Project NSX: Dreams Of Power – Part One

At the end of my last Project NSX update, the car was deregistered at the Saitama Land Transport office and ready to make the journey Down Under to its new garage in Melbourne, Australia.

The car wouldn’t be rolling onto the ship straight away, however. I’d been collecting parts for a final bit of work while we were still on Honda’s home turf.


The guiding principle for Project NSX has been to bring the performance of the car up to modern standards, without compromising the original sporting spirit. More specifically, I’m targeting the power-to-weight ratio of the explosive Porsche 996 GT3 without resorting to forced induction.

For the most part, the decision to abstain from boost is just personal preference – I love boosted cars but the NSX feels so right with a naturally aspirated powerplant.


The Advance workshop down in Yokohama had looked after Project NSX’s upgrades until this point, but their packed schedule meant they wouldn’t have space on a lift for engine-out work for months. As a result, I asked my car buddies if there were any shops recommended for this sort of work, and good friend Masa came to the rescue.


Masa introduced me to Suga-san, who runs the Motor Club workshop just north of Tokyo. Suga-san has a history in factory motorsport teams, but recently opened up his shop to give himself more freedom.


Suga-san and Masa both race Honda S2000s, and Suga-san’s had just received a fresh lick of orange paint prior to our visit. The S2000 and NSX share a surprising amount of DNA beneath the surface, not to mention that completely unfiltered NA driving experience.

The S2000’s F20C famously held the production car record for power output per litre at a zinging 123.5hp/litre. Although groundbreaking at the time, the NSX’s C30A is a bit of a slouch in comparison at 90hp/litre from factory. But let’s see if we can’t close that gap a wee bit.

Pieces Of The Puzzle

It was time for a little bon voyage gift to 27 – a premature 100,000km service as well as some performance goodies from some of Japan’s stalwart tuning brands.


Nothing beats unwrapping new power parts. Project NSX’s origins as an automatic car will soon be all but forgotten – the ever-so-slightly less powerful AT cams finally banished to the landfill and replaced with some spicier offerings from Toda.


Toda offers three versions of their performance camshaft for the NSX with varying levels of lift and duration, of which I selected the ‘one chilli’ Spec A. I toyed with the idea of running something more aggressive, but making the most of the Spec B or C profiles really requires an expensive bottom end build which is simply out of scope for this project.

Cam lift

The Spec A camshaft keeps a near-stock cam profile for the non-VTEC lobes but ramps up on VTEC, targeting uncompromised low-speed drivability with improved top-end power. Like an OEM, Toda designs the cams with an asymmetric profile for a quick valve opening and a slightly slower closing to reduce stress on the valvetrain. You can also notice how both Honda and Toda design their camshafts with a different profile on the primary and secondary cam lobes to encourage better swirl and mixture of air/fuel in the combustion chamber.


These parts aren’t cheap, but Toda parts have a strong reputation both in Japan and abroad, and the fact that these are brand new camshafts cut from blanks (as opposed to regrinds) gives me a bit more confidence in drivability and surface durability.

After all, this is the company that built the 500hp engines that powered the NSX to victory in the JGTC.


Complementing the camshafts is a full set of Toda valve springs, which allow the valve more lift while preventing float (when a valve is thrown off the end of a tall cam lobe).


It also seemed sensible to install Toda’s reinforced timing belt, designed to be stronger and more resistant to the additional stresses of a tuned engine.


In the spirit of improving reliability, I also placed an order for one of Toda’s upgraded oil pump gears. The stock design is fine but can shatter during an over-rev, which can obviously be catastrophic for the rest of the engine as oil flow is replaced by a shrapnel grenade of metal pieces.

These Toda parts are hardened to resist shattering even in over-rev situations. A 5th to 2nd gear down-change might still bend some valves, but the damage will be contained.

ProjectNSX-blakejones-speedhunters--23 copy

Also on the oiling side I decided to replace the stock pan with something more track appropriate. This is not a ‘must-do’ modification for the C30A, but there have been reports of NSX’s running grippy modern tyres pulling high Gs and starving for oil on right-hand corners, which this baffled pan is designed to prevent.

This Racing Factory Yamamoto unit is cast aluminium, holds 1.3L more oil than the stock oil pan, and features enlarged cooling vanes, which will both assist with keeping the oil cool on track.


While the engine was out, it made sense to replace the perishables that would usually be tucked into difficult-to-reach corners of the engine bay. Suga-san placed an order with the local Honda dealership for all hoses, rubbers and gaskets.

Before work began, we applied some protective film to the car’s exterior to protect it – not so much for the workshop, but for the international journey ahead.


Removing the NSX engine’s is undoubtedly a big job, but with the right workshop space it’s not as complicated nor time consuming as you might think. Once harnesses, fuel and vacuum lines, exhaust and so on are disconnected, the entire rear subframe is lowered out the bottom of the car with the engine still attached. It took Suga-san about four hours.


The C30A is a surprisingly large engine once removed from the engine bay – the 90-degree angle between each bank exacerbates the girth of the quad-cam heads. In the car the rear bank sits almost completely out of sight in its hiding spot against the rear firewall.


The chance doesn’t come up too often for this photo opportunity, but it took a bit of convincing to get Suga-san in there with me…


I wasn’t just horsing around – it was a rare chance to see some parts of the NSX that are usually completely hidden.


Like the cast aluminium subframe and wishbones, which have an almost organic H.R.Geiger quality to them.


I’ve been loving the sound and power provided by the Fujitsubo manifold, and it was interesting to see the heat-induced colour change from when they were installed last year.

For any NSX owner considering a header upgrade, I really recommend considering the Fujitsubo units. Little details like the titanium heat shields add up to an unrivalled package for the Honda.


The Spoon Rigid Collars were also visible on the rear subframe. A few were split/damaged during the removal process (this is expected with their design), so will need to be replaced in the future.

Where’s Your Head At?

While I was bouncing back and forth between Tokyo and Melbourne in preparation for my impending move, Suga-san got straight to work tearing into the C30A’s valvetrain. He kindly photographed much of the process so I could share it with you here.

This won’t be the most detailed chronology, but I’ve tried to capture the main points here through the photos from Suga-san.


Here’s the disassembled cam holders for one cylinder bank. You can also see the billet cam plugs in the left of frame that we chose to replace the stock rubber units.


A view inside the head with valve cover, camshafts, rocker arms removed. Everything seemed to be in relatively good condition.


Here you can see the standard valve spring (left) and the stiffer Toda unit (right) which replaced it.


Another small upgrade worth making to the C30A is upgrading the lost motion assembly or LMA. The LMA cradles the large VTEC cam lobe while the car is in non-VTEC range.

The early cartridge design on the right is known to get sticky with age, symptomised by a noisy clicking from the top end of the engine at idle. The simpler spring design was introduced for ’99 and newer NSXs, and Honda later designed an upgraded LMA for the earlier cars, available from their parts department. It’s pretty neat that Honda are still improving a spare part for a car which is out of production.


I found some new VTEC rocker arms that have been slightly lightened and polished, so those went in too. For those unfamiliar with a VTEC head, the two outer arms actuate the intake (and exhaust, on the opposite side) valves independently. When in VTEC, the centre arm is engaged and locked to the two outer arms. They then operate according to the VTEC cam profile, which has higher lift and longer duration, meaning more power at peak RPM.

Yes, VTEC just kicked in, yo.


Rocker arms go in and valves are adjusted.


Then the camshafts and holders are reinstalled, and clearances checked to ensure everything is exactly where it is meant to be. We decided not to open the can of worms of degreeing the cams and simply used the same setting as the factory camshafts. Considering that everything else internally is still to factory specifications this won’t cause any problems. In a perfect world I’d have installed adjustable cam sprockets and used an engine dyno to tune for maximum power, but the cost involved would have likely been enough to buy a nice S2000.

Being a V-configured engine, it means twice as much work for Suga-san as the same process is followed for the other head.


Moving around to the front of the engine…

Water pump bearings degrade over time, so a new one was installed while we had the timing belt off. You can also see a common sight on all NSX engines in the top right corner of this photo – black sludge coming from behind the cam gears.


The culprit is this crankshaft position sensor that inexplicably melts over time but continues to function perfectly. It’s not really a problem, but Suga-san cleaned up the area around the sensor and installed new ones (which will eventually melt).


Now that the camshafts have been changed, the final difference between an auto and manual C30A is the harmonic balancer pulley you see here.


All engines create potentially damaging resonant vibrations by their very nature. Factors like stroke, cylinder pressure, crank-train stiffness and inertia all culminate in what is known as the ‘mass elastic performance’ of the engine. Basically speaking, how much shock runs through the engine during operation. A harmonic damper or torsional vibration damper (TVD) helps to convert the energy associated with the undesirable torsional vibrations into heat. A well designed or matched TVD makes this energy conversion at the resonance point that is most damaging to the engine. The energy is stored in the viscous fluid in a fluid damper, or the rubber element of a rubber type damper.

The mass of the inertia ring, along with the viscosity of the fluid (or stiffness of the rubber) is tuned to the resonance of the cranktrain in order to be most effective. It is a very simple looking part but has a very clever interplay in the engine itself. The automatic pulley (on the left) is actually lighter than the manual variant, presumably due to the flywheel assembly affecting either the stiffness or inertia values of the crank-train, changing the frequency output and retuning the energy required to be ‘buffered’ by the TVD. Installing a lightweight pulley might seem like a good idea to free-up the engine, but could potentially reduce engine life, as the torsional vibrations of the crank-train would be left unchecked.

I was considering an aftermarket fluid-damper replacement, but Suga-san talked me out of it since the OEM unit is tuned to the correct frequency, and as we are not changing any components in the bottom end, the stiffness and masses would remain the same, so the frequency that the TVD is tuned to would remain correct. Something I’ve come to appreciate lately is the amount of design and testing that goes into OEM parts – it’s very rarely something the aftermarket can compete with.


T3 Craft Works, another NSX specialty workshop, makes this aluminium shield that mounts in front of the balancer pulley. If the pulley ever suffers a de-lamination failure it can blast into the timing belt, causing a skipped tooth or worse – and potentially a fatal meeting between piston and valve. Circuit driving increases the chances of this unlikely failure, so it’s nice protection to have.


The stock cam gears were reinstalled into their original positions.


Then it was all tied together with the very red Toda timing belt and tensioned to spec.




Underneath the engine, the stock oil pan was removed.


Here you can see the replacement RFY pan (grey) next to the stock pan (black with gold).


Suga-san attached the sump with liquid gasket and applied heat-reflective material to the raised section of the oil pan under which the front exhaust header runs to minimize heat soak.


He used more of the same material to add some extra protection to the two titanium Fujitsubo heat shields that protect the rear firewall and trunk, as seen above.


The front firewall shield received the same treatment. You can see the fuel tank just in front of the shield, mounted as low and central to the car as possible to minimise balance changes due to varied fuel loads.


A few finishing touches, replacing the alternator with a new unit…


The alternator sits on top of the C30A, so visually it’s nice to have a fresh one.

The starter motor was similarly replaced with a fresh OEM component.

image1 (2)

The photos from Suga-san stopped pinging into my email inbox for a few days, and I started to get a bit nervous. Then this: “Project NSX all back together and ready for collection.”

Note the raised suspension; who knows what the approach angle to a 60,000 ton roll-on-roll-off shipping vessel is? I sure don’t.


With a clean bill of engine health and a few special parts the car was finally ready to leave the motherland. As the title suggests, this engine work is but the first step of a longer hunt for power, and at the very least the car requires a tune once in Australia to make the most of the new camshafts. Then we can see what progress we’ve made towards hunting down those six-cylinder rivals.

I’ll take bets on final peak flywheel horsepower* in the comments section of the next dyno session. Do you think we will just scrape by and achieve the Stage One 290hp goal, or break the 100hp/litre barrier and go to 300hp+? Closest guess gets a Speedhunters sticker mailed to them.

*measured wheel horsepower x 1.15 (drivetrain loss) 

Blake Jones
Instagram: blaketjones



Comments are closed.


by Oldest
by Best by Newest by Oldest
Brennan McKissick

As always, I love the thought process behind the build! I'm gonna go with 292hp.


292 was gonna be my guess.... :P


299hp. Close but no cigar..

Those purp disc bells look sweet!


Thats some great progress here, on my side i'll go with 297hp.. so close to the 300 barrier.


Wild guess on my part without a reflash: 284 at the crank. With a reflash: 293 at the crank. Ignition timing becomes more critical once another camshaft is chosen. As for the 1,15 rate: Drivetrain losses should be at a constant, so aren't based on a percentage. Only replacing something in the drivetrain will alter that. Adding more power does add more heat, but so long as there isn't anything slipping there shouldn't be more drivetrain los in a manual....


Huh, I'd always believed it was a percentage and just used 15% as a generic figure. But I know you know your stuff Kevski, and now I think about it as long as RPM and the drivetrain are constants, there shouldn't be more power used/'lost' by the drivetrain to turn the wheels. For the sake of this fun experiment I'll stick with 15%, but point well noted!


I'm pretty sure he just has the 1.15 drivetrain loss figure so that when he has a dyno number he can find an estimated flywheel HP. 15% is a very commonly used generic drivetrain loss number. I don't think he means anything of a change of drivetrain loss. Besides that way we all know how he is getting his flywheel hp number after getting the dyno done (let's not go into differences between types of dyno, humidity, elevation etc - this is just for fun)
I'm guessing 282hp (crank of course)


@ Josh D. Lets say its 15% of 200 hp. Thats 30 hp. Then you up the power to lets say 500 hp. If it where a percentage that would mean the same drivetrain would absorb 75 hp. And thats out of the question. A drivetrain takes up a set amount of horsepower. If a drivetrain takes up a percentage that would mean there is a 15% slip involved in the drivetrain. A percentage is only used on a stock car, but should, in any other case, be avoided....


Hmm I get what your saying, but isn't that exactly what we see? For example just did a random search on Mustang GT stock dyno and I saw numbers between 360-380ish whp while the factory rated HP is 435 and at 15% drive-train loss the 380whp ~ 437hp which is close to the 435hp factory number - there is a ~57hp loss, then looking at an hypothetical '95 US E36 that might dyno at 200-210whp with a factory rated 240hp there is ~30hp loss which is less of a hp loss than the Mustang (without getting into types of dynos etc). I thought the parasitic loss of the drive-train is usually calculated at a percentage?

I think your saying if I had a stock car and I dyno it, at say 200whp, and it was rated factory 230 there is a drive-train loss of 30hp so if i slap a turbo on it and make 350whp I should only add 30hp for the drive-train loss since that should not have changed? That seems really interesting but I'm just curious how all I have ever dealt with or heard has been a percentage of loss, never a specific number for the drive-train?


I second what Jackhead says.

And to further go into the matter: The problem arises with the difference inmanual and auto transmissions. Or more precisely: a Torque convertor. A torque convertor uses a fluid coupling which always has some form of slip. Especially when you ad more power to it. A clutch on the other hand does not have slip. Its either not slipping or slipping. The only increase you get from sapped horsepower through the drivetrain, is in the form of heat buildup. That should equate to a couple of HP at best though.

As for the conclusion of Jackhead: Thats in a sense right as well. But it can be simplified. What does a stock dyno pull result at the wheels for a NSX? And what does the same car have as claimed hp for that specific engine?

Subtract the claimed hp from the dyno pull wheel results, and it should be a negative number. Thats your drivetrain loss. Lets say it is 15% on a 3.0 for the sake of argument?

3.0 crank HP: 270 HP
3.0 wheel HP: 229,5HP (which would seem a bit low? I would estimate it at the 235HP mark, but I have no experience wit this specific engine/gearbox)

229,5 - 270 = -40,5HP of drivetrain los. With the kind of gains your aiming at that wouldn't change at all.

The main problem with it all is all of the following though: Wheels and tire combo (weight, width and circumfence. Can have any sort of impact depending on how much you change them), final drive (lot of impact actually), Oil (almost no impact in itself, but can have an impact if weather/running temps change.).

you never heard of it before, as almost no people have the opportunity to get an engine on an engine dyno, and have a car with the engine installed on a rolling road. Thats the only real way to judge drivetrain losses. Thats mostly exclusively an OEM/high level motorsports practise.

Easiest way to find out the OEM drivetrain losses would be to just contact Honda and ask them directly. They should be able to provide that info as long as you provide Engine/Gearbox/F.D. and car info. But it still doens't take tire size into account, allthough it should be fairly easy to extrapolate that data.


@Kevski-Style Cool! Thanks for the explanation, it is always good to learn more!


That's because the internet is doing ricer math...

You lose X amount of HP for example through the diff. It doesn't drastically change if you doubled the engine power. It would largely remain the same.

He won't ever be able to find true crank HP because he had no baseline to apply the delta to after all the work.


**282hp before being tuned**


I'd say it would be around 307hp.
Take an AMG for the next project speed hunters!!!! :)
Keep up all the awesome content and coverage.


307 would be mega. Fingers crossed! Thanks for the support.


Great article with some excellent info. I'm gonna guess 288hp. Looking forward to the next chapter!


Thanks Wires, glad you enjoyed it.


I think with all upgrades 305 HP


To-da! :)

great thread. i'll go with 291hp.


Bring the car up to modern standards?

So, you'll be adding a thousand pounds of lead ballast, then?


And then brake-by-wire and a CVT!


And an i-VTEC system with fifteen more ECU modules and a couple of extra electric motors to make it all work.

My guess is probably around 294 engine horsepower...


Seeing those headers made me wish you went with a dual exhaust exit system. IMO it looks better than just a single exit in the middle. Having each bank have its own system might be a tad heavier, but looks better and sometimes you can notice differences in pitch and sound between cylinder banks idk


Hmmmm I'd go with 297.6


That engine will begin to discourage you as you near 325-335hp mark since the budget rules the build, but the body and chassis can do much better than 1270kg if you focus from that direction with the budget.
After around €20k with your long term approach, you'll be wanting to aim at 997 GT3's after being bored with aged yet modded 996's... The spiral that follows doesn't have limits!


Yes, there's certainly diminishing returns when it comes to extracting naturally-aspirated power from these engines! To go any further would require a lot of machine work - more displacement, compression and porting. At the end of the day it's not about beating this or that modern car (there's always something faster!), but taking the platform to a level of performance that I feel it can handle without compromising drivability or reliability. You're right about the weight - probably a few more kilos to be found with composites.


295.7bhp at the flywheel

Also what's happening with the store?? It used to have a countdown and we're well passed the end of that now.


We had hoped to have it re-opened by now, but we've decided to go for a full design refresh and all new products, so it's just taking a little longer than we originally planned.

We hope that you think it will be worth the wait!

Tigo Koppelman

To-da! I will go for 296hp


I think hp will be 303


Great update. I always wanted to do this with my NA1 when I had it.

Excellent work on the harmonic balancer, mine came apart on my 60k km NSX while cruising on the highway. It caused $6k of damage


Thanks Joe. Wow, sorry to hear about your harmonic balancer disaster - 60k is still well within the recommended service interval, right? Was that new valves/pistons for your car?


I believe that the valves were closed when the belts snapped so I got away with just a head rebuild.

The Honda tech said it was just age, the rubber had deteriorated over the cars life and ironically the harmonic balancer became extremely unbalanced.. I’d say this is crucial service part for any NSX.

It was a 91 and happened about eight years ago.


Lucky outcome then. I'll be keeping a close eye on the balancer and the timing belt. Funny how such a simple part plays such a complex and important role.


I'm going for 298 hp at the crank
Nice work and article ! So eager to see what following !


I love this update, going to say 311hp.

Richard Chavez

ill say 298hp.


Nice Giugiaro designed alien watch sneaking in there. 305 hp.

George Norman

As a car mechanic I love people like you who just want it right. No, how much is this, how much is that. Just do a good job. Been following the builds on speedhunters for years this and project gti are my favourite. I’ll go with 301 hp.


Man Speedhunters must really pay well..


That noise you hear is the the entire staff of contributors laughing themselves to death.

If you want to be rich, this isn't the game for it but it's by far one of the best jobs going :)


I'm going to guess on the optimistic side. I'll say it makes 313 hp!


294 for me




287 hp before tune is my guess






309 ^_^


306 hp

Oskar Pettersson

Nice post! Love to see some work done to Project NSX.
Since most other numbers has been taken I'll go for 300 flat ;)


Good things are worth the wait they say and that certainly rings true for this update! Fantastic as always.

Lots of variables to consider but given how I've come to know your philosophy, I'm sure the tune will be for pump fuel 98 RON. My prediction is 296hp at the flywheel.


Yup, only 98RON for Project NSX. Thanks Martin!


I guess it's going to make 296 hp


@blake this is a proper project car update. The detail you go into makes for great reading. Can't wait for the next one.


Cheers Ben!


It'll be 264 to the wheels so 303 crank hp.


I want you to break 300bhp so bad!

Jay Soh Tsu Chung

I would say it'll be producing 294 hp on flywheel.


The development of camshaft angles, not to mention valve polishing, has always amazed me; how the difference of mere micrometers can make such a difference. Great article!


let's be optimistic, 301 hp. Some motors just have something special and exceed expectations.


302 hp

Alistair Lawes

Great work there, going to be far sharper on VTEC and much more enjoyable. I reckon it'll make 302hp.


I think I say the same on every Project NSX article, but I will say it again. It's my favourite project. I wouldn't do anything any other way than what you have been doing. I love the goals you've set for the car, and yourself. OEM+, adding lightness, a bit more power, and Porsches and M3s to beat.
And when I saw the picture with all new hoses, gaskets, etc., brother from another mother!!!!

Can't wait to see the next post.

PS: To-da! got literally laughing out loud.


Glad you're enjoying the updates so far, there's a few more in the pipeline I think you'll like!


308! Good luck man, thanks for the awesome content!


As an optimist and fellow brooklands green owner I’m gonna say 315hp!

Love the build and the updates!


BGP represent! 315hp with this setup would be amazing.


I'd say 291HP


Good stuff! I’ve gone with the TODA upgraded oil pump gears, Spoon for the baffled oil pan and I’ll be going JUN for all of my internals. GReddy timing belt as well.


I'm going to go with 294, right in the middle of most of the guesses