E38 idle tuning

Ahh, yes.....
This is another one of those times when the descriptions for the tables don;t seem to match real life. It's not EFILives fault as I believe that descriptions are taken directly from the GM documented specs.

Think of these as follows...

{B1651}Max Commanded Idle Area - Not Stationary : That's right, this table will limit the maximum amount of idle control air when the car is not in "stationary idle mode" as defined in {B1817} and {B1818}. This is a % setting for throttle opening and may come under the old "TPS logarithmic" function. It has more of an effect at cold idle, where you command more air than normal.

{B1652}Max Commanded Idle Area - Stationary : This is the maximum amount of idle air when in stationary idle mode ( again {B1817} and {B1818} ). The description of "preventing vehicle creeping forward when the brake is pressed" is correct, but it it basically just plain old idle control.

To make it seem more "logical" use {B1652} as the primary idle control limiter. Normally you wouldn't think of having different tables for moving idle control, so it's easier to start with the second table. Also remember that it's simply an upper limit to the air control tables. Once you have controlled the idle here, moving idle control should not need a lot of adjustment. The one exception is if you get light throttle, low speed surge. The E38 seems to get a bit confused ( it's more likely me getting confused.... ) and tries to adjust the airflow when moving slowly. I'm assuming that the throttle position is not enough to exit "idle mode" and the ECM wants to try and control the engine speed. As Hymey has said, because the minimum control amount is quite large, you get surge when the correction amount as a large proportion of the current throttle opening. Setting {1651} to a lower figure than stock simply tells the ECM to give up on moving idle corrections at a lower throttle %.

To give an example, I have..

{B1651} = 0.75
{B1652} = 2.00

If you set {B1651} too low, it will stall when cold, and just idling along. I've had mine at the point where the idle speed will change as the vehicle speed transitions the settings in {B1817} and {B1818}.

Finally, remember that EFILive will not clear any learned settings in the ECM for the E38. You have to do this yourself with an ECM reset. Also, the Idle proportional step size may be too great for a cammed car. I've mentioned before that the engine will want to surge too some degree. All engines do it, even stock they do it, you just don't notice it unless you log the data and look for it. Trying to correct the "surge" can simply end with over correction surge, which is what I think most people end up with. With my idle proportional settings {B1845}, I have "0" in the normal idle control for the +/- 64RPM cells, then corrections ramp up past those limits. This limits the amount of "drift over time" that idle correction can cause. The engine idles at 700RPM +/- 50RPM normally, it's smooth with a 220/224@114 cam so why try and correct ? That's my thinking anyway, the same thing goes with the spark timing control. Too much too soon simply leads to over correction.

Last tip on finding airflow..... ( I may have already said this as well ). I log commanded spark, and average it over a period of time in ScanTool ( trying to eliminate where fans and aircon might be on ). If the averaged value is over the tune commanded idle spark, then I add more base air, if the average is under the tune settings, I reduce idle air. Once it's spot on, I add some more base air to help with stability and then use the correction tables to keep it in line. The main tables are the spark correction ones.

Simon.

Oh My God!
that explains a lot...a subtle but very important clarification lolol
thanks heaps, I could probably lower B1651 now knowing what it actually does :) most of my idle testing was done stationery u see and I've been altering the wrong settings
Mike

Very interesting, thanks :cheers:

Yeah this is the beauty of the e38 ecm, cars with overlap camshafts of around ten degrees at 50, bugger all really but where most people don't want to go with daily drivers.

B1651 is definately for when the car is moving, but more importantly I have found that B1652 is basically stationary idle and even on coast down the TP% will find themselves at where they are when vehicle stationary, When the throttle is tipped in, pedal position percentage plus B1651 is total throttle percentage. So what Simon explains is definately correct.

A few weeks back I tuned a VE with a large cam, basically I did the following.

Drilled a 7mm hole in TB.

Zeroed out the integral and proportion tables and set min idle air to 63 g/s.

Set B1651 and B1652 to 1.10 and tuned as previous procedures to obtain desired TP% and idle timing, then reduced min idle air(with factory values populated in integral and proportional tables) until I started to see the throttle trim at idle. Then raised the value by .5 of a second( just to make sure I was bang on) and created a slight upward taper in the table.

The max idle area I achieved was 1.08, So set B1651 and 52 to 1.80, this then allows enough room for the car to have plenty of air cold start.

The car surged badly from the overlap. The trick here is to reduce B1651 to 0.15 below what was the correct max idle air in testing. So I ended up with .93(1.08-0.15) in B1651 and in B1652 I used 1.8, please take note that this exceeds the hard limit in most e38 pcms anyway.

The car then would drive much smoother on the throttle, ie around roundabouts there was zero surging say whereas before at 2% pedal position I would get bad surging now it was eliminated as it was building on the B1651 limit, but as I got off the gas the coast down idle percentages would raise back up, So it took no effort to take off from stand still etc.

Bare in mind this car was tuned in full open loop SD, running lower values in B1651 kind of works like a d port ex port in the way it reduces reversion, and I must say the e38 does it better then any other LSx pcm to date hands down. This car cruises at low 15 afrs at 100 km/h in 6th gear and puts around like a stock cam and yet has near 20 deg of overlap!

Again I was able to achieve this combined using, open loop speed density, working on dynamics and disabling power enrichment using the open loop fuel multiplier, After a lot of stuffing about I have also eliminated the lean spike issues on gear changes entirely by going away from PE and using STFTs to using more a LS1-cos3 approach.

For start up(cranking idle air-B1832), simply add your min idle airflow values across the board. When tuning fast to achieve a goal I found that if required idle airflow is 8g/s I add 8g/s everywhere and normally just massage it slightly here and there, it doesnt have to be fancy, infact if you seen some of my tables you would be shocked how bland but effective they are.

Cold start, As long as the above procedures are in place, I set...

B0141 through to B0144 all to 1.00 everywhere,

B0146 is populated to make fueling richer cold, Aim for 13:1 cold up to around 60 degs IVT, then lean off to 13.5 too 100 degrees before going to 1.00 above 100 IVT which will allow some enrichment on hot starts which is important.

So 1.15 to 60 then 1.10 ish to 100 IVT some smoothing there, your done.

I never add timing on cold start, it is not required, just airflow.

B5116 on a cammed car is a no no, set these to 0, that is all thats required for a largish cam to make them drive cold, optimal timing (same as hot) and remove the cold noise reduction values and there fine.

Again, I'm no good at tutorials, I am as clear as a Mathematics uni lecturer lol, You have to request a full tut by Swingy lol.

Also most of my testing is done in speed density, I have done a lot of MAF cars too,

Setting up B1651 really helps MAF cars are they are much more sensitive to reversion with large cams.

Normally at idle hz values, I will set the values the same across 5 or so cells, say from 3500 to 4500 odd I will add a fixed value of say 14.8 g/s as an example.

This does wonders for idle, As I have in the past set the idle parameters in a sensitive cammed car and it ran fine, I then went back and changed idle hz values and it started hunting, not from extra fuel but the TB started trim severely. It appears altering the g/s values in the MAF table also has an effect on the g/s values in min idle airflow.

I then decided to lock the idle airflow out get the afrs(g/s settings in MAF table) where I wanted them set them at a fixed value of where the hz will oscillate at idle, and then went from there. This was an experience for me I tell you lol.

Light loads at say 40 kpa I run around 14.0:1 afr around 50-60 I go low to mid 14s with cams sometimes 15s above 3000rpm I never go leaner then 13.0.

Getting B1651 right helps a lot with MAF tunes, then on slight throttle movements I look at the MAF oscillating and also keep the values the same in those cells as rpm increases I just ensure there is a smooth transition between cells. MAF tuned e38s with correct vves are magic, but in my belief done right is more work, Normally I tune speed density unless the customer requests. I have had cars have the afrs move over time with MAF cars whereas SD tuned cars never move. In a situation where a car is heat soaked and ramped on the dyno, the afrs will be more consistent with the MAF and mafless can move a little from heat soak, This is where the MAF shines. When changing altitude a MAF tuned car will vary afrs a lot more then SD car, and overall as said previously, I tuned two HSV's same owner similiar mods similiar kms before and after, 2 years later the SD car has a rock solid afr line, The MAF car did run a fair bit leaner I had to clean the MAF and retune. I presume it may have something to do with the total amount of airflow that is drawn into the engine has changed over time. I write about this as there is overall really less work in the SD tune to set up including idle and drivability of a big cammed e38 powered car. But having said that excellent results are obtainable with both.


This was just a quick update as a part of my blog, and again thank Simon for his relentless work also on e38s. Most tuners don't realise the work he has put into it. Above all dont make radical changes to compensate from errors elsewhere, Start from scratch keep it simple and keep it as basic and bland as possible. I just don't think there is enough info out there on e38 pcm tuning and believe me I could right a book on it this is just the icing on the cake.

Cheers

Joel

Quote:

---

Originally Posted by The Alchemist

on topic with idle settings I have found this works well for me, engine has a mild cam 225/230 591lift and a supercharger:
b1651 1.8
b1652 2.5
min idle airflow in park neutral at 800rpm around 6.8g/s and ingear around 8g/s

following your other suggestions for the integral/proportional stuff seemed to work well and relaxing the spark idle control.
average readings were as follows:

idle neutral 800rpm > SAE.TP 18.4% GM.ETCTP 9.3% MAF 11.96/s calc MAF 10.63g/s spark 20.5deg average
in gear 800rpm > SAE.TP 21.8% GM.ETCTP 13.5% MAF 15.57g/s Calc MAF 12.82g/s spark 22.5deg average.

what I could not seem to do was go open loop by maxing out minimum idle air to 63g/s and zeroing out the correction pids b1843/1845 as per this thread.
When I did this it didn't matter what I commanded in b1651 (I went from 0.5 to 3) all I got was a SAE.TP of 22% flatline and a high idle of about 1200rpm???
What have I done wrong?
I ended up finding the sweet spot by putting min air really low at 4g/s and B1651 at about "2" and slowly narrowing the gap until the idle swings stopped.
Cheers,
Mike

---

Mike I wish I was over there with ya having a fiddle with it. B1652 is definately the parameter that will prevent it from going to the hard limit percentage.

The reason why you may of been having issues with the min idle air is because the MAF g/s values definately play a role and are linked to min idle airflow. I get the idle afrs right before tuning and flat line them. across 4 or so cells so I get consistent airflow readings at idle. As long as B1652 set at say 1.00 and min idle air is 63g/s it will definately stop it from going to 22%. The MAF cars can be tricky at idle. Once I go blower or big cam the MAF goes and it eliminates the issues you get with it from idle tuning. I should have mentioned this earlier, just another glitch.

Once you get min idle air right, and you no what B1652 value you used to get reduce that value by .15 and place it in B1651 and then set B1652 to around 2.00, 2.00 and around 0.90 is around normal, you can experiment going down to 0.7 as Simon has done 1 car I have 0.5 in there.

Hope this helps

Joel

I should also mention that the factory style efi manifold does play part in some weird idle tuning, Ie idle timing corrections and airflow corrections are largely related to the style of the factory bunch of bananas manifold.

A single plain traditional manifold with equal runners not going into a large plenum area makes a massive difference. I run idle timing locked out and airflow locked out hot and cold, I just add more choke at start up it fires up and runs to commanded rpm, Also drives smoother then the factory manifold.

From my testing its not only cold moving parts and oil that create the need for more air at start and also power steering air con and alternator load variances that require the need for timing adjustment while idling. It has appeared to me the design of the manifold with some cylinders readily getting more air at idle then others that creates the need for a complex idle control system.

My new setup, Some nice ported L92 heads....GMPP manifold, 12.2:1 compression, E85.

Camshaft...248-254 on 106 LSA, from testing the efi 112-116 LSA cams aren't working with this manifold going to just a 110 LSA showed a massive improvement in torque, So I am going to 106LSA to make the most of the manifold a traditional carb manifold grind, and with the manifold having better idle manners it should work out just fine.

Cheers

http://www.youtube.com/watch?v=-Ye_IGa3i2c

Here is an example of a customer's vehicle. It is a VZ SS sedan with a 6L L76, 4-2-1 pacemaker headers and a dual 2.5 inch exhaust. I fitted a 22x 22x 112 LSA camshaft with overlap exceeding 0 degrees. The first owner wanted it choppy, the new owner purchased the vehicle fitted a 4000rpm stall and 3.9 diff gears. I retuned the idle and it reels off 11 sec passes easily at full weight. Here you can see car idling at 800 rpm and 40 kpa both in drive and neutral, including loading the car in drive and letting it return to idle using only 15 degrees of idle timing with very little spark correction. I also have a 6mm hole in the throttle blade and it uses entirely stock proportional and integral correction tables. The car is tuned in full open loop and speed density. Has only been road tuned for drivability and track tuned for top end. I love this car as it idles stock and has no troubles hosing the likes of a boosted typhoon on the street as its so snappy, I have never achieved such a smooth idle in earlier LS cars using the same camshaft. Gives an idea of what can be achieved.

cheers

Wow, good job mate :cheers:

Yes...what Joecar said. Great job there hymey. You are leading the way there delving into E38 tuning with cams and idling with Mr Swingtan. Wow that's smooth as !!!!! ;)

Well I think we are missing a couple of tables still :)
Like an ECT modifier for adding g/sec for cold starts.
Min airflow is fine once warmed up. But not for cold starts.
Sure there is ECT "Correction" multipliers. But nothing for base idle airflow addition vs ECT.

The reason I mention this is because my 440 cube lump doesnt like cold starts. It fires and dies immediately no matter what min idle airflow I use.
Im going to have to join the rest of you lot and drill the blade. Im sure if there was an ECT adder to base airflow then that would not be necessary.

Cranking idle air is also fine for the initial firing, its when it blends into idle airflow at like 3*C it just dies. Ramping correction multiplier up doesnt do anything useful either.
At around 40*C ECT she settles down and idles much better.

Anyway, the drill is coming out this week hopefully that will give me the base airflow g/sec I need to prevent the cold start stalls.