I had an "Oh NO!" moment yesterday morning. Loaded the cooler & chair into the trunk to head into the local Ford dealer's place for their annual All Ford show, hit the starter and it gave me the "extended crank, no fire" syndrome once again. When it lit after three or four tries it was loaded up pretty good, so I just noted the exact coolant temp (84F) on the ECU programmer and headed into town. Later, after sitting at the show for seven hours I checked the coolant temp before cranking (102F) and again it took three or four short starter bursts, and seemed too rich. Had to make a quick stop on the way home, and at coolant temp 189F it started right. Looking at the Crank Fuel vs Coolant Temp correction curve, the area from 73F up to 125F is the steepest part of the curve, where a few degrees in temp makes pretty big changes in the injector pulse width. I had to wait for the engine to cool down completely, so this morning I hooked up the laptop and hit the starter. Again, it did not fire immediately at 83F so I dialed back the closest value on crank fuel a hair and it hit immediately. As I warmed it through 110F I stopped near each break point on the curve, and ended up leaning out three points to get the best starts. What's amazing to me is that these "changes" are soooo small. The injector pulse is in micro-seconds (seconds x 10-6), and by changing my correction curve value from 1.508 down to 1.5 it reduces the pulse width by 0.000028 seconds! That tiny change makes a huge difference in response!

Thinking about it, during crank the air coming into the throttle body is flowing into the huge 'tunnel ram' plenum area, and the injectors are squirting right into the intake runner ports at the heads. Charge velocity is pretty low (near stagnant?), and mixing turbulence is almost nil so I guess it makes sense that a very small change in fuel dumped in can make a pretty big difference in the cylinder mixture. One thing that really helped me out was Jeff at Edelbrock pointing me toward re-defining the breakpoints on my crank fuel curve, eliminating temperature extremes that could never, ever be seen (like -58F low, +320F high) to get improved resolution in the middle of my curve. Once again I think I've got the starting mixtures pretty much dialed in, but I'm pretty sure that I will be tweaking things again in the Fall/Winter as ambients fall around here.

Sorry to get into so much detail, but its intriguing to me how all of the curves inter-relate, and it's a rush when things start coming together!