[mbtoloczko]

Over the last year, I've been trying to figure out how I can get some control over the air-fuel ratio and stop the MAF from maxing out quite as soon. I've come up with a solution involving two items:

1) 2000 Impreza RS fpr
2) A MAF "bypass" tube.

The Impreza RS fpr, which costs about $75, raises the baseline fuel pressure from 36 psi to 43 psi. This, in turn, richens the air-fuel ratio. Comparing my wideband O2 measurements to other people's (see Chiketd's and SVXRide's dynosheets), it looks like the Impreza RS fpr richens the afr from about 14:1 to about 12.5:1 over the 3000-5000 rpm range, and above 5000 rpm, it richens from about 13.5:1 to 11.5:1. This is a bit too rich.

So, I got to thinking about how I could lean out the afr. My first thought was an SAFC-II, but the darn things are expensive and it doesn't do anything to keep the MAF from maxing out at high airflow. I got to thinking about other ways to lean out the afr, and I realized that if I could just add small intake tube that allows some air to go from the airbox straight to the intake plenum at the entrance to the throttle bodies, then some air could get to the engine without going through the MAF. So, the total amount of air getting to the engine is the same, but the MAF signal is reduced because not all the air is going through the MAF. The result is that the ECU sends less fuel to the engine, and the afr gets leaner. Seemed like a cool idea to me. So I tried it. The picture below shows the MAF bypass installed. Its 3/4" ID rubber tubing. The fittings at the airbox and the plenum were bought at Lowe's. Total cost for tubing and fittings was $10. The fittings are cool because they have nuts to hold the fittings to the airbox and to the plenum. I had to use a bit of silicon sealant at the plenum because the plenum is not completely flat. The configuration in the picture will probably not be the final configuration. The elbow significantly reduces airflow.




Next, I got out my LM-1 wideband meter and took some afr readings before and after installing the MAF bypass. The first graph below shows the baseline reading with no MAF bypass. You can see that the afr is around 12.5:1 from 3000-5000 rpm, and then goes even richer after that. You can also see that the MAF voltage is reaching an *average* peak value of about 4.4 volts (which is still less than the 4.8 volt max that it can produce). The second graph shows the result for an experimental bypass that I didn't end up using. The result is quite good though. The afr is leaned out to about 13.2:1 from 3000-5000 rpm and then sits at about 12.2:1 at peak rpm.




Here are the results for two other MAF bypass configurations that I tried. The first config works a bit better than the second. The second one is the configuration in the picture above. You can see that not as much air gets through the bypass because the afr is a little richer. Both configurations give about the same result though. Overall, I'm very pleased with the result. The engine is much more peppy when I step on the gas, and now don't have to worry quite so much about the MAF maxing out with further NA mods (ram air is next). If anyone is interested, I can post more detailed pictures and descriptions of the parts for the MAF bypass mod.