In reply to Toyman01 :
Not likely to be an a/f problem.
If premium fixes it, then the knock is octane related, which means you are getting too much spark. For your truck, look into the ignition section, and see if there's a window where knock is or isn't working (assuming it's there). If you have knock control, then you may be running in an area where knock is supposed to work differently because it expects AFM to be running at that point.
If you don't have knock, then you need to log where it's happening (which you should be able to do) and take a few degrees of spark out that specific area. Logging the data will also give you an idea what the CAI is doing- and it *might* be doing the opposite of what you want- causing the knock.
BTW, nominal fuel in the US is E10, not pure gas, so stoich is 14.3:1.
In reply to Toyman01 :
Thre "Stoich AFR" table (I wonder why it is a table!) should not really affect things like that. The car doesn't really "care" what it is, because the fuel trims are based on the lambda (or equivalency ratio, to be pedantic). That is to say, the computer knows the injectors flow THIS many grams under several different tables' worth of conditions, and the MAF is reading THAT many grams/second of airflow, which gets condensed down to grams of air per cylinder per cycle.... divide that by the stoich figure and it knows how much fuel to inject. If the stoich figure is wrong, then there will be a fueling error and that is what short term and long term fuel trims are for.
AFR will generally always be correct once in closed loop, in other words.
Where you might see knock issues is if the MAF is reading incorrectly, because ignition timing is based on airmass grams/cyl/cycle.
Keep in mind that GM has a high octane and a low octane ignition map, and they will ramp up to the high octane map to see how much the engine knocks, on purpose...
On further reflection, 90% of the time cold air intakes seem to make MAFs read a little low, which also would make ignition timing advance relative to ideal. (GMs do have tables for mean best torque, that I assume is the ultra-hard limit for advance - going further is by definition worse)
The worst MAF transfer I had to fix was on a Ford that was underreporting by 40% in spots. This was one of those deals where the sensor element bolted into a new tube. Of course the vendor told the car owner that no tuning was necessary. The second-worst was an LS1 car that was up to 25% low. A whole Summit catalog was thrown at it and it felt about as fast to me as a V6 car. Client was very happy and all I did was correct the MAF and VE tables.
Knurled. said:
On further reflection, 90% of the time cold air intakes seem to make MAFs read a little low, which also would make ignition timing advance relative to ideal. (GMs do have tables for mean best torque, that I assume is the ultra-hard limit for advance - going further is by definition worse)
The worst MAF transfer I had to fix was on a Ford that was underreporting by 40% in spots. This was one of those deals where the sensor element bolted into a new tube. Of course the vendor told the car owner that no tuning was necessary. The second-worst was an LS1 car that was up to 25% low. A whole Summit catalog was thrown at it and it felt about as fast to me as a V6 car. Client was very happy and all I did was correct the MAF and VE tables.
And a low MAF will under estimate the load on the engine, which would over spark.
In reply to alfadriver :
From my side of the calibration equation (dealing with wear/age, and modifications), engine controls are extremely fault tolerant on the fuel side. That Ford drove great aside from throwing lean codes because long term fuel trims were off the charts. (So the CARB-approved CAI was making it fail emissions, and I had to to a technically-illegal recalibration on the PCM to bring it back into spec...) That Chevy managed to have the VE skewed AND the MAF skewed in the same directions so it wasn't throwing any correlation errors, but it just drove like crap...
