S-Prihadi

Guys,

I never expected this kind of positive result.
I replaced all four sensors last Thursday, here : https://mbworld.org/forums/e-class-w...-bank-2-a.html

Since my Banks Gauge can not interpret Bosch wide band aka Lambda aka AFR sensor correctly, I never log the data.
The rear switching type oxygen sensor is generic 0.1 to 0.9 volts, so reading it is easy and I been logging it.

Yesterday I went out for dinner and test the car while at it with a detour to my favorite test paid highway, which unfortunately has so much road repair ( corrupted contractor bas-tardz, under 2 years old road !!!! ),
So I can't run as fast as I wanted to, to also test the repaired boot of the propeller shaft , here : https://mbworld.org/forums/e-class-w...ml#post8835941
Just in case vibration exist at 100MPH+.

Along the way I realized there is an obvious improvement on the throttle response for casual driving, its like how my tranny improved its response to minute throttle modulation after replacing
its 10,000KM old oil but of 4 years old ( 2014 to 2018 ), some months after I bought the car in 2018.

I logged this test + dinner combo trip with Xentry 50% of it and 50% of it with my Banks Gauge, but I was not actually thinking of throttle modulation improvement,
I was actually wanting to see the Lambda sensor response only. Thus I did not add live data of the throttle body opening angle while in Xentry.
I can only use either one of these loggers/scanners, they can not be connected both at the same time to the OBD2 port.

I been spending hours to figure out where the actual improvement came from ?
My Lambda ( front ) sensors are in good health, but is of 9 years old and 38,000KM.
The rather lazy one is the rear LEFT or Bank 2 oxygen sensor which would not have any impact for performance, it is for emission and it always pass OBD Mode 6 test and never
ever produced DTC of any sort. I would slap myself silly if any AFR or oxygen sensor can produce DTC on my watch

...

I was not yet driving aggressive, hence that was the surprise. I was avoiding potential dumb-azz which most likely would "eat" my lane.
Below is the dash-cam capture and the Xentry , for the 8.8 seconds I speak of. Where there is a time code on the dash-cam video, that 8.8 seconds are what I am talking about.




.
.
So I took a deep look at these 8.8 seconds worth of data , but only the 1.5 seconds of it which is important for seeing what is actually happening.


The yellow box DOWNSHIFT is my own estimate , as there is no way RPM can change/rise that fast by near 1,000 to 1,500 RPM within 0.1 second without tranny downshifting.

Delay in seconds means the time it took from previous data point to current data point ( screen capture ). Example data point 0 (zero ) to data point 1 took 0.217 second.
Lambda values 1.0 = Stoich or AFR 14.7 for gasoline. Perfect. Below one means RICH, for acceleration. Above 1.000 means lean, usually when coasting or deceleration.
To get AFR( Air Fuel Ratio ) from Lambda, for gasoline engine is : Lambda * 14.7. Example data point 5, Lambda 0.910 x 14.7 = 13.377 AFR

Fuel injectors open duration, the longer the duration of opening, it means it is pumping out more fuel. However, as RPM increase, the time or window for fuel spray get shorter and shorter before TDC.

Engine load for turbocharged/supercharged engine is expressed as forced induction plays a part when engine load above a 100%. So in theory 143% means 43% is from forced induction of the turbo and
the 100% is from the engine, as in a normally aspirated engine. Some call this total load as absolute load.


AA.
Lambda OFF means fuel system is in Open Loop, meaning ECM does not take into account AFR sensor value. This is coasting or when engine/car in deceleration mode.
In my case I was coasting as seen by 15% engine load at data point/screen capture -2
I think by data point 1 and at 86% engine load, ECM starts to go into Close Loop and hence we can see later at data point 1 and 3, both the Lambda of LEFT and RIGHT bank is operational.
I do not know why the data point 3 which is Bank 1 or Right Bank is slower by 0.1 seconds to be ON for its Lambda compared to Bank 2 LEFT bank.

Also bear in mind that there will be slight lag in data transfer from ECM N3/10 to Gateway N93 at front SAM and into CAN D ( OBD2 port ) and into the C4 multiplexer and to the laptop via RJ45 Ethernet
cable and the Xentry software itself. The time stamp I use is from a 60FPS video, that means I only have 60 data points per second albeit the time code counter of the video editor I use has 1/1,000 speed
or 1,000 data point per second. Last known data speed of the Xentry I can confirm on my hardware is 7.8Hz for monitoring VVT operation and has CSV file as log.
No 3rd party scanner to date I own can exceed 1Hz live data in its MB software environment.
My Banks Gauge can do 20Hz or 20 data points per second , but is limited to OBD2 data only and will not work on proprietary data type of MB-Xentry .

BB.
I never knew that KNOCK monitoring is only turned ON by ECM when engine load is probably 50% or more, in my case I see 69%. I have always thought that it is a permanent monitoring.


CC.
What I mean by improvement in throttle response is obvious at data point 2 to 3 and 3 to 6.
When at data point 2 we see engine load at 102%, within 0.083 second later ( data point 3 ) the tranny already downshift and then downshift again at data point 6 and
the total time taken was data points : 4 + 5 + 6 or
0.084 + 0.150 + 0.050 = 0.284 second which is crazy impressive for this "slow" tranny.
I did not even full kick down to the final DOT/Click at the throttle pedal. This is why I said I was surprised.


CONCLUSION
My guess is, virgin new AFR reacts much faster than my so called healthy but 9 years old and 38,000KM tortured AFR.
Since AFR value is action and reaction ( closed loop ), this reaction of AFR is what ECM is waiting for as confirmation that fueling quantity is correct.
Hence I get smooth but fast increase of power with little throttle modulation.

Also since ECM is the King of the car drive system, tranny speed of gear shift will be dependent of ECM too for the go-ahead, not the mechanical or hydraulic side of the tranny which will benefit so much
from fresh approved oil and often replaced, as in 20,000KM or 2 years for me.



THE GOOD NEWS
This means I can expect even snappier response when I am done with the oil valve by-pass. That will create faster breathing command & execution

.
This improvement is not about brute power, its about smooth and fast state of change, where its a proof that I can maximize the engine intended design goodness when components or sensors
can be made to response as fast* ( being a virgin ) as possible.

And this will also mean, like fuel filter where I decided 25,000KM is as far as I would use it pressure drop related reason, and 20,000 KM for engine mounts + tranny mount for maintaining highest smoothness.
Therefor next on the list will be 2 of AFR and 2 of oxygen sensors limited probably to no more than 30,000KM or 6 years.
Blo-ody hell, good stuff do cost money....just like fine wine and very old 30+ years fine Scotland selected whisky, once you drink it you then know what good drink should taste like.

======================



Too bad this live data part of Xentry can't be logged as CSV file. Unlike the 60 seconds window I can log the VVT system test and at 7.8Hz.


The data point I marked -2 to +13 page by page, all below.
Red arrows with green dot means good/positive change.
Red arrows with BLUE dot means reduction of value

CaliBenzDriver

I am a huge fan of clean Lambda sensors. they drive the whole engine performance or lack there of with LEAN TRIMS... that cause greater engine heat. They deliver the ultimate old car feeling with unbalanced crankshaft vibrations.
Even L4 engines use two separate banks! What for ??
Whatever is necessary to get a lean engine to shake from misfires.

It's almost like you'd want to empty the oil out of the intake plenum.

How much oil gets stored in there, unable to go hang on the intake valves ?

S-Prihadi

My CAT honey comb is very dry, no oil.
No oil trace on them sensors too.
I think the M276 small secondary fine mist oil vapor separator really does a GOOD JOB.
This one : The long box shaped one.



Your M276 3.5NA uses the whole intake assy as also vapor separator right ?, my 3.0 does not, instead it uses this long box shaped 2nd fine mist oil separator and it is located LOWER than the intake manifold.
So by gravity alone and better swirling ( I am guessing ) , this does so much better.
I was surprised that my intake manifold was SO OIL-LESS and so was my throttle body when I did that intake valves carbon removal.
Unlike all 3.5NA I seen on youtube with oily mess when they remove that big & wide black intake manifold assy.