>>>I wonder if the differences in HP could all be attributed to drivetrain losses. Can't see anything else, really, but as I think I can only hit about 160HP in 1st, that would mean a 33% loss. This could actually be useful and even challange the common wisdom that shifting at the rev limiter is the best way to optimize acceleration. <<<

As gears are further from 1:1 losses rise. Also, MOI effects become more of a factor.

For best accel you stay in a given gear until the next gear is better or you run outta revs. For example if you plotted the gs for each gear versus road speed , where they cross would be a good shift point. Cuz then you see that you'd enjoy more gs than if you didn't shift. If they don't cross, you stay until just before the limiter, which is likely the case for an s2000, at least in the lower gears. For best 0-60s it is sometimes best not to shift but to hold a gear. On my M3 I can hold 2nd to just past 60 for a best 0-60 result. But for best accel through that speed, I shift slightly before 60 as my shift RPM is achieved. That gives me the best time to distance. In other ords getting the best time to a speed may require different tactics than getting ahead of the guy next to you.

Stan

 

If you look at the dyno plot (below), you'll see that we don't suddenly gain torque, it's gradually built after the "angry" cam kicks in. Luis's G-force plot just confirms what we all know: your butt-dyno measures torque.

 

Am I the only one wondering how these plots relate to throttle position and actual time rather than RPM's?

If you look at the torque curve posted above, you don't see a huge increase, but the curve is graphed across RPM's, not time. If you were to tie each RPM to a millisecond time with all your weight on the throttle, I think you'd notice that the torque curve when plotted over time in ms would look slight different- there would appear to be an abrupt climb in torque because we've all seen how quickly the RPM's increase from 6000 to 9000 when compared to 3000 to 6000.

Maybe I'm missing something, but the feeling of torque is related to the rate of change in RPMs. You obviously won't feel the torque increase if you increase RPM's slowly. And I suppose it could be my perception only, but it does seem that the RPM's increase much more quickly after the cam changes.

-B

 

[QUOTE]Originally posted by Ludedude
[B]Cool "scientific" analysis.

I'll tell you this, from a purely "seat of the pants" feeling...my previous car was a 1999 Prelude (H22A I believe)and the VTEC changeover was much more pronounced, almost brutal, by comparison to the S.

 

My old ITR had a very pronounced kick when Vtec cut in. You had to watch it on the track, or it would catch you out. The S2000's much smoother - seems to build up puff over a few thousand revs, rather than getting a big kick that only lasts a few revs.

One interesting thing, since the ECU adjusts the vtec point a bit, have you though of going into vtec on part throttle and seeing what happens? My butt dyno says that going into vtec like this is much smooth, but I don't trust it

I was thinking for a second that Honda tuned the airbox, to get a disproportionate increase in induction noise when you hit vtec. But then I remembered Luis's has a Mugen CAI.

This is an interesting test. I wonder what things are going to be like with Toda's cams?

-Brian.

 

[QUOTE]Originally posted by bguernsey
[B] ... And I suppose it could be my perception only, but it does seem that the RPM's increase much more quickly after the cam changes.

 

[QUOTE]Originally posted by TXRacer
[B]If you look at the dyno plot (below), you'll see that we don't suddenly gain torque, it's gradually built after the "angry" cam kicks in.

 

[QUOTE]Originally posted by Elistan
[B]

A dynojet measures wheel torque in much better controlled conditions than a car-mounted accelerometer measuring first gear.

 

[QUOTE]Originally posted by Luis
[B]

I think you are missing two facts:

 

You need to think about how an accelerometer works. For these specific measurement events a sample was taken every 10 miliseconds. As the car accelerates it experiences all sort of bounces, and dips, that will be recorded by the accelerometer, and that can, over multiple samples, produce swings of say 30%. However, in the end, it has to even out, as the time it takes to accelerate is pretty even over the same distance. I could have used the smoothed data that the unit computes and you'd see two perfectly matched curves.
If I measure the time it takes to go from x to y krpm, on the same stretch of road, I get results repeatable to the 1%. That's how accurate the instrument is.

[QUOTE]Originally posted by Elistan
[B]
...
2)