NefMoto

Reaching out to some of the more experienced guys... Has anyone played with EOI when switching to larger/different injectors?  It can be quite helpful on other platforms when dealing with non-stock injectors, especially large ones that might not have the same/similar atomization characteristics.

Map KFWEE (Kennfeld Winkel Einspritzende) I believe is the one.

MED9/GDI?

I've played with it, but there's not much angle left to spray without smoking. Have you fully advanced SOI?

i think VR means ME7 and you MED9 TFSI.
Don't think it will make a big differents with bigger injectors.
Maybe with crazy cams it will make some differents..

Johan

port injection in my focus yes....

It can make a big difference with very large injectors actually.

Can you tell us about the effects you have tested?
Does it have effect powerwise to give more time to mix the fuel with air?
Or low load/idle that the fuel is injected in too short time to get good mixture?

With both GDI and port injection, I think the benefits / effects are similar.

It can help get a little extra fuel if you're nearing the end of the spray window and need that little extra fuel. Too much will cause poor mixing and smoke. The side effect of this is that the O2 will see a rich mixture, but the actual "burnt" mixture is potentially lean. Creating hot spots, instead of a homogeneously rich mixture which has a uniform cooling effect.

I run a hybrid on my Golf R with high exhaust backpressure and thus high EGTs. Once I got my fuelling where I wanted it, I added a little extra trailing spray towards the top end to try and cool the valves and turbo. Without an EGT probe, I can't be sure how effective it is.

Why does a R have 400 deg start angle and older GTI's 420 deg.?
Of course injectors are bigger, but also makes a lot more power.
430 deg start than start of injection is with still open exhaust valve, so you  have got chance that you throw fuel straight into exhaust.
But i haven't tested a lot with it yet, so can't tell what real live results are..

Good question. The whole point of GDI is to try spray the perfect amount of fuel, at the latest time possible creating a focused combustion in the piston bowl instead of an all-over cylinder combustion like a traditional port injected engine. Larger injectors you can spray later.

If the injection event was instant, then yes, I believe it would spray straight into the exhaust, but it's the point where the injection event starts. Bear in mind the ECU still needs to charge the booster circuit (40V to initially open the EV for peak and hold) and then have flow start. It's not inconceivable that at high engine speed, the preparation to spray would take up to 70 degrees of engine RPM.

I've found that too much SOI advance in the midrange causes smoke, which I think might back my theory of the ECU needing time to actually start the injection

I never checked but i think ecu calculates lag time for injector and the 400 deg is realy the startpoint of injection.
But of course you have some time from start injection before fuel reaches exhaust valve.
I will do some more testing with advance and retard injection timing.
But even on 8 ms @ 7000 rpm with 400 deg soi you still have EOI @ 40 deg btdc and have another 20 deg before ignition.

Not had much experience with this but im sure you need to account for fuel pressures as well as they can affect the timings.

I have done hours and hours of thesting this.
Few things what i have seen:
1 If your injector duty is over 60% it dont make lots of different in full power pulls, but you can set it right to get good idle and emission.

2 If you have huge injectors like 2000cc for 1.8 4 cylinder it is super important to adjust this and it make huge different if you have 
lots of pulsation in your intake channel, like na car with rare cams.

3 if you have  long distance between intakevalve and injector it makes more and more difference.

4 if you have duty more than 70 and spraying angle is about right compared huge injectors and absolute right spraying angle  i am not shure witch is better, we are testing this still.

5 also some ecus use start or end of injection tables, bosch uses end of injection because thing what changes is most is intake cam and they do it right and if you change bigger injectors end of injection is still right and it means most.
Idea behind this is that you put all your fuel in cylinder and if it sprays when valve closes fuel are there for next cylinder charge.

I am not engineer or so but this is what i have learned for hours and hours of dyno testing different race engines and also standalone and bosch ecus.

I hope it helps you guys a bit .

next project is use 2 different size injector in one cylinder and another one is outside of intake like f1 engines etc , lets see what happens =)

Following this up a bit more : Don't know if its any help for anyone but from the AWEA section of FR

FB AWEA 27.20.0 Functional description
1 General description of the "injection-type-based structure"
In the% EAKO function, the injection-type based structure is described.
2 The output angle of injection control
2.1 Functional description% AWEA:
Note: The injection angle refers to the ZOT, with positive angles being a position before and negative being a position according to ZOT.
The function AWEA calculates the position of the injections for all injection types of the petrol direct injection. The position of an injection is determined by the crank angle
Start of finish or finish.
2.1.1 Central size
vwikuwe_w
Is the angular velocity in ◦KW / ms and is essentially used for the conversion of the injection duration into an angle.
B_noesab
Is used to control the abort function in the HW-related software and to enable compression stroke injections.
wesabr
Is the angle at which a current injection is aborted before the end defined by the injection duration if B_noesab = false.
wrlbhom_w
Is a reference quantity for the prediction angle calculation and gives the angular distance from the reference mark at which the suction stroke injection for the cylinder i is calculated up to
To the ZOT of the cylinder i.
wrlbsch_w
Corresponding to the compression stroke injections in the layer injection modes.
2.2 The calculation of the injection position (s) in each BDE injection type is described below.
2.2.1 Injection type Homogeneous single injection HOM_SINGLE_HO1 in homogeneous mode:
In HO1, a suction stroke injection is carried out per cylinder and working clearance. The position of the suction stroke injection is defined by its starting angle wbhxxs1_w. While
Of the start in the injection mode HO1, the injection start angle is determined in the hierarchy WBHO1SS, dependent on anztib, tmot. The starting angle in HO1 is for the low-pressure start
Provided. After starting, the injection start angle from KFWBHO1SW / -L (for a warm engine is dependent on a threshold of the charge movement)
certainly. For systems with turbo (SY_Turbo> 0), the characteristic fields KFWBHO1SWE and KFWBHO1SLE with a higher value range (0 ... 431.25KW) and more protection zones are used
provided. The values ​​from these characteristic fields are added with a deltawinkel. This deltawink is generated by the weighting of the Delta-KF KFDWBHO1SK with the
Temperature-dependent characteristic KLWFWHXXS.
In order to reduce the runtime, KLWFWHXXS should supply the output value 0 for the operating-warm motor so that an interpolation of the KF values ​​and the multiplication no longer occur
Are required.
If the injection stop is not permitted (B_noesab = true) or is started (B_stend) and there is no fault in the high-pressure supply (B_eprail), the starting angle
From wbhls1_w in wbhxxs1_w.
If this is not the case, the injection angle and the current speed are used to calculate the angle of the injection. In the next step, the end angle becomes
Is calculated and compared with the angle of incidence taking into account an angle reserve DWEHO1DY (for rotational speed dynamics). If the end angle threatens the termination angle
The start angle is increased. For this purpose, the angle is added to the angle of abutment, which is covered by the injection and, in addition, an angle reserve
DWBHO1DYN for speed dynamics. The displacement in the direction of spring is limited by the motor-temperature-dependent angle KLWBHO1SMX. This can lead to an over-
Wetting of the piston bottom. This functionality makes the probability of an emaciation by injection interruption occurring
Reduced. Such crashes occur mainly during operation with pre-pressure (fuel line) and during the start-up overrun.
2.2.2 Injection type Homogeneous knock-protection double injection HKS_DOUBLE_HK2 in homogeneous operation:
The starting angle of the suction stroke injection is stored in KFWBHK2S1.
For the compression stroke injection, the injection end angle is stored in KFWEHK2K1 above the operating point. From this, the angular velocity vwikuwe_w of the
Injection start angle wbhk2k1_w. While B_hk2z = true, HKS is not yet active (B_hk2w = false), the end angle obtained in the time raster becomes the starting angle
Copied. Thus, a plausible center angle wmhk2k1_w can be provided in the time grid of the backpressure calculation as long as no injection time to the end-
Start angle conversion.
2.2.3 Injection type Homogeneous split dual injection HSP_DOUBLE_HO2 in HSP mode:
The starting angle of the suction stroke injection is stored in KFWBHP2S1.
The calculation of the injection start angle wbhp2k1_w is based on the end angle characteristic field KFWEHP2K1. While B_hp2z = true, HSP is not active yet
(B_hp2w = false), the end angle obtained in the time raster is copied into the starting angle. Thus, a plausible center angle wmhp2k1_w in the time grid can be the backpressure
Provided that no injection time is yet available for the end-to-start angle conversion.
2.2.4 Injection type high pressure start HOM_SINGLE_SHX:
In SHX, a compression stroke injection is performed per cylinder and working cycle. The injection angle weshxk1_w is stored in the characteristic field KFWESHXK1. Depending on
B_shearlca is the starting angle either with B_shx = true (calculation and output of the injection at zzylh) or with B_shxvw = true (calculation and output of the injection
At the zzyls).

part 2

2.2.5 Single-injection injection-type layer SCH_SINGLE_SC1 in shift operation:
In SC1, a compression stroke injection is carried out per cylinder and working clearance. The injection angle angle wesc1k1_w is calculated from a basic angle (KFWESC1K1)
Which is added with an offset angle. This offset angle is based on the multiplication of a deltawinkel (KFDWESC1K1 dependent on nmot_w and dpsmscs) with a
Weighting factor (KFFDWESC1K1 dependent on nmot_w and rk_w). For SY_STERVK> 0 a deltawinkel KF (KFDWESC1K1B) can be applied for the 2nd bank. measurements
Have shown for the air- and air-wall combustion processes that the end angle is determined not only by the throttling, but also by relative fuel mass and
Speed. (Accurately, a "3-dimensional characteristic space" would be stretched from these 3 parameters) To take account of this, KFDWESC1K1 is calculated by a
Weighting factor from KFFDWESC1K1.
While B_sc1z = true, but SC-1 is not active (B_sc1w = false), the end angle obtained in the time raster is copied to the starting angle. Thus, a plausible
Angle wmsc1k1_w are provided in the time grid as long as no injection time is yet available for the end-to-start angle conversion.
2.2.6 Injection Type Layer Catheater Double / Triple Injection SKX in SCH mode:
In the operating mode of the layer catalytic converter, two types of injection are operated simultaneously. Since the same boundary conditions are used for the compression stroke injections in the SKH mode
One of the injection modes SC1 is active. The working stroke injection is controlled by the additionally active injection type SC1.
The starting angle wbskxa1_w is stored in KFWBSKXA1.
2.3 Block mitibgr:
In the event of a fault in the high-pressure system, high pressure can no longer be produced under certain circumstances. In this case, B_eprail is set. Only the HOM mode is available. The
Injection time is so long that it can no longer be accommodated in any case. The available injection window is determined and from this a speed-dependent
A max. Torque mitibgr_w calculated. The air mass is limited by the torque limitation.CWAWEA FW Codeword AWEA: Bit3: Influence of wkrmav in HKS
DWBHDY FW Delta angle for switching off the angle correction
DWBHO1DYN FW Dynamic range start angle 1. Suction stroke ES in injection type ho1
DWEHO1DYN FW Dynamic response angle end 1. Suction stroke ES in injection type ho1
FKTTIMI FW Conversion factor for mitibgr calculation
FLBWBSW FW Threshold flb for changeover. The map injection angle
KFDWBHO1SK nmot rl KF Deltawinkel Injection start for 1.Saughub-ES cold in injection type HO1
KFDWESC1K1 nmot_w dpsmxscs KF Deltawinkel Injection for 1st compression stroke ES in injection type SC1
KFDWESC1KB nmot_w rkm_w KF Deltawinkel Injection for 1st compression stroke ES bank 2 in injection type SC1
KFFDWESC1K nmot_w rkm_w KF Factor for weighting Deltawinkel Injection of the 1st compression stroke ES in SC1
KFGRPWBHDY nmot rl KF gradient threshold for correction start 1. Suction pipe ES warm in the EA HO1
KFTWBHDY nmot rl KF Time constant for correction angle Start 1. Suction pipe ES warm in the EA HO1
KFWBHK2S1 nmot_w rl_w KF Beginning of angle 1.Screw stroke ES in EA HK2
KFWBHO1SLE nmot rl KF Beginning of angle 1.Shaughub-ES warm into the EA HO1 during charge movement> Threshold, -
expanded
KFWBHO1SS anztib tmot KF Starting angle 1.Saughub-ES during the start in the EA ho1
KFWBHO1SW nmot rl KF Angle Start 1.Saughub-ES warm into the EA HO1
KFWBHO1SWE nmot rl KF Beginning of angle 1.Shaughub-ES warm into the EA HO1, extended value range
KFWBHO1SWL nmot rl KF Beginning of angle 1.Shaughub-ES warm into the EA HO1 during charge movement> threshold
KFWBHO1SWS anztib tmot KF Start of angle 1.Saughub-ES Start in the EA ho1 for repeat start
KFWBHP2S1 nmot_w rl_w KF Angle Start 1.Saughub-ES in EA HP2
KFWBSKXA1Q nmot rkkhm KF Beginning of angle 1.Workshift ES in the EA skx (pos_quantization)
KFWEHK2K1 nmot_w rl_w KF Angular end 1.Compression stroke ES in the EA hk2
KFWEHP2K1 nmot_w rl_w KF Angular end 1.Compression stroke ES in the EA hp2
KFWESC1K1 nmot_w rkm_w KF Angle Injecting 1st compression stroke ES in EA SC1
KLDWBHDY nmot KL Correction angle Start 1. Suction pipe ES warm in the EA HO1
KLSWBHTKR nmot KL Temperature threshold for correction start 1. Suction pipe ES warm in the EA HO1
KLTDGRDPS nmot KL Time delay Correction angle Start 1. Suction pipe ES warm in the EA HO1
KLWBHO1SMX tmot KL maximum angle start 1. suction stroke ES in EA ho1
KLWESABR fktprps_w KL Characteristics for injection abort wingle
KLWESHXK1 tmst KL End angle Injection type SHX for 1st compression stroke ES
KLWFWHXXS tmot KL Weighting factor warm / cold KF Starting angle in the EA ho1 / 2, hp2
KLWTDY tmot KL Weighting Correction angle Start 1. Suction pipe ES warm in the EA HO1
PRNOESAB FW Rail pressure threshold for injection interruption at BDE
SDP06ZUUB dpsmxscs SV (REF) Support location distribution Intake manifold pressure difference
SNM08PS3UW nmot_w SV (REF) Support location distribution speed
SNM10FSUB nmot SV (REF) Support location distribution speed
SNM10ZUUW nmot_w SV (REF) Support location distribution speed
SNM12FSUB nmot SV (REF) Support point distribution speed, 12pcs.
SRK06ZUUW rkm_w SV (REF) Support location distribution layer fuel mass
SRK12ZUUW rkm_w SV (REF) Support location distribution layer fuel mass
SRL08FSUB rl SV (REF) Support distribution rl_w Support
SRL08ZHKUW rl_w SV (REF) Support location distribution Relative airflow for HKS with 8 support points
SRL08ZHPUW rl_w SV (REF) Support location distribution Relative airflow for HSP with 8 support points
SRL12FSUB rl SV (REF) Protection distribution relative airflow, 12pcs.
WBHXXS1AP FW Beginning of angle 1.Swing stroke ES for application purposes
WBSKXA1APQ FW Beginning of angle 1. Working stroke ES for application purposes EA SKH (pos_quantization))
WEHXXK1AP FW Angle end 1. Compression stroke ES in the EA ho2
WESC1K1AP FW Angular end 1.Compression stroke ES in the EA SC1 for the application

Just to clear up some false statements - a lambda probe reads oxygen, not gasoline content, so it does not matter how much fuel in there, only how much is combusted.
Hence it will not show rich when the burn is lean.

I know this, but I'm assuming the dive in the AFR is because of fuel being added too late to a a hot mixture on its way out and burns up some of that air.

Lambda readings can be affected by when it burns as well as by aggressive camshaft profiles. Hence why monitoring EGT is always a good idea.
However... if your cams are not in overlap, then it does not really matter.

When the spark comes in, you have already stopped spraying.

Indeed, but if you're spraying right up to ignition, the end of spray won't have had time to homogenize so will end up in the exhaust manifold along with unburnt oxygen from other charges. This is only my theory, but it's based on observations from my own car and seems to correspond with my big orange book of GDI. I agree that EGT could be a concern here, but I've not got a probe and hybrid I'm running just now is considered "disposable". I use the car as a practical learning platform. Omelettes and cracking eggs etc.

This is not how it works at all. If there is burning the oxygen will be consumed, if there is no burning it won't. Any fuel injected and not burned will reduce EGT due to latent heat capacity.