- Joined
- Nov 27, 2015
- Threads
- 99
- Messages
- 911
- Reaction score
- 2,499
- Location
- Torrance California
- Website
- www.hondata.com
- Vehicle(s)
- 23' Type R,22' Civic Si,18' Accord 2.0T,17' FK8
- Vehicle Showcase
- 1
- Thread starter
- #1
Now that people have the Civic Type R FlashPro on their vehicles, here's some information to help those who are interested in tuning the ECU.
First in the series is the fuel pump. The high pressure fuel pump is a fixed volume pump driven from the camshaft. The high pressure pump output pressure is high (200 bar / 20 MPa / 2900 psi), and the inlet pressure from the in-tank pump is around 7 bar.
The pump can only pump a certain volume of fuel each camshaft revolution, which is almost exactly 1cc or 0.25cc per cylinder cycle. Since the fuel pump delivery volume per cylinder is fixed then the fuel pump delivery characteristic is tired to torque rather than power. From stock the pump is fairly close to its limit.
Here the datalog from a dyno run. In the bottom half of the image the purple line is the fuel pressure, green line is the target fuel pressure and the red line is the fuel pump duty cycle.
The fuel pump duty cycle is an important number to watch on the dyno. This channel did not exist and we added it in to the ECU to help everyone tune these engines. The duty cycle runs from 0% to 100%, with 100% being maximum output.
In the above datalog the fuel pump peaks at 95% at turbo spool and then tapers off to around 85% at peak power. This is fairly typical of a non-stock calibration.
Here is a datalog from a dyno run where the fuel pump is asked to deliver more fuel than it can provide. The fuel pump duty (red) is maxed out at 100%. The fuel pressure (green) drops. The ECU will tolerate a certain fuel press drop and compensate by increasing the injector duration without any ill effects, but at a certain difference (about 30 bar drop) the ECU will run a backup strategy. This increases the injection window to all the intake and compression stroke, which tends to make black smoke, an error code is set and the boost is reduced. Obviously we want to avoid this.
If, when you are tuning, you want to reduce the demand on the fuel pump, you can do one of three things:
1. Reduce the torque demand in the throttle tables.
2. Make the lambda leaner by using the minimum lambda table.
3. Reduce the torque/boost/air charge using the maximum air charge table. This is the best method.
In the above example I would reduce the maximum air charge in the region where the fuel pump is 100% by around 5% and then re-test. This is an interactive process.
What should the target fuel pump duty be? Because we are working with the air charge, not boost, it is largely invariant with temperature etc. I'm comfortable with peak duty cycles being around 95%. If the duty touches 100% very briefly then it is not necessarily anything to get too concerned with as long as the fuel pressure does not drop.
I may to revise this post If needed, so consider that if quoting the whole post.
Questions?
First in the series is the fuel pump. The high pressure fuel pump is a fixed volume pump driven from the camshaft. The high pressure pump output pressure is high (200 bar / 20 MPa / 2900 psi), and the inlet pressure from the in-tank pump is around 7 bar.
The pump can only pump a certain volume of fuel each camshaft revolution, which is almost exactly 1cc or 0.25cc per cylinder cycle. Since the fuel pump delivery volume per cylinder is fixed then the fuel pump delivery characteristic is tired to torque rather than power. From stock the pump is fairly close to its limit.
Here the datalog from a dyno run. In the bottom half of the image the purple line is the fuel pressure, green line is the target fuel pressure and the red line is the fuel pump duty cycle.
The fuel pump duty cycle is an important number to watch on the dyno. This channel did not exist and we added it in to the ECU to help everyone tune these engines. The duty cycle runs from 0% to 100%, with 100% being maximum output.
In the above datalog the fuel pump peaks at 95% at turbo spool and then tapers off to around 85% at peak power. This is fairly typical of a non-stock calibration.
Here is a datalog from a dyno run where the fuel pump is asked to deliver more fuel than it can provide. The fuel pump duty (red) is maxed out at 100%. The fuel pressure (green) drops. The ECU will tolerate a certain fuel press drop and compensate by increasing the injector duration without any ill effects, but at a certain difference (about 30 bar drop) the ECU will run a backup strategy. This increases the injection window to all the intake and compression stroke, which tends to make black smoke, an error code is set and the boost is reduced. Obviously we want to avoid this.
If, when you are tuning, you want to reduce the demand on the fuel pump, you can do one of three things:
1. Reduce the torque demand in the throttle tables.
2. Make the lambda leaner by using the minimum lambda table.
3. Reduce the torque/boost/air charge using the maximum air charge table. This is the best method.
In the above example I would reduce the maximum air charge in the region where the fuel pump is 100% by around 5% and then re-test. This is an interactive process.
What should the target fuel pump duty be? Because we are working with the air charge, not boost, it is largely invariant with temperature etc. I'm comfortable with peak duty cycles being around 95%. If the duty touches 100% very briefly then it is not necessarily anything to get too concerned with as long as the fuel pressure does not drop.
I may to revise this post If needed, so consider that if quoting the whole post.
Questions?
