I was out messing around with open loop tuning a few days ago and wanted to see how the predicted AFR compared to my measured AFR with the intention of scaling my MAF so they would be the same.

While I know it is well established that the AFR MAP does not reflect true AFR values, it's surprising how far off they are and it makes me wonder if the relationship chosen where 128=14.7 is actually what the factory ECU is meant to be scaled as?

Also, is MUT32 a value that has all of the fuel corrections accounted for? Essentially making it a calculated value based on injector pulsewidth and airflow?

 

It is somewhat surprising that its off so much, but MUT32 is the target AFR and is used in the IPW calc. The scaling, AFR = 14.7*128/x, is correct. Its explicit in the IPW calc, and its implied from the closed loop AFR map value of 128.

 

I noticed the same thing when logging a WOT pull in 3rd gear. It seems that the higher the RPM the more off the AFR is compared to my ZT2. The Fuel table is far more rich in the higher RPM range than the actual O2 my wide band sees.

What gives???

 

FYI I used a similar method you are describing above to "calibrate" [the MAF to the tables] in the low/mid areas on my 2g and it worked wonders. I gained ~5mpg on that car by doing so and the drivability was much better.

 

As a follow up, logworks seems to be a little interesting on setting up new channels. I thought adding the channel in the protocol.xml file was all that was needed and the scaling would be correct in the logs. This isn't the case however. You also need to configure the channel from the dashboard application to match the scaling. Otherwise, it takes your scaling and messes it all up in the data logs.

I rescaled the MAF by minimizing the error between MUT32 and actual AFR while running open loop. By adjusting the MAF Smoothing table, I was able to greatly reduce the average error. I found that I needed to reduce the low end values ~25%. Beyond 800Hz, I found that the actual AFR and MUT32 had almost zero error with the factory values, so I left them alone. My table AFR values now reflect what I see on the wideband under most conditions.

The results are mixed. With one fuel tank of gas gone through, my mpg has gone from 18 mpg to 22 mpg. I've been beating on it pretty heavily this last tank of gas as I've been messing with the WOT tune as well, so it seems the improvements would not be from a change in driving really as my driving should have reduced mpg.

Warmed up, I haven't really noticed a difference in drivability or idle quality. Cold drivability has gotten worse. The car has always ran a little lean when cold, and now it is even leaner at the low end.

If the mpg does show to be a consistent 4mpg improvement after multiple tanks of gas, I would say a little warmup driveability issue is worth it. Once I got logworks loggging correctly, it was pretty straight forward on getting the tune setup.