What about timing vs. torque? Would that narrow it down more? Is there a timing point that gives higher/flatter torque curve even if lower peak hp?

 

Since HP is a direction function of torque, at the same rpm, you don't trade torque for hp or vice-versa. Now, since torque peak is at a different rpm than peak power, you certainly will have different optimal timing at that point (and thus need to shape the timing curve, not simply move it flatly up & down).

Sorry I didn't update this thread after the dyno - I had done it in several other threads.

I have reasonable confidence in the wideband products- I had the LM1/Innovate & it is not auto-recalibrate, as you say-- so if one was to forget, after a long while maybe you'd be off. Now, since cars vary in what AF they make best power, I see the wideband as a first indication and then you need some way of tuning for best power (in a reasonable AF range), whether dyno and/or track.

 

Timing is a "moving target" - aside from the"static" loads that a dyno provides (relative to the "real world"- on the road), picking a value and placing it in a table is only good for that specific time domain, and IT IS STILL AN AVERAGE PCM BEST GUESS! One cylinder may be lean one rich, and the advance may be to far for one and too retarded for another - but only for that few RPM's or if the fuel, load, or conditions change with the next intake charge, or the spark moves to a different table in a fraction of a second later, where now you have a completely new baseline.

Most of what is being done with all these edit programs is just "timing by eye" - only a Neanderthal improvement from "timimg for ping" as was common in the 1960's.

Just my $0.02.....

 

For any given mixture and engine geometry and
charge temperature there exists a MBT point
(maximum best torque) for spark advance.

The delivered torque from a single power stroke
is the integrated cylinder pressure on the piston
area, applied through the crank lever arm at its
sine projected length. This is a dynamic length
and a dynamic pulse. Contribution is maximized
at 90 degrees from TDC by the lever arm angle,
by pressure peak and by time-near-peak for the
burn. Engine components' stress limits cap the
livable pressure peak. Compression ratio, lean
mixture, higher charge temps speed the burn,
octane and lower temps slow it. Many variables.

Knock tells you that you have fired too soon.
Some of the energy is being wasted on hammering
the moving parts. Firing later (less advance) will
move the pressure pulse toward the "sweet spot"
of maximum torque contribution.

So long as the pressure pulse (mostly) develops
after TDC and the burn is overwith before BDC
of the power stroke, the expansion energy
captured to the flywheel will be somewhat less
timing sensitive. But another 10HP stands to gain
from the last 3% worth of spark timing optimization.
Particularly if the spark timing positions the pulse
peak way late from 90degATDC, pullng it forward
can really improve the power added efficiency.

Conventional wisdom is that MBT lies just prior to
onset of detonation. It's a "horse race" that keeps
the piston's expansion just ahead of a detonation
back-wave and in plain combustion instead.

If you can log your engine's delivered torque PID
you can evaluate best timing on a relative basis,
yourself, against any spark dithering you might
try.

 

As to delivered torque, I monitor it, but I have a tough time deciding on how accurate it is. It is better than most think, but I don't feel comfortable with it until I know how the computer calculates it.

 

I don't trust the absolute accuracy (in any of
the scan tools or in the PCM's ability to make a
super accurate computation). But from one run
to the next, there is very little to change and
I think the relative difference is very usable as
tuning feedback.

 

Yes, but what, for example, if the PCM's calculation assumes a certain amount of fuel (say inj pulsewidth) makes more power to a point... and you go in that direction, the pcm gives a higher delivered torque, but you in reality are running too fat...