M383
M383
(*)Lambdakennfeld bei Teillast (KFLF) ---> (08848) ---> Lambda map at partial load
Map for lambda under partial load conditions. Note that KFLF's description as
"lambda" is a misnomer. It is not lambda at all. It is an injector on time scaling
table where larger numbers are richer, smaller numbers are leaner
You should not use KFLF, it has to do with fuel pulsations and other things. See
this note from Bosch here...
"The KFLF map should not be used for intervening in the mixture, because the
relative air charge rl must be balanced to the engine
demand via the pulsation map KFPU. If rl is corrected by KFPU faktor fgru has to be
1.0
The two interventions vsfrk and FRKAP are multiplying interventions via application
devices. FRKAP is changed by VS100 adjustment;
vsfrk by VS20 or via the serial interface. These two interventions must be set to
neutral during definition after start of
production"
So you should use LAMFA, LAMBTS or LAMKR. There is a good thread about LAMFA,
LAMBTS and LAMKR already.
Remarks: the map KFMIOP is used in the funktionsrahmen module MDMAX 1.40 with
inputs of engine
speed and maximum allowed cylinder charge to calculate the maximum allowed indexed
torque mimax_w. It
is also used in MDBAS 8.30 with inputs of engine speed and relative cylinder charge
to calculate the
optimum torque variable �mioptl1_w� at lambda = 1. The optimum torque is corrected
for the influence of
lambda by multiplying by the lambda coefficient variable �etalab�. The lambda
efficiency is obtained from the
characteristic line ETALAM. The basic torque variable �mibas� is obtained by
multiplying by the ignition angle.
This corresponds to the indexed torque produced if the combustion takes place with
the basic lambda
variable �lambas� and the basic ignition angle variable �zwbas�. There is no
difference between the BAM &
BFV maps for optimum engine torque. It is stated in module MDFUE 8.50 Application
Notes that the map
KFMIRL is the inverse of the map KFMIOP in module MDBAS. Therefore, any changes
made to
KFMIOP should be reflected in KFMIRL and vice versa. If the difference is too
large, problematic
operation/torque intervention will be experienced. It is accepted that this can be
achieved by scaling the
load axis instead of altering individual table addresses. Inverse in this sense is
assumed to mean
�complementary�, not the exact arithmetic inverse
(*)LDR-Sollwertbegrenzung (LDSMXN) ---> (093D4) ---> Maximum Load for Boost Control
LDRXN - maximum specified load / Maximum cylinder charge / maximum cylinder charge
for charge pressure control
Specifically, on a full throttle pull, your boost profile will follow LDRXN. K03s
and K04s have some severe flow limitations, so unlike big turbos, you will want
your boost to taper (not ramp up) to redline. ECUxPlot has a pressure ratio/flow
plotter that you can use to compare against your turbo's compressor map.
Remarks for LDRXN: these maps contain the most significant differences between the
two states of tune
(BAM & BFV) of any of the primary maps discussed so it seems reasonable to assume
that adjusting these
profiles will bring about the most noticeable effect. Note that the peak in LDRXN
is around 5500 to 5700 rpm
which corresponds to the point at which peak power is achieved in the BAM and BFV
engines. LDRXN can
be set to near maximum by some professional tuners as a means of sacrificing some
safety margin in the
pursuit of higher output. Although this represents the short-circuit of a safety
feature, it is perhaps deemed
sufficient for only timing to be retarded during knock rather than both timing to
be retarded and charge
pressure to be reduced. However, considering that the output profile I logged on my
car using lower octane
fuel than the remap was designed for is actually close to stock (peaking at only
234 bhp), it appears likely
that a reasonable safety margin has been retained.
The LDRLMX module calculates the allowed maximum cylinder charge. In the main path,
the engine speed
dependent maximum cylinder charge is determined by the characteristic LDRXN
(Maximalf�llung LDR,
maximum cylinder charge for charge pressure control). There is a similar maximum
specified charge profile
during continuous knock: LDRXNZK (Maximalf�llung LDR bei Dauerklopfen)
Remarks for LDRXNZK: It is noted in module LDRLMX 3.100 that the values should be
set about 15% lower
than LDRXN which would be reasonable from an intuitive understanding; i.e. overall
ignition advance is
reduced during knock so maximum allowable cylinder charge should also be reduced.
However, in these
examples, the values in LDRXNZK are actually higher from around 2,300 to between
5,000 and 5,500 rpm.
The only plausible explanation is that a slightly larger cylinder charge is allowed
to counteract the suboptimal peak cylinder pressure and consequential loss of
torque that would result from sub-optimal ignition
timing due to knock. Otherwise, the driver would feel a noticeable step-change in
power during the onset of
knock. If in doubt, don�t try to second-guess the OEM calibrator; follow the advice
in the funktionsrahmen and
set the values in LDRXNZK a bit lower than LDRXN.
KFMLDMN
(*)TV - Kennfeld f�r LDR (KFLDTV.0 / KFLDTV.1) ---> (09144 / 091D4) ---> Map for
linearization of boost pressure / Duty Cycle Open Loop Lineariz Boost Map
KFLDRL - Map for linearization of boost pressure = f(TV). This is the post-PID
waste-gate duty correction table. Calibrating this correctly is time consuming, but
worth the effort.
KFLDRL can also be used to get open-loop type behavior for operation past the MAP
and requested boost limit by making the output duty cycle unresponsive (flat) to
uncorrected duty cycle (from the PID) at various RPM/DC points. Again, if you do
this, make sure to leave DSLOFS at the stock value. This way, requested boost
will always be higher than measured boost, and you will stay in open loop control.
One undesirable side effect to leaving DSLOFS stock (and requesting more boost than
the MAP can read) is that the ECU will continuously trim the I-limiter upwards to
try to get actual to meet requested. This means you cannot rely on KFLDIMX to limit
WGDC; you MUST limit duty via KFLDRL, or numb the positive I adaptation.
Aftermarket/external wastegates
With after market or external waste-gates, flat line the map at something like 25%
until your turbo should be spooled. Then taper off to 10% at areas of higher load
and RPM. Log requested vs. actual boost to see where you need to adjust KFLDRL to
line things up. The higher your after market waste-gate spring pre-load the
lower the WGDC you will need to accurately control boost. You may need to numb
KFLDRL response a bit since very tight wastegates need very little DC change to
make a big difference in boost. Or, if you have time, do it right, and use CWMDAPP
= 8 along with KFLDRAPP to calibrate it correctly.
KFLDRL is you duty cycle correction map, so i guess that can help smooth out some
spots
KFLDRL (Kennfeld zur Linearisierung Ladedruck = f(TV) (map for linearization of
charge pressure = function
of duty cycle). Units (x,y,z): % charge, RPM, % charge
The S4 wiki comments that KFLDRL can be used to get open-loop type behavior for
operation past the MAP
and boost limit by making the output duty cycle unresponsive (flat) to uncorrected
duty cycle (from the PID)
at various engine speeds and/or duty cycle values.
There are basically two distinct operating modes:
1. Quasi steady-state operation with PI control which gives a relatively weak
control action. Derivation of the
control parameters is carried out using an engine dynamometer test according to the
Ziegler-Nichols tuning
method.
2. Dynamic performance with PID control which gives a strong control action.
Derivation of the control
parameters is carried out using the transient oscillation method using an engine
dynamometer.
These operating states are distinguished via the MV/DV error, i.e. above a positive
deviation threshold, the
dynamic control action is activated and only stopped at the change of sign of the
deviation (DV > MV). The
S4 wiki notes that if actual boost is not meeting target boost, the PID integral
limit between 2200 and 5000
RPM for 850 and 1000 mbar might need to be increased.
The charge pressure profile �pvdkds� as a function of duty cycle is determined on
an engine dynamometer.
This procedure is performed starting at 1500 RPM in 500 RPM steps to the maximum
engine speed. The
necessary linearization values at any given speed are subsequently determined
graphically (or calculated) as
follows. In the diagram 'pvdkds' which is a function of 'ldtvm', the first
measuring point (0%) and the last
measuring point (max. 95%) lie on a straight line. Then, at 10% duty cycle for
instance, the corresponding
pressure value is determined from the line and then the pressure associated with
this value from the
measurement curve �ldtvm�. This 'ldtvm' value is now addressed into the map KFLDRL
at the corresponding
interpolation point (here, 10%) �
And what about FKKVS - Correction factor for fuel supply system(from me 7.5)?
I can't find it in me 3.8.3
I'm going to replace my 260cc(AGU) injectors with 317cc(AUQ).
I've changed
1. Fueling constant: Grundanpassungsfaktor 0
2. Start fuel inj time(ms): Startgrundmenge (2 1x1 maps) - copied from AUQ
3. Spannungskorrektur - copied from AUQ
And should I change Spannungskorrektur if I'll just install 4bar fpr with 260cc?
Can't find specs for these injectors(bosch 0280150464) wth 4bar pressure.
We recently changed the injectors from 210cc to 315cc and I am not happy with the
way the fuelling is at the moment.
I did reduce the maps for close throttle and part throttle but at WOT it goes all
the way down to 9.8.
At cruise it's at 14.8 and idle at 17
how you calibrate injectors? there are 2 1x1 maps to calibrate injectors, one
multiplier and one for load
Change only the fueling constant 1x1
Also change fueling(ms) for engine start.
Fueling constant: Grundanpassungsfaktor 0
Start fuel inj time(ms): Startgrundmenge (2 1x1 maps)
The constant differs on M3.2 from the M3.8, the damos from M3.8 cannot help on that
constant.
I have found only the multiplier and it is not in the same same place as M3.8
If you calibrate injectors correct you need not touch Startgrundmenge. Last week I
calibrate deka 630cc @ 4bar for a Precision setup and I need not touch
Startgrundmenge, engine starts perfect at cold or hot
The fuel trims block and the lambda controler on VAGcom will be your best friend to
fix perfectly injectors
Which one is the injector multiplier ? Grundanpassungsfaktor 0 at 0x7507 or
Grundanpassungskonstante at 0x7536 ?
and which one is the injector deadtime ?
I have a stock 1.8T@180hp with me3.8.2 with the values :
0x7507 = 128
0x7536 = 127
---------------------
So if I install Siemens 630cc, what values shoud I make these approximately ?
Any help will be greatly appreciated.. I've been working on this for 1 week, and
since I can't flash the ecu instantly without taking it out and putting it in, it
takes LOTS of time..
On 3.8x ecus I change Grundanpassungskonstante. Always worked fine. Never had to
touch anything else.
For 3.2 it's a different story though.
It a ball braker to scale/graft a 2.75" or bigger maf into the m3.8 AGU code, I
have tried 100s of combintaions over the last 7 year, from adding X% to the full
maf scale, copy and past oem maps, interpolating % scale like -7% to +32% ect...
All my methods and trick have worked and kept me and my customers happy.
BUT... The truth is the best way and the best results come from off setting the
map, fitting bigger injectors and basing the maf size on the same or simlar %
increase, you will need to set minTE and start injection to set a base line then
the tweak the 16x16 lambda at part load as a whole % based on fuel trims. But now
you have load maps wrong and timing load axis out of sync, all you need to do is
write a new timing advance map from data based on VCDS logs and det cans. Then last
tune the 9x16 LDR load maps as you would at any other stage of tune to get a base
line map. Once you got your basic fuel map done, timing and load you just need to
fine tune fuel via Lambdakennfeld bei Teillast and timing advance maps.
By going from a stock maf to 2.75" with unscaled MAF Linearisierung the timing will
end up around 5* to 7* advanced.
Nick
Maybe not the correct way in theory, but it works 100% and produces much smoother
results. I have tuned this way From k03s 310cc 240bhp/ K04 hybrid 330bhp 440cc/ to
T304E Stage3 630cc 430bhp.
Just a lot of logging and in most case I keep all load axis stock and edit the
maps.
The best way to start would be fit MAF and injectors set TVUB based on Bosch data
sheet and reduce min start injection 1x1s. This is enough to get the car starting
hot and cold without soiling your spark plugs/flooding engine. Then work on part
throttle map as a whole % rescale to get your idle fuel trims about right. With n75
unplugged drive car on light load logging via VDCS and edit points on part load
lambda map unti you get close on to the fuel stratergy you after. Reduce timing
maps over 2000rpm by around 5*.
Then run car with the stock N75 and load maps and log to get a picture of whats
going on and make sure the ecu has control of the base line boost and its safe.
Then start by tuning the LDR map as a whole % you might have to take out some
midrange load ms to make the power linear for the bigger turbo as the k03 maps as
for max load around 3200rpm then tails off. Set you self a basic torque curve at
lowest boost u can get (hard without dyno) but on the road just set it based on
boost gauge reading and blocks 025 via vcds.
Once you got torque curve/boost profile contolled how you want as basic power then
the LDR maps become very easy to alter/tune for final power. Now is just a
matter log winding the timing up and logging load, actual timing and timing
corrections to get a base maps which is safe and controlled.
Now its just the case of adding small % gains to the LDR maps and working up in
load to target torque curve., while logging and making corrections to lambda and
timing maps which will relate to actual load ms logs.
It takes a lot of time to get right, I tend to spend a 2-3hours a day over a week
to do a big turbo map for an AGU, I always start from scratch using a stock factory
map. I aint the best person to explain things so I hope you can get your head
around it.