saw this on qw, I figured I’d post it here so CV, jfonz, snowtrooper and some of the other guys who have experience tuning bosch ecus could comment:
http://forums.quattroworld.com/rs4b7/threads/5405.phtml#79503
[QUOTE=Mickf29]Hey guys, some good data off my car on the lambda, AFR…
http://mickflanigan.smugmug.com/Cars/Data-and-Plots/i-hJcmNvs/0/XL/lambda-XL.jpg
So you think I am lean? I will replace the O2 sensors and make sure all the exhaust fittings are tight, etc Thursday. Next week will put the SAi system. Then the determination will be made on the tune. i will be able to get past DEQ at that point though, but if the Lamda values don’t come down after the O2 sensor replacement, tune is getting yanked. Right now this is showing an AFR of over 15, unacceptable.
but from the data, the upstream sensor is reading too rich. Its leaning out the car and the second o2 sensor is reading to lean because of that. Tells me the upstream sensor is bad. Interesting thing though, the bank 1 sensors upstream is bad today, yesterday was bank 2 sensors upstream. Can’t believe all 4 sensors went bad, but perhaps both upstream sensors were severely fouled due to running too rich? I do feel good that the replacement of sensors will fix this and that the advice I got about the tune was most likely wrong. At least int eh case of the O2 sensors. Not the SAI codes.
Generally blocks 30, 31 and 31 are used to determine A/F mixture.
Had a look at 31 first. That will give you the Lambda values around 1.
Lambda of more than 1 is Lean
Lambda of less than 1 is Rich
A/F ratio = Lambda x 14.7
My readings:
And my fault codes:
ECU #1:
4 Faults Found:
001169 - Secondary Air Injection System: Bank 1: Insufficient Flow
P0491 - 002 - Lower Limit Exceeded
Freeze Frame:
Fault Status: 01100010
Fault Priority: 0
Fault Frequency: 1
Mileage: 86854 km
Time Indication: 0
Freeze Frame:
RPM: 763 /min
Load: 23.9 %
Speed: 0.0 km/h
Temperature: 36.0°C
Temperature: 22.0°C
Absolute Pres.: 990.0 mbar
Voltage: 13.843 V
008342 - Post-Catalyst Oxygen (Lambda) Sensor Correction; Bank 1: Too Lean
P2096 - 002 - Lower Limit Exceeded - MIL ON
Freeze Frame:
Fault Status: 11100010
Fault Priority: 0
Fault Frequency: 1
Mileage: 86862 km
Time Indication: 0
Freeze Frame:
RPM: 2308 /min
Load: 13.7 %
Speed: 61.0 km/h
Temperature: 93.0°C
Temperature: 14.0°C
Absolute Pres.: 1000.0 mbar
Voltage: 13.462 V
008598 - Oxygen (Lambda) Sensor B1 S1: Signal too High (Rich)
P2196 - 002 - Lower Limit Exceeded - MIL ON
Freeze Frame:
Fault Status: 11100010
Fault Priority: 0
Fault Frequency: 1
Mileage: 86906 km
Time Indication: 0
Freeze Frame:
RPM: 2322 /min
Load: 13.3 %
Speed: 61.0 km/h
Temperature: 83.0°C
Temperature: 17.0°C
Absolute Pres.: 1000.0 mbar
Voltage: 13.589 V
006435 - Please Check DTC Memory of ECU Number 2
P1923 - 008 - Implausible Signal
Freeze Frame:
Fault Status: 01101000
Fault Priority: 0
Fault Frequency: 1
Mileage: 86918 km
Time Indication: 0
Freeze Frame:
RPM: 0 /min
Load: 0.0 %
Speed: 0.0 km/h
Temperature: 55.0°C
Temperature: 53.0°C
Absolute Pres.: 990.0 mbar
Voltage: 11.938 V
ECU #2:
2 Faults Found:
005137 - Secondary Air Injection System; Bank 2: Insufficient Flow
P1411 - 002 - Lower Limit Exceeded
Freeze Frame:
Fault Status: 01100010
Fault Priority: 0
Fault Frequency: 1
Mileage: 86854 km
Time Indication: 0
Freeze Frame:
RPM: 753 /min
Load: 23.9 %
Speed: 0.0 km/h
Temperature: 36.0°C
Temperature: 22.0°C
Absolute Pres.: 990.0 mbar
Voltage: 13.970 V
008344 - Post-Catalyst Oxygen (Lambda) Sensor Correction; Bank 2: Too Lean
P2098 - 002 - Lower Limit Exceeded
Freeze Frame:
Fault Status: 01100010
Fault Priority: 0
Fault Frequency: 1
Mileage: 86912 km
Time Indication: 0
Freeze Frame:
RPM: 2608 /min
Load: 24.3 %
Speed: 68.0 km/h
Temperature: 93.0°C
Temperature: 18.0°C
Absolute Pres.: 1000.0 mbar
Voltage: 13.843 V
[/quote]
[quote=Mickf29]Yes, which is what I was trying ot figure out yesterday…
Why is my target lambda at 1.0 which would be AFR of 14.7? Why would JHM set it up to be like this? What do people see with a stock tune? When I ran on dyno my AFR was never 14.7, it was 13.1, with this same tune.
[/quote]
[QUOTE=beemercer]x-axis is engine speed, correct?
it looks like it stays between idle an 4700?
was the car under load during this? was it WOT during the engine speed blips?
iirc the factory tune always aims for 1.0 until it cannot match requested load at that afr. it uses egts or a modeled egt (on wideband cars) and knock, probably other parameters, to ensure the motor is running safely. It increases throttle angle, boost and timing to meet the requested load (how far you have the pedal pressed). It dials back timing and afr when the egts get out of control.
it aims for 1.0 to meet emissions requirements iirc.
[/quote]
[quote=Mickf29]yes it is engine speed…
You will notice the spikes happen when load goes to nothing, and the RPM tails. Also you will notice the specified does dip on the second rpm ramp, which was wide open for a short period, did not run to max rpm. So you can see it is responding to the change in load and dialing back from 1.0 to correct the lambda actual spike.
BTW the time span is about 45 seconds.
[/quote]
[quote=beemercer]to me it looks like
the lean spikes are happening in conjunction with the throttle plate closing after you rev the motor. Engine vac should be highest when the motor is revving down with the throttle plate closed, maybe its a small vac leak that gets stronger when the engine is under the strongest vacuum.
The richening after the 2nd blip(desired lambda) may be in part a result of the tune looking to cool the cylinders after going WOT (my car does this), in addition to reacting to the lean condition.
it looks like a hardware issue to me, the tune is doing what its supposed to (lambda specified = 1.0, richening when needed such as high egt, knock, lean condition), but the car can’t match it.
from bosch:
The overview LAMSOLL shows the formation of the nominal values for Lambda in the combustion chamber lamsbg w, lamsbg2 w and
Lambda at the sensor lamsons w, lamsons2 w. A nominal Lambda limited lamsbg is calculated for the advance control of a
combustion chamber Lambda from the different absolute Lambda requests lamfa w, lamnswl w, lambts w lamkh w by a minimum selection
as well as the additive delta Lambda requests in the function Lambda coordination LAMKO. The limitation is set by the running
limit of the engine for rich (lalgf) and lean (lalgm) mixtures in the respective running state.
The incorporation of the nominal lambda value lamsbg in the computation is shown in the function Mixture Control GK.
All interventions for Lambda which are not equal to Lambda 1.0 are realized by means of this intervention.
Lambda post start warm up lamnswl is active for the start and during engine warm up.
This is 1.0 at the end of the engine warm up cycle.
The nominal requirement for the Lambda control is lamsons = lamsbg. Regulation is to this nominal Lambda if continuous Lambda
control is present, and even if Lambda is not equal to 1.0.
If secondary air is injected during the warm up, then the transition is made by the factor flamkh from lamnswl w to lamkh w. The
return to lamnswl w is made following termination of secondary air injection. During this time, lamsbg and lamsons are
different because the Lambda sensor detects “thinner” exhaust emissions due to the secondary air.
For a large charge and high speeds, the engine must be run with a Lambda < 1.0 for thermal reasons. The intervention foreseen for
this is lambts w and is generated in the function LAMBTS ( Lambda component protection ). lambts w is an absolute Lambda, i.e
Lambda < 1.0 leads to enrichment.
An additive enrichment is realized by means of dlambts w (Function LAMBTS) as a function of the change in the ignition-timing
efficiency so as not to give any thermal damage from exhaust emission temperatures which are high and attributable to poor
gasoline quality and knock combustion resulting from this.
This additive Lambda nominal shift is taken into the computation with correct observance of the sign, i.e. negative values
lead to a more enriched mixture.
A further enrichment additive dlamatr w,dlamatr2 w is also realized for a given exhaust emission temperature control ATR by the
exhaust emission temperature controller.
By the function LAMFAW, an additional engine torque request can be realized by the driver by a Lambda < 1.0.
Switchover is made to the fixed value LASOAB at the respective bank which is independent of the nominal Lambda for Ev cutoff’s by
the function AEVAB.
A nominal Lambda > 1.0 is given by this fixed value to ensure with certainty that no overheating of the catalyzer occurs.
what does your timing correction at idle look like? when my S4 had a tiny (did not throw a dtc) vac leak that was killing performance(20% drop in rpm/sec) my timing angle at idle was retarded a few degrees.
have you logged under load? this looks like idle and free revving. it would be interesting to see how much quicker the tune richens up wot under load.
have you tried swapping the o2 sensor bank to bank to see if the lean codes move with the sensor?
[/quote]
thoughts? I’m not an expert in this, just started reading through the big Bosch ME manaul
