I would actually recommend that you leave the closed-loop ON while you dyno tune. Provided that you have enough time, do the following procedure:
1) Do one pull and discard the numbers. Your first pull will always show more power than subsequent pulls, so the number cannot be used for tuning (but you can brag to your friends).
2) Set your target A/F to 12.5:1 and do a "baseline" run in closed-loop, and do a log that includes the % of WB correction.
3) Note the % correction, and make slight tweaks to the VE table. This should reduce the % correction on the next run.
4) Make a "trial and error" adjustment on the target A/F ratio. For example if you were at 12.5:1, try changing to 12.2:1.
5) After a few minutes of cooldown, do another pull. Now take the GRAPH of this pull (not just the peak number) from the dyno printout and overlay it against the "baseline" pull to see whether you are making more or less torque (which is more important than HP) at each point on the graph. Figure whether torque went up or down at each point.
Note: You *might* want to have different A/F at different RPMs. That's why you look at the whole curve, and not just the peak number.
6) Tweak the VE table again slightly to reduce the O2 correction %.
7) Keep making pulls and overlaying the graphs until you are satisfied that your A/F is making maximum power (and your oxygen sensor is making < 5% correction).
Now move to timing.
8) Starting around 34 degrees, try adding one degree of timing to see if torque increases or decreases.
9) After a few minutes of cooldown, adjust timing by another degree to see if torque increases or decreases.
Note: Optimum timing may or may not be just less than where you don't get pinging. For example if you get knock at 37 degrees but not 36, you might still make optimum power at 34 degrees. (This is especially true for nitrous engines)
Note: Again, you might want different timing at different RPMs. In my car, I run 1 degree less timing above 4400 RPM than below.
10) After you have found optimum timing, try one last run making the A/F ratio slightly richer (i.e. move from 12.2 to 12.0) and see if that helps your torque. If it did help, it means that you were slightly lean when you were advancing your timing, and the slightly richer mixture will help you get a tad more torque.
This whole thing will take you 10 - 15 runs, but you'll get your timing, VE settings, and A/F ratio exactly where you want them.
A few notes about running on the dyno:
- Allow the engine temp to stabilize between runs. At least, you should have a big fan blowing on your radiator constantly. The dyno I go to has my radiator sticking right out the front of the garage door, so the operator lets me spray water from a garden hose on the radiator. (The engine creates of lot of heat during a run, and you want this to fully stabilize before your next pull so you don't have excess heat and get some abnormal detonation.) The engine should be idling during cool-down so that the water gets circulated.
- Run the same fuel as when you run the track (or your illegal street races or whatever). Make sure it is fresh.
- Make adjustments to your transmission if necessary to avoid accidental downshifting during the run. Downshifts can create "spikes" and false readings because of the sudden change in gear ratios (the dyno's torque calculation depends on a constant gear ratio during your pull).
- If possible, do your pulls in 3rd gear (for automatic transmissions) so that you have your most efficient 1:1 coupling.
- If the dyno has a Wb sensor, use it only as a point of curiousity. Since your built-in WB sensor will probably not match the dyno's, and your car will only be using its own sensor in the future, that is the sensor that is most important for tuning.
EDIT: I just saw that you said you don't want to run 50 times on the dyno- therefore just adapt the above procedure as best you can during your limited runs. I think you'll end up around 34 degrees of timing and 12.2 A/F ratios, but every engine is different. I'm guessing you'll also want 96 - 98 octane because of your 10.5: compression ratio, but this depends on a bunch of factors (like cam design).
-Bob Cunningham
