Long Post
Oh, what a tangled web us NSXers weave for that endless search of forced induction power optimization. I was driving home from the dyno shop last night with the biggest question mark hovering over my head. Here goes.
When I first got by car back from MB, it dyno'ed at 383 RWHP. The air-fuel ratio was very rich - dipping down in the low 10's in the middle portion of the RPM range then coming back up at redline. It even flat-lined to ten, suggesting to me that it has actually richer than that, but the dyno shop's wide-band oxygen sensor could not read that low (rich). The conventional wisdom suggested that my AFR needed to be leaned out for two reasons - avoiding detrimental cylinder wash over a long period of time and picking up some horsepower. I certainly agreed with the former; however, part of me always questioned the benefit realization of the latter. I adjusted the fuel map of the SS box to eventually yield an average air-fuel ratio of 12.0:1. And talking about a smooth / flat curve, you would be impressed! I'm having trouble opening the Dynojet file, I think I need to download an upgrade.
OK OK, get on with your story, boy. The car now dyno'ed at 342 RWHP. While I was perplexed as to why, and initially I was disappointed, I now am not. Speaking at length with the dyno guy (their shop specializes in Japanese cars) we superimposed the before-and-after dyno curves with the corresponding air-fuel ratio curves. You can see that in the early segment of the RPM range, the horsepower curves followed each other identically. But when the new air-fuel ratio begins to go lean from the original one, that is when the new horsepower curve begins to drop off. He also said what has been mentioned here – Honda engines respond better to timing changes than to fuel changes.
So, what does this all mean?
In my opinion, although I dropped 10.7% on paper, the "seat-of the pants" measurement of horsepower is virtually zero. To soothe the pundits, the reason I can make this statement is because the gradual reduction of horsepower of 10.7% occurred over a time span of many months. It is not like the car was making 383 RWHP on Saturday then 342 RWHP on Sunday.
How do I get the "lost" horsepower back? I have three options. Restore the fuel map to the original rich condition, install smaller pulley to increase boost or dial in some timing advance.
Rich Fuel Map - an option not to be considered because of the aforementioned concern of long term cylinder wash.
Increase Boost - changing pulley and belt takes about twenty minutes. Oh, but now I will have to re-tune. Remember what I said, this is a tangled web weaved.
Add Ignition Timing - uncharted waters for most, especially with the SS box. How much is too much?
I will begin with option #2 followed by option #3.
In closing, I'll offer an anecdotal opinion. I am not inviting flames, only healthy "debate" and exchange of ideas. I think that NSX engines boosting 6-8 PSI and making upper 300 RWHP to 400-ish RWHP, have too much timing advance to be generating that much horsepower. Obviously, this comment is predicated on an average air-fuel ratio of 12.0:1. Kendall - are you reading this? I know, I still owe you the promise of the Honda scanner results.
Why do I make this comment? A few weeks ago in another thread, Miami's NSX with Novi 2000 and AEM produced "only" 340-ish RWHP. The pundits landed on him pretty hard by saying that the engine should be making some 15% more power. Armando defended his configuration along the lines of, "Hey, I'm happy so why do you guys have to critique my results?" Interestingly enough I was wondering with an air-fuel ratio of X and boost of Y and timing of Z, just what should a forced induction NSX engine be making?
Finally, with regards to timing. Armando's AEM has the timing map set to be at 18-19° of advanced timing at redline (peak boost). I believe that most will acknowledge this is a conservative value for NSX forced induction. As I increase the boost and trim the air-fuel ratio to maintain 12.0:1, by holding the timing advance to 18-19°, the horsepower will be what it will be.