Afr adjustment problems E38 L98

[Mark300]

Hi All,

Over the last view weeks/months I have been working with weatherman Shaun’s concepts on MAF and VE calibration and have been having some good success. However I made a change yesterday and i got some unusual results. I changed my AFR from 14.63 to 14.00 (table B3671) and the results were unusual. On my scanner the command fuel is showing 14:00, but my LC-1 shows no change from before. However my LTFT change by up to 5% from what they were.

How is this possible. I am working in AFR and not EQ, could this be a factor

regards

Mark

[odd boy]

Post the tune

[mr.prick]

Are you using serial or analog connection?
If you are using serial you will need to change the fuel setting with the Logworks software to match {B3601}
http://forum.efilive.com/showpost.php?p=112020&postcount=11

For analog you will need to match the output range with a new calc_pid expression.

There have been countless posts on this the past 2 days. :)

[Mark300]

Hi All,

I will follow up on those points.
The issue that is confusing me though, is why did the LTFT move by the same % I change the AFR in table (B3671). When I set table B3671 to 14.63 my LTFT are averaging +/- 0.5% across the MAF Hz range. Which is good, but when I change my AFR from 14.63 to 14.00 which is around -5%. My LTFT went from an average of -0.5% to -5% across the Hz range on the MAF.
So what I am trying say is that even though I change the AFR, the LTFT are moving me back to 14.63. why is this happening. I even tried changing table B0141-open loop normal command fuel in gear, but I keep getting 14.63 on my LC1. I believe the LC-1 is correct as it is support by the fact that the LTFT are adjusting by the -5% to bring me back to 14.63. why does the car not go to the set AFR in table (B3671). The command fuel pid is showing 14.00 in the scanner

Regards

Mark

[gmh308]

Hi All,

I will follow up on those points.
The issue that is confusing me though, is why did the LTFT move by the same % I change the AFR in table (B3671). When I set table B3671 to 14.63 my LTFT are averaging +/- 0.5% across the MAF Hz range. Which is good, but when I change my AFR from 14.63 to 14.00 which is around -5%. My LTFT went from an average of -0.5% to -5% across the Hz range on the MAF.
So what I am trying say is that even though I change the AFR, the LTFT are moving me back to 14.63. why is this happening. I even tried changing table B0141-open loop normal command fuel in gear, but I keep getting 14.63 on my LC1. I believe the LC-1 is correct as it is support by the fact that the LTFT are adjusting by the -5% to bring me back to 14.63. why does the car not go to the set AFR in table (B3671). The command fuel pid is showing 14.00 in the scanner

Regards

Mark

Assuming you are running closed loop then the O2/LTFT system will always trim to whatever stoich needs to be for the fuel you are running....which must require around 14.63:1 AFR for your fuel. Change the fuel to 14:1 simply means that your O2/LTFT relationship says it is incorrect and will trim 5 rich to meet the stoich objective.

Open loop is a different matter.

[Mark300]

Hi Gmh,

Thank you for your reply. So in a nutshell, what you are saying is that I can't change my AFR from 14.63 in closed loop.

[gmh308]

Hi Gmh,

Thank you for your reply. So in a nutshell, what you are saying is that I can't change my AFR from 14.63 in closed loop.

Thats correct.

[swingtan]

In the E38, no.

In CL, the final AFR is controlled solely by the switch point set for the O2. While there is a table for this ( B1515 & B1516 ), there are missing calibrations still that result in these tables having no effect. So the actual switch point, or "rich/lean" crossover point, is fixed at a voltage that represents 14.63:1 AFR

Please note though, that altering the Stoichiometric setting in the tune will effect your PE fueling. PE fueling is calculated as an EQ ratio from the Stoichiometric setting. Altering the Stoichiometric setting by 0.5 AFR richer, will also move the commanded PE fueling by 0.5 AFR richer. This will occur with no adjustments made to the PE tables. So be careful if you decide to try leaning out the Stoichiometric setting, as you will need to richen up the PE settings to remain safe.

Simon.

[VEGASROBBI]

Exactly my observation, the E38 is obsessed with the O2 swithch point. Change fuel, offset MAF, change stoich settings, change O2 switchpoints, all have little to no effect; in CL the E38 will migrate to stoich for that fuel and look for the NB O2's to switch.

I've tried to modify the O2 signal with piggyback voltage and the E38 knows, gets irrational. I do not believe WB simulation of the NB will work correctly due to issues with the lo reference circut(as GMPX mentioned) as well as input from the rear O2's, cat temp and protection, etc..

Any chance of identifying the missing calibrations or create a semi-OL COS for the E38?

[Mark300]

Hi All,
Thank you for your help. The answers is a bit disappointing. I was trying to richen up to 14.00. I feel the car does not like 14.63 on high altitudes and hot temperature. It feels better on the throttle and less knock, lower chamber temperature.

Regards

mark

[swingtan]

Try altering the PE settings so PE comes in sooner. Then alter the PE commanded AFR's so that at "cruising" speeds, you run a "leaner " PE. For example, you might set PE AFR's to be 14:1 at 1000 RPM and slowly richen up to 12.4:1 at 2500 RPM. This way you will drop into a richer mixture when you load up the motor more for cruising, but still have plenty of fuel available for WOT runs.

Simon.

[Mark300]

Hi Simon,

Good idea. I have reset all my settings. My LTFT are back to +/-0.5%, so i am calibrated on MAF and VE.
I have set my pe to come in at 60kpa and throttle at 40%. My PE is starting at 14.00 moving towards 12.4 at 2500RPM. if this works then i can play allittle more with the spark table.

regards

Mark

[tadmur]

what about changing {B3671} by -5%

[swingtan]

what about changing {B3671} by -5%

How is altering the Stoichiometric Ethanol blend value supposed to help here ?

Simon.

[tadmur]

if the car is flex fuel enabled {B0178}, shouldn't the ECU look up {b3671} to decide what is Stoich. and adjust fuel to get that reading from O2s?

[swingtan]

Did I miss where Ethanol was mentioned? I thought the issue was just commanding additional fuel at altitude because the car ran better.

If you read the rest of the thread, the O2's are controlling the AFR's at the the correct stoichiometric level for the fuel being used. So trying to fool the ECM simply won't work as the O2's will pull the AFR's back to stoichiometric.

Simon.

[tadmur]

i'm new to OEM computer and EFI... things are done a little different with stand alones..
but i get it... the ECU is using {b3671} to estimate (calculate) how much fuel to use but still trimming using O2 to lamda 1.. no matter what...
unless it is in PE or Open loop.... right??

[VEGASROBBI]

i'm new to OEM computer and EFI... things are done a little different with stand alones..
but i get it... the ECU is using {b3671} to estimate (calculate) how much fuel to use but still trimming using O2 to lamda 1.. no matter what...
unless it is in PE or Open loop.... right??

Kind of, the ECM primarily looks the the O2 sensor feedback in CL. The crossover rich/lean (stoich) is abot .450V. Regardless what fuel is in the tank or what settings you make to B3671 the ECM will strive to switch the O2's around stoich.

If I remember correctly from my stoichiometry class while the A/F ratio changes with the fuel used; the NB O2 looks for 1.5% oxygen (upstream no AS) in the exhaust.

[swingtan]

Close.......

The ECM uses the existing fuel maps to calculate the required amount of fuel to add to the incoming air. The calculated amount of air, is then manipulated by the NB O2 readings to either make the mixture richer, or leaner than stoichiometric. The O2's will always give the same voltage output, no matter what fuel is used, as it looks for the oxygen left in the exhaust gas. So the NB O2 sensors will always read "Lambda" and not AFR.

So to control the fueling in CL, the ECM will look at what the NB O2's are reading, and add or remove fuel to drift the lambda reading back and forth from rich to lean. The average of the amount of correction needed to do this, is known as the Short Term Fuel Trims ( STFTs ). Changing any stoichiometric setting in the tune, only effects the base fueling levels, and not the NB O2 corrections. In CL, the O2 readings will force the average fueling back to stoichiometric and only the STFT's with change.

So to alter the amount of commanded fuel, you need to get past the O2 sensor correction. This can either be done by running full OL ( because we don't get a COS with semi OL in the E38 ), or you mess with the PE settings so yyou can retain CL for cruising / idle etc, but can command a different AFR when needed.

Simon.

[VEGASROBBI]

So, would you say B3671 is the blend factor the ECM applies to modify the ethonal tables? Rather than name it AFR Commanded would it not be better to label it Blend Factor Applired and use Lambda as the units?

[Mark300]

Hi All,
After many attempts to change the AFR from 14.63. I have come to terms with myself, (Ha ha) that it cannot be done in closed loop. I have fiddled with the PE and that seems to meet my needs. It is a fine line to where i map WOT and general good throttle response. I have keep the values above 2,500rpm for WOT. I also made PE come in quicker when PE is enabled and at lower MAP. It still needs a little fine tuning. I am also working on the spark table.
Another interesting issue I noticed, is the high level of intake valve temp on cruise. On a really hot day I got up to 175 DegC. This made the engine very susceptible to knock. So I have drastically reduced spark at lower MAP range from 50 Deg to 38 Deg. This has made the INVT drop to 155 DegC, which has allowed me to increase spark on the acceleration. This was only done once I had properly calibrated my MAF and VE. The car is very smooth at cruise and seems to be better on throttle response. This was where I came up with trying to increase AFR to 14.00 so I could increase spark at lower MAP and keep temp in Chamber down.
please could some comment on this as to whether I am opening myself up to issues I am not aware of when reducing cruise spark

Regards

Mark

[gmh308]

So, would you say B3671 is the blend factor the ECM applies to modify the ethonal tables? Rather than name it AFR Commanded would it not be better to label it Blend Factor Applired and use Lambda as the units?

Its actually an AFR table though to reference the base fueling from. :)

i.e. on gas if the MAF says 14.7 parts by weight of air per unit time then it looks up this table and injects 1 part by weight of fuel for normal gas.

Lambda stays the same. AFR changes with different fuels.