top of page
  • Writer's pictureRoger Maeda

Building a Honda Prelude Race Car - Part 4

The Engine

This is probably the question that you've all been waiting to have answered. What does the engine bay of a Honda Prelude racecar look like? Well, this is what you see when you open the hood on my car. 

Those of you who are familiar with Honda engines have probably immediately realized that this isn't a hybrid motor or even a VTEC-equipped H22. It's actually a run-of-the-mill H23A non-VTEC motor from the USDM 4th gen Prelude Si. Yes, the H23 has a lot less power out of the box than the H22. And there's no doubting the power potential of a H22, especially when you are free to tune the ECU as you are in SCCA Improved Touring.

Yet there are quite a few reasons to choose the lower output H23 over the h22 for a racecar:


1. Overall Car Performance

The main reason is, susrprisingly, performance. Although Improved Touring isn't a power to weight restricted class, the minimum weights and performance allowances are set using a formula which takes into account the stock power of the engine, drivetrain layout (FWD vs RWD) and suspension layout. Since many cars in Improved Touring S are capable of producing upwards of 200hp at the crank, the H23 powered 4th gen Prelude gets a generous minimum weight of just 2555 lbs with driver.

While the H22A powered Prelude VTEC is fairly nippy at 2850 lbs, the H23A powered Si is a rocket that stops, turns and goes with the best of them. It also helps that the H23A is actually very responsive to the engine modifications allowed in Improved Touring.

2. Cost

Cost is always a factor and of course, the H23A is the cheaper option all around. Not only is the motor plentiful in healthy, low-mileage condition, everything from headwork to internal work is cheaper and easier. Even the transmissions are easy finds. The transmission that's in the car now is a $100 unit from an ordinary salvage yard in Newark.

3. Drivability

Unless you have been following all of the drama around the new hybrid motors they introduced in the 2015 Formula 1 season, you probably haven't heard this term before. Drivability is a generic term that racers use to describe the manner in which a car outputs its power.

Generally speaking, a car with good drivability characteristics delivers more power as you rev it higher and gives you the feeling that applying a sudden increase of throttle will give you an instant response from the engine. A big displacement V6 or V8, like what you'll find in a new Acura TL or a Chevrolet Corvette will have excellent drivability.

Conversely, a car with poor drivability is one that has the driver guessing how much power he'll get when he floors the throttle. Laggy four cylinder turbo motors and the VVTL-I equipped Toyota 2ZZ-GE in the 7th gen Toyota Celica are examples of motors with poor drivability.

Drivability is critical in wheel to wheel racing, simply because there's so much that you have to worry about when you are racing alongside other cars at high speed. If I'm fighting for position through a 90-mph sweeper, the last thing I want to worry about is whether I'm going to get a sudden, unexpected kick of power when I open the throttle at the exit. The big displacement and mild cams in the H23A take all those worries away. The motor comes with excellent drivability from the factory and it's very, very difficult to ruin it.

So how do you build a race-ready H23A? Let's start backwards, from the outside in.


The Bolt-Ons

Improved Touring is a very easy class for the budget-minded tuner to work within. You are allowed to change the intake, header, exhaust, catalytic converter and change consumables like spark plugs and the air filter. There's definite value to doing each component separately, but the modifications work best when you do them all together.


Cold Air Intake

The Improved Touring rules require that the engine air intake pull its air from the stock location. The rules were worded that way to prevent builders from designing odd-looking ram air setups that pull from outside the engine bay. Fortunately, the stock air intake goes to the inside of the front passenger side wheel well.

The popular AEM long tube Cold Air Intake is both legal and effective, so that's what I've kept on my car (kept, because the car came with it when I bought it as a street car). In order to give it as much fresh air as possible, I cut away the plastic fender liner on the passenger side wheel well. If you look through the lower foglight hole in the front bumper, you can now see the air filter. It's not quite ram air, but it's as close as I can reasonably get within the ruleset.

In order to keep the heat of the engine bay from affecting intake temperatures, I've also taped parts of the header with aluminized fiberglass insulation wrap. Underhood temps in the H23A tend to run extremely high, so this makes a difference over longer races.

That being said, there may be even more power to be gained here. Jeff at Evans Tuning once told me that the cheapest way to get power out of 4 cylinder non-VTEC motors from this era is to run as long an intake tube as possible. Robert Oliver, formerly owner and proprietor of North End Performance (now turned my current go-to engine expert), tested that exact theory for me by varying the length of the intake on my car.

Take a look at these results:

The blue line shows the power vs RPM with the AEM intake. The red line is the same run under the same conditions using a short ram intake with an overall length of about 12" (including the filter). Both runs were done with the hood open to limit the variance from the heat coming from under the hood. If you're having trouble reading the tiny text, click on the dyno graph to see the whole thing in a higher resolution.

Under identical conditions with the same air filter, adding 24" of piping to the intake length causes a noticable power difference between 3500 RPM and 4500 RPM. WinPEP 7 (the dyno display software that comes with DynoJet dynos) shows a difference of between 6.5 hp and 7.2 hp between 4100 RPM and 4400 RPM. This might not sound like much, but low RPM torque is vital in a road racing car. This is exactly where the tachometer is when I come out of a slow 3rd gear corner like turn 1 at Summit Point or turn 1 at NJMP lightning. 7 hp at the front wheels is the difference between me having to fight for space alongside RX7s and 240Zs versus being able to get ahead of them coming out of the first corner.

My engine experts are both giving me very strong indications that if it were possible to fit an intake tube with another 8 to 12 inches of length over the AEM and fit it into the space available, that I could earn an even bigger boost in the midrange. Unfortunately, I haven't been able to figure out how to stuff 4 feet of piping under the hood of the Prelude without adding some seriously flow-restrictive bends. If anyone has any ideas on how to achieve this, give me a shout. I would love to hear your ideas.


There's a surprising amount of power to be found in attaching a well-designed, equal-length exhaust header to the H23A. Unfortunately, almost all of the bolt-on header out there barely make a difference in the overall power output of the non-VTEC motor.

After trying every option from generic eBay to the ubiquitous DC Sports 4-1, I finally broke down and spent the $1250 to have a header built by Hytech Exhaust in Irvine, California. Boy, was it worth it. even from my terrible photos, you can tell that it's a completely different design from every other header on the market. The huge pipes coming out of the head are almost 2" in diameter and the collector for the 4-2-1 setup sits far back, behind the oil pan. The primaries are almost identical in length, down to the inch.

In fact, the Hytech header flows so well that it actually confuses the stock ECU at higher RPMs. This leads to some highly entertaining off-throttle backfires which are both very cool looking and incredibly unhelpful as far as generating power to the front wheels. While there's nothing wrong with using this header with the stock ECU, you really want to invest in a programmable ECU in order to get everything you paid for out of these headers. Believe me, it's worth it. Between this header and a quick tune of my Hondata equipped ECU by Jeff Evans, I saw a 15whp gain at 6000 rpm. In ITS, 15whp is the difference between a front running car and one that can barely keep up with the lower midfield. When I think of it that way, I say that it's worth every penny.



Part of the reason the Hytech header makes such a huge difference is that the H23A pushes a lot of exhaust out of the back end. Not only does backpressure become a big deal, factors like exhaust reversion (a power-robbing phenomenon which happens when the engine sucks hot exhaust gas back into the combustion chamber from the exhaust side) become a big issue.

The guys at Hytech knew right off the bat that this would be a problem and gave me very specific instructions on how to build the exhaust. Through the course of a 20-minute phone conversation with the foremost Honda header experts in the US I was told that the best way to produce power from an H23A is to:

  1. Run a 12" straight pipe from the 2.5" ID outlet on the end of the header collector

  2. Expand the pipe out to as big an expansion chamber as would fit for the next 18-24". I was half-jokingly told to go look for a 6" straight muffler out of an 18 wheeler and install it here.

  3. Build the least restrictive, mandrel bent exhaust aft of the expansion chamber. The pipe diameter has to be a minimum of 2.5" ID. 3" would be better.

Being the cheapskate and skeptic that I was at the time, I ignored this advice at first, instead choosing to stay with a 2.5" ID mandrel bent cat back exhaust that I had on hand. Until Robert coerced me into building a custom exhaust to these exact specs (which, by the way, he arrived at independently without any detailed knowledge of my conversations with Hytech), I had no idea what I was missing out on.

The exhaust setup that I have on the Prelude today sounds like this:

  1. Hytech 4-1 race header from the head to an area behind the oil pan

  2. A 12" segment of 2.5" ID stainless going from the header to a stainless steel flex coupling from Summit Racing

  3. A 22" long expansion chamber that tapers from 6" to 5" in diameter. The expansion chamber is literally just a huge hollow tube with adapters on both ends to connect to the 2.5" ID pipe on both sides. The 12" segment in part 2 actually extends into the expansion chamber to provide an anti-reversion effect and keep hot exhaust gasses from re-entering the engine.

  4. A section of 2.5" mandrel pipe to get the exhaust around the fuel tank and out the back of the car, where the stock exhaust exits

  5. A Burns Stainless 17" lightweight muffler

  6. A 4.5" ID piece of pipe leading from the muffler out the back of the car so it doesn't set the rear bumper on fire

Not only did this setup produce instant power gains on the dyno (something like 8hp just from these modifications), it is significantly quieter than any of the bolt on exhausts I have ever had on the car.


The Borla stainless steel exhaust that came with the car would easily exceed 103dBA at 6400 RPM from 50 feet away and it had two resonators before the muffler. This new exhaust barely registers 88dBA at 6400 RPM from 50 feet. If you know anything about the Decibel scale and its non-linear scaling, you'll realize that the noise level coming out of my new exhaust is a tiny fraction of what the old 2.5" straight through exhaust would belt out. Noise is a serious distraction and a cause for major fatigue when you're racing wheel to wheel, so even if this didn't produce significant power gains, I would still recommend this exhaust setup wholeheartedly to anyone working on a Naturally Aspirated setup for their Prelude.


*I realize this is hard to visualize without photos and I promise I will get you guys photos. I just need to get the Prelude up in the air and get the lighting in the right spot for clear under-chassis photos. I'll also record an audio clip if things are quiet this weekend, so you can get an idea of what this setup sounds like.


Next Up...


So what's next? Before I continue onwards towards the ECU and some of smaller items that went into this racecar, I'm planning to get some more details and photos around the header and exhaust. I firmly believe that modifying the header and exhaust is a critical step towards making the usable power that will help you go fast on a real racetrack. Once I have that, I'll move on to some of the other components that make this Prelude Si such a force to be reckoned with on track.


Commenting has been turned off.
bottom of page