For what it's worth, here is the dyno plot of the best of four runs today. HELP!! I can't post attachments. Until we get this resolved, here are the results:

HP Torque
Stock 205.7 136.8
Mugen Header/Exhaust/Injen CAI 197.7 138.4
Mugen H/E/ and ECU / Injen CAI 197.5 140.2 with Mugen thermostat/fan switch/rad cap

These runs were all measured on a DYNOJET Model 248C dynamoter at EIP Tuning, 1203 Baltimore Blvd, Westminster, MD 21157, Phone 800 784-8100, featured as Baltimore's best tuner by Baltimore Magazine. All runs were duplicated in both 3d and 4th gears and multiple runs at each session returned repeatable results. Large fans were arranged at the front of the car. While the last run was about ten degrees F higher, the barometric pressure and CF were essentially identical.

So, I'm kind of shaking my head in final understanding that the S2000, at least my MY 2002, is hard to improve without forced induction. Hams has written that his Toda cams make good power and I may try them. Or I may take Secret's advice and go with a VAFC.

To bolt on expensive parts and progressively lose HP and gain only 4 pounds/feet of torque is kind of demoralizing, especially since I ran the tank dry and filled up early this morning before the dyno with Sunoco's 100 octane "Race Gas".
I'm not fixing blame on anyone, hell no one made me buy the parts, but it's an expensive lesson to learn. Stock is good: spend your money on brakes, stereo, body kits, or a hard top instead.

 

Was the engine thoroughly heat soaked at any of the occasions that you dynoed it? If your engine (particularly the intake) was hot when you did the runs, it is entirely possible that the additional 10 degrees of ambiant temperature might have caused the ECU to retard timing. The dyno will "correct" the figures for difference in ambient temp, but not for retarded timing.

 

gernby, according to Scott the Mugen ECU is not so sensitive to heat in retarding ignition timing as the stock ECU. And it was a cool morning, I had not run the car hard and there were fans blowing over the engine the whole time EIP strapped it onto the dyno. And remember, I had a tank of supposed to be 100 octane gas. I posted an analysis of octane vs load/temp/detonation a couple days ago temp should not/not have been an issue today. Thanks for asking.

 

Well, it was just a thought.

I am not really surprised that the mechanical mods robbed a few points, but it REALLY seems odd that the Mugen ECU resulted in a loss (albeit insignificant). I would have expected that the Mugen ECU would have optimized things at least a little bit for the added air flow!?! Doesn't it ignore all the OBD-II sensors?

 

[QUOTE]Originally posted by gernby
[B]Well, it was just a thought.

I am not really surprised that the mechanical mods robbed a few points, but it REALLY seems odd that the Mugen ECU resulted in a loss (albeit insignificant).

 

What does the area under the curve look like? How much was gained/lost in the mid-range ...that is what truly makes the difference.

 

Sorry to hear that the Mugen ECU didn't solve your HP loss problems
For what it's worth, I'm sure you could sell your Mugen parts used and get a good portion of your money back. Heck, I would take the Mugen ECU off your hands at a reasonable price if I had any money right now.
BTW, if you do get cams, remember that you can't run the Mugen ECU with them (I think you already knew this, but just to be safe).

 

seriously, people should be crapping their pants at this news. If this is true (no reason to believe otherwise) this shoud make people really think twice about modding the S2000 in N/A form. Buying headers and seeing no gain at all is just inexcusable. If your headers do nothing don't make them. I can't believe mugen has pawned these off as some kind of performance upgrade. If it had added nothing that would be bad enough, but to lose 10 HP at the top end is terrible. Unless the cuve added like 10% area or something I can't imagine anyone being happy with this situation. If something like this happened on any other car from a reputable tuner company people would be screaming bloody murder. Maybe in the Jap. import market it is ok to spends 100's if not 1000's of dollars and lose performance, but it isn't OK anywhere else!!

 

Lots of things to talk about.

First of all, coquinn brings up an important point. You hve to look at the overall curve. Peak numbers are good for bragging, but average area under the rpm range where you spend most time accelerating (6000-9000) is what's most critical. When I put the Mugen header on my car (no other mods) I consistently lost 2 hp at peak (202 to 200). However, I gained 4 lbs-ft of torque at peak, and picked up as much as 6 lbs-ft elsewhere. Torque gains of 3-5 lbs-ft were evident from 2500 rpm on up. From 7000-8000 the curves were identical, and then header curve fell off a little faster at the top. Quite a worthwhile tradeoff in my book, but your needs may vary.

Second of all, anytime you change more than one variable you risk messing up the experiment. Now, this isn't critical scientific stuff, but to be truly accurate, you've got to check it. I found it interesting that the injen intake showed no benefit. With just the headers and exhaust, I can easily believe losing a couple hp up top and gaining midrange. But I would expect the Injen to show large gains based upon previous dynos.

For me, the change in gasoline is a big one too. Good race gasoline can really help power, even if you don't need the octane increase, but you have to see what works for you. If you didn't baseline with race gas (and I'm making the dangerous assumption you didn't 9906, correct me if I'm wrong), then the results of the mod tests are less useful.

Finally, even though the Mugen ECU may be less sensitive to heat, you have to be veeerrry cautious about coolant and intake temps on the dyno (unless it completely ignores temperature - can that be true?). I spent 5 hours last night tuning Team Hybrid's Speedcraft turbo S2000 on my Dynapack. Things were pretty messy to begin with, but we got it working right. However, during the test, we carefully watched IAT and coolant temps using an OBDII scantool (a must for proper tuning IMO). If we let coolant temp climb above 190 F, mixture would get rich fast and power would go down (190-195 F could richen the mixture by a half point easily). These results were consistent and repeatable, as they were on my normally aspirated S2000 tested earlier in the day. Even with 70F ambient temps and a large fan, 3 runs in the space of a minute or two would send temps well over 200F. So, we tested only when coolant temps were between 183 and 187 F. Took longer, but made things much easier. BTW, underway on the road in 65-70F ambient weather, we never saw coolant temps above 190F until we made a 1st to 4th gear full throttle run (in the turbo car), so we felt the temp we chose on the dyno was appropriate.

In summary, I'm not saying that the Mugen parts do or don't work. I like the header for the midrange boost it gives, but I haven't used the exhaust or ECU. What I am saying is that you have to look carefully at all mitigating circumstances when testing parts on the dyno (or track).

UL

 

....what UL said

on the subject of header (singular form please), different header designs have different impact on the overall power curve. comparing the spoon header with mugen header (generally speaking) you'll note the difference in mid and top end. mugen header REALLY shines in the mid range where normal driving are usually done, spoon header's top end power (post peak) is great, and you can see clearly see why spoon chose that type of design on their header.