The content below is a mirror of http://www.dalmotorsports.com/cars.asp. I did not author the content. I'm just providing it here since the original site does not seem to be up much of the time.
						
						


First I want to thank everyone who has contributed to this information; race teams, friends, news groups and my own R&D by product failure or success. I’ve listed topics in no particular order, other then thinking through the car from front to rear. None of the opinions within this document should be considered or used without personally assuming full responsibility of the potential consequences related your actions or use of knowledge.
Use at your own risk
The NSX community is unique; it’s made up of a high percentage of active car enthusiasts. I frequent other groups because of other interests, and have found that most of them are not as strong. Be grateful that we have a place like Prime that is so active you can almost consider it live, rather then needing to wait weeks or even months for answers or parts. I’m writing this because I want to give back to the community what I learned. My Race partner Vaughn Duarte shares this sentiment and has encouraged me to write this. It took a considerable amount of time to recall, review and write, so try not to get offended as this only needs to be a guideline of my opinions… and if yours are as strong, you can draw your own.
[NOTE: this document contains information about the DAL Motorsports Grand-Am Cup NSXs and the RacerXtream World Challenge Racing NSX. Both cars and series follow a totally separate set of rules. As not to be confused, we have tried to note as often as possible what areas apply to just one series/car.]
						
The front and rear bumper skins do not need the heavy metal crash bumpers for structural stability or mounting. (For racing only, and then this still assumes personal risk) Weight is the enemy of a good performing car. Even if you’re under weight in class, a rule of thumb is to whenever possible keep the weight between the wheels and as low as possible. I have witnessed scenarios where this metal structure caused harm. If removed, the front and rear bumpers are soft and collapsible to light love taps by a competitor, virtually un-noticeable… keeping the bumpers you literally increase the length of the car and taps result in a spin. In a race the metal doesn’t move back easily from a tire or radiator. If you stop 6 inches after the wall, your bumper will permanently push the frame 5 of those inches… with no bumper nothing hard gets permanently relocated. It’s arguable that I’m talking a fine line, and on the other hand I can imagine many scenarios where a bumper could help protect the car, but most of those scenarios would put me out of a race anyway. I also consider the safety angle, but I’m in it to win… and you don’t see GTS cars or F1 cars putting big heavy bumpers on because the sport is dangerous. Something to note: an aluminum rear bumper replacement exists as an OEM part number.
							 Ride 
							height 
							is an important challenge that plays a critical role 
							in the handling ability of the NSX. Lowering the 
							center of gravity on any car will aid its handling 
							ability, but some adjustments must be made to 
							restore Hondas intended suspension design.
							
When 
							lowered, the rear tire’s increase in negative camber 
							and 
							the toe arm rotates to a less then optimal angle 
							that produces what’s called “Bump steer”. This is 
							simply defined as; when the tire bumps up under the 
							load of lets say a turn, the hub attached to the 
							tire rotates further from the point that the toe arm 
							is linked to, so it turns the wheel similar to a 
							steering arm. In the rear the toe link is toward the 
							back of the car, and will develop additional toe out 
							as the tire bumps further. This is not a good thing, 
							you want a car to keep its optimal settings, and the 
							NSX likes a toe-in condition. The answer is to shim 
							the toe arm near level, and where you get the least 
							bump steer during usable suspension travel. When 
							lowering, the front wheels lose their ability to get 
							the negative camber they need, so you will have to 
							buy the Comptech camber kit. Our class requires 17” 
							rims and a spec tire. The car looked like an off 
							road truck(left picture), and we were far from 
							getting down to legal ride height. The problem was 
							two fold.  We are limited to -3 deg camber so 
							lowering the car to legal 
							ride height meant rubbing the inside wheel well and 
							fender. In search of a
							
solution, 
							other then sectioning the NSX into a Trans-Am race 
							car, Vaughn had an epiphany.  He knew that our 
							problems included a couple of issues: a) Our ride 
							height was too high because we couldn’t get the car 
							lower with the 17”, and b) Our ABS was not as 
							accurate because we were using oversized wheels/tires(discussed 
							next).  
							He knew that the early stock NSX came with 15” 
							wheels and with them we wouldn’t have the two 
							issues… but we also knew the stock wheel did not 
							allow the contact patch(width) we needed. We 
							subsequently stumbled onto a miracle. The Touring 
							class had a 15” rim with 
							our spec tire (That measures exactly the same height 
							as OE but also gave us the width of the 17”). We 
							covertly tested this combination in a regional event 
							at Barber. We had something… without even fixing the 
							alignment the car was noticeably better handling. 
							Armed with the fact that the Cadillac team was 
							allowed to pay Hoosier enough money to make them a 
							special 18” tire (While the rest of the class had to 
							run 17’s), Pete Halsmer lobbied our request to allow 
							the 15” rim and tire. The handling dynamic on the 
							car is now incredible. Note; we don’t need to use 
							the lower lip spoiler.
Tire/rims are critical on the NSX because the ABS system is one of the best production based designs. The ABS system is a simple design that makes complex adjustments based on both assumptions and feedback from your wheel sensors. When one or more tires stop rotating by a specific programmed percentage, the system reacts to control the braking faster and better then most of us. Now, depending on the percentage of speed differentiation, it will time or control at different rates as needed. This is a simplified explanation of what it does, but will help explain many of the wrecked NSX’s at track events. If you buy aftermarket rims and tires that don’t perfectly fit the originally designed ratios from front rolling radius to rear radius, then the ABS brain already sees a percentage of variation even when grip is perfect. It will first react too early because if it was designed to react at let’s say 5% wheel speed variation, and your rolling radius is already off by 3%, then it will be reacting at only a 2% variation in traction. It then takes control and makes adjustments based on the tire’s it thinks is on the car. Brake coefficient of friction, and tire traction also contribute to the ABS getting confused, but this scenario is not your typical street car. Self programmable ABS systems exist for race cars, but they are expensive. Large rims are cool, but be happy with your super light 15’s in front… DAL Motorsports uses SSR Competition wheels which are only 8 lbs. And un-sprung rotating weight is the worst place to have weight. On a large percentage of the televised race cars you will see 3-part BBS style rims that look very heavy. The reason for this is that in any racing and especially at the pro level you get a lot of rubbing and bumping. The lower cost 1-part wheels like the ones we use will bend very easy. The BBS wheels look heavy because they are, or at least the part you see. The center section is likely magnesium and the outer is a thinner and lighter aluminum. Ultimately they are nearly the same overall weight, but the 3-part rim is much stronger.
							
							Compliance 
							is best defined as rubber bushings flexing and 
							compressing when put under a heavy load like 
							cornering. The negative result is that all your 
							perfect alignment settings flex to less then perfect 
							because of this rubber. The rubber is there to 
							absorb bumps so the driver doesn’t fell like they 
							are in a junk box, but the rubber doesn’t belong in 
							a race car (Get in one… they sound 
							like rattle boxes). I will say that the NSX is 
							stiffer then most,
 
							but eliminating this compliance can do wonders for 
							your handling. We have 16 suspension locations that 
							fit this compliance category. Comptech is marketing 
							the 4 most important locations. The rear beam and 
							the vertical compliance lock. I’m going to start 
							with the vertical compliance lock on the front 
							suspension; the part is designed to stop this NSX 
							unique assembly from rotating. If you read the hard 
							cover NSX book this assembly is described as a 
							feature that improves braking by counteracting the 
							toe out compliance by producing toe in. You see, the 
							NSX also has another unique suspension design by not 
							only incorporating Non-compliant ball joints where 
							the “A” arm connects to the wheel hub, they also use 
							ball joints where the front of the A arm connects to 
							the frame. This eliminates all but one compliance 
							joint on each arm. Under heavy braking the front 
							tires want to push back and use up all the rubber 
							compliance. Since this compliance is only on the 
							rear of the upper and lower A arms (where it 
							connects to the frame), the tire wants to toe out as 
							the front side of the arm stays stationary. The NSX 
							featured vertical compliance rotates the tire back 
							in to minimize the differentiation. So one might say 
							that it may be counter productive to lock one area 
							without the other. (More on this later) In the rear 
							both sides of the A arms have compliance, in 
							addition to the toe arm. The Marketed Non-compliant 
							rear beams give less compliance, but are not fully 
							Non-compliant. They use urethane that is harder in 
							durometer then rubber, so it flexes less = 
							less-compliant rear beam. Most compliant locations 
							move spherically, not just rotationally. They rotate 
							and twist because they are on different planes. 
							Because of this spherical motion, the harder the 
							rubber, the more it will bind the suspension from 
							working properly… that’s as bad as having 
							compliance. Now just the rear beam shouldn’t bind 
							your suspension, but a job worth doing is a job 
							worth doing right. On a race car all of the 
							compliance points need to be replaced. On dedicated 
							race cars all of these suspension points are either 
							uni-ball joints, heims, or spherical bearings. It 
							takes a lot of engineering over 100 parts and $4000 
							to make an NSX non compliant. But the results are 
							jaw dropping. Note: Be sure to check your rear ball 
							joints for play or slop. If you have play, they 
							should be replaced for better handling.
							
							

Radiators When you’re trying to keep your engine revving between 7000 and 8000 RPM every possible second in a three hour race, the car is past its intended cooling capacity. The first thing we tried was playing with the thermostat, then we purchased a replacement radiator that claimed double the cooling capacity, we added oil coolers, fans, and even built an under-tray to reduce turbulence behind the radiator. Right now we are just under the danger zone with temperature. The best solution is a vented hood, but our class wont allow it. I’m not comfortable sticking with what we have. The way I see it, we have two more options A) I’m not to confident in the OE water pump, so I will try to increase flow with a secondary electric pump. B) I can still replace our radiator with a version where the manufacturer claims 4 to 6 times the cooling capacity. If you have the intentions to get an RTR nose, splitter/under tray, and hood, they have incorporated an angled radiator (Near 7” thick) for more down force. Note: Don’t trust your OE gauge, maybe not even your OE sending location. I’ve been nearly out of water because of a leaking head gasket, and the OE sender/gauge will only see the steam temperatures.
							
							Sway Bar 
							We find that with good horse power the NSX wants a 
							lot of front sway bar. We also run no rear sway bar. 
							If your running a race car I suggest buying the 
							biggest front bar you can get your hands on, because 
							it’s not big enough. The largest bar I’ve seen on 
							the market is 1” or 1-1/8” bent bar, and I know of a 
							few one-off 1-1/4” bars made for racers (See DALI). 
							Realtime Racing used a 1-1/2” straight bar called a 
							knife edge bar (Because of the rotating
							knife-edge arm that acts as adjustability).
							
These 
							bars are commonly used in Trans-Am cars, and are 
							welded to the front part of the roll cage. When I 
							first saw one, I didn’t even know what it was. A 
							straight bar has much more torsional rigidity then a 
							bent bar, and the thicker you go, its strength rises 
							incrementally. When building my World Challenge NSX, 
							Realtime advised me that 1-1/2 Inches wasn’t enough 
							(I’m sure a joke fits in here somewhere… anyway) so 
							I built one out of 1-3/4”. This is the WC car. The 
							Nickel bar running across the front of the car is 
							the sway bar. It’s held in place with a billet 
							bearing block and large stainless spherical 
							bearings. Needless to say again that the 1” bars on 
							the market are not enough, but buy and put on what 
							you can get. 
							Our Grand-Am class posed another problem; we could 
							only have a bar that fastened to the stock location. 
							More hours of Engineering- instead of using the 
							weaker bent bars, I relocated some components and 
							designed overcomplicated billet bearing blocks to 
							utilize a 1-1/4” 
							straight bar. The blocks house spherical bearings so 
							the bar doesn’t have any compliance loss, and won’t 
							bind under chassis flex. The ends are splined for 
							custom bent billet arms. The bars cost me about 
							$1300 in parts,
							
							so 
							I’m not thinking it’s very marketable, but then 
							again winning is priceless to those of us addicts. 
							With all of the suspension 
							points and sway bar having bearings, I can jack the 
							front of the car off the ground and easily make the 
							opposite wheel rise as I lift one side. That’s how 
							it’s supposed to work, if you try this on a stock NSX with compliance resistance and paper clip thin 
							sway bar… it doesn’t move.
							
							
The picture to the left also shows the vertical 
							compliance lock
							
							Brakes 
							are mysteriously good on the NSX, the reason I use 
							the word mysterious is that the fronts are single 
							sided dual piston with relatively small rotors, and 
							the rears are single piston. Compared to the 
							competition this is one of the worst combinations 
							(in theory). Many cars (Even cars like the RSX) have 
							dual sided pistons, 4 pistons, and larger rotors. 
							It’s a wonder we consider the brakes good? I think 
							it is related to the mid engine design, coupled with 
							the fact that we have a race level 4-channel ABS 
							system. Unfortunately when you start racing the car, 
							the weaknesses revile themselves. After you add 
							larger slicks, race pads, and higher speeds, it 
							becomes apparent that the NSX is lacking brake 
							potential! Partially due to the loss of a modern 
							crutch… the ABS system, and partially due to the 
							fact that our “wusy” brakes should have never felt 
							as good as they do.
							
If 
							your rules allow and you can afford it, see RTR for 
							a Bosch ABS system. I will add that one of the first 
							mistakes people make is to fit the largest brake 
							system they can find, 6 to 8 pistons with monster 
							diameter and wide rotors… many times with equal size 
							in the rear. Brake systems should be only as much as 
							you need for the worst of conditions, because the 
							brakes are one of the worst places to have weight. 
							In motorsports like NASCAR, they will change rotors 
							and brake
							systems depending on what they need for that 
							particular track. This is more important then you 
							think. Dyno your car before and after a big brake 
							upgrade, and you will see a horsepower loss at the 
							wheels… and that’s only measuring the smaller rear 
							rotors. Were not all brake experts (I’m not either), 
							but if winning is more important then just racing, 
							you should consult an expert or fund the 
							experimentation yourself to get only as big as you 
							need. Back to racing- I find that when set-up for 
							the best handling, the NSX will want to lock the 
							rears first. You may be able to find a front to rear 
							pad combination that will change the bias in the 
							rear… if not, you will need to add (2) brake bias 
							levers in line for the dual channel rear brakes. 
							Fortunately the brake lines go right past the 
							driver’s seat on the inside of the car. Brake line 
							bias is the wrong way to adjust bias, a balance bar 
							is optimal, but that requires a complex addition of 
							dual masters. This eliminates the power assist, 
							requires bracketry, linkage, and a new way to 
							incorporate ABS. If you’re building a World class 
							car, and the rules allow this mod, it is 
							unavoidable. (If you’re that person, I can provide 
							billet brackets for Tilton dual masters w/balance 
							bar). 
Another subject is Brake pads… The track is no place of OE brake pads. There are a lot of aftermarket pads that are better suited for autocross or tracking. DAL uses Cobalt Friction blue front with Hawk blue rear. The Cobalt’s are a better on the rotors than some aftermarket pads. Speaking of rotors- we can crack the familiar aftermarket rears in one race… don’t even think of using them in the front. Use OE, or even better a two part rotor like Stoptech with curved cooling veins. When you go to a race setup, investigate other rotor manufacturers with the same bolt patterns. I know a famous racer who doesn’t use his *Free* Brembo rotors (About $300 each), he pays near $800 each for Alcon rotors. Material, design, and treating are very important in a super brake system. Brakeman rotors are an in-between to Brembo and Alcon leaning closer to Alcon, and are at near Brembo prices. If allowed, carbon rotors are even better. Fluid- I use Motul if I don’t have heat issues and Castrol if I do. DAL uses ATE Super Blue Racing Fluid. Brake ducts are important in the front (Usually can’t get enough heat in the rear). You want to point the air in the center of the rotor, to be drawn out with the veins. Not an easy task, until you have a race car and can remove some things in the way, and add an under-tray for protection. Back to the ABS system- In a race car that allows ABS, unfortunately you must use it to be the most competitive (even if you need to add weight)! As mentioned earlier they are expensive $10,000 and up, but are unbelievable. When programmed right with good brakes, you can use both feet on the brake pedal as hard as you can, and slow before a turn as good as if you were trying. And if it rains, competitors won’t understand why you and the other ABS cars are lapping them. *Recent information: I’m looking at brake modulation test sheets from Tex Racing (They do more then 50% of the Winston cup cars). Their brake dyno runs from 125 to 80 mph consecutively and records speed, brake torque, air pressure, hydraulic pressure, and rotor temperature. The Lighter and less expensive Brake Man F4 caliper out performed the $2000 each Brembo GT monoblock caliper (Previously the industry standard for best) The F4 ran 30 degrees cooler, required 100lbs less line pressure (On average), made a flatter torque curve, and most important had a much more consistent torque vs. line pressure between runs.(this means less caliper flex)
Brembo monoblock Brake Man
							

							A common misconception is that if you can lock-up 
							your brakes, you have as good of a brake setup as 
							you can get. That’s not a true statement- Here’s an 
							analogy; When using the heater controls in many cars 
							it’s difficult to get the setting from not being too 
							hot or too cold. But in many current cars you have 
							increased adjustability that makes it easier to get 
							what you want. Brakes are the same, all of them can 
							lock the wheels… the better ones have more 
							adjustability before they do lock. If you have two 
							knobs; Knob (A) representing OE or bad brakes, 
							(having) 3 clicks before full. Knob (B) a good 
							performance brake has 10 clicks before full. Think 
							of full as lock-up with your brakes. With knob (A) 
							you can go to click 2 before lock, but one click 
							more and your skidding. Knob (B) you can safely go 
							to click 8. That’s 4/5 braking capacity vs. 2/3 
							braking capacity. In addition to the higher 
							percentage of usable threshold, you have 2 clicks 
							before a locking condition, giving you a 50% larger 
							safety margin.
							Mathematically you can get infinite variability 
							before lock-up, but using realistic technology we 
							have limited solutions to work with. I’m not sure if 
							most people know that lock-up is actually first due 
							to an imperfection of rotor surface, tire flat spot 
							or road surface variation. Your pressure threshold 
							can stay consistent until it hits a high, low, rough 
							or slippery spot. Now with the road surface or 
							tires… just avoid water, oil and flat spotting your 
							tires. But with brakes, the HP manufacturers go with 
							larger diameter rotors so they can benefit from 
							leverage, and don’t have to push as hard because of 
							being towards the center of rotation. They also 
							spread the clamping between two or more pistons. And 
							last, they float the rotors from the hubs so they 
							can absorb imperfections of the hub or the rotor 
							surface. All of these methods help miss the high 
							spot. More driver adjustability (Like mentioned in 
							the first paragraph) will also allow you as a driver 
							to closer flirt with that threshold. ABS technology 
							aids after the lock-up has already happened. A new 
							technology is a relief valve that absorbs spikes in 
							the brake line pressure from those high spots. Some 
							are experimenting with this now.
Clutch Street- I like the OE clutch for its mild manner. Note: if your reinstalling the OE clutch, it not only needs to be initialized, it first needs to be “loaded”. Initialization is the spring loaded release for proper seating. A new clutch is already loaded, so even most dealers don’t know to do this step. They typically install new clutches. This tip might describe why so many who installed good condition clutches had problems putting the trans in gear. One would assume their clutch was bad, and replacing it with new would solve their problem. The 1991 manual explains how to “load” the springs, but it didn’t clearly explain why.
							
							Track car/Race- I like the Exedy single or 
							dual Cerametallic for its strong grip (It may be a 
							little aggressive for many streeters), low weight, 
							balance and long lasting. (Some have complained 
							about T/O bearing noise)
If weight and size is unlimited and you have 
							high horsepower, I would use a mini Tilton triple 
							disk. The track/race clutches are already borderline 
							for using the OE clutch master and slave, so I would 
							suggest replacing them if reliability is important. 
							For the Tilton you will need a new Tilton slave and 
							master. The slave is slick… it replaces the throw 
							out bearing and is easier to do trans swaps. The 
							master requires new bracketry and linkage. (If 
							you’re doing this expensive conversion, I have the 
							brackets. I also have the custom order Tilton part 
							numbers for the slave, flywheel, master and clutch)
Exedy Single w/flywheel
							
Exedy Dual w/light flywheel
							
Tilton Triple w/light flywheel
							
							
							Roll-Cage 
							The NSX is state of the art in design and can only 
							be compared to exotic limited production cars like 
							some Ferrari’s. That’s not to say that other cars 
							like Porsches, or Corvettes aren’t as competitive or 
							even faster, but that’s only because of Horsepower 
							or aftermarket modifications. One of the features of 
							the NSX is the all aluminum chassis, frame, and 
							cockpit.
							
This 
							light weight design is 
							made up of extrusions, and formed aluminum sheet 
							welded together to create a strength far greater 
							then the sum of just one of its parts. The NSX is 
							one of the stiffest chassis sold, and one of the 
							contributors to it’s good handling characteristics. 
							Quoting something I read. An NSX is so ridged that 
							if you put it on 4 jack stands, it will rock because 
							the floor is likely not level enough. Though that 
							has since been proven, the NSX is not ridged enough 
							to fully support race conditions especially with 
							high horsepower… No car is. With this in mind, a 
							cage needs to be designed not just for safety, but 
							for chassis strength. Since your NSX is likely much 
							lighter then your weight requirements, add extra 
							reinforcement to improve rigidity. Especially in the 
							front. I wish I had the picture in mind to insert 
							here- It’s a straight-on shot of the Realtime NSX 
							coming out of a turn with good traction during a 
							race. If you look at the plane of the rear wing 
							compared to the twist of the roof and front nose, 
							you would think it was a top fuel drag car. 
							Supporting evidence would be the new NSX that 
							Realtime built for the 2003 season. Most of you have 
							never seen this car because it never made it to the 
							track, as Peter transitioned to Nissan. The cage in 
							this new NSX was nothing like the first car;
 
							its focus was similar to mine.  (World Challenge NSX 
							pictured) 
Notice the braced tube; this is used as a jack point. If the tires are flat a jack won’t fit under the car, so we keep a steel insert in the pits, that can be put into this tube for jacking the car up.
Another benefit… that apparently happened during a race with Peter Cunningham as a driver- A competitor started pushing or “rubbing”, and this cage mounted device refused to move like the thin aluminum door and cut that competitors tire. This gives a new meaning to the saying “He rubbed me the wrong way”
							 A part I could supply to any future cage builder is 
							a bolt on plate kit. My cage fabricator told me that 
							it would cost me as much money to make all 
							
the 
							mounting plates, as it would to build the cage.
							So I measured and drafted professional looking 
							mounting plates that are laser cut. I put (4) 
							rectangular plates in the cockpit, with equal 
							thickness backing plates to sandwich through the 
							extruded frame rails. Some NSX’s have hole-sawed the 
							bottom of the frame and only bolt through thin 
							sheet. (I don’t believe this is safe or good for 
							performance from rigidity). The JGTC cars are 
							required to weld the cage to the car, so they use a 
							combination of bolts carbon fiber plating and epoxy 
							to get a chemical bond with the chassis. Our plates 
							for the (4) strut towers are also well thought out, 
							and can be ordered for OE style or Moton style 
							shocks. The front struts tie into (2) front frame 
							plates. All plates are pre-drilled for a perfect 
							fit. (See pictures under Sway bar (6)). Here’s a 
							picture showing the detail that went into the 
							external safety controls that we put on both sides 
							of the cars. Corner workers have complimented us on 
							this feature. The third component in this picture is 
							an external port for our data acquisition.
Suspension- If you can afford them, Moton Triple or Quadruple adjustable shocks are the best. I can’t take credit for finding these, but I can tell you why the claim of best is more then opinion. I will start off by stating that more then 50 cars in the World Challenge use them. Truthfully many shocks can be adjusted to work equally, but the difference is what it takes to get that adjustment. Most shocks have a small range of adjustment within the good working range you need for a particular track or condition. Penske brand for instance is an icon in racing shocks. They can be adjusted like most others, but within a short range, beyond that they need to be disassembled, re-shimmed, and then checked on a shock dyno. The Motons creator who started with JRZ (The only other shock with this design), developed a valving using an adjustable orifice rather then shims and shim pressure. The end result allows you to fully tune your suspension, rather then just to the best place before re-valving. So simply… the Motons are better because they are adjustable in the real race world. I’m not Dale Earnhart Incorporated, and have many other things to tune in a short time. And if it rains, you can adjust further and faster then many of your competitors… If it rains during a race, they are dead in the water (While we can adjust while changing tires). Rumor has it that Evringham uses them to tune with, then dynos them on a full range shock dyno, so he can duplicate the valving in the shocks that everyone sees the cars racing with. Motons use spherical eyes at each end, so you will need conversion mounting brackets. I can supply these, along with the eyelet bushings.
Alignment- In the beginning I use to change toe for super speedways that favored long straights vs. tight road courses with lots of curves. But I systematically have proven to myself that the best overall performance for the NSX was to favor the handling vs. top speed. After all it is a handling car, and the alignment changes hardly affected top speed, but drastically effected handling. On a race car- Max out your caster (Max is most amount of angle), because max will increase your negative camber under turning conditions. Also Max out your negative camber… well this is relative to what numbers you can get. Even your average weekend warrior would benefit from –3 degrees in the front and rear. But if you start using offset ball joint bushings and offset spherical bearings, you will see gains all the way to as much as –6 degrees in the front. I must warn you… don’t get fooled by tire temps, or conflicting advice. Try it yourself. As you start going beyond –3 your insides will show high temps because of the straights, but if you take the angle back out to even the temps, you will be hurting your cornering ability. The skid pad will tell the truth. Toe- in the front I suggest 3/16” total toe out, and ¼” toe in at the rear.
							
							Aero (World Challenge) 
							- A big rear wing with a true airfoil design is the 
							highest priority. The wing design will give down
							force at lower 
							speeds, 
							and actually have less drag. Put it higher then the 
							roof, and extend it beyond the bumper 
							
if 
							you can. 
							I would suggest the wing from RTR if you have high 
							HP. While you’re at it their trunk lid also gives 
							down force. Next, I would focus on getting the car 
							low especially in the front, and incorporate a 
							strong splitter (preferably a carbon fiber sandwich 
							honeycomb structure). If your splitter doesn’t 
							double as an under tray, that should be next for 
							extending the flat floor of the NSX.
I also add a solid steel or aluminum plate (Depending on weight), as a further flat bottom extension under the gas tank. This adds additional safety and is the best place for weight. Angle it up towards the back for added down force. A rear angled diffuser would also be good, as the OE bumper is like a parachute without the OE muffler. A less obvious modification would be a vented hood and front nose combination for more down force. And the least obvious are side skirts with a large side extending lip for down force in the turns. Everything I just mentioned has been mastered by RTR, so that is where I would start for the splitter/under-tray, nose, side-skirts, trunk and rear wing. The fuel cell plate I can provide. I would try Science of Speed for the diffuser or rear wing if not Realtimes.
							
							Fuel System- 
							Our stock NSX cell is better then any other OE cell 
							I have seen. It has hinged doors to trap fuel in a 
							small compartment
							
around 
							the fuel pump. This design is made to capture fuel 
							in the turns when the tank is low. Race cells are
							made like this. We can drive past “E” on the gauge… 
							even in 
							race conditions. I have a championship SSC 
							Neon that starves above a 1/4 tank. If maximum fuel 
							level isn’t an issue, I would add fuel cell foam to 
							help minimize fuel sloshing (Not in the small hinged 
							compartment). This sloshing can get a race car 
							loose. For most applications I would suggest the HP 
							Walbro pump, unless you’re above 500 hp. At this 
							point I would use the Expensive Bosch pump. Not many 
							of the drag race pumps will handle continuous road 
							race duty. (They overheat) If you’re running more 
							then a single shot 
							of Nitrous, defiantly replace the fuel pump. This is 
							what happens to a piston when you
							
run 
							out of fuel capacity while using Nitrous (left 
							picture). 
The WC car made big HP and could use as much as 25 gallons in a single race, so the stock 18ish gallons wouldn’t cut it… even if your class did allow stock. I developed a Kevlar design with all the best features that utilizes every inch and fits as much as 26 gallons. It’s fabricated from my drawings by the best in the business. (This cell would require the floor plate mentioned in Aero(12)). If you’re running High HP, use bigger fuel line. A twin pump design would decrease your chance of a DNF. You should cross over your fuel rails in more then one location, to insure even fuel delivery. Return to the tank through a quality regulator and use #8 line. Replace your injectors with 440’s even with minimal mods, go higher when needed. My favorite unleaded fuel that’s commercially available is Phillips B-42. I had a sponsor that funded a dyno test with all the unleaded fuels, the B-42 was best for detonation, and was tunable to make the best HP. It’s also inexpensive, but not likely at your local road course.
							
							Engine Mounts- 
							Even the stock HP will eventually tear the OE engine 
							mount.
							
They 
							are a webbed rubber design that won’t handle much 
							abuse. I have seen 
							continuous exhaust cracks due to their 
							flexibility. You can easily make your own 
							polyurethane mounts, by purchasing a two part 
							polyurethane, sealing one side with good duct tape, 
							level them before pouring, and let them dry. The 
							tape will come off easy, and you can trim any 
							overflow with a razorblade. Don’t forget to put 
							something down to protect from runoff. Etching the 
							OE rubber before pouring would be a good idea, but 
							not necessary as the liquid urethane will seep and 
							harden on both sides of the mount. We can do this 
							for you, but it would cost about $200 a set because 
							of time and about $100 in materials.
							
							Oil Conditioning- 
							Honda uses a specific viscosity oil that works best 
							for street driving, you may have noticed that they 
							suggest a thicker oil in California and in warmer 
							climates. In a racing engine, especially road 
							racing, your temps sustain high heat (As high as 
							350+ degrees) that thins your oil beyond what the 
							factory wants. You may have noticed your pressure 
							decreasing as your engine gets hotter (This is 
							normal). But when racing the oil can get so thin 
							that you risk it’s ability to maintain the 
							pressure/volume needed to do it’s job as designed. 
							Many of your engine parts never touch metal to 
							metal, they actually ride on a thin layer of liquid 
							oil, since this oil is incompressible, all you need 
							is a thin amount provided the volume and pressure 
							stay constant. When the oil thins, it flows out of 
							its restricting gaps too fast. If the oil isn’t able 
							to reach one of the bearing surfaces for whatever 
							reason, the metal to metal will burn before you can 
							say Oh Shi?. This is why any performance engine 
							builder will suggest 20/50 (ie. Valvoline Racing 
							Oils) or other synthetic 15/50. When up to operating 
							temperature it will be the viscosity the engine was 
							designed for. Note: always bring your engine up to 
							operating temperature before putting it under heavy 
							load. I didn’t always have the patients to wait, and 
							would fool myself into thinking the engine is warm 
							already, so I use a 
							unique sending unit in my radiator (A trick I got 
							from a veteran racer), the sending unit senses 6 
							pounds of pressure before the light (I keep on my 
							dash) goes off. See, water temp isn’t always a good 
							indicator that it’s ok to push the engine, but 
							you’re OK when the water actually gets the engine 
							and system hot enough to raise the pressure.
							
Note: 
							The OE oil pressure gauge is another one that I 
							wouldn’t trust. The temperature of your oil can also 
							climb way above your water temperature, and 20/50 
							can even get too thin. The other side effect is that 
							the hot oil helps over heat the water cooling 
							system. You should defiantly add an oil cooler. My 
							favorite choice is an 85,000 BTU water to oil 
							cooler. It’s both expensive, and will require custom 
							engine adapters. But it’s the only cooler that will 
							allow me to over cool, so I can adjust the 
							temperature depending on track conditions. See, too 
							cool will rob you of power, and too hot will do the 
							same. My target is around 250 degrees. The cooler 
							I’m using is the only water to oil cooler that has 
							the capacity to handle the NSX. And any air to oil 
							cooler wouldn’t allow me extreme adjustability.
							
I would also highly suggest an Accusump to any NSX’er that tracks their car. When in sustained V-tech, most of your oil is in the engine and not in the pan. If the oil pick-up in the pan starves for a second… your engine is done. Preventative insurance (Especially needed if you track High banked ovals) get an Accusump! An Accusump holds more oil (our cars have 3-qt versions), the oil is pressurized by the stock oil pump against a piston and spring (Like a syringe). If the pump stops pushing for even multiple seconds, the spring will push back to keep your oil flowing with pressure. Genius design. Shutting a valve before shutting the engine, will allow you to maintain pressure in the canister, so that you can open the valve and pre pressurize before starting your engine. With our oil cooler, #12 lines, Accusump, and filter adaptor with Knob adjustable oil flow rate (For cooling), gives us near a 10 quart oil system. Another strong suggestion, but one I can hardly suggest unless your engine is out anyway, is aftermarket oil gears. A known weak link in engines running in the upper RPM range. We also use a hinged baffled oil pan, but that’s probably overkill. Unfortunately our oil system with AN plumbing and block adaptors costs near $4000. I would also use synthetic oil or race proven non-synthetic (ie. Valvoline VR1), however if building a fresh engine first break it in with non-synthetic oil. That’s how good synthetic oils work! they wont allow the rings to seat properly. Something else I do when I change pump gears, is add a .060” (or less) washer on the inside of the spring loaded piston that governs the oil pressure. The washer acts as a shim to slightly increase your oil pressure about 10lbs. Be sure to use a washer that is nearly an exact fit.
Rerouting- A lot of stock NSXs have had problems with oil spitting out of the rear valve cover under heavy tracking at high RPM. Honda acknowledged this design flaw by redesigning the rear valve cover on the later cars. Some have never experienced this... some don't drive hard enough. I have permanently solved the problem on all of the cars including our race cars. The problem- (Likely accentuated by your new driving capabilities or style) In V-tech your oil is routed to the heads (Via solenoid valve) to facilitate the hydro lock of the rockers below the more aggressive cam lobe. This design requires massive amounts of oil that leaks profusely like the oil pressure at the crank bearings. The rear head actually fills with oil (The reason why an Accusump is needed for extended V-tech and high G turns... Not enough left in the pan) When the rear cover is full, high-G right hand turns push pure oil out the tube... Not what a breather was designed for. In the OE design, minimal oil is routed directly back into the engine at the throttle body and burned. A breather can be bolted horizontally on the ledge behind the drivers head (In the engine compartment)
You have two options (One takes longer then the other)
(Short)
Plug the rear valve cover oil out
Plug the hole in the intake hose before the throttle body
Plug the vacuum line (engine side) going to the front valve cover PCV. (discard PCV)
Run a hose from the front valve cover (PCV hole) to the new catch can. Preferably as high as possible.
I promise, no one needs to re think this, it will work.
(Long-This one requires valve cove removal, but is the best solution)
Drill the side of the front valve cover (Similar location as the hole on the rear valve cover)
Remove the pressed in tube sticking out of the rear valve cover (It’s pressed in)
Tap both covers for AN fittings
Run a hose from one valve cover to the other
Plug the hole in the intake hose before the throttle body
Plug the vacuum line (engine side) going to the front valve cover PCV. (discard PCV)
Run a hose from the front valve cover (PCV hole) to the new catch can. Preferably as high as possible
Both variations would be more professional if the valve covers were removed for tapping to AN fittings. The PCV hole will require disassembly and reassembly of the internal baffle in order to use a bulk head fitting. Note- fittings can not be welded, as the valve covers are Magnesium.
Transmission- Use a 
							6-speed if you can afford one. The more gears you 
							can get in the same ratio span, the less likely you 
							are to have a track with a turn where you can’t get 
							the right RPM. At one time I contemplated having 
							gears made spanning 2nd gear to just over 4th gear, 
							using all 6 gears. That would
 
							be optimal short of the $85,000 cost wanted to build 
							the custom sequential for the NSX. The other 
							advantage the 6-speed has over the 5-speed is the 
							Differential. The 5- speed has an adjustable limited 
							slip, that needs to be adjusted very high to put the 
							power to the pavement. Unfortunately because of 
							trans temps (Up to 450 degrees), the high clutch 
							pressure set at the beginning of a race compared to 
							the measurement at the end of a race is usually 
							half. In addition, the high clutch pressure is not 
							good for the turns. The 6-speed has a helical 
							differential, that helically puts more pressure on 
							the clutches as you put the power down. The benefit 
							is that you can set the clutch pressure less then 
							stock for great cornering, and when the car has 
							traction off the corner the horsepower will push the 
							clutches tighter then you can adjust the 5-speed. 
							The best of both worlds. The adaptation of a 6-speed 
							in a 5-speed race car is easy, use a toggle switch 
							for reverse. The shifter is the same, the solenoid 
							in the trans differentiates between 6th or reverse. 
							The spline count is the same on both, but the spline 
							length is shorter on the 6-speed, so the dual disk 
							clutch wont work in the 6-speed without a main shaft 
							replacement. Temperature along with angle cut 
							gearing is a weak transmission link in high 
							horsepower cars (If I made gears I would make them 
							straight cut for racing, but they would be loud). 
							Angle cut gears put pressure rotationally and 
							linearly… they push sideways and because of this 
							they break. 
In the WC car, and soon to be in our Cup cars, we 
							put a large air to oil cooler with a fan (In the 
							back of the trunk). When
 
							the oil gets over heated the clutches act different 
							and consistency is important in road racing. To 
							utilize the stock transmission oil pump, you need a 
							complicated adaptor because the oil goes in and out 
							the same hole in the trans.
Notes: If you do run a 5-speed… the 6-speed dif can 
							be modified to fit. Also, if I needed to pick one 
							set of gears between, stock, short gears, or a 
							combination with the three common sets of ring and 
							pinions; I would run the short gear set with the OE 
							ring and pinion. It has had the best average gearing 
							for the tracks you will encounter. This combination 
							is far better for road racing then OE, and if you’re 
							driving on the street it will still have the same 
							final drive ratio. I’m told that the aftermarket 
							ring and pinions are less durable then the OE when 
							used with high horsepower. The short gears even your 
							ratios and bring them down from 2nd to 4th. 1st and 
							5th stay the same. The ring and pinions lower 
							everything, making 1st unusable, and you still have 
							the poor 2nd gear transition. A combination of short 
							gears and the R&P bring your gearing way too low.
Weight- It’s not too difficult to get the race NSX to 2300 or 2400 lbs (Like our cup car), but most race classes require a heavier weight then that. I still take the time to trim the NSX, so that I can add the weight where it will do the most good. Earlier I mentioned the Fuel cell protection plate that doubles as a flat floor for better (under car) aero. It actually has a third function; it fits the race rule for weight: Between the axels and as low as possible. So when I need to add weight, I use ¼” thick steel plate (or thicker) instead of aluminum. This is an optimal location for weight, the NSX is not your typical car that wants a 50/50 weight distribution. It is designed to be closer to 40/60 especially with high horsepower and class limiting tires. Many useless things get removed when building a race car like power seats and AC, they will make up the majority of the weight loss. One of the best places to go next is the rear hatch, it’s quite heavy and sits high on the car. I don’t suggest replacing the front windshield unless you plan to replace it twice per season, because they fog quickly from rock chips. The OE glass is actually not too heavy. Another trick that a veteran racer showed me, was to use good and thick foam tape around the window seat, and then use the tabs to hold the windshield in place. Windshield glue is very difficult when removing a windshield. The original thought was really quick replacement (We always keep one in the trailer), but you would be surprised how often I remove it now that it’s easy, for working on something in the dash. I would also work on removing most of the harness and electronics. You can easily get 50 lbs here, but you better know what your doing. If you’re building an all out race car with aftermarket engine management, you’re better off just starting with a Painless brand harness for the basics. I’ve heard of people striping and acid dipping their cars for weight reduction, but that’s more of a cheating technique for All Stock classes like SSB or Super stock drag cars. It’s also been suggested (By those who have done this in the past) this seriously compromises the structural stability of the chassis, as it dissolves the seam sealing and gluing that bond the inner panels. I don’t suggest it. You can however take the labor intensive time to chip away all of the brittle floor insulation and you can scrape off the undercoating. Put all of this material in the same garbage bag and weigh it… you will be surprised at the results. Another place to look is the flip up headlights (Have you ever lifted this assembly?) Unfortunately I’m not sure if a lightweight and low cost solution exists. I have drawings for the bracketry that hold my –40lb solution, but they are also labor intensive. (This may be an exaggeration, I never weighed them)
Air filter- I have heard that tests have been performed with various air filter styles, and none of them drastically out perform the OE box. In the same breath they say “and the K&N can allow more dirt”. Hogwash! I have personally seen minimal improvements with the K&N and even more with no filter at all. If your too lazy to wash and lube a race filter, then you should probably use the OE filter… I however need every horsepower I can get. As for the filter housings- of course you’re not going to see an improvement on a stationary chassis dyno, ram air designs need movement. The design I’m fond of is the Cantrell ram air. I’m too poor to afford a pit bike cost worth of air filter components, but I think you will get the majority of the gains by utilizing the OE box with the Cantrell ram air nozzle. My thought is that the air still has to go through the same size filter regardless of the box it’s in… it does look nice though. (DALI gives you the choice and sells it this way for us non Ferrari owners).
Throttle Bodies- These were a logical test, when I discovered that our 300 RWHP version of the stock 3.2 would draw a small amount of vacuum again at high rpm. This tells me that I have a restriction limiting optimal air flow. I tried removing the air filter and that didn’t solve it. I then spent 20+ minutes swapping out my throttle body with the beautifully machined unit I got from SOS. The 20+ minutes was due to using dyno supplied tools, and changing over some of the emission/cruse control parts from my race modified unit. I did see gains, but they were minimal. So minimal that it could easily be written off as a change in conditions due to the 20+ minutes. It also didn’t solve the mystery vacuum clue. This leaves three places that could be restricting optimal air flow; 1) The intake manifold 2) The head flow or 3) the valve train. The intake manifold is already improved and I doubt it’s restricting flow (I will cover this later). The heads also flow about 3% better within the stock specifications and the rules. For now I’m guessing the valve train. So this is my current project that will also be discussed later in this engine section. Though our 300 RWHP out of a stock spec engine is impressive, it’s far shy of the hp that my supercharger engines have made. I’m guessing the throttle body would be better suited for higher HP. I almost forgot to write about those cruse control components I mentioned earlier. On the side of the Throttle body is a stepper motor that has the ability to close the throttle plate even if your foot is planted on the floor. For a self demonstration have someone hold the throttle wide open (Engine off!) and reach in and push the throttle closed. You will see that its not directly attached to the throttle cable, it’s only spring loaded. This does close some on it’s own, so I remove this assembly and tig the throttle bell directly to the butterfly shaft. However, this assembly is what your throttle position sensor is attached to, luckily the NSX has two TPS’s. I extend my harness 6” or 8” to reach the other TPS on the lower left of the throttle body. I also remove the vacuum idle adjuster. New thought- Don’t forget to plug any vacuum lines.
							Intake manifold- First 
							remove and plug the EGR, this is an emission 
							component that also helps heat the manifold in cold 
							conditions. Not needed or wanted in a race car, even 
							the smallest amount of  Carbon Monoxide entering the 
							intake will reduce the burnable oxygen content and 
							reduce your power. The next item to go is the 
							butterfly assembly plate known as the variable 
							runner system. This vacuum controlled contraption 
							along with v-tech is an innovative concept that in 
							theory will give you both long and short intake 
							tracts for better high and low RPM horsepower. 
							Unfortunately when removed both high and low RPM 
							horsepower are improved. In trying to figure out 
							it’s intent, I believe Honda engineers implemented 
							this device for better drivability. Better 
							drivability is defined not only in terms of 
							reliability, or fuel mileage, it’s how smooth a car 
							functions or starts etc. This assembly may have also 
							played a roll in emissions. The reason for its 
							over-all performance loss is likely related to its 
							restrictive throttle plates that would
							
flow 
							less air then the manifold runners. In addition the 
							plate assembly is less then smooth and would surely 
							cause more turbulence then smooth runners. The 
							turbulence slows the potential air flow into the 
							combustion chamber. The other notable draw back is 
							that this device is vacuum controlled and not 
							designed for boost applications. Anyone with 
							aftermarket turbos or superchargers will benefit 
							from removal. It’s a simple modification:
-Remove the intake manifold (The gaskets are 
							O-ringed and re-usable).
-Turn the manifold over and remove the nuts 
							retaining the belly of the manifold. (try not to 
							damage the paper based gaskets, it may be a good 
							idea to have new ones on hand)
-Slide the Variable runner assembly off the studs 
							(Protect the gasket here also)
-Next remove all the studs. (A stud puller will work 
							best for potential reusing them in the future, but a 
							pipe wrench, vice grips, or channel locks will work.
-Now cut one of the gaskets so that it only makes an 
							outer ring to seal the belly pan without the runner 
							assembly.
-Measure for the length of bolts you will need in 
							each location to fasten the belly without the runner 
							assembly. -Take one of the studs and go to the 
							hardware store to buy exactly what you need. (They 
							are obviously metric- 6mm if I remember)
-Now assemble using some Honda bond, vacuum leaks 
							cause many problems especially with Nitrous. 
-Then re-install the manifold, and anything else you 
							removed.
One step further but will require heavier tooling, 
							is to remove the divider wall on the inside of the 
							manifold. Be sure to smooth out all of the 
							transitions. This combination will maximize your 
							stock manifold. If porting is allowed you can port 
							match and smooth out the runners. I have been 
							threatening to make a sheet metal manifold… even 
							started the drawings. Maybe soon.
If Nitrous is your thing, the threaded boss for 
							holding the throttle body strain relief is a good 
							permanent location for a single fogger nozzle. If 
							you want to go big and don’t mind destroying an 
							intake manifold, the EGR ports act as a great 
							location for a direct port nitrous system. (You will 
							however need to mill the top off the manifold). 
							Don’t forget to plug any vacuum lines.
Balancer- I’m not sure 
							if this issue has been previously addressed, but I 
							have seen many harmonic balancer failures with other 
							NSX’s, and one of our Cup cars. Some engines are 
							internally balanced and some use the balancer to 
							counterweight the rotating assembly of the pistons, 
							rods and crank. Externally balanced engines need a 
							balancer. Internally balanced engines do not… hold 
							on and read more. The NSX is very well balanced from 
							Honda, but can be improved even more. On an 
							internally balanced engine like the NSX, the 
							balancer is not really a balancer and serves no 
							balancing function. It’s a pulley that’s weighted 
							like a flywheel. Flywheel weight is used to retain 
							load torque when on or off the throttle by utilizing 
							the inertia weight that wants to keep spinning. but 
							it takes more torque to get and keep that weight 
							spinning, thus sacrifices overall torque and 
							throttle response. Like a weed eater, it spins at 
							high RPM with little torque and would stop if it 
							encountered a finger sized twig. If you put a 
							flywheel weight on it, it would torque through that 
							twig… but may need more torque to get to speed. The 
							other purpose of the Balancer is to absorb harmonic 
							vibrations, tiny vibrations are measured in 
							frequency like sound, and all high rpm motors will 
							have increased activity at a specific frequency. 
							This harmonic frequency can do damage similar to the 
							results seen from detonation or cavitation, or 
							simply explained as more damage then seems feasible. 
							The way the balancer combats this Harmonic is by 
							insulating it with rubber. The balancer is made of 
							three parts; an outer and inner ring pressed 
							together with a thin
							
rubber 
							ring sandwiched between. Similar to cutting a brake 
							rotor with a band wrapped around it. This is a low 
							cost way of combating harmonics within our rpm 
							range. You can also minimize harmonics within this 
							range by doing a good job balancing… as Honda does. 
							I think this balancer feature is there as insurance 
							incase time or unknowns change the tolerance of the 
							internal balance. Many race engines eliminate this 
							weight as a performance gain, without problems. In 
							fact other factors such as a flywheel, clutch, or 
							driveshaft. Would have a greater impact then this 
							balancer. In any pro level race car, this balancer 
							would be replaced with something more reliable. I 
							can tell you that all of the Famous super NSX’s you 
							know have replaced or modified this part to be 
							solid. On the World Challenge engine, I turn down 
							the outer ring and the compliant rubber, machine a 
							high tolerance centering lip, tap holes and bolt an 
							off the shelf 8-rib Vortech pulley for a solid part.
							
The reason- Heavy loads produce heat, heat melts the 
							rubber and causes the OE balancer to fail. The DAL 
							Motorsports cup engine had a balancer fail because 
							the alternator belt was tightened too much. It will 
							also eat up more horsepower. The tighter you go the 
							more heat you will produce. This is even more common 
							with the supercharged NSX’s everyone knows that when 
							the belt isn’t tight the SC slips and wont make as 
							much boost. I tend to only tighten the belt until it 
							stops making noise, and if supercharged I tighten 
							the belt while on the dyno, until I stop seeing a 
							rise in boost. I then feel how tight that is for the 
							next adjustment without a dyno. If you are doing 
							this without the dyno, lean towards too loose rather 
							then too tight. In fact tonight I need to make an 
							adjustment because my wife’s car is screeching at 
							start-up. Another way to increase your belt strength 
							is to go with more ribs. The cheapest way for 
							minimal improvement would be to use the Goodyear 
							Gatorback belt. They have a patent on a ribbed 
							design that gives more grip then the rest of the 
							brands. Most of the failures cause minimal damage; 
							Balancer, belt, and usually some cosmetic damage to 
							the front cover… But recently the balancer on Wei 
							Shen’s car (NSXCA President), separated in the 
							opposite direction towards the engine (Likely had 
							something to do being in a high G turn). Well it 
							went through the plastic front cover and made the 
							timing belt skip, bending valves, hurting pistons, 
							etc. I wont divulge the actual cost, but I will say 
							that if you are unfortunate this lesson can cost you 
							over $10,000 to fix it the right way.
Exhaust- The 3.0 
							manifolds are lacking compared to the 3.2 OE 
							headers. So get them while you can… In fact you will 
							see more then 2/3’s of the HP gain (on the dyno) 
							with the 3.2 manifolds compared to aftermarket 
							headers. Another interesting
							
note 
							is that many, if not all of the marketed headers for 
							the NSX are made by the same place. If you do go to 
							the 3.2 manifolds you will also need the remaining 
							3.2 exhaust or aftermarket components designed for a 
							3.2. Comptech use to sell conversion pipes, but I’m 
							not sure if they still do. The 3.2 manifolds are 
							welded internally and this weld restricts its flow. 
							My tip is to remove this bump with a die grinder and 
							clean them before installation. If racing, we can 
							make a good stainless steel exhaust that scavenges 
							from each bank. Not recommended for the street 
							unless you like tickets and dislike your neighbors.
							
							
The next picture (right) shows 
							the 3.2 manifolds/headers. (3.0 are castings) The 
							ones on the left are the same as on the right, 
							without the insulation sheeting. See they are 
							headers.
 
 
 
  
 
 
 
 
This is a picture of the hand built Inconel headers 
							for my WC car from RTR.
							
They are huge. If your running boost, use mufflers 
							with stainless steel sound insulating material. The 
							high heat and close proximity to the engine will 
							burn up anything else. I would also suggest that you 
							warn the dyno operator about the excessive heat, SC 
							or turbo NSX’s tend to burn up the dyno exhaust 
							tubes.
 
 
Valve covers- Nothing big here, but I can help solve the oil release problem many have found while tracking with early model NSX’s. While in VTEC, oil is rerouted up to the heads via an electronic solenoid valve. The oil pressure is used to push 24 pins for locking the rocker arm assembly into using the higher lift cam lobes. Like the oil that goes to the bearings, the oil is free to seep out of all the gaps and flow back down to the pan. This oil can seep faster then it leaves the head so at certain tracks that have conditions where you sustain VTEC for extended periods of time the heads can actually fill with oil. Now, you will never pressurize the valve cover because the holes going to the pan are so large, but as I said the head can fill to a certain level. The rear valve cover is tilted lower then the front and when filled with VTEC oil, will allow unrestricted oil to flow out the tube going to the air filter bellows while in higher G right hand turns. In stock form you will just burn this oil as it goes through the throttle body. This is not good for performance, but at least it wont get you kicked off the track for oiling it. Those of us that do away with the stock configuration will usually get a breather tank, but that breather tank can overflow before much time can be spent on the track. Before I give you the fix, I want to state that many of you will never have this problem. It depends mostly on the track, and somewhat on the driver. It’s an easy fix, that doesn’t require modification of the valve cover internals. I had this problem and took my cue’s from the revised valve covers of the 3.2 (Honda knows they had a problem). They no longer exit the rear cover, they now exit on top of the higher front cover. I think I took it one step better. First I pull out the tube that sticks out of the rear valve cover (Its just a press fit). Then I tap this hole and install an AN fitting. Then I drill and tap the same location on the front cover for the same AN fitting. Use hose to connect the two. This will equalize them, if one fills first it will feed the other. I then use a bulkhead fitting through the PCV valve hole on the top of the front cover. A bulkhead fitting is an AN fitting with a nut and Teflon washer for sealing fluids. This type of fitting is required because the manifold is made of Magnesium. Magnesium would be difficult to weld… if you didn’t ignite it. Pure magnesium when it catches on fire will burn very hot until it’s gone. Not even a fire extinguisher will put it out, or at least that’s what I remember from science class. Interesting- Original drag cars used Magnesium for wheels until it was outlawed after flat tires would ignite the inextinguishable element that burned so hot it would melt steel. It also burns bright, so like a welder… don’t look at it. Anyway back to the bulkhead fitting; run a hose to a breather tank as high as you can get it, so that any oil would flow back down the hose into the valve cover (Even know the oil never comes out anymore). Boosted cars need to remove the PCV anyway. And don’t forget to plug any vacuum lines. By the time I’m done with this, I’m going to remember how to spell vacuum.
Ignition- This may be a 
							long one. The OE Honda ignition (Even the pre OBDII) 
							was ahead of its time. It tries to think of 
							everything. I have been curious enough to see three 
							different “LIMP” modes not counting the cold engine 
							failsafe. A limp mode is when the ECU sees something 
							wrong and sets a safer environment for the engine to 
							“limp” home and get fixed. You can monitor this 
							adjustment with a wideband sensor, heater control, 
							and an Air/Fuel ratio gauge. Lets say you stall at a 
							stop light, restart your engine and drive off. If 
							you didn’t turn the key back to off, all the ECU 
							knows is that something stopped the engine and 
							that’s not normal. So in the engines best interest 
							the ECU retards the timing, and riches up the air 
							fuel ratio to be safe. For the next 5 or 10 minutes 
							until the ECU verifies everything seems OK, your gas 
							mileage is cut in half and your performance is 
							decreased. Some limp modes start to reduce the rev 
							limiter also, one time we were limited to 5500 RPM. 
							And some of the limp modes require ECU resets, while 
							others require only the ignition key be moved to 
							off. I’m sure I have only scratched the surface of 
							complexity built into this computer. Many things can 
							trigger limp modes. I have seen stalls, temperature 
							readings, shorts, something unplugged, and something 
							like the VTEC solenoid not working. (I’m doing good 
							with the “VTEC”, I haven’t missed one since). I 
							would love to have access to the information within 
							this computer, and also be able to program it for 
							racing. This computer has all the capability you 
							would need for tuning, but the information has yet 
							to be shared (I guess because Honda doesn’t want 
							stock cars changed). Unfortunately this state of the 
							art ECU can not be used for racing. As a racer I 
							want to decide when to lift, or stop a race. 2/3’s 
							into the car’s build-up I chose to go with an 
							aftermarket ECU that when researched offered as much 
							and more then it’s more expensive competition. That 
							brand is AEM. First I want to say that the software 
							that was developed for this product offers any 
							combination you would ever need in tuning your 
							engine, enough choices to get most people in 
							trouble. Right off the start it fired up my engine 
							and got me to the dyno. The first pull saw about a 
							15 horsepower gain over the stock ecu, without even 
							tuning it. The problem I was seeing was a misfire 
							above 7000 RPM, you could hear it on the dyno and 
							see it as choppy fluctuations on the dyno curve. At 
							first I was excited to know that I had even more 
							power on tap when I solved this problem. I did 
							research on problems surrounding its use of “Wasted 
							spark”. The AEM doesn’t have enough coil drivers to 
							run all six of the NSX coils like the OE ECU, so 
							they fire two coils at the same time with one 
							driver. That’s common practice, but that means that 
							each coil has to charge and drop 2x what’s needed or 
							in my case at the equivalent of 16,000 RPM. I found 
							that Supra's have had similar problems with wasted 
							spark. I buy the MSD coil enhancement wire it up as 
							described in its instructions, but it wont work. 
							After a call to MSD they think it’s because of the 
							wasted spark, this was an issue that required me to 
							reroute my coils in series rather then parallel, I 
							chose not to hack my harness until I get to the 
							bottom of the problem, so I had my cable department 
							make a few harnesses for me. In addition to the coil 
							routing, it turned out the signal coming out of the 
							AEM wasn’t what the MSD is looking for. After some 
							help we wired the MSD to the OE ignition module, and 
							were now ready for another run. Success this helps, 
							but it’s only a bandage and not a permanent 
							solution. Most of my missing is gone, but not a full 
							100%. I think a great solution will be to also use 
							(3) MSD twin blaster coils (made for GM engines) and 
							use two sets of good plug wires designed for the 
							VTEC 4 cylinder engines. A better coil and coil 
							enhancement. Since my trouble shooting started AEM 
							now mysteriously sells a coil enhancement product 
							that they recommend for this application. For now I 
							can’t help but to recommend the AEM because of its 
							price, hell I already solved their problem.
 
Cam Gears- The 
							adjustable cam gears are expensive for the NSX 
							because three of the gears are different, and one of 
							them acts as a trigger for your ignition pick-up. 
							Expensive, but be grateful that someone makes them. 
							I have talked to most of the manufacturers who at 
							one time made them, but dropped this part from their 
							line because the volume wasn’t good enough.
Cam gears not only allow you to change the 
							centerlines of cam’s, they allow you make the cams 
							equal to each other. The 4 cams in the NSX are 
							ground and indexed by a keyway, they then sit in 
							bored cam journals in the heads, the heads are 
							surfaced and sit on the block that’s also surfaced. 
							The cams need to work in relationship the machined 
							or casted pistons, that connect via a bored hole, a 
							machined pin, two bored holes in a rod, and a ground 
							crank, sitting in a bore in the block. All of this 
							is indexed by a series of gears on the cams and the 
							crank via a ribbed belt. Sorry you had to go through 
							that read, but it illustrates all of the tolerances 
							that can stack up to give you different cam 
							location. The odds of having all four cams where 
							Honda preferably wants them is very low. This is one
							
of 
							those variations that randomly make engines 
							different in horsepower. Now go mill your heads or 
							deck the block… you will be no where near even with 
							the cams and bank of cylinder flow. The adjustable 
							cam gears allow you put each cam where you want it 
							on a particular engine. Degreeing the cams on the 
							NSX is very taxing. First off it can’t be done in 
							the car, the engine needs to be on a stand. 
							Degreeing the cams take me 3 to 4 hours… I know you 
							may be superman and you can do it in less time, but 
							I am self proclaimed anal about getting the cams 
							exact. To start with you need to remove the 
							balancer, valve covers, plugs, front covers, belt, 
							and cam gears. Install the new adjustable cam gears, 
							new belt and lower front cover. Followed by a large 
							degree wheel (Mine happens to line up with the 
							pointer on the cover. You may need to install your 
							own pointer. Then screw a hard but rounded piston 
							stop in the number one spark plug hole. The stop 
							will allow you to rotate the piston until it bottoms 
							out in the clockwise and counter clockwise 
							directions. Document the degree wheel readings at 
							both crank rotation extremes. This will now allow 
							you to adjust the degree wheel center to be exact. 
							Recheck your work. This is how you get the exact 
							center of the crank. If you just measure the piston 
							at top dead center, you will notice a dwell and 
							multiple degrees of crank rotation without piston 
							movement. Now remove the piston stop. Install a dial 
							indicator on the rocker arm of the number one 
							cylinder (This is another special tool because of 
							the depth and inconsistent surface for mounting), 
							and bring your #1 piston to top dead center. Now, 
							the NSX has three different cam lobes, the only one 
							your interested in is the bigger center lobe that 
							you will be using when in VTEC. Unfortunately this 
							means you need to be in VTEC on the engine stand 
							while your Degreeing the cams. First you will need 
							to plug an oil port in the appropriate head, your 
							manual will better describe where. I use a very long 
							bolt with a big red zip tie on it, because if I 
							forgot that bolt in the engine I would likely be 
							building a replacement. Then remove a plug on the 
							side of the head. Another special tool is needed 
							here, I made mine using an AN fitting, hose, and a 
							male air hose connector. You will need regulated air 
							pressure for pressurizing the VTEC. What the manual 
							doesn’t explain is that first you will need to 
							squeeze oil in that hole in the side of the head. 
							This is how I know that most of all the NSX engine 
							builders have never degreed an NSX cam (Not 
							including Comptech or RTR), I called around to find 
							out why I wasn’t getting VTEC… nobody knew. Anyway I 
							put two and two together to figure out that a system 
							designed for oil wouldn’t seal with just the air. So 
							I used a gear lube bottle with non synthetic 20/50 
							oil, I squeeze oil in, hurry up and screw the home 
							made fitting in, pressurize the system and rotate 
							the cam near the highest visible point on the cam 
							lobe. Note that before you pressurize, make sure the 
							cam lobe your working on is on the base circle and 
							not on a lift portion of the lobe. If the rocker is 
							lifted the VTEC pin will not engage. All of this 
							should be done quickly, and quickly disconnect the 
							air pressure as soon as the cam lobe is at the high 
							point. The spring pressure from the valve will stop 
							the pin from going out of VTEC. The reason for the 
							rush is all that oil seepage I mentioned in the 
							valve cover section. Now that it’s in VTEC (I typed 
							vtech a couple of times since the 1st post, but 
							caught myself), you need to zero the dial indicator 
							at the highest point of the lobe. (Remember not to 
							rotate the cam all the way around, or it will come 
							out of VTEC) When you get this exact, rotate the 
							crank counter clockwise until the dial indicator 
							travels about .100” before peak of the cam lobe. 
							Then rotate the crank clockwise (as if it was 
							running) slowly and smoothly (A long breaker bar 
							helps) until you get to exactly .050” before peak 
							(Or the zero on the gauge). If you pass the .050”, 
							you need to rotate back to .100” and re-approach the 
							.050”. This is to eliminate any slop and to get an 
							exact reading rotating as the engine will run. When 
							your comfortable with how exact this is, look at the 
							degree wheel and jot down the number. If the pointer 
							is in the middle or I actually divide the space into 
							quarters, then note that also. Now rotate the crank 
							slowly and smoothly past the highest point (and past 
							zero), to .050” after the highest point exactly. 
							Same thing if you go past .050” you will need to 
							rotate counter clockwise to zero or more and get to 
							.050” while rotating in the right direction. Jot 
							that number down then add the two numbers and divide 
							by 2 this will equal the center line of this cam. If 
							it equals 110 degrees and you want 104 degrees you 
							will have to adjust the cam gear 6 degrees to get 
							here. Now to complicate things the crank is a 2 to 1 
							ratio, and to further complicate things the bank of 
							cams on the other side will require an opposite 
							direction of rotation to go up or down in number. 
							Not too important if I didn’t describe this clearly… 
							you will understand if you do this. Not done yet. 
							Now that you know what you want to adjust to, you 
							move the cam gears by loosening the multiple 
							retaining bolts enough for the assembly to slide, 
							then remove the small screw that acts as a locater 
							for estimated cam adjustments. Now rotate the cam 
							using a wrench on the center cam bolt. Hold the 
							wrench with resistance in both rotation directions, 
							or sometimes the load of the valve spring will 
							quickly force the cam to rotate. (This would cause 
							you to start over) Align the hole that best 
							represents the choice you made, and screw in the 
							alignment bolt. Now snug the retaining bolts to 
							remove any slop and they securely hold the assembly 
							from rotating. Unfortunately this is really high 
							tolerance and due to fabrication tolerances and the 
							slop built into the cam gear locating holes, you can 
							have between a 1 and 3 degree variation from where 
							you thought you were going. (This depends on where 
							the pin was when you first started). Now you need to 
							re-zero the dial indicator (if moved), and recheck 
							at .050” before and .050” after zero. Your numbers 
							should be different, added and divided by 2. The 
							number may or may not be what you want because of 
							the slop. If your lucky and close to what you want, 
							you can loosen the bolts and try to maximize the 
							slop in the proper direction. If your not lucky, you 
							need to go to the next hole. Again you need to 
							re-zero the dial indicator (if moved), and recheck 
							at .050” before and .050” after zero. If you did 
							move holes you will likely have to work with the 
							slop for another pass, or even move back to the 
							first hole because that will get you closer. Now 
							start this whole process over including the oil, air 
							pressure, and the dial indicator on the exhaust cam. 
							When that bank is done remove the long bolt with the 
							big red zip tie, put the bolt and crush washer back 
							in the side of the head… then move to the other head 
							and make those cams exactly the same. The number 5 
							cylinder has the same top dead zero, so the degree 
							wheel doesn’t need to move. This is a very tedious 
							process and you might be tempted to say well that’s 
							close enough. Give yourself enough time and talk 
							your self into not taking short cuts (before you 
							start). This would also be a good time to adjust 
							your valve lash. Another unfortunate with the NSX is 
							that if you want to change your cam timing while the 
							engine is in the car, I would only do it in an 
							emergency because you cant check your work, (and it 
							wont likely be where you think it is). This is why 
							you will need to experiment on an engine dyno, or 
							listen to advise from someone you trust. The 
							centerline is important because it dictates how much 
							overlap you have between the intake and exhaust. 
							This amount can effect the drivability, power, 
							mileage or even if it runs. The proper overlap will 
							scavenge and help flow. I want to check my notes 
							before I publish where I think the stock cams should 
							be. I will say that I have seen as much as an 
							8degree variance between cams, and 6 degrees from 
							where the centerline should be. I usually see an 8 
							hp gain on a stock engine, and more can be gained on 
							performance engines. The big potential gains are 
							from changing the centerline. Now some of you 
							experienced engine builders are going to find this 
							hard to believe… In a high hp supercharged engine 
							(over 500hp), with Comptech cams in the stock 
							location. After testing and degreeing with help from 
							someone that has lots of experience with this set 
							up, I moved the centerline as much as 14+ degrees 
							and found 100 horsepower. My guess is that these 
							cams were not designed for a boosted application. 
							Note; Here’s a warning I ultimately ruined that 
							engine, because I never took the time to check the 
							piston to valve clearence after moving the 
							centerline 14 degrees. (I did put more clearence in 
							the aftermarket pistons, but it wasn’t enough) Oh, 
							don’t forget to back each of those cam gear bolts 
							out one at a time and put blue lock-tite on them. I 
							made valve covers for the dyno that seal the oil, 
							but expose the cam gears. There is some debate about 
							exposing these in the car, but the NSX wouldn’t make 
							them that visible anyway, so it’s probably not worth 
							it. That was longer then I thought it would be.     
Timing Belt- This can be done in the car, but it’s a pain. The main thing I want to convey here is that the spring used to tension the belt is not a universal solution. If you are using better springs, higher lift cams, or even have more resistance then when new, you risk the possibility of slipping teeth on the belt. This equals bent valves, money and down time. Some big teams use 10K devices to adjust perfect belt tension, you have the same problem here like the balancer belt. Too tight can rob power, and destroy parts, too loose will destroy parts and rob power. Either way I’m sure the spring is too loose for my applications. My other problem is using one bolt that applies tension by rotationally stretching threads, to hold and locate the slotted tensioner pulley from moving linearly. (Not the smartest engineering in this area) Measure the distance that the spring would stretch, armed with that info go to Home Depot and buy a $3 galvanized turnbuckle (Very small), also buy the appropriate nut for the right hand threaded side of the turnbuckle (For permanently tightening it) (Maybe the least expensive insurance you can buy for an NSX). When tightening the belt lightly snug the center bolt and install the turnbuckle where the spring was located. The turn buckle has very fine left and right handed threads, so this makes it very easy to slowly and accurately adjust the belt tension. In addition this aids the tensioner from never moving back. A side benefit is that you can carefully back out the main tensioner bolt to put blue locktite on it (Check the torque specs). Now put a drop of locktite on the turnbuckle nut and tighten it. I apologize this doesn’t tell you how tight to make it, but that’s feel.
Head Gasket- The 3.0 head gaskets are not made of the best materials and have had problems that mostly seem to develop when overheated. If they start leaking that means the heads need to be removed, and that’s usually best if done out of the car. So if your ever doing engine work, I would replace the head gaskets. Honda is aware of this problem and has revised the head gaskets for the 3.2 using a layered steel shim type. The 3.2 version is thinner, and will raise your compression in a 3.0 from 10.2 to 10.7. I wouldn’t recommend this unless you can adjust your engine management. If you want a stock or different thickness or different bore size, Cometic has the drawings and can make any thickness you would want (Above .030”) and any bore size within 7 days. They also emboss around all of the water and are better then the OE. They cost near the same price. ~$200 for a set. I can order for you. With OE ECU’s on any car, I wouldn’t suggest more then 10:1 compression on pump gas, but with ECU adjustability you can supercharge 10:1. Oh when building my wife a new engine for boost, I added gasket thickness to reduce the compression rather then change the good pistons. Another suggestion you wont see in a Honda book, is to spray the head gaskets with a gasket tack/sealer. If you can’t find this product at your local auto store, Dodge sells it. Another part to spend your money on is ARP head studs, this will improve the clamping force. If you don’t buy studs, you should replace the OE bolts, clean the threads and don’t torque them dry. Do what’s listed above and you will never have gasket problems… provided your sleeves don’t move too much.
Heads- This is a very 
							important subject and a necessary mod when ever you 
							have an engine out. After the second engine loss, I 
							did some further forensics. My original assumption 
							was that like before the engine must have spun high 
							enough to float the valves and contact the pistons. 
							I didn’t disassemble the first engine, but this next 
							one I built and had an attachment to. Sure enough 
							everything was junk, the pistons hit the valves, the 
							valves bent, and since we have very little squish 
							area the hard valves pound the heads, and pound the 
							piston. With no where to go the piston rotates and 
							cracks the block. This was a special engine, and sad 
							to loose, only the crank and some of the rods were 
							salvageable. Anyway after over revving happens the 
							valves break or get so deformed that the springs, 
							retainers and keepers are usually loose in the head. 
							No different this time, but this time I looked very 
							close. None of the valves made contact with the 
							pistons other then the one cylinder. That’s not 
							really possible if it over revved. So now I know 
							that the valve train failed, and one of the 
							retainers has a clean break in two. I got on the 
							phone and talked to a couple of valve train experts. 
							The first one said that yes, the NSX keepers are 
							single radius grove, stamped steel, and the same 
							part number used in the B-14 or 18? engine. They 
							distort and fail. He also mentioned that his company 
							makes hardened machined ones. The NSX valves have 
							very thin stems, saving weight for high RPM. This 
							small stem requires a small keeper. The second 
							expert I talked to said yes the retainers fail a lot 
							of the time with the Honda engines he builds, and he 
							makes Titanium retainers. But I would need custom 
							for the NSX. I’m not sure what part to blame, all I 
							know is that I replace both in every engine I build… 
							even stock. If you need them give me a call. Oh, I 
							looked at the junk head I didn’t disassemble earlier 
							in the year… same thing, not an over-rev. The 3.0 
							valve springs are not too good, Honda changed them 
							to a double spring replacement on the 3.2, and they 
							are not bad for a street car. I change them for 
							aftermarket, especially if using custom cams. I can 
							get good springs at good prices. The 3.2 also has 
							bigger valves that flow considerably better. .040” 
							larger valves can flow as much as 20% better. Some 
							new heads I have at the shop now have larger valves 
							then the 3.2- I will tell you how they work. I’m 
							also anxiously waiting for new custom cams
 
							from SOS. The Comptech cam is a reground OE cam with 
							OE lift and potentially different, ramp, duration 
							and/or center line. Your average engine builder may 
							not be qualified to build an NSX engine, And I 
							defiantly wouldn’t trust most mechanics. The NSX has 
							a few head components that can be missed during 
							assembly. I have rebuilt failed engines with VTEC 
							pins missing, and missing O-ring and oil transfer 
							sleeve under the cam journal cap. Most of the 3.0 
							owners are in luck, The 3.0 heads with 3.2 valves 
							can easily be made to out flow the 3.2 heads. I’m 
							going to tip toe through this subject a little ; ) 
							When putting 3.2 valves in a 3.0 head you can 
							machine the bowl just like Honda did, but better. 
							But even more gains can be made by de-shrouding the 
							intake valves like Honda did but better, especially 
							good flow when you can go out to the 3.2 bore size.
							
From right to left; 3.0 intake valve, 3.2 intake 
							valve back cut, and the new larger valve. Below is a 
							cracked retainer.
 
Block- The 3.0 block has steel sleeves for the cylinder bore, and the 3.2 has an aluminum bore that is Nicasil plated (Or Hondas variation of Nicasil). The 3.2 is an awesome improvement. This process is used on Indy engines and high performance motorcycle engine. This technology is where the whole performance industry will be in the future. The concept is that the cylinder walls now have closer expansion rates as the piston material, so tolerances can be made tighter without sticking a piston, and the tighter tolerances reduce power loss. Another benefit is that the aluminum walls dissipate heat at a far greater rate allowing a leaner more efficient engine that burns less gas and makes more power. One small difference is that using this technology may require your engine builder to do things that go against what he learned over the years, like piston to wall clearence. Another thing you need to be conscious of is the piston, to utilize the OE clearances I would use a brand like Mahle, if you use the common performance pistons like JE, Ross, or Weisco. You may have to use normal tolerances. I have been exposed to this process from my motorcycle days, if you oversize the cylinder walls more then a couple of thousandths, the walls should be re-plated. I warn you that every plater will say that they can do this, but I cant tell you how many failed motorcycle engine I have seen from failed plating. You can bet they wont eat your loss. It’s a gamble really, when I moved to Georgia I needed a new location, asked questions, used my gut, and got lucky. Everything is the same on the 3.0 and 3.2 blocks, even the crank journals. The only difference that will affect bolt-on’s is the bore size.
Cylinders- If you view a 
							running engine with inferred heat detection, almost 
							all of the heat is at the top of the cylinder wall, 
							so I’m guessing this is why we have an open deck 
							design for better water flow and heat absorption. 
							But this open deck has horsepower limitations. Many 
							deck strengthening variations have been tried. I 
							don’t care for the 1 part insert designs, because of 
							what I know about fitting something that has to be 
							perfect with internal and external dimensions. This 
							perfection needs to be maintained on the block and 
							the insert it self. With zero tool deflection or 
							wear. Even the thin but large shape lends it self to 
							difficult tolerance. And I’m sure grinding is out of 
							the question.
 
							The next choice is sleeving with a shape that spans 
							the deck gap. I like this design because each cyl is 
							separate and requires a round cut. Both the size and 
							shape make the tolerance easier. This application 
							would be best for high horsepower drag racing, and 
							maybe street driving. The sleeve material adds 
							strength, and the deck support is superior, but 
							cooling gets reduced both at deck area because less 
							water can flow there. If your running a 3.2 you will 
							also have less cylinder wall heat absorption because 
							the materials have changed, and wont benefit from 
							the current technology for tighter piston to wall 
							clearence. For street and drag racing this wont 
							likely matter, but for road racing I would do my 
							best to retain the aluminum cylinder walls. This 
							application may be best solved by pinning the gap 
							with a tight slip fit dowel made from aluminum. The 
							holes are bored with a CNC machine using a flat 
							bottom end mill, so that a slightly chamfered dowel 
							would have a nice floor to sit on. The block is then 
							decked with the pins for a perfect surface. The 
							heads will hold the dowels in and they can be free 
							to expand or not expand. My idea was to retain as 
							much water flow as possible, retain the same 3.2 
							bore material, and to help the cylinder with the 
							minimal support it needs from the deck. I can’t tell 
							everyone to go do this because I have only done one, 
							but I don’t think there are any holes in the theory.
							
Mains- Some people have 
							seen main bearing failure with the NSX engine. My 
							most reliable resource with high horsepower NSX 
							engines suggests that they solved this problem with 
							Billet replacement main caps. It make since… the NSX 
							engine i
s 
							well built like most top fuel drag engines with 6 
							main bolt. The difference is the material; the NSX 
							engine has cast caps. The 6 holes in this cast cap 
							contribute to its weakness. Because of the price, 
							and my past relationship with arguably the best main 
							cap builder in the world, I decided to have my own 
							caps made (Nicer and less money). They are very nice 
							parts that retain the OE oil passage differences 
							rather then making them all the same, he even CNC’s 
							the partial threads for easier disassembly by using 
							the bolts for leverage. Some of the famous drag 
							racers you know use my caps for over 1400 HP. If you 
							do this mod, find a competent shop that can line 
							bore and line hone. If you cant find a shop with a 
							line bore, they can be honed without one… but the 
							operator will need to hone very slow, or the harder 
							cap material will push rather then cut. This will 
							move the crank center up in the block. That’s not 
							the end of the world, but it causes additional 
							complications like piston height or seal alignment. 
							With the caps I would suggest APR studs.
Rods- Rods are theoretically another weak link, and the stock rod bolts are likely torque to yield and are not meant to be re-used. Torque to yield isn’t made to stretch back, and is usually torqued by bringing it to a specific torque then rotated X amount of degrees (90 degrees is common). Normal bolts are just set to a certain torque. the rods you see above have the ARP wave bolts, they require some machining but are the strongest. If you went this far, you are using aftermarket pistons and will need to bush the little end to be free floating.
Boost (World Challenge) - First I want to help my crusade of stopping people from using boost pressure numbers as a barometer of potential horsepower gains. Boost is irreverent… Cubic Feet per Minute of air flow, being able to sustain that air flow, and at what RPM is what makes power. You can have a tiny little turbo making 30 lbs of boost with 500 CFM, and a supercharger with 7 lbs of boost that makes twice the horsepower because it flows 1100 CFM. Boost is awesome! It gives great power all of the time. I’m a proponent of supercharging vs. turbo's but maybe that’s because of my ignorance. Turbo’s can in fact make maximum horsepower, so for drag racing within a certain rule set a turbo would be my choice… Oh I would probably use them for salt flats also. But I can already make more power then I want with a supercharger, and this to me is a better choice because; Less heat, less lag, better lower RPM power and most of all (Here’s where the ignorance comes in) a supercharger is more reliable and easier to tune. Not to say that you shouldn’t go out and buy an Audi, because they have spent countless hours tuning for drivability. But I cant tell you how many people I have witnessed with problems… so I never felt it was necessary to go that route. In the NSX the turbo’s have to go below the engine, and not one of the turbo manufactures suggest that. I can make 625 at the wheels reliably with a modified SC NSX engine, and could easily make another 200 with a supercharger change. Why would I want to mess with a Turbo? If your going all out I would also suggest water to air intercoolers vs. air to air, as big as you can get. Water volume is the key so you will need to get creative with the hose size and water pump you use. After a certain temperature the gains from cooling can equal a 50% increase to the power added- so if you made 100 hp more you have the potential to make another 50 hp by cooling. The system developed for my WC car cooled the air to 5 degrees above ambient. I prefer to drive the SC with a small pulley, so it will bring me boost earlier in the RPM range, and even if you wanted to use only 8 PSI, you can just pop off at 8. The centrifugal SC is superior in efficiency and heat, it will also allow after cooling because it’s not sitting directly on top of the intake. The RTR designed location in my WC car requires you to remove a chassis brace. And is hard to work on. When using a cage this is not an issue, but then it won’t fit a street car. I would like to think that instead of finding a suitable place to locate the SC, you could remove the alternator and locate the SC there, and now have a much smaller alternator to locate… even buy a mini alternator if needed. Unfortunately that’s on the list like the sheet metal intake, they all take time.
Future and current projects- My current project is a super 3.3 N/A engine with stock compression that will make an estimated 350 at the wheels. This engine is being developed for the Cup car with parts from Science of Speed. I once worked for a famous engine shop that has a great history of world records in drag racing along with many magazine featured accomplishments related to drivable cars that were capable of great track times. This set my standards high, so if any NSX owner is interested in a project, I gravitate towards special projects regardless of the difficulty. A concept I have toyed with is a “Super Turbo” that used a gear increased centrifugal supercharger (for low RPM), that shared the same shaft as an exhaust driven turbine for maximum boost at high RPM. The blower shaft could go faster then the belt that drives it, by designing a one way drive similar to the rear sprocket of a bicycle. Another less challenging project is to build my own SC system with a bigger SC unit and after cooler (By bigger, I mean more CFM), I don’t care for the shaft drive, And the RTR designed location in my WC car requires you to remove a chassis brace and is hard to work on. When using a cage this is not an issue, but then it’s not a great choice for a street car. I would like to think that instead of finding a suitable place to locate the SC, you could remove the alternator and locate the SC there, and now have a much smaller alternator to locate… or even buy a mini alternator if needed. Unfortunately that’s on the list, like the sheet metal intake- they all take
The recommendation to Honda for the future NSX (With all due respect)- First fix the oil pump gears. Then fix the retainer or keeper weakness in the heads. Offer dealer sold performance parts… here in the US, the racing associations allow dealer sold upgrades even if they are aftermarket parts (As long as they have Honda part numbers). Examples would be Nismo, Mazda Speed, or Mopar. These dealer offered parts like cams, brake upgrades, and ground effects are why cars like the 350Z can out perform the NSX in pro competition. For improved performance the runner up would be Brakes, please offer 4 piston dual sided calipers or a factory offered aftermarket upgrade (Like the RSX). Use industry standard rotor sizes so that when the consumer upgrades, the cost is competitive with other cars. The number one improvement that would most improve the ability of the NSX in competition with cars in this category is (Both on the track and in sales); Horsepower potential. Yes the potential is more important then actual HP from the factory. The addition of boost via turbo or supercharger allows the customer to increase the performance for much less money and without engine removal. Offer programmability with your ECU and ABS systems, they are race breed equipment programmed for street applications. I don’t buy the need for secrecy to minimize OE re-programming. It takes me 5 minutes to change back to a stock ECU and still request warranty coverage. Honda is the #1 Japanese manufacturer for performance, and contribution in motorsports. Please continue the effort, and re-visit the supercar category.
All content and text are protected by copyright laws and are considered property of DAL Motorsports LLC. Content as a whole or part may not be copied without the written permission from either Rob Morrison or DAL Motorsports.