- Sun Jun 01, 2008 7:35 pm
#1920467
Driving around in my CTR for a couple of months now. love every minute of it, but there is room for improvement. Especially when you look at the inlet side of the engine. There are some CAI setups available, but they seem to lack at some point in the rev range. While you can gain some BHP up top, driveability can suffer below 4000 rpm. One reason for this is the k20z4 engine is running pig rich with average air/fuel ratios between 13-12:1. This was done as a safety precaution to keep the engine from melting a piston at high revs in the unlikely event of a lean burn mode.
Ik was never happy with the way these CAI systems function, so I set about figuring out how to improve the induction setup of the engine. I spent a good deal of time experimenting with my diy induction setup and I think I have it worked out.
In short, the desigin of the stock airbox is a disaster.
It`s a box with a panelfilter sandwhiched in between where the engine sucks the air from. This air becomes turbulent after it passes the filter and is immediately drawn through the mass airflow sensor (maf). The maf is one of the main sensors controlling the right amount of fuel delivered to the air the engine gets.
There are two problems with this system:
-The turbulent air makes it difficult for the maf to measure it correctly. Especially under heavy loads the air passes so quickly through the maf tube that the sensor cannot keep up with it. It constantly tries to correct for the right air/fuel ratio but when it falls ouside the stoichiometric range of 11-15.5:1, it compensates by changing to full rich mixture.
- Turbulent air in the maf tube can become so violent that ignition timing is reduced by one or more degrees, as a safety precaution, but with additional loss of power as a result.
I decided to make a modified induction setup giving the engine stable air/fuel ratios (AFR), improving driveability and give better performance.
Suspecting most of the gains were to be found in the maf I searched the web and learned that Subaru Impreza turbo`s had maf also. Some tuners had figured out how to modify their maf, leaning out the fuel curve, allowing the car to run better.
I took some bits of information here and there and used it to modify the maf tube of the k20 engine.
Below are some random pictures of my mods.
I started by cutting the maf tube of the airfiltercover
These are the parts for my induction setup. There is also a 75mm pvc tube for connecting the maf to the filter but is not shown.
The key factor getting the engine measuring the correct afr without going to lean lies in the correct modification of the inside of the maf.
By trial and error I managed to find the correct shape of the tube`s inside contours resulting in a stable, and at the same time leaner afr signal from the mass airflow sensor.
The panelfilter is being replaced by an open filter with radiused curves inside. The air being drawn inside is forced by the radiuses into a laminair flowing airstream pointed directly at the mass airflow sensor inside. This clean flowing airstream enhances the afr stability and provides maximum ignition advance for better power.
The Apexi dry filter is rated amongst the best among èm. But it is a bit pricey.
There are a few alternatives:
-Air dynamik
-Fujita
These filters do their job.
Inside of the maf with filter and mass airflow sensor attached.
In order to monitor my afr reading from the modified maf I decided to invest in a afr meter. But the exhaust had to come of the car in order to fit the bung for the lambda widebandsensor. It turned out that the downpipe consists of an inner-and outer tube, the outer tube`s diameter measures 70 mm, the inner tube 60mm.
The header bolts were done up so ridiculously tight, it was a matter of time before we knackered them. I had to fetch another pair. Take care not to strip the threads from the header collector eyes.
This is an overlay of the afr readouts taken from my afr meter of an fn2 with stock airbox and my modified induction setup. The dotted lines represent the modded induction.
Purple= afr curve
red=throttle
green=torque
blue=horsepower
black=rpm
In stock form the afr curve sharply falls off into the 13:1 region, at the same time retarding ignition timing, holding back torque during acceleration mode.
With modified setup torque curve is higher througout the powerband up to 5200 where vtec sets in.
The improved airflow from the maf tube together with stable afr signal allowes the vtc system to induce higher gasspeeds. At idle, vtc is zero degrees, afr can be as high as 15.5:1. In acceleration mode vtc kicks in, steadily increasing valve overlap to a maximum of 50 degrees at approximately 5200rpm where vtec is gradually taking over.
Note that this setup leans the engine out by a full point, improving torque and bhp throughout the rev range compared to stock.
I must point out that this readout is not from my car, it`s from a customer of mine who was trying out his setup. I also borrowed him my afr meter, but he didn`t found it necessary callibrating the torque, bhp parameters for true representation. Hence no real world torque and horsepower reading, except for the fact that the curve overlay shows the difference between stock and modified. The afr curves however are callibrated for true readout.
My own car already has some further mods, making a direct comparison between stock and modified inconsistent.
My increased torque output necessitated stiffer engine mounts to keep the engine from rocking back and forth under load. I took the upper and lower mounts and filled the holes in the rubber big end bearing with some chunks of rubber and poured some polyurethane glue over it. Works pretty well and as a bonus, shifting improved too.
My Innovate motorsports lc1 controller with Bosch lsu4.2 wideband lambdasensor. An invaluable aid for monitoring afr on the engine.
I couldn`t get the lower airfilterbox off, so I left it in place.
Above are some pictures taken during the time I was constantly experimenting with my setup, so everything was temporarily fixed with tie wraps.
I found a piece of metal bar in my garage and sawed/filed it in shape so it would support the maf and filter.
As far as I`m concerned you can remove the insulation mat attached under the bonnet. It serves no function at all. This way, more air will find its way into the engine compartment. The bonnet is specially shaped so that it allows an airstream over the top of the engine and behind it, taking the warm air from the exhaust manifold with it and disperses it through the opening in the undertray underneath the car.
The grille area is completely sealed by a plate behind it and will not pass any air into the engine bay. I cut some holes in the mounting plate for the grille, allowing cool air to cool the intake manifold, injectors, fuel rail as well as providing more air for the filter. In the process I removed the intake manifold cover aswell, as I believe it can contribute to heatsoak on warm days/city driving.
Up until now I`m very pleased with the way my car drives know. With this modified intake setup I have experienced the following results:
-More torque/horsepower throughout a wider powerband. Because of the leaner condition my engine runs in, I think I have traded in a few top end horsepower, but gained substantially more in the midrange where it counts.
This is all subjective, totally based on arse dyno. I have not had the car on the rollers yet, don`t really believe in them as readings can be rather inconsistent.
-Mpg improvement because of the leaner afr. Engine warms up very quick, due to the catalytic converter reaching closed loop operating temperature faster.
-Driveability greatly improved, less bogging, better throttle response. Very smooth idle and acceleration pickup. No heatsoak noticed.
-Sound rivals any aftermarket CAI. Because the open filter is placed directly under the cowl it amplifies the typical engine wail. The vtc system gives an agressive growl, with a noticeable change in tone when vtec takes over, then it`s LOUD.
Yet in daily traffic the engine is whispering at idle and runs smooth an quiet like any stock FN2.
I`m still not done with my car. In the meantime I have found a few things that can be improved. When I have the time I will post an update. B.t.w excuse my lousy English, I`m doing my best.
Enjoy, Paul
Ik was never happy with the way these CAI systems function, so I set about figuring out how to improve the induction setup of the engine. I spent a good deal of time experimenting with my diy induction setup and I think I have it worked out.
In short, the desigin of the stock airbox is a disaster.
It`s a box with a panelfilter sandwhiched in between where the engine sucks the air from. This air becomes turbulent after it passes the filter and is immediately drawn through the mass airflow sensor (maf). The maf is one of the main sensors controlling the right amount of fuel delivered to the air the engine gets.
There are two problems with this system:
-The turbulent air makes it difficult for the maf to measure it correctly. Especially under heavy loads the air passes so quickly through the maf tube that the sensor cannot keep up with it. It constantly tries to correct for the right air/fuel ratio but when it falls ouside the stoichiometric range of 11-15.5:1, it compensates by changing to full rich mixture.
- Turbulent air in the maf tube can become so violent that ignition timing is reduced by one or more degrees, as a safety precaution, but with additional loss of power as a result.
I decided to make a modified induction setup giving the engine stable air/fuel ratios (AFR), improving driveability and give better performance.
Suspecting most of the gains were to be found in the maf I searched the web and learned that Subaru Impreza turbo`s had maf also. Some tuners had figured out how to modify their maf, leaning out the fuel curve, allowing the car to run better.
I took some bits of information here and there and used it to modify the maf tube of the k20 engine.
Below are some random pictures of my mods.
I started by cutting the maf tube of the airfiltercover
These are the parts for my induction setup. There is also a 75mm pvc tube for connecting the maf to the filter but is not shown.
The key factor getting the engine measuring the correct afr without going to lean lies in the correct modification of the inside of the maf.
By trial and error I managed to find the correct shape of the tube`s inside contours resulting in a stable, and at the same time leaner afr signal from the mass airflow sensor.
The panelfilter is being replaced by an open filter with radiused curves inside. The air being drawn inside is forced by the radiuses into a laminair flowing airstream pointed directly at the mass airflow sensor inside. This clean flowing airstream enhances the afr stability and provides maximum ignition advance for better power.
The Apexi dry filter is rated amongst the best among èm. But it is a bit pricey.
There are a few alternatives:
-Air dynamik
-Fujita
These filters do their job.
Inside of the maf with filter and mass airflow sensor attached.
In order to monitor my afr reading from the modified maf I decided to invest in a afr meter. But the exhaust had to come of the car in order to fit the bung for the lambda widebandsensor. It turned out that the downpipe consists of an inner-and outer tube, the outer tube`s diameter measures 70 mm, the inner tube 60mm.
The header bolts were done up so ridiculously tight, it was a matter of time before we knackered them. I had to fetch another pair. Take care not to strip the threads from the header collector eyes.
This is an overlay of the afr readouts taken from my afr meter of an fn2 with stock airbox and my modified induction setup. The dotted lines represent the modded induction.
Purple= afr curve
red=throttle
green=torque
blue=horsepower
black=rpm
In stock form the afr curve sharply falls off into the 13:1 region, at the same time retarding ignition timing, holding back torque during acceleration mode.
With modified setup torque curve is higher througout the powerband up to 5200 where vtec sets in.
The improved airflow from the maf tube together with stable afr signal allowes the vtc system to induce higher gasspeeds. At idle, vtc is zero degrees, afr can be as high as 15.5:1. In acceleration mode vtc kicks in, steadily increasing valve overlap to a maximum of 50 degrees at approximately 5200rpm where vtec is gradually taking over.
Note that this setup leans the engine out by a full point, improving torque and bhp throughout the rev range compared to stock.
I must point out that this readout is not from my car, it`s from a customer of mine who was trying out his setup. I also borrowed him my afr meter, but he didn`t found it necessary callibrating the torque, bhp parameters for true representation. Hence no real world torque and horsepower reading, except for the fact that the curve overlay shows the difference between stock and modified. The afr curves however are callibrated for true readout.
My own car already has some further mods, making a direct comparison between stock and modified inconsistent.
My increased torque output necessitated stiffer engine mounts to keep the engine from rocking back and forth under load. I took the upper and lower mounts and filled the holes in the rubber big end bearing with some chunks of rubber and poured some polyurethane glue over it. Works pretty well and as a bonus, shifting improved too.
My Innovate motorsports lc1 controller with Bosch lsu4.2 wideband lambdasensor. An invaluable aid for monitoring afr on the engine.
I couldn`t get the lower airfilterbox off, so I left it in place.
Above are some pictures taken during the time I was constantly experimenting with my setup, so everything was temporarily fixed with tie wraps.
I found a piece of metal bar in my garage and sawed/filed it in shape so it would support the maf and filter.
As far as I`m concerned you can remove the insulation mat attached under the bonnet. It serves no function at all. This way, more air will find its way into the engine compartment. The bonnet is specially shaped so that it allows an airstream over the top of the engine and behind it, taking the warm air from the exhaust manifold with it and disperses it through the opening in the undertray underneath the car.
The grille area is completely sealed by a plate behind it and will not pass any air into the engine bay. I cut some holes in the mounting plate for the grille, allowing cool air to cool the intake manifold, injectors, fuel rail as well as providing more air for the filter. In the process I removed the intake manifold cover aswell, as I believe it can contribute to heatsoak on warm days/city driving.
Up until now I`m very pleased with the way my car drives know. With this modified intake setup I have experienced the following results:
-More torque/horsepower throughout a wider powerband. Because of the leaner condition my engine runs in, I think I have traded in a few top end horsepower, but gained substantially more in the midrange where it counts.
This is all subjective, totally based on arse dyno. I have not had the car on the rollers yet, don`t really believe in them as readings can be rather inconsistent.
-Mpg improvement because of the leaner afr. Engine warms up very quick, due to the catalytic converter reaching closed loop operating temperature faster.
-Driveability greatly improved, less bogging, better throttle response. Very smooth idle and acceleration pickup. No heatsoak noticed.
-Sound rivals any aftermarket CAI. Because the open filter is placed directly under the cowl it amplifies the typical engine wail. The vtc system gives an agressive growl, with a noticeable change in tone when vtec takes over, then it`s LOUD.
Yet in daily traffic the engine is whispering at idle and runs smooth an quiet like any stock FN2.
I`m still not done with my car. In the meantime I have found a few things that can be improved. When I have the time I will post an update. B.t.w excuse my lousy English, I`m doing my best.
Enjoy, Paul
Last edited by culasse on Mon Aug 17, 2009 1:25 pm, edited 8 times in total.
