About to throw in the towel nailing down idle

[5_Liter_Eater]

My pleasure. Great to hear!

[rio95]

Alight what the heck am I missing here? I tried your procedure, but don't get the idle adapt adjustment step. I started by zeroing out my airflow table and then logged airflow with your config with adapt turned off (your procedure doesn't say to turn it off, but one of your posts after said to do so). I then pasted the airflows for 650, 800, 100, and 1200 into my tune and estimated the lower cells and then multiplied everything by 0.8. Then I turned adapt back on and logged adapt and got averages very close to zero in the same RPM cells. By your explanation doesn't this mean my airflow is good? I then decided to see what would happen if I doubled my airflows in my tune. Then I logged again and once again got average adapt values very close to zero. When you log adapt and use it to adjust airflow isn't it half the range you are interested in rather than the average? For instances, if adapt goes fluctuates from say +2 to -2 it averages zero. Do you do a percent multiply with zero or 2?

log with no airflow and no adapt.hpl
log with new airflow.hpl
log with new airflow times two.hpl

[5_Liter_Eater]

That's interesting that your adaptive ide spark values are similar when you doubled the airflow numbers. Doesn't make sense. How does it drive with the doubled numbers though? I'd expect it to hang coming back down to idle. What's also interesting is that your max and min idle spark values are similar between the new airflow and doubled airflow logs. The theory has been, and it's been a while since I've done it, that if the airflow number is too high then idle RPM will want to rise and the adaptive idle spark will average negative in order to bring the RPM down.

The second part of the procedure is more of a sanity check. See how it drives with the calibrated numbers and the doubled numbers.

[rio95]

That's interesting that your adaptive ide spark values are similar when you doubled the airflow numbers. Doesn't make sense. How does it drive with the doubled numbers though? I'd expect it to hang coming back down to idle. What's also interesting is that your max and min idle spark values are similar between the new airflow and doubled airflow logs. The theory has been, and it's been a while since I've done it, that if the airflow number is too high then idle RPM will want to rise and the adaptive idle spark will average negative in order to bring the RPM down.

The second part of the procedure is more of a sanity check. See how it drives with the calibrated numbers and the doubled numbers.

OK I thought I was understanding you correctly.... Your theory sure doesn't seem to hold true on my car. Is there something else I'm not doing? Do you do it in closed loop like I did?

Yes I can get the rpms to hang. I didn't drive the exact doubled tune referenced here but I did increase my airflow significantly the other day just to see what it'd do and I got hanging rpms. The annoying part is I still had some of the rpm dips I'm trying to get rid of....

Guess at this point I have no idea how to tune the airflow table. Both times I've played with this things are different than you explain.

[GHuggins]

OK I thought I was understanding you correctly.... Your theory sure doesn't seem to hold true on my car. Is there something else I'm not doing? Do you do it in closed loop like I did?

Yes I can get the rpms to hang. I didn't drive the exact doubled tune referenced here but I did increase my airflow significantly the other day just to see what it'd do and I got hanging rpms. The annoying part is I still had some of the rpm dips I'm trying to get rid of....

Guess at this point I have no idea how to tune the airflow table. Both times I've played with this things are different than you explain.

Typically and I mean that - "typically" if your having idle dips you need to go in and address your idle proportional and integral airflow tables... Increase your proportional airflow table by 1.1 for starters. You'll probably need to go higher than even this. Then for your integral airflow multiply it by .5ish. If you increase your proportional table again make sure to decrease the integral table at the same time. I also like to start by copying the higher of the two tables whether it's the drive or coast into the other table then modifying them if you have a manual trans. If it's auto leave the tables different then modify them from there. You can over modify these tables to the point that the idle hangs, so keep an eye out for that.

Then something that gmtech likes to do is increase the torque follower vs gear table. Aparently does the same thing as the formentioned proportional and integral adjustments or atleast something like it where it forces the tb to close slower once getting to where you've jacked the numbers up. I've never tried this, so I can not comment on how to do it or what the negatives of it would be if you went too high or anything.

Hope this helps you out.

[rio95]

Typically and I mean that - "typically" if your having idle dips you need to go in and address your idle proportional and integral airflow tables... Increase your proportional airflow table by 1.1 for starters. You'll probably need to go higher than even this. Then for your integral airflow multiply it by .5ish. If you increase your proportional table again make sure to decrease the integral table at the same time. I also like to start by copying the higher of the two tables whether it's the drive or coast into the other table then modifying them if you have a manual trans. If it's auto leave the tables different then modify them from there. You can over modify these tables to the point that the idle hangs, so keep an eye out for that.

Then something that gmtech likes to do is increase the torque follower vs gear table. Aparently does the same thing as the formentioned proportional and integral adjustments or atleast something like it where it forces the tb to close slower once getting to where you've jacked the numbers up. I've never tried this, so I can not comment on how to do it or what the negatives of it would be if you went too high or anything.

Hope this helps you out.

Thanks that's worth a try. I'll report back.

[gmtech16450yz]

Hey guys the reason the ECM has those tables is because they do different things. It may seem like they're having the same effect, or a different way to do the same thing, but they're not. I'll try to explain...

This actually is easier for us old guys that have been around carburetors for a few decades. The throttle on a carb has the same issues like closing too quickly, not opening quick enough for transient loads like A/C, not closing quick enough after the load is taken away, etc. To solve the problem of the throttles closing too quickly, they put "dashpots" on carbs. When you open the throttle the dashpot extends. When you close the throttle, it hits the dashpot instead of the hard throttle stop or idle speed screw. The dashpot then slowly retracts until the throttle is sitting against the idle speed screw. THAT'S the 'Throttle Follower"! Think of the throttle follower as kind of like a dampener for the pedal MOVEMENT. Even though your foot lifts off the pedal quickly, you don't want the throttle itself doing that. The throttle follower, like the dashpot, can't actually move the throttle on it's own, it only controls how "fast" the plate moves. When you add torque (raise numbers) to the follower table, you're telling the throttle you want more torque left on after you let off the pedal. If you lower the numbers, it means you want the torque to drop quickly after letting off the pedal. These tables are VERY important for tuning throttle response and throttle "feel". They will also effect throttle opening. Lower the numbers and the throttle will have a sluggish feel. Raise the numbers and the throttle could be too touchy on a high hp application.

Can you see how this is very different than setting min air rates? The minimum air settings are like the throttle stop screw, it's simply the lowest amount of air, or blade opening that can be used. It doesn't really control how fast the plate closes, just how far it closes. The adaptive idle stuff is another story, but for the most part that stuff is only when the car is stopped anyway. And yeah, there's a bunch more tables that control throttle up and down steps, total range of control, etc. Another vital table for tuning this throttle response stuff is the ETC Scalar. The tuners in here with Engineering backgrounds like to calculate exact numbers to change the scalar to when changing throttle bore sizes, porting throttle bodies, etc. I like to use that table to also control throttle speed and I don't worry too much about if the numbers "compute" to what they "should" be.

This throttle/ idle stuff is pretty darn complicated but not impossible to understand and control. Just try to remember it's basically a balancing act between airflow and ignition timing, with fuel mixture playing a smaller role. The goal is roughly to have your ign timing as high as possible (I run most LS engines at ~26 degrees at idle) and your TPS as low as possible (most stock cam LS engines should be ~6-9% on a warmed up engine in neutral). The recent posts in this thread and the other thread about the rpm's dropping too fast have more to do with the Throttle Follower than the airflow tables. That's why I figured I'd jump in because it looks like the Throttle Follower tables don't get the attention they deserve. They are very powerful tables for making these high hp cars drive nicely. I know those tables are FAR from stock in my tunes.

Does that help or did I make it worse? lol.

[edge04]

^^^intersting. I use my car for road racing and I always want RPM's to slowly come down, seems like factory they come down quickly to 1500 or so and then slow down. Well in road racing you're rarely at or below 1500. Having them come down slower would be similar to or helpful for rev matching and have better/more throttle response exiting a corner.

Do adjustments in the throttle follower tables work with both the clutch in or not?

[edge04]

all, maybe a dumb question but does the exercise in post 85 work on E38 computers? I ask because I tried it and not getting anywhere. Maybe its the "rich after flash"???

I idle good and find idle good but when at a stop idling it is real rich. Let me sure I'm doing this right. Say I want to remove 10% fuel from the airflow final minimum I would multiply the table by .90?

[schpenxel]

You're confusing two different topics.

To remove fuel you need to adjust either VE or MAF tables, depending on which you are/aren't using

And are you running open loop or closed loop?

How rich is "real rich"?

[edge04]

sorry, your right.

I tune in open, running both VE and MAF and running 10-13% rich. I have further work to do before applying these tips!

You're confusing two different topics.

To remove fuel you need to adjust either VE or MAF tables, depending on which you are/aren't using

And are you running open loop or closed loop?

How rich is "real rich"?

[Michael_D]

http://www.masterenginetuner.com/ls-big-cam-idle.html

[GHuggins]

http://www.masterenginetuner.com/ls-big-cam-idle.html

Step 5 - Just to further help things. The integral airflow table is probably just as important if not more important than the proportional from my own experiences although the greater the proportional adjustment the less you need to adjust the integral.

I will now however combine Gmtech's and this articles way of adjusting the throttle follower tables with the adjustments to the proportional and integral airflow and min airflow tables. I'm still just not sure I'll be adjusting them as high as Gmtech as I can foresee throttle hang and dangerous situations resulting from too high of numbers in those tables. Plus as I've stated already, I've never needed to adjust these even with some serious cams. Perhaps I was just overcompensating like it would seem Gmtech is with just adjusting part of the tables? In other words had the proportional and integral tables adjusted too much?

We'll see on the next one I guess

Thanks Michael for the article.

[schpenxel]

Good article.

[Michael_D]

I like Dan Maslic. He helped me out when I first decided to try this new fangled 'tune via laptop' crap..... I see he new has his book on line, free to anyone who wants to read it. Not sure what his angle is, but he seems to be giving away a lot of stuff he could be making money on.

[shoaf85]

I'm trying to fix a car for a buddy that has been previously tuned. The car had an awful hanging idle (which I've fixed) problem during coastdown and also a major idle fluctuation when coming to a stop ONLY IF THE CLUTCH IS OUT AND CAR IN NEUTRAL. If I keep the clutch pressed in it will settle into its idle no problem.

The car does fair without the ac on but with ac the idle goes crazy again.

Here is the tune and two log files using the config in this thread.

First log is sitting still checking the error as described in this thread at 600-1200 rpm, I actually went up to 2k just to see.

idle tuning.hpt
adaptive spark test.hpl
drive1.hpl

Once the car comes to a complete stop it will settle in so I'm guessing it has something to do with a coastdown table or follower table that I haven't tried yet


Thanks for any help!

EDIT: Idle still hangs around 2k rpm at a slow rolling idle and drops to commanded idle as soon as car stops.

[Michael_D]

Read the article I linked to above.

Fine tune idle air using Bill's adaptive idle histo. The combination of both techniques works amazingly well.

Also, keep a close eye on actual idle spark verse commanded spark. If it's high, you need more air, if it's low, you have too much air. Start there.

[shoaf85]

I used the idle cfg and I'm within 1%+/- in his #11 histogram. Idling at a stop is good it's just the transition from rolling to a stop that is giving me problems.

[Michael_D]

You can be less than .0000001% error, with timing pulled down 8 deg.......or above commanded.

You need to verify where you are at with idle spark FIRST, and get that where it needs to be. I bet you will find it being pulled at idle, which means too much min air, which will cause rpm to hang.

If not, then go more negative above 1000 rpm in the idle torque table.

Read the article. It is all explained very clearly.

[rio95]

You can be less than .0000001% error, with timing pulled down 8 deg.......or above commanded.

You need to verify where you are at with idle spark FIRST, and get that where it needs to be. I bet you will find it being pulled at idle, which means too much min air, which will cause rpm to hang.

If not, then go more negative above 1000 rpm in the idle torque table.

Read the article. It is all explained very clearly.

In your example are you saying the average idle adapt would be 8 degrees or do you use something else to compare actual timing to commanded? Is there a commanded timing pid?