Parts Changing Experiment Results and Questions

Yeah I know, don't be a parts changer!!  What happens if you do swap a part and then want to test to see if it is an improvement?  What I had noticed with my 2AZ-FE was it was using oil at a rate of 1 quart in 1600 miles, so I started monitoring very closely.  I also started watching fuel trims and noticed TFT was biased slightly positive around 5% total.  I had a spare brand new MAF sensor and swapped it.  Oh, the sin.  With the new MAF, I collected some data.  The new MAF is made by Wells, I believe.  The old MAF is a Denso.  What I notice is total fuel trim for the new sensor is closer to zero, but the range of highs and lows is twice as big.  I based these conclusions mostly on the trend lines I got from the data after putting in a spreadsheet.

One question I have concerns what effect will the bigger swings in TFT have on say MPG?  Is the bigger range of high and low TFT a concern? 
   

Here are two charts of the same data with different trend lines used.  I think the new charts show the swings in the TFT better.
   

I'm trying to determine which sensor is better, more accurate, but I have not posted any actual data from either MAF.  The first attachment is the Wells sensor and the second is the Denso.  It looks to me like the Denso under reports air flow in comparison to the Wells, but maybe the Wells over reports.    

It finally dawned on me that I should reset the computer so it can calibrate to the new sensor.  I realized this when I did a Volumetric Efficiency test and TFT was biased about +10% and the only thing that has changed is the sensor.  The MAF g/s is a large factor in the VE calculation and it hit 115% with this new sensor while 105% is the most it ever hit with the Denso .  I think 115% unlikely from a 2AZ-FE plus the wild swings in fuel trim hints that the computer is not getting values it expects from the MAF.

I saw a video from The Car Care Nut and he talks about how the MAF is fired from the parts cannon more than any sensor.  He goes on to explain how to diagnose a MAF with any scanner that will show live data.  If calculated load with foot to the floor goes above 90% then the MAF is good.  Scanner Danner has tons of videos on diagnosing a bad MAF and one condition he mentions is that the sensor can be skewed which I understand to mean it could be good at lower RPMs and bad at higher or vice versa.  In this case, I'm pretty sure both sensors are good but the Wells is fresher.  

 

MAF diag can be tricky depending on just how bad the sensor.
I've recently had a run of bad new MAFs that seem normal at one range but misreport at another causing fuel trim im balance codes.
I agree, the wells looks to report over a wider range conditions.
Thanks for sharing your findings!

Be care full using the CALC load PID, depending on induction design (turbo, supercharged) you can get 100%, but the ABSOLUTE LOAD, which is the engine's VE% can be say 60%

Absolute Load is the best indicator of breathing efficiency when available.

 

Thanks for the warning. 

I calculated Volumetric Efficiency before and after the parts swap, just didn't post it and showed Calculated Load instead.  I've been driving it since and it seems the ECU has adapted.  I checked fuel trims yesterday after driving about 30 miles and STFT was 0 - 1.6 and LTFT was -2.3 at idle.  I was seeing some big swings but Car Care Nut's video convinced me not to sweat 5-10% swings in fuel trim.