[tjts1]

I decided to copy and paste the article because they have a lot of annoying pop up ads. But if you must, you can read the original here:
Fuel Economy Tuning - Tech Review - Honda Tuning Magazine

Usually you'll only modify your engine's computer to achieve higher engine output, but with recent gas prices it's worth looking at your ECU to see what kind of gains in economy you can get. Since the ECU controls fuel and ignition events (as well as cam timing for the K-Series engine), we can optimize these engine parameters in order to maximize fuel economy and power.

There are, however, a great many factors which affect fuel economy, including vehicle size, shape, drag, rolling resistance, engine efficiency, engine size, gasoline composition, driving patterns, driving style, and even the time of year, so ECU tuning is not a silver bullet when it comes to fuel consumption. Small efficiencies, however, do add up.

Tools

We used an RSX to get an idea of the optimum ECU settings for economy. A common way to get a rough idea of fuel economy is to use a pressure gauge to monitor manifold pressure while driving. The lower the manifold pressure (or the higher the manifold vacuum), the less fuel is consumed. Manifold pressure only gives an approximation of economy however, and we decided to use a more controlled environment for measurement.

We placed the RSX on a Dynapack chassis dyno set to a constant load, equating to the approximate load of the RSX cruising at 65 mph in top gear (about 50 kPa manifold pressure or approximately 15 inches manifold vacuum). On the dyno, we ran the engine at the matching rpm (3000) and the fuel consumption was measured and averaged over a time period.

It must be emphasized that every car is slightly different, even for the same make and model. The test results for this vehicle may not necessarily apply to every other vehicle, every other Honda or even a Honda with the same engine and setup. It should also be noted that there are significant differences in fuel, both worldwide, from state to state within the US, and from summer to winter.

Air/fuel ratio

Because of the requirements of a three-way catalyst, almost all vehicles run as close to stoichiometric as possible at light and medium load. Stoichiometric is the 'chemically correct' ratio involving the complete combination of all the fuel with oxygen from the air. We can guess that running a leaner air/fuel mixture will result in less fuel consumption, but there are several factors against doing this.

First, the actual fuel savings are fairly slight. Secondly, the catalyst will not reduce nitrogen oxides efficiently with a mixture leaner than stoichiometric. Lastly, the combustion and exhaust gas temperature increases to the point where it may damage the engine.

This graph shows relative fuel consumption and manifold pressure versus air/fuel ratio. Note that the manifold pressure increases as the air/fuel ratio is made leaner. To obtain the same power output with a leaner mixture, you have to use more throttle. Generally, fuel consumption decreases as the mixture is made leaner. Eventually the increase in manifold pressure offsets the reduction in fuel flow and fuel consumption increases once the mixture is leaned past 15.7:1.

At the stoichiometric air/fuel ratio, fuel consumption is around two percent higher than optimum. Air/fuel ratios richer than stoichiometric sharply increase fuel consumption so that running 13.5:1 results in eight percent more fuel consumption than stoichiometric. This is why it's important to have a correctly functioning oxygen sensor.

Timing

Ignition timing is also subject to restrictions from both catalyst requirements and engine longevity. Decreasing ignition timing results in lower combustion gas temperature but also raises exhaust gas temperature. On the other hand, increasing ignition timing lowers exhaust gas temperature, but eventually gets into the realm of pinging. Extremes either way also run the risk of misfiring and incomplete combustion.

This graph shows relative fuel consumption and manifold pressure versus air/fuel ratio. In this case, both manifold pressure and relative fuel consumption react in the same way to changes in ignition timing. The optimum timing is in the low to mid 40 degree range, with a much steeper increase in fuel consumption from too little timing compared to too much timing. For reference, the stock ECU runs 47 degrees timing at this load, which is close to optimum.

Cam angle

With the Honda K-series engine, the intake cam angle can be varied for both maximum power and, in our case, to reach maximum economy. Unlike the ignition timing and air/fuel ratio, cam timing is not subject to restrictions for emissions or engine longevity reasons.

This graph shows relative fuel consumption and manifold pressure versus intake camshaft angle (in crank degrees). The results of this test were not as predicted. The manifold pressure responds to cam angle be decreasing from full retard (0 degrees), reaching a minimum pressure at around 15 degrees and then slowly increasing as the cam reaches full advance (50 degrees).

Fuel consumption, however, does not follow the same curve as the manifold pressure. Full retard results in the most fuel consumption and full advance appears to have the least fuel consumption. But for everything else, there isn't a smooth relationship between cam angle and fuel consumption. We believe that this is from standing waves on both the intake and exhaust side creating differences in the way the engine runs.

It's also possible that in this case high manifold pressure doesn't necessarily mean that fuel consumption increases. Higher manifold pressures require wider throttle openings, which can reduce pumping losses for the engine. The bad news is that it's also certain that these results are fairly specific to the exhaust setup on the test vehicle. More than likely every other vehicle will be slightly different.

Conclusions

In our case, we could make up gains of around two percent with fine tuning and cam timing. The factory settings were not too bad. If we could run the engine at 15.7:1 air/fuel ratio, we could realize about two percent (as outlined above however, we don't feel that this is a good idea). There might be additional gains in areas where we did not make measurements (cold start air/fuel ratio, for instance).

Is it possible to tune the ECU both for economy and maximum power? Certainly. This is a point of common misconception. Since the ECU uses different areas of the fuel, ignition and cam angle tables, it's possible to run different settings for normal running and full power.

I found this article interesting because for the last few months I've tried various air fuel ratios on my megasquirted mercedes with little success. Everything from 15 to 16.5 afr in .2afr increments during cruise conditions (1500-3500rpm, below 60kpa manifold pressure). While there was a pronounced loss in throttle response and increased manifold pressure (as expected), there was no FE improvement above and beyond 14.7 afr. MS can data log the the injector duty cycle very accurately in these situations. This article sort of confirmed my own real world experience and eventualy I settled on 14.7afr during cruise. Even at the same air fuel ratio as the original Kjet mechanical injection system was running, the car burns less gas for other reasons (no Kjet mechanical air flap, much faster switching time, more accurate and balanced injectors, improved multiple nozzle spay pattern etc). I'm hoping once MS has full authority over ignition timing and I upgrade to coil on plug ignition, I can try to lean out the AFR again with better results.
cheers
Justin