My blackwing has preignition knock (the marbles in a can sound that most of us haven't heard since the 70s or 80s). It doesn't happen all the time - it's only at wide open throttle under high load, and not every time I use WOT. I am convinced this is NOT normal as this is my second 5 Blackwing. The first 5 I owned NEVER did this under any circumstance, nor did the car I drove at Spring Mountian. On my current car, I noticed this phenomenon on the drive home from the dealer (with 3 miles on the car). I figured the dealer put 87 octane in it, but after running 5 tanks of 91 octane the problem has not gone away.
I had one of my hardcore race guy friends ride along, as I thought I was loosing my mind. He agreed - it's preignition of the type you get from low octane. So we went to a local gas station that sells 101 octane unleaded race gas (at $14/gal) and mixed that with what was in the tank for just over half 101 octane / half 91 octane. Problem is totally gone with what is probably about 96 octane in the tank. I need to run some more tanks of higher octane to confirm, but given the results of this one tank, it does seem to be classic preignition from gas that is too low in octane for the compression / boost level.
Here's my theory: The car is tuned for 93 octane, but all we can buy in California at the pump is 91. My car doesn't like to run on 91 for some reason (if I go out of state I'll try 93). Best case scenario, the knock sensor is bad, so the car doesn't realize there is 91 in the tank and runs the tuning map/curves for 93 octane, hence preignition knock. Worst case, there's some hot spot in the cylinder head or other defect internal to the engine that causes preignition in one cylinder. If that were the case, I don't think the dealer could ever figure it out (until the motor blows up at a track day). Alternatively, there might be a fueling problem in one injetor or one bank (a la Car and Driver engine failure). But high bandwidth O2 data might show a lean mixture on one bank.
Now the question is how to diagnosis this / get it fixed. It doesn't happen all the time, but the most consistent way to make this happen is WOT in 3rd gear at 5000 RPM. Unfortunately, that happens at 100mph plus, and I don't think the dealer is going to be able to repeatedly do 100+ MPH WOT runs while reading a scan tool to diagnosis this. Finding ANY stretch of road except for a race track that is long enough and empty enough to use WOT for more than a second or two is extremely difficult. So my thoughts are: Find a dyno tuner with experience with the LT4 that can pull out knock sensor data, timing, boost, and air/fuel ratio with a high end scan tool while on the dyno to try to figure out what is causing this. Once diagnosed, take it to the dealer to replace the affected parts. Alternatively, I see the Cosworth toolbox reports max boost. Anyone know if I can get a plot of knock, timing, and boost vs time out of the PDR via Cosworth toolbox (or some other tool)? This could also be used to confirm the ECU is (incorrectly) running the same timing and boost levels regardless of what octane fuel is in the tank.
Anyone know how the ECM in the LT4 manages lower octane fuels (again, the best we can buy at the pump here is 91 octane)? What happens if you accidentally put 87? Does the computer pull timing and boost, and, if so, how does it know 87 is in the tank?
I had one of my hardcore race guy friends ride along, as I thought I was loosing my mind. He agreed - it's preignition of the type you get from low octane. So we went to a local gas station that sells 101 octane unleaded race gas (at $14/gal) and mixed that with what was in the tank for just over half 101 octane / half 91 octane. Problem is totally gone with what is probably about 96 octane in the tank. I need to run some more tanks of higher octane to confirm, but given the results of this one tank, it does seem to be classic preignition from gas that is too low in octane for the compression / boost level.
Here's my theory: The car is tuned for 93 octane, but all we can buy in California at the pump is 91. My car doesn't like to run on 91 for some reason (if I go out of state I'll try 93). Best case scenario, the knock sensor is bad, so the car doesn't realize there is 91 in the tank and runs the tuning map/curves for 93 octane, hence preignition knock. Worst case, there's some hot spot in the cylinder head or other defect internal to the engine that causes preignition in one cylinder. If that were the case, I don't think the dealer could ever figure it out (until the motor blows up at a track day). Alternatively, there might be a fueling problem in one injetor or one bank (a la Car and Driver engine failure). But high bandwidth O2 data might show a lean mixture on one bank.
Now the question is how to diagnosis this / get it fixed. It doesn't happen all the time, but the most consistent way to make this happen is WOT in 3rd gear at 5000 RPM. Unfortunately, that happens at 100mph plus, and I don't think the dealer is going to be able to repeatedly do 100+ MPH WOT runs while reading a scan tool to diagnosis this. Finding ANY stretch of road except for a race track that is long enough and empty enough to use WOT for more than a second or two is extremely difficult. So my thoughts are: Find a dyno tuner with experience with the LT4 that can pull out knock sensor data, timing, boost, and air/fuel ratio with a high end scan tool while on the dyno to try to figure out what is causing this. Once diagnosed, take it to the dealer to replace the affected parts. Alternatively, I see the Cosworth toolbox reports max boost. Anyone know if I can get a plot of knock, timing, and boost vs time out of the PDR via Cosworth toolbox (or some other tool)? This could also be used to confirm the ECU is (incorrectly) running the same timing and boost levels regardless of what octane fuel is in the tank.
Anyone know how the ECM in the LT4 manages lower octane fuels (again, the best we can buy at the pump here is 91 octane)? What happens if you accidentally put 87? Does the computer pull timing and boost, and, if so, how does it know 87 is in the tank?