The computer..smart or not?

[JAP63]

Example
A standard Vr4 with the standard 14B @ 10 psi of boost.
Remove the 14B and install a 16g..still on 10psi of boost.

14B-405CFM
16G-505CFM
*Ratings at 1 Bar

Now the engine will be receiving around an extra 100 cfm*

Will the motor then be running leaner?
Or
Will the airflow meter see the larger charge of air being drawn in and give it the correct amount of fuel...?(of course within the limits of the fuel system..still only @ 10 psi as an example)

Can it see it or not?

 

[BennyGee]

yes it can see it, but as you say, your fuel system has to be able to handle it.

 

[JAP63]

yes it can see it, but as you say, your fuel system has to be able to handle it.

So if i had an safc i would need to crank more fuel in on that..the computer would do it for me?

 

[vertex_vr4]

The fact that more air can move through the turbo is unrelated to the final Air/Fuel mixture.

The ECU "sees" the amount of air entering the system at the MAF and adds an appropriate amount of fuel. This amount is affected by a number of factors but is calculated to be approximately stoichiometric (actually a bit richer).

Thats the theory anyway.

In practice the following can happen:

1. The MAF cannot "read" the amount of air coming in at high air flow - the karmen vortex signal breaks down (MAF overrun) = more air than "seen" = lean
2. Fuel system unable to supply the calculated Injector pulse frequency/widths or dirty injectors = less fuel than needed = lean
3. The bigger turbo heats up the air and the intercooler is unable to cool either because its too small or its heat soaked. Now although the correct amount of fuel is being injected, the air/fuel mix is more likely to auto-ignite = knock = less timing

Number 3 is nasty as there is no way of knowing (with a stock system) what the final temperature is. The guys on the dsm-ecu forum suggested using a GM hotwire setup at the intake to work around this.

 

[JAP63]

The fact that more air can move through the turbo is unrelated to the final Air/Fuel mixture.

The ECU "sees" the amount of air entering the system at the MAF and adds an appropriate amount of fuel. This amount is affected by a number of factors but is calculated to be approximately stoichiometric (actually a bit richer).

Thats the theory anyway.

In practice the following can happen:

1. The MAF cannot "read" the amount of air coming in at high air flow - the karmen vortex signal breaks down (MAF overrun) = more air than "seen" = lean
2. Fuel system unable to supply the calculated Injector pulse frequency/widths or dirty injectors = less fuel than needed = lean
3. The bigger turbo heats up the air and the intercooler is unable to cool either because its too small or its heat soaked. Now although the correct amount of fuel is being injected, the air/fuel mix is more likely to auto-ignite = knock = less timing

Number 3 is nasty as there is no way of knowing (with a stock system) what the final temperature is. The guys on the dsm-ecu forum suggested using a GM hotwire setup at the intake to work around this.

So if i have an SAFC2 and i put it into the + values im wasting fuel because the computer will have worked out the right amount anyway?

 

[vertex_vr4]

No and yes.... in closed loop you wouldn't be wasting fuel because the computer has already worked out the correct amount anyway.

More drivel:

The VR4 ecu in closed loop can compensate for up to 40% lean or 40% rich conditions. The O2 sensor in the exhaust tells the ECU if it is rich/lean and the ECU compensates. This compensation is slowly trickled into the long term memory.

If you "lie" to the ECU about the amount of air entering the system - the ECU will eventually learn/compensate for this.

In closed loop, unless you are "lying" to the ECU about other things (such as injector sizes) or there is some reason you would be more than 40% richer/leaner than factory I can't see a point of an SAFC. I do concede however that it could have more granularity than the ECU which only really has 12 points (i think) and interpolates from there.

In open loop the SAFC is very useful as alot of the sensors are ignored by the ECU - including the O2 sensor. It mostly works on predefined lookup maps. Here an SAFC could help you get out of trouble if you are running lean or topping out off the maps.

Check out this: http://forums.evolutionm.net/showthread.php?t=116519 particularly the comment from EVOTEXAS.

My advice is to get your ECU socketed and start writing your own maps. It's really not as hard as it sounds. 8)

Cheers,
John