For those that may be interested in MAFt vs. Stock 2G MAS comparisons, I had done one in the 2003 season. The info is all in a 4 page thread on DSMlink, but I will provide the graphs here as well as some of the discussion, within reason. Its also worth noting that this comparison was done around 45 lbs/min, and doesnt really show some of the choppiness you can get in the MAFt's output. At 65 lbs/min it is much more pronounced. I will add a snapshot of the airflow signal later when I can get the logs off the laptop again, or you can check the DSMlink screenshot in the AFR page. Here is the DSMlink thread info:
Graph of stock MAS and MAFt
I recently put the newer version of the MAFt in my car. I had calibrated the original one I had to the stock MAS, and didnt think I would need to make much of an adjustment for the new one. Wrong I had to turn the knobs in the opposite direction I had them before.
I went to the dyno, and my AF ratio was very odd. From when boost hit (4500) to 6k, it was too lean, and from 6-7k it was too rich. It looked like an S curve. So I had fuel sliders matching that curve, with plus 3% at 5k dropping down to -3% at 7k. It was much better, but still not quite there.
I turned down the WOT knob one click (5%) to get both sensors to peak at the same airflow, so they are "scaled" the same. I logged both the Stock MAS, and the MAFt (with both sensors physically installed to hold the intake path constant) in third gear. Though I think I started at different rpms, or perhaps its the difference in lag. In fact I'll bet that was it. At any rate, it was interesting to see what some time with Excel revealed. I chopped it down to a simple graph, starting at 5500 rpm, where both logs showed that airflow was fully ramped up. (Perhaps tonight I can post the full chart that starts at a lower rpm)
So before it would read low belw 6k, but after 6k it rose above where the stock MAS plot is. I had dropped the WOT knob 5% to get them both scaled the same, to 45 lbs/min peak. Putting in the correction numbers on the bottom in DSMlink should get me close to where the stock MAS was... But I'll be borrowing a WB this week to get it fine tuned.
Alright, heres the full one. It also occured to me that the rpms rose at different rates, and I just scaled it so the redlines match up. That could also explain some of the midrage discrepancy.
There is so much info in that thread I just realized it would be impossible to get it all captured here. For those with DSMlink access, here is the thread. It was in this thread that myself and a few others found the baro issue with the MAFt, tested it, and found two solutions (clamping Baro/IAT in DSMlink, and sending an artificial (but correct) Baro signal to the ECU). The MAFt should have been "fixed" since this discovery, but it's important to be aware that this issue exists, in case it pops up again. I was ultimately able to correct this issue with the baro signal. By getting that right at 3.9 volts/~30"HG, everything fell in line. With the MAFT zeroed out except for going one click leaner at WOT, it emulated the stock sensor almost perfectly. So the MAFt is quite close to the stock sensor in my case, as long as the baro signal is what it was intended to be.
The MAFt is a great product, and I couldn't run without it (Tial BOV and well over the stock sensors limitations). But flaky baro signal and choppy airflow signal seems to be the norm. The benefits easily outweigh these minor issues though, in my opinion.
