Sign in

View The Full Version : Bosch MED9.1


Exelion1986
hi guys, I open this discussion with the objective of getting to know and learn the Bosch MED9.1
I know it's not ? easy thing, but ? gi? by a p? that I am studying and I would like to begin to share with you that for now I understand, and not ? said that what am thinking is correct
I appeal to the experts, maybe for some straight and confirmation, both beginners who would like to explore the topic and maybe they are available to any test
in particular, I would like to be able, in small steps, to map my car, fitted with engine 2.0 TFSI

first of all, I am attaching the file that I am studying, I would say to refer to this during the discussion, I do not know? which car belongs



in fact, the MED9.1 inherits from ME7 most of the logical and of the names of the maps that user?, a big difference ? the management of the direct-injection, and then the high-pressure pump and of the corresponding rail
unfortunately, not knowing the ME7 I had to start from scratch in this study

let's start with the maps pedal
in fact, if you do not want to vary the response of the pedal should not be touched

1CAD50 KFPED
map pedal true, 2 similar
16-bit
size 12x16
x = value * 0,25 = g/min
y = value * 0,001526 = % pedal position
z = value * 0,003052 = % torque request
1CB652 breakpoint x
1CB66C breakpoint y

1CB050 KFPEDL
map pedal for low speed, probably for low gears, 3 related
the same conversions and the breakpoint of the previous

1CB4D0 KFPEDR
map the pedal for reverse
the same conversions and the breakpoint of the previous

now, let's see the maps related to the pressure of the turbocharger
these ECU's did not have the management of the turbo as a function of pressure, but are based on the specific load in %
this parameter represents the ratio between the mass of air in the cylinder and that it would be contained in the same volume in the environmental conditions (the cio? with density? environment)
then a specific load of 200% means that in the cylinder ? request (or ? present) a mass of air equal to the double of the one that would contain the cylinder itself
this formula converts the specific load pressure of the turbo

pressure = (load %) * 10 + 300 = mbar absolute

? note that this formula is an approximation and subject to the conditions of operation may not provide a real value

the maps in question are

1CBAF2 KFMIRL
map the desired load, converts the request torque data from the pedal in load demand, specific
16-bit
size 12x16
x = value * 0,001526 = % torque with respect to the maximum
y = value * 0,25 = g/min
z = value * 0,023438 = load % specific

1CBCAE KFMIRLS
variant of the previous operating in the stratified charge, it is not used and may not be present

1CC3D4 KFMIOP
converts the specific load in % of torque, practically does the job inverse of the KFMIRL
this value is used for other functions of the ECU that they have to think and/or act on the pair, such as the ESP, the DSG etc.
? this to save computational work
16-bit
size 11x16
x = value * 0,023438 = % load air specific
y = value * 0,25 = g/min
z = value * 0,001526 = % torque with respect to the maximum

1C959C KFMIOPS
should be the same as the address of the KFMIRLS

if we modify the KFMIRL ? of obligation to reflect changes in KFMIOP, worth the chance? interventions do not correct of the systems based on the couple
in other words, the KFMIOP must be the inverse of the KFMIRL, at least in the area of high
also ? needless to take a specific load in the KFMIRL beyond a certain threshold, why? the axis of the KFMIOP the limit to its maximum value
so if you want to push the map over an increase in the basis you need to change the axes of the above, but there? introduces complications since? there are other maps of important use to the axis, the specific load and it is necessary to intervene in those resizing and reinterpolando values (we will discuss later)

now let's see the maps that could be limiting factors for the specific load, or pressure, or that may give errors

1C4696 KFMLDMX
the maximum flow mass air flow sensor for diagnosis, it may give an error, only change in this case
16-bit +/-
size 9x8
x = value * 40 = g/min
y = value * 0,024414 = % butterfly
z = value * 0,1 = kg/h

1CD2B6 KLDLUL
the maximum deviation of the pressure turbo with respect to the request for diagnosis, to the max in order to avoid errors
8-bit
size 8x1
x = value * 10 = hPa
z = value * 5 = hPa
breakpoint in the normal position, but the x +/-

1CD538 KFTARX KFTARXB
correction specific load as a function of the temperature of the inlet air, 2 consecutive
16-bit
size 10x12
x = a * 0.75 - 48 = ?C air intake
y = value * 0,25 = g/min
z = value * 0,000031 = correction factor
1CD9C5 breakpoint x 8-bit
1C9480 breakpoint y in 16-bit

1CD718 KFTARXZK
variant of the previous one during the detonation of the engine
the same conversions and breakpoint

these three maps can be repeated with similar alternate in the same order

1CD858 LDORXN
the maximum specific load during an overboost error
8-bit
size 8x1
x = value * 40 = g/min
z = value * 0,75 = % specific load maximum
1C9363 breakpoint x 8-bit

1CD860 LDPBN
limiter pressure turbo
8-bit
size 8x1
x = value * 40 = g/min
z = value * 10 = hPa pressure max turbo
the same breakpoint of the previous

1CD87A KFFKRXTM
limiter-specific load engine temperature
8-bit
size 4x4
x = value * 1,25 = km/h
y = a * 0.75 - 48 = ?C
z = value * 0,007813 = correction factor specific load

1CD89C KFLDHBN
the maximum pressure ratio of the compressor
is used to avoid overloading the turbo in some conditions
depends on the map of operation of the compressor
8-bit
size 8x8
x = a * 0.75 - 48 = ?C inlet air temperature engine
y = value * 40 = g/min
z = value * 0,015625 = pressure ratio turbo

1CD8FE LDRXN
limiter specific load
16-bit
size 16x1
x = value * 0,25 = g/min
z = value * 0,023438 = load % specific

1CD940 LDRXNZK
variant of the previous detonation
the same conversions

these two can be repeated with similar, not sure if the alternate or no
however, it would seem that only the first variant is used by the ECU

for basic changes, in fact, it would be enough to just increase the LDRXN as the KFMIRL requires gi? a load more? the top, and the other described if necessary

this ? pi? whether or not the theory
with regard to the practice, as soon as I have time, I try a couple of these changes, accompanied log diagnosis, with the aim of understanding the influence on the turbo pressure of these maps and to understand the factors that may limit

the next episode ? theory of operation the PID of the wastegate for better control of the profile of the pressure trend
cawadany
Mooltoooo interesting! I have a TDI and I would like to better understand how it works, the management of the MED9.1. I follow with interest!
frantik3
very good really
Exelion1986
in annex a guide for doing a log with vag com



also useful for those who use the tools of diagnosis different
describes the parameters that must normally be kept an eye on
what is pi? important to find a complete list of the blocks of the measurement
frantik3
great guys and great forum you these things I like very much
Exelion1986
then today I made a first test
first of all, the parameter in the diagnosis called engine load (engine load) ? the specific load
for various problems are not able to make a log for? I have seen that in the load. arrived at 178 from the original, and in my file the LDRXN limits to 175
in KFMIRL to the maximum ? required 200, and then I set the LDRXN to the same value as climbing a p?, also for safety, I increased a p? the LDPBN
gi? cos? you feel a sharp improvement, I need to see what is the value of the specific load is cut, in theory, 191%, even if requested 200 to the axes of the KFMIOP
however, after writing the ecu has happened a couple of strange things
the fuel gauge ? went to zero, and then when I tried on the road slowly ? back to the correct value
also, especially when I do the maneuver, I hear a strange noise coming from the steering wheel, as if something grattasse
maybe nothing to do with it, for? I parked the car in the garage which he did not, I mapped it, and pull it out it started...
any idea?
Exelion1986
small update
unfortunately in this period I had little time, also I just can not operate the flight recorder of the diagnosis, so no log for the moment
at least I bought a car charger for the laptop cos? I see a p? the trend of the values
then first of all no matter how much load you ask in the map, the axis of the KFMIOP limited to its maximum value, in my case 180%
in an overboost of course, get a p? pi? high
I have prepared another version by changing the axis to get to 200%, today the load and see what happens as there are other maps with the axis, the load
for now I try not to change them, I want to understand if limited or no
in particular, we have:

1C7DBF KFZW
map advance
8-bit +/-
size 12x16
x = value * 0,023438 = load % specific
y = value * 0,25 = g/min
z = value * 0,75 = ? advance
1C953C breakpoint x 16-bit
1C94DE breakpoint y 16-bit
are more? a, the first met? are variants, and the second related to the first according to the variable valve timing

1C7FFF KFZWLB
map advance with flap closed suction
the same conversions and breakpoint
it is the same as on the variations and the like

1CC534 KFZWOP
map advance for comparison
need to determine how much load is used to generate the required torque on the basis of the advance
in theory, should not be changed, but only reinterpolata if you change the axis to the KFMIOP
it has two variants
the same conversions of the previous
1CC7F6 breakpoint x 16-bit
1C94BC breakpoint y 16-bit

in my case, the maps advance up to 190% of the load to the axis, while those for the comparison of 180%
I want to know if these I limit the load
Exelion1986
another test done
now the specific load stops 186-188%, so I think there is a map that I am limiting myself, and not the axis of the advances
I have to investigate better, maybe try to increase the maximum pressure ratio of the turbo
anyway, I'm attaching my golds and a couple of mods done up to now if you want to give us a look

frantik3
good study continues cos?!!!!
frantik3
you can get the maf?
alxcrs
Hello Exelion, bravo, really good job....
Exelion1986
thanks to all slowly, slowly, I hope I get to get a good complete map

*frantik3 what do you mean by picking up the maf?

however, the new episode: management of the performance of the turbo pressure
there are several maps to manage the performance of the turbo, or better than the PID control of the wastegate
some are for specific functions, other than the limitations of the maximum values, and still others linearization...

1CD9D2 KFLDIMX
the maximum function of the PID of the wastegate
16-bit
size 8x16
x = value * 0,039063 = mbar difference between turbo pressure request and the current
y = value * 0,25 = g/min
z = value * 0,005 = % maximum function
1CDEDE breakpoint x 16-bit
1C949A breakpoint y 16-bit

1CDAD2 KFLDRL
linearization of the previous map
in theory, that does not change if you change the wastegate
16-bit +/-
size 10x16
x = value * 0,005 = % function
y = value * 0,25 = g/min
z = value * 0,005 = % correct
1CDEF0 breakpoint x 16-bit
breakpoint y the same as the previous

1CDC12 KFLDRQ0
function P of the PID
16-bit
size 4x16
x = value * 10 = mbar difference between turbo pressure request and the current
y = value * 40 = g/min
z = value * 0,05 = % function
1C9349 breakpoint x 8-bit
1C93A7 breakpoint y 8-bit

1CDC92 KFLDRQ1
function of the PID
all the same as the previous

1CDD12 KFLDRQ2
function D of the PID
all the same as the previous

1CDDF2 LDRQ1ST
function of the PID scheme
16-bit
size 16x1
x = value * 40 = g/min
z = value * 0,05 = % function
breakpoint x the same of the previous

1CDE12 LDIATA
the correction function of the PID as a function of the temperature of the inlet air
must reflect any changes to the KFTARX
8-bit +/-
size 8x1
x = a * 0.75 - 48 = ?C intake air temperature
z = value * 0,64 = % correction
1C93EC breakpoint x

1CDEC8 LDRQ0S
function P of the PID scheme
16-bit
size 1x1
z = value * 0,05 = % function

these should be all the maps related
really? I have only used two, I don't know if my approach ? theoretically correct, for? works
first of all, there are two modes of operation depending on the turbo pressure, the cio? when the actual pressure ? lower than required and when ? top
in the first case, we have the contribution of all the 3 functions of the PID, in the second, the function D is turned off
I have only used the KFLDIMX (maximum function) and the KFLDRQ2 (function (D)
in particular, avoid spikes too high of pressure you need to decrease the last columns of the first and increase the second
to make it more? simple as possible I copied the penultimate column in the last two of the KFLDIMX and copied to the second to the last in the last of the KFLDRQ2
trying various combinations of the kind you should get to the desired result, cos? as I have done, it seems to me that the pressure never goes beyond the request
I have yet to verify good performance and then if you do? still some testing to understand the influence of the changes
for further information I recommend going to read some explanation on the PID in general
Exelion1986
after a bit of absence due to various commitments, we continue with new maps, this time related to the rail pressure
having the high pressure pump changed, my goal was to get to 130 bar against the 110 original
by modifying these maps (not all, of the list for completeness) I verified that the pressure rises to the desired value and remains stable

1D097E KLPROV
rail pressure limiter
16-bit
size 8x1
x = value * 0,25 = g/min
z = value * 0,0005 = pressure MPa

1D5DE6 RQUAHDPK
radius of the high pressure pump to the square
original diameter 2.0 TFSI 8 mm
original diameter 2.5 TFSI 10 mm
autotech/kmd 9.8 mm
apr 9.5 mm
16-bit
size 1x1
z = value * 0,001 = mm2

1D5DE8 SKHDPMX
stroke of the high pressure pump
stroke 2.0 normal 5 mm
racing engines CDL 5.4 mm
run 2.5 TFSI 4.5 mm
16-bit
size 1x1
z = value * 0,001 = mm

1D617E PGBDVHDO
pressure opening the valve of the return of the rail
16-bit
size 1x1
z = value * 0,0005 = MPa

1D6254 PRNL1
maximum rail pressure (I think for diagnosis)
16-bit
size 1x1
z = value * 0,0005 = MPa

1D6328 KFPRSOLHKS
rail pressure during the beating operation in the homogeneous
16-bit
size 8x8
x = value * 0,001526 = % couple
y = value * 0,25 = g/min
z = value * 0,0005 = MPa rail pressure

1D63A8 KFPRSOLHMM
rail pressure in the homogeneous lean operation
16-bit
size 8x8
x = value * 0,001526 = % couple
y = value * 0,25 = g/min
z = value * 0,0005 = MPa rail pressure

1D6428 KFPRSOLHOM
pressure in the rail operation in the homogeneous
16-bit
size 8x8
x = value * 0,001526 = % couple
y = value * 0,25 = g/min
z = value * 0,0005 = MPa rail pressure

1D64A8 KFPRSOLKH
rail pressure during catalyst overheating
16-bit
size 8x8
x = value * 0,001526 = % couple
y = value * 0,25 = g/min
z = value * 0,0005 = MPa rail pressure

1D6528 KFPRSOLOFF
offset rail pressure during shifting operation in the homogeneous
usually post all to 0
16-bit
size 8x8
x = value * 0,001526 = % couple
y = value * 0,25 = g/min
z = value * 0,0005 = MPa offset rail pressure

1D65A8 KFPRSOLSCH
rail pressure during shifting
16-bit
size 8x8
x = value * 0,001526 = % couple
y = value * 0,25 = g/min
z = value * 0,0005 = MPa offset rail pressure

the breakpoints of all of these pressure maps request are shared and located immediately after the last of them

1D6636 KLLFPRSG
volumetric efficiency of the high pressure pump
16-bit
size 6x1
x = value * 0,0005 = MPa rail pressure
z = value * 0,000061 = coefficient, volumetric efficiency

1D664A KLPRMAX
limiter rail pressure for the engine temperature
16-bit
size 6x1
x = a * 0.75 - 48 = °C engine temperature
z = value * 0,0005 = MPa rail pressure
breakpoint just before as normal, but x in 8-bit

1D66A0 VHDP
the volume of the high pressure pump
calculated with the data of the radius and the race that I listed before
16-bit
size 1x1
z = value * 0,01 = mm3
frantik3
very good really
ferchu2323
very good info
tomatos666
thanks so much!
munro
wonderful guide, however, you skipped the part about the "supply"
Exelion1986
wonderful guide, however, you skipped the part about the "supply"

thanks
oh I know the guide is not finished yet
the portion of the fuel mixture is right the next, I'm still studying it, and then will arrive
munro
if you can be of help...
always relative to the file studio posted by exellion we have maps related to "supply"
fuel.we find then:

1D43F4 driver wish lambda
LAMFA
15x6 8-bit
conversion factor
0,007813
Y-axis rpm x 40,000000
8-bit
x-axis % pedal relative to the driver's wish KFPED
16-bit x 0,003052
this map from the name determines the lambda desired of the driver who requires through
the pedal in the ecu,even if the map is small enough,behind it there is an algorithm
very big and complex for the wide choice of afr.I would like to remember that the me7/med9 work constant
mind in cloosed loop so setting this correctly map the ecu autoadatta to give
that afr.this is just one of the many ways that you can use it to give more fuel via the ecu
to the engine.

1D2852 lambda component protection
KFLBTS
12x16 8-bit
conversion factor
0,007813
y-axis rpm x 40,000000
8-bit
x-axis % relative to the fill the air
8-bit x 0,750000

1D2912 lambda component protection with the swirl manifold open
KFLBTSLBKO
12x16 8-bit
conversion 0,007813
y-axis rpm x 40,000000
8-bit
x-axis % relative to the rimepimento air
8-bit x 0,750000
these 2 maps that titanium called limiters, lambda, A and B, in reality, do not limit anything in fact, from the name
the same "take" through a trigger called TABGSS in the ME7 otherwise in the MED9(more forward
I'll write the address) that is the egt computed or real(depends if there is a egt probe or not in the ecu)which
precisely does "take" and work these 2 maps to enrich the afr and avoid "sciolgiere" everything.
it is worth adding that when this condition occurs the ecu in addition to the enrichment also decreases
the timing of the ignition since when persists high temperature of the exhaust gas.
small clarification, the model egt systems with egt calculated is strongly affected for example
if you replace the downpipe with one more "permeable" and then it may happen that the egt computed are
the most high of the real ones, then it would be appropriate to review both the trigger that these 2 maps.

1D27B9 factor delta protection component
KFFDLBTS
12x16 8-bit
conversion x 0,007813
y-axis rpm x 40,000000
8-bit
x-axis % relative to the fill the air
8-bit x 0,750000
this map is a multiplication factor of the calculation algorithm that the ecu adopts contemporary with
the 2 maps above for a given afr, not yet, I was not very clear when you change what you do should be a
further enrichment but there would be to do testing and logs some say that disabling the maps
BTS main through this you can increase the wide choice in collaboration with the other 2 maps when the
the engine and on the threshold of knock to ignition..


1C474C the map correction function MAF
KFKHFM
14x14 8-bit
conversion x 0,007813
y-axis rpm x 40,000000
8-bit
x-axis % relative to the fill the air
8-bit x 0,75000
this map driver titanium defines it as "injection base map" in reality the name is only part
of that big algorithm that I said and the map conversion and calibration function of the maf.
the map injection base is another,this map, according to me, should not be touched for changes
standard you should touch only
for example, when you replace the maf with a bigger and linearize the same correctly in the ecu
however, it is one of the many ways that you can give it gas but here you go, to vary the trim of
adaptation this fact can be changed, with care to fit the correct percentages
adaptation trim fuel.

1C7AB8 correction Factor of the power system alias, map the injection base
FKKVS
conversion 0,000031
8x8 16-bit
y-axis-efficiency injection time
16-bit x 0,001000
x-axis rpm
16bit x 0,250000
this is the map injection time main I do not hack ever, according to me, should be touched when you replace
the injectors in series with the other more big.




for the moment I will stop here but there are really a caterba maps and algorithms that are cell explain in this ecu
sum it all up in these few maps I think not..d altronte there is to say also that there are people to walk
these ecu's/cars touch more of a 50ina maps, and others have obtained the same result if not better by tapping to the maximum
2 or 3, so the gist is, are the ecu is very complex...I have more than three years that I'm studying but still not
I can understand the logic I would like to add that for the me7 for example, there are over 1800 pages of funktiionsrhamen from
put up to understand the above for the med9 pages of FR become 4000 and beyond as the various med17 are still
more complex.
is not a polemic and just a clarification...sorry.
isfedale82
from where you take all of this information
munro
from the study by the logs from the damos...
alxcrs
Congratulations on the studio... I'm following, really carefully...!!!

What I like of this world, and this is when you think you've got a little something, turns the corner and still find a highway....!!!!!!!!

I don't have a TDI available with which to do my tests, but I follow you wheel to wheel.
munro
exellion could you post a log with the times of injection??
munro
...sorry....
also other users if they have available to log injection time for a tfsi can post....
Romanoalessio
mmmmmmm but with the offset I am not well at least I think....