease and only when the AF got below 12 did I start to see noticeable drops.

Conversely, once the AF got higher than 13.4 I start to see power drop off quickly. It seemed that one is better to error slighly richer than 12.8-13.1 than leaner.

As I said, this is with "dynos". But, it also seems that tuning at the track (with a WB) yielded similar results as well (in terms of MPH), so there does seem to be a co-orleation between the AF & power. And if I was hard pressed to pick my "preferred" AF ratio, I would say 12.9-13.0:1

 

Everyone should do their own as I found there were difference at various rpms. But this formula worked fairly well - though not perfect. It's algebraic so best of luck:

Current_VE_Value x (1-((1-(Current_INT/128))/2.3) = New VE_Value

There is still some "slop" but it's as close as I can get without complex math.

I set it up as a couple of Excel Tables with some math.

Table 1 = Current VE Value, Table 2 = Current_INT_values & Table 3 = New VE_Value with the calculation in Table 3. Also, set up 2.3 as a "separate factor" and play with it and you will probably dial in a better factor that gives even better results.

Also, if you use TunerCat, you can "Copy" the table straight off TunerCat into the Excel Spreadsheet...and then paste results straight back. (I got TC to work with Excel a number of years ago when I used this method).

HTH