Optimize injection timing

[robespierre]

Hi,

How does one optimize the injection timing? Would these values remain similar for cars with the same engines but using injectors of differing sizes?

JR

[RossB]

A few notes I have made about injection timing:

Timing Table
The injection timing table has values in Degrees BTDC (compression stroke). When using an M400/M600/M800 you can choose to have either a 2D or 3D Injection Timing table (20 x 11 sites). The Setup Parameter "Injection Timing Position - (itp)" selects Start of Injection or End of Injection. End of Injection is normally best. When using an M4/M48 the Injection Table is a 2D table (20 sites) for End of Injection Timing. The table is found in the Adjust menu under FUEL / Injection Timing.
The Injection Timing may be adjusted at various RPM points and optionally at various Efficiency Points depending on the state of the 'Miscellaneous Setup - Injection Timing Table (itt)' parameter. The injection timing position must be selected to either start of injection or end of injection in the ECU Manager Fuel Setup (M800). It is most common to select the end of injection point as it is easier to relate this to the time the inlet valve closes and it means that the values in the table will have smaller steps. At idle speed a 5ms fuel pulse will have duration of 18 degrees of crankshaft rotation, a 10ms pulse at 3000 RPM will last for 180 degrees and a 10ms pulse at 6000 RPM will last 360 degrees. So for the same end of injection time we would have to have a table with very large differences from one point to the next if we were to use beginning of injection as a timing position. It would also be necessary to have different timing values depending on the length of the injection pulse.

Normally the ideal aim is to inject the fuel with the inlet valve open but we have to take in to account that there is a delay between the time the injector closes and the time the fuel passes through the valve, this delay will vary depending on engine speed and manifold design. The closing rate of the valve and the amount of valve overlap will also have an effect on the ideal fuel timing position.
The best and easiest way to find the ideal fuel timing is by tuning to the richest lambda value. Moving the injection will have an effect on the lambda reading because in-correct timing will cause more of the fuel to condense on the port wall and so will not be burned. As the timing approaches the ideal point the lambda reading will show richer because more if the injected fuel is being burned. A good starting point is to have the timing at around 270 degrees at 1000 RPM and increasing by 20 degrees every 500 RPM until about 4000-5000 RPM where the injection timing is less critical (depending on injector sizing). It is reasonable to expect that this would give the most efficient timing point whether or not there is any gain in performance. The fuel table can then be re-trimmed to give the desired air-fuel mixture. When injector pulse times are long the injection timing will have less effect. To avoid rapid transitions in Injection Timing it is advisable to limit variations between adjacent sites to less than 100 degrees. A value of 0 is the same as a value of 720 i.e. both are Top Dead Centre. The value may be increased past 720 degrees if necessary which allows a transition back through 0 degrees without causing a transition problem. In other words a value of 820 deg. is the same as 100 deg.
It is reasonable to expect that this method will give the most efficient timing point and provide the most accurate control of fuel mixture but there may be applications where this is not going to give the best performance. The main reason for this is fuel atomisation. If the injector is not particularly good at atomising the fuel better results may be found by injecting on a closed inlet valve. This can also be beneficial for engine starting. This will allow the fuel to atomise better in the inlet port prior to induction but will lead to a less precise control of fuel mixture.
When the Injection Timing screen is open and the engine operating point is close to the current cursor position the Injection Timing is forced to the cursor value rather than being interpolated from the adjacent sites, this allows adjustments to be made without the adjacent sites affecting the adjustment value. The quick lambda function may be used to adjust the main fuel table when the Injection Timing screen is open. Some engines (e.g. Subaru) have tumble valves in the inlet manifold to improve the fuel mixing at los speed. Where these are disabled or removed timing the end of injection at about 450 will give best results at idle speed and starting.

Injector Sizing
Typically in a production car an injector is selected to run at around 85% duty cycle at maximum fuel demand. This will normally allow enough resolution in the fuel table at idle speed to provide accurate fuel mixture adjustment and injection pulse times that are not too short. In this case the injection timing will have little effect at high RPM / Load because the injector will be open most of the time, by the time it closes it is nearly time to open again. Bigger injectors can be used which would give shorter injector duration at high speed and take advantage of the benefits of timing the fuel, however this will cause a loss of resolution and very short injection pulses at low speed. This will cause poorer idle quality and starting, particularly when the engine is cold. An alternative to this is to use secondary injectors (Hi/Lo).

Fuel Conditioning and Targeting
Most injectors deliver an atomized fuel spray, and some engines have injectors with a spray pattern specifically tailored to suit the port / valve / injector location for that engine. Some motor sport injectors do not have a spray pattern at all and just squirt fuel in a straight beam. This may not seem like a good idea but where high quantities of fuel are required (particularly when using methanol) it is an effective way of getting a large quantity of fuel into the engine but it relies on swirl in the combustion chamber for good mixing.
Injectors which target fuel directly on to the back of the inlet valve will provide the best control of fuel mixture and response to transient changes. Engines with this type of injector layout will respond very well to injection timing where the injection takes place with the inlet valve open. The deviation from desired air fuel ratio will be reduced with this layout because there is little or no port wetting. The injector’s ability to atomize fuel is critical in this instance because there is no time for the fuel to atomize in the intake manifold. Although this will not effect the mixture stability it may mean that there is higher levels of un-burnt hydrocarbons in the exhaust emission. Timing the injection pulse with the inlet valve closed (exhaust stroke) will improve fuel atomization but deviations from the desired air fuel ratio will be greater because of the effect of the fuel layer that is created on the port wall.
Some engines used in Motorsport have injectors mounted on the air intake side of the throttle. Typically these injectors would be mounted roughly 250 mm away from the valve. Gains in engine performance can be found at higher RPM with this arrangement, this is partly because the fuel has more time to atomize before it reaches the engine. There is also a slight cooling effect generated by the atomized fuel (stand off)in the intake. The disadvantage of this is that there is a large area of the inlet covered by a film of fuel which leads to problems with deviations from the desired air fuel mixture when accelerating and decelerating. Other common problems with this arrangement are fuel stand-off (reversion) and the risk of air box fire. Injection timing can be used to reduce and help control stand-off but is more effective with shorter injector durations.

Idle Stability
Injector timing will have an influence on idle stability as well as tail pipe emissions. It is expected that fuel mixtures need to be a little bit richer at idle and at low load than at cruising speed but this can be minimised by injector targeting and timing. This will vary depending on engine design but is significant particularly when using a closed loop idle control system.

[Alex B]

Thank you Ross for the detailed explanation on the injection timing. It is really helpful since this subject is not much written about in the literature or the net.

What timing would you recommend to improve cold starts on 4G63 Evo 9 engine with 1000cc injectors, 272 cams and ported head?

Thank you
Alex

[RossB]

For cold starting it would help to inject fuel on to the closed valve, try something like 450.

[Scott@FP]

Another thing you might notice is when injecting on an open valve and timing is set to EOI, you need very little coolant temp correction once running.

FYI the Mitsubishi OEM way of doing injection timing is SOI tied to the crank sensor 2nd trigger edge, SOI is always 5 ATDC power stroke. This method may ease initial tuning and driveability but won't be 100% optimal, you can always change it later.

[msgofast]

What would one do in the case where the inlet cam can be adjusted some 45degs to then test best injection timing and cam timing?

Different cam timing results in a different injection timing for light load cruising for best economy, question is which to adjust first and which gives the best results?

Also is the base injection timing map for a ej207 in the start file a tested map or is a copy of oem maps?

cheers
Michael

[RossB]

What would one do in the case where the inlet cam can be adjusted some 45degs to then test best injection timing and cam timing?

Different cam timing results in a different injection timing for light load cruising for best economy, question is which to adjust first and which gives the best results?

I would adjust EOI timing agter finding the ideal cam timing. The cam timing is going to have a more significant effect on performance, the EOI timing will allow you to run a leaner mixture with smoother running and better drivability. This is especially true with the Subaru engine.

Also is the base injection timing map for a ej207 in the start file a tested map or is a copy of oem maps?

I am not sure of the origins of the values in the start file but I would treat them as start up values only. As with any aspect of tuning the more time you spend on tuning the better the end result will be.

Another thing you might notice is when injecting on an open valve and timing is set to EOI, you need very little coolant temp correction once running.

The only reason we need enrichment as a function of engine temperature is to compensate for the fuel that is condensed on to the side of the inlet manifold/port/cylinder walls which will not burn and ultimately ends up in the oil. If the fuel is targeted and atomised well a greater portion of it is burnt so less compensation is needed.

[Scott@FP]

You could always use the cam position as an axis in your injector timing table and eliminate differences due to cam position change.

[whrdnsx]

Regarding this subject, has anyone come up with a basic formula to work out injection timing based on Cam timing & injector size & distance from the valve?

[terryo]

my Toyota 4AG race engine has values all over the place. ranging from 100 at low RPM to 700 mid-range, to 300 at rev limit. I suspect this is a result of incomplete or even erroneous tuning efforts over the years

i would be really pleased if anyone with a 4AG or similar race 4 cylinder , could comment on what they have used

thanks