Power Modifications

Guideline: Estimated power 'caps' are ~230whp NA and ~400whp turbocharged. That being said, getting past ~215whp NA will require a >$5,000 engine rebuild and/or drivetrain friction reduction treatments. Both if you want over ~220whp. Exceptions are possible, but don't EXPECT it. Getting past ~300whp with FI will usually also involve an engine rebuild, either because you want to do it properly in the first place, or the engine you have shattered while trying. Crossing the 300whp mark also means that you can't expect that your engine will last more than ~10,000 miles. More is possible, but you can't EXPECT it. To get to 400whp you will require a complete top notch engine rebuild with significant changes to increase the strength of the engine, including the block itself. Don't expect your engine to last long.

Guideline: If you can't find a dyno to prove the gains, assume that claims of power gains are 'Willful misrepresentation of the truth in order to bilk gullible customers out of their money'. Also known as LYING. The sole exception to this is if it's an actual race team making the claim as they typically don't sell their developed products, and won't share their exact results. But then, they aren't trying to scam you either at that point. Any company that is trying to sell to the public has every interest in showing their product in the best possible light. So if they can't produce actual dyno sheets (not just photoshopped ones), then it's either because they have them and it proves their product is useless or they don't have them because they never tested. And if they never tested, how can it be trusted to produce the results advertised? It can't.
Also, remember that unless the dyno comparison is done on the same car at the same dyno in the same environmental conditions, comparing the numbers is meaningless. Comparing PEAK numbers is also largely meaningless, as it's worthless to gain 10hp in a 200rpm band by dropping 10hp across the rest of our 5-6,000 rpm band.

Guideline: The relative "high power" limitations to the Renesis are largely due to the exhaust ports. They can not be increased in size by much since they are run through the side plates, between the oil and coolant passages, and the studs, etc... Prior RX-7s had the entire end of the rotor housing to work with for exhaust dumping. Adding exhaust ports to the housing is viable, though imperfect and makes the engine no longer a Renesis, rendering this point moot. Any high power goals will have to take this fact into account with some serious consideration. There are some methods of increased exhaust porting, though none known have performed flawlessly, usually introducing different issues that have to be addressed. There are also other points that produce capping effects on power, but the exhaust port size is the the one that people run into first and is the hardest to get around.

Guideline: The more power you are making, the more heat you are pushing through the exhaust ports. Consistently high EGTs (exhaust gas temperature) have been shown to cause problems with the side seals and side seal springs. There are resolutions to this, if you are having the engine rebuilt. High power cars with a reliable and stable tune and fueling will likely have their side seals fail before the apex seals, however when a side seal fails, it's because of the spring deforming, pushing the seal out until it clips the exhaust port, shattering it and the seal fragments then destroy the apex seals, so most people might assume apex seal failure when it's actually the side seals that went first. High power with an unstable tune or fueling moves the likely failure point to a detonation that destroys the apex seals first. As far as the engine is concerned, the results are the same between these.

Guideline: If you live in California, you have CARB certifications to check for, consider, and either include or discard depending on your own decisions of how to proceed.


The OEM intake is really VERY good. The only 2 aftermarket intakes that can boast a power improvement are the Mazdaspeed and AEM. Which happen to be the exact same thing, just with different logos. That intake can give you a very slight bump in power. And I mean very slight. Within the 2% error margin of modern dynos. (2% of 180rwhp = 3.6hp). The only other viable intake for power reasons is the Racing Beat intake. There are a few other intakes that are net equal to the OEM intake, but not many. Every single other intake than this handful, from K & N to ebay kits, cold air and ram air, complicated and simple, every single one will cause you to lose power. This is because they do not have the proper R+D into how the air flows into our engine. They just fabricated a tube, made a port for the MAF, and jammed an air filter on the end. The air becomes heavily disrupted, the MAF can't read it right, and the ECU believes what the MAF is incorrectly telling it, causing the ECU to over and/or underfuel the engine all over the rev range. Even removing the OEM screens can cause power loss, since those screens are very well designed to straighten the air flow and smooth it down considerably. They aren't there to 'catch rocks'. Just to straighten the air flow.

The filter isn't even bad. There is FAR more breathable surface area on the OEM air filter than nearly every intake kit out there. The OEM filter is NOT a restriction on the air flow. The K & N drop-in panel was dyno'ed by Speedsource to show a consistent 1HP increase vs OEM. However, the K & N's filtering is substandard, and in Daytona races (Grand-Am ST class), they were losing engines to sand ingestion. (reference) 1HP is not worth that risk, especially if you are paying for it. Stick with OEM. The only solid reasons to go after market are:

  • the subjective reason of "I like the sound"
  • or the improvement in how easy it is to get to any items under the OEM air box.

Exhaust Header

The header is one of the few areas where clear power gains can be had. Unfortunately, it's very hard to find isolated dyno charts of just the header gains. Pretty much every good header out there CAN make more power than stock. But it needs to be mated to the right combination of midpipe and catback. You need the entire system working together for the best gains. The header itself likely only has better power gains from OTHER power modifications. The OEM header is sufficient for OEM power levels, but can be quite terrible for higher power levels. Just replacing the header itself will give you a slight bump, but nothing really noticeable. With more power comes more exhaust though, so at some point, the header DOES become a restriction and you can free up that restriction with most of the aftermarket headers out there. The exhaust theme will be "no restrictions" as you will see. We don't want or need 'backpressure'. "Long tube" headers are better than shorter ones, but to date, yo will have to have them custom made. BHR has been working on one for release for a while, but it is still not yet available.

Note:Using a header without a heat shield will dramatically increase underhood temperatures, and it is strongly recommended that you use a heat shield. We have enough heat problems already, don't add to it.

Midpipe / Test Pipe / Race Pipe

If you take the admittedly illegal (for any country that has emissions regulations) step of removing the cat entirely, (which is also removing a dangerous failure point as well as giving you back some power) you have quite a few options.

There are a variety of other midpipes you can buy, including the RX7Store midpipe, Agency Power, Rotary Performance, BHR, Racing Beat, some rarer race developed midpipes, and a host of nameless ebay brands. What you need to know is that the OEM cat kills a TON of noise from the exhaust. Removing the cat introduces this noise back into the exhaust stream. Note that I said "noise", not "sound." An uncorked rotary is a beautiful thing to hear. From a distance. With ear plugs. Rotaries are inherently raspy at full throttle and partial throttle. The kind of raspy that makes your brain bleed as if it was being sandblasted. As one owner put it: "It's like attacking a tin shed with a chain saw, from inside the shed." Rotaries also have a much different pitch of exhaust note than piston engines. It's higher, more piercing. Rotaries are also much much louder than even moderately more powerful cars. This is a side effect of how much energy (heat) we have getting dumped into our exhaust. The 787B 4 rotor being tested and measured for inclusion into Forza 4 was described as being "the loudest thing ever heard" by the sound guys. So loud that they had to use special equipment buffered and layered multiple times to try to dampen down the volume to prevent severe distortion. A 2-rotor isn't quite that loud, but it's certainly loud enough to cause physical pain. Perhaps you don't care, but the local cops, friends, family, and other enthusiasts will NOT share your opinion and cause you much trouble.

In order to combat the noise and tame it down, these midpipes use a variety of methods trying to accomplish the sound reduction. Only 3 midpipes out there use resonators that are easily accessible by consumers: BHR, Racing Beat, and Agency Power. These resonators are physical welded obstructions in the exhaust stream to chance, dampen, or muffle the sound. The rest use some sort of packing. Glass or steel wool are the two more common ones. Steel wool can melt/burn if you hold a match to it. Glass has a melting point between 1,400F and 1,600F We can produce upwards of 2,000F exhaust temps. See where this is going? Midpipes that use steel wool or glass as packing material to dampen the noise WILL fail. The packing will melt and blow out, leaving you with an unresonated, uncorked, and unrestricted straight pipe, full of rasp, drone, and pain. It doesn't matter how cheap it was, you wasted your money. An exception to this is Coast Fabrication. They produces work mostly for race teams, but their midpipes are accessible to the public as well, and their packing is a stainless steel mesh that does not melt or blow out. Racing Beat also says that their midpipes are packed with stainless steel wool, presumably in addition to the resonators.

The key point to remember here is to investigate the type of packing that is used. It has to be able to withstand our exhaust temperatures, otherwise it's a waste of money.

The three resonated midpipes are the BHR single resonated, the Agency Power dual resonated, and the Racing Beat dual resonated. Don't automatically assume that more resonators is better. Agency Power and Racing beat added the 2nd resonator because using only one resonator of a more conventional design was simply too loud. So they added a second one. Racing beat also has a bit of a hybrid, as they use packing as well. It helped enough to make it viable. But there is little R+D into the resonator as it relates to the rotary. As a result, they can still gets raspy, can increase how much a catback drones, and can play havok with the tone you wanted out of your catback. And it can still fail under our heat as our exhaust temps weaken the metal of the resonators, increasing rasp. It just takes longer than any packing. However, these pipes are often more cheaply obtained and so are still popular.

The BHR midpipe is the ONLY consumer midpipe out there specifically designed to work with a rotary engine. It's single resonator is not only capable of eliminating the drone, the rasp, and the excessive volume, it does it without power loss, it does it without structural weaknesses that can fail due to heat or rust, and it does not mess with the tone created by your selected catback. No other marketed consumer midpipe can meet all of these points. None. Some "race" products made by race teams may as well, but they won't be streetable. If you don't want BHR's full pipe, they will sell you the resonator itself for you to put in whatever pipe you want.
Note:This isn't a marketing ploy. If another company wants to step up and create a competitor, they are more than welcome to do it, and I will add a recommendation for that pipe. But none have, so BHR's remains the top choice.

Removing the cat will probably cause a CEL. (P0420 Catalyst system efficiency below threshold ON 2 Catalyst) However, it might not cause it continually, or at all. Our ECU isn't always consistent in detecting when the cat is missing. If you get the CEL for no-cat, then there is only 1 way of clearing it permanently, and that is to block it in a custom ECU flash. The most common way is using an AccessPORT map, however any flashing tool or service should be able to accomplish it as well once the flash has been properly edited. Anti-Foulers and other tricks simply do not work. They work to stop P420, but the ECU is smart number with random testing and looking for expected behavior, that instead of P420, you will end up with any one of 9 other codes instead, and you will get them reliably, vs only periodically or unreliably with P420. So don't bother with those tricks. If you want to block the CEL, get a reflashing device and block it from appearing.

Alternate Catalytic Converter

The OEM cat is quite expensive, but if you must have a cat, there are a few other options. Most of the cheap cats out there will melt and blow out in short order. A Davesport cat installed in a BHR midpipe failed in 1 day of autocrossing, as an example. Largely how fast they fail depends on how much heat and fuel you are pushing through them. Your tune will certainly impact this. One of the only long-term cheap cats that hasn't seemed to fail yet is on HiFlight's car, a metallic-substrate Magnaflow #59959. It has survived daily driving and auto-cross, but has not yet been tested on the track. Other metallic substrate cats might fair a bit better than most as well. You are largely on your own when trying cats though, so I recommend researching properties and limitations, and not just picking at random. A cat option that is expensive, but not OEM, is HJS. They have a cat that can survive our exhaust temps about equally as the OEM cat, but the price is comparable.
A good thread to read on options here: RX8Club.com: Catalytic Converters - Cats Cats Cats!


This is the section of exhaust from the rear tip of the midpipe to the exhaust tips under the rear bumper. Catbacks are largely personal preference on tone and note. Do your research, listen to sound clips. Some will drone badly under cruise, but you might be willing to deal with it. If you increase the size of the tips to 2.5" or higher, expect to have trouble with melting your exhaust shrouds. The only real power difference from one to the next is single tip vs dual. The single tip catbacks tend to slightly edge out the dual tip catbacks on the dyno, usually because the single tip catbacks streamline the exhaust flow a bit better, creating a bit less pressure buildup. Neither will produce significant power gains.

ECU Tuning

ECU tuning is often recommended, as it can improve a wide range of dynamics about our car, including throttle response, overall power, ECU triggered safety items such as OMP injection rate and fan trigger temp, etc... Power gains are limited to about 10hp maximum, though your gains are largely dependent on how bad the OEM tune is for your particular combination of engine and sensors.

Other debatable items such as increasing the rev limit are possible, though should be thoroughly understood before pursuing.

As for HOW to modify the ECU, there isn't much for options. There are some aftermarket full ECU replacements out there, but they are expensive and quite a bit of work. ECU tuning has some options, with the best option being through the COBB AccessPORT or MazdaEdit. A few additional companies put the effort into hacking the Series 1 ECU and were successful, though only COBB actively marketed a product for it full scale. They did not hack the Series 2 ECU however, and Mazda changed it enough that a full hack was needed. Too few cars were sold, too small of a possible customer base, for COBB to see the pay off in doing the work. The AccessPORT was also discontinued for all naturally aspirated cars in 2010. Mazdamaniac was able to negociate a deal with them where he would become the sole dealer for the RX-8's AccessPORT, COBB just needed to supply the hardware and he handled everything else. Unfortunately, even this route has come under trouble, as Mazdamaniac was banned from RX8Club.com in March of 2012 due to vendor related complaints. His business still exists at Mazdamaniac.com. The tuning through the AccessPORT does not require his tuning services however, as there are other tuners that are successful at producing viable tunes for our car, and the AccessPORT is just a tool. The only exclusivity he had was his remote tuning service (no one else decided to offer it), and after the discontinue, the dealer sales of the AccessPORT itself. Be very wary of tuners that claim to be able to tune "because they can tune piston engines" or "because they can tune RX-7s". There are differences between older MAP based and MAF based tuning, as well as dramatic differences between piston and rotary engine tunes. Make sure that they know MAF based tuning AND rotary tuning, and then you should be good to go.

The Racing Beat reflash is also a viable option, and follows a similar concept as the AcccessPORT, only you send your ECU to them, no hardware to purchase, but you don't get the benefit of the extra hardware either.

MazdaEdit is the newer up and coming challenger to the Cobb AccessPORT. It functions on the same basic principles of flash based tuning, however it requires a patch cable and a laptop. The personal version is a bit cheaper than a Cobb, and the Pro version is far more powerful than anything Cobb can offer, including the ability to handle flash based tuning on more cars than the RX-8. For a comprehensive list of the differences between the Cobb AccessPORT and MazdaEdit, see this comparison put together by Brettus: MazdaEdit vs Cobb

Piggyback ECUs are generally all a bad idea. The factory ECU will eventually learn what it needs to do to counteract the piggyback's effect, and does so. Adaptronic Select is the only real option marketed as a "piggy back", and while it is in the technical sense (it still needs to plug into the factory ECU), it completely overrides the fueling control, which is something that other piggy backs do not do. It thus functions more like a replacement ECU. It is also priced accordingly, at $1,600+.

Every single "ECU chip" is a complete scam for our car. At best it's the same as taping a couple $100 dollar bills to your ECU, at worst it can fry your $1,800 ECU. Our ECU is too smart to be fooled into doing anything more than a few miles, and the OEM tune is already close enough to perfect that the ham-fisted approach of trying to fool it in a narrow set of circumstances while crippling it in the rest is a recipe for engine destruction.

MazdaEdit or AccessPORT

This is more of a sub-topic of the ECU tune, but it seems that it needs it's own section, as it really isn't clearly spelled out anywhere.

A reflashing device is a tool, much like a wrench. Think about this question: "Can I remove my engine with this wrench?". A wrench is part of what is involved, but it certainly isn't the whole story. You won't be able to make any progress without it, but even if you have that wrench, if all you do is hit yourself in the face with it, you won't get anywhere toward your goal. It has to be used appropriately, and it is simply one of many pieces that you will need to get there.

In more practical terms, a reflashing device is nothing more than a communication device, a phone if you will. It can communicate with the OEM ECU. What it tells the ECU is entirely dependant on what information you are pushing through. If all you are pushing through is requests to clear codes, then that is all it's going to do. In order to increase the rev limit, you need to have the device "flash" a map to the ECU.

But you have to get this map first.

Well, what is a map? A map is basically a digital copy of the ECU's parameters and tables that it uses to manage the car. Much like if you right click on a picture on the web and click "save-as" and save it to your computer. Just like that. You can upload that picture to the web if you want, and everything is exactly like it was.

But, you want a change. Lets say that the picture was of someone with red eyes from a camera flash. You wanted to touch that up and remove the redeye. Well, you will use some other software to make those edits. Then you have to upload the new picture to replace the original. That's all the flashing device is doing when you flash a map. It's uploading the digital copy of the ECU's data files.

What tuners do, is they make edits to the data files to adjust fueling, timing, CEL blocks, rev limit changes, etc... as needed. Then, they use the communication cables to upload the new map files to replace the existing ones in the ECU.

So you can get a reflashing device to help you toward your goal, but you will also either need the software to make the edits AND the knowledge of what you are doing (because it's not a simple process of point and click), or you will also need the services of someone that does have the software and does know what they are doing. If you pick the 2nd option you will also need to pay them a fee of some sort, and probably work within their scheduling.

Warning! If the ECU loses battery power for any reason while a reflash is in progress, you run a high risk of 'bricking' the ECU, which will be big money and/or big trouble to fix. Make sure your battery is fully charged when you go to upload a map to the ECU.


Don't. You might ask "why?", so here is an explanation:

There are 2 "benefit" points to changing to aftermarket pulleys.
The 1st "benefit" is a reduction in the weight of the pulley. This is a similar concept to going with a lightweight flywheel to assist with a more responsive engine. The problem is, the weight you are saving is very very very small, and it's close to the center of rotation, so there is very little gain by reducing it. It takes roughly 10lbs of weight saved at inch from the center of rotation to equal the interia savings of 1lb of weight saved at 10 inches from the center of rotation. Lightweight flywheels are still rather large in diameter (almost larger than the core engine block), and much of the weight stripped is near the outside of the flywheel. Pulleys have only a few inches and at most maybe a pound, and most of the weight to remove is near the center. It just isn't going to give noticeable gains for the price.
The 2nd "benefit" is to underdrive one or more accessories. The problem with this is that under driving your alternator is bad for your electronics and battery, and underdriving your water pump can have significant penalties to your ability to keep the engine cool. And all it takes is 1 overheating, or even undetected localized hot spotting, and your housings can warp and your engine is done. Underdriving your A/C is without penalty, but removing the AC system completely would accomplish the same AC effectiveness as underdriving it, and leave your car lighter and without the AC compressor drag at all, so it's the better choice (and cheaper).

Even the large main pulley should be left alone, as it needs to act like a stabilizing influence on the e-shaft, like the flywheel does to the other end. You can lighten the flywheel without adverse affect on the e-shaft, because it is still the heavier end. A greater difference between the two ends increases instability and flex. Adding weight with a pulley that is also a proper harmonic damper is common for high power piston engines, though a rotary is compact enough with lower lateral stresses (lateral in relation to the e-shaft rotational axis), that this hasn't been seen as necessary.

Engine Rebuild

Proper engine rebuilds can be a source of significant power gains. Not as dramatic as you can achieve with a piston engine rebuild, but there are certainly benefits that can be realized. This can get quite expensive, to the tune of $5,000 or more (a "basic" engine rebuild would be more like $2,500 to $4,000). Most of the advantages over a factory engine or a Mazda reman are simply in the quality of the build. The basic points are:

  • Seal clearance / tolerances: The Mazda remans, and to a lesser degree the factory engines, have quite a bit of improvement available in the seal clearances. Too great of a clearance and it allows combustion blow-by that greatly reduces compression and greatly accelerates seal, housing, and/or side plate wear. Properly clearanced seals is critical to making good power AND having longer engine life.
  • Intake porting: The OEM intake ports are good, but there are areas of improvement, and the rebuilders are learning more on a regular basis
  • Exhaust porting: There isn't as much available to do here as you might expect, due to the need to keep the exhaust ports from breaking into the oil or coolant passages or having to remove studs. It's possible to use RX-7 housings or peripheral port the Renesis housings, which would open up a new world of possibilities for your engines. This is not something that a street driven RX-8 with a factory ECU can really deal with however, so it's development is largely relegated to race teams. Although race regs usually keep many teams from pursuing this. Bridgeporting has only been done successfully a few times, and the long term penalties are still largely unknown.
  • Taller Apex seals: You will find much debate around this. Some people swear by the benefits, others point out the trouble to get them to fit and how the apex seals aren't really our problem. There is no direct power gain to be had from taller seals, but the theory behind the practice is that it will allow you more safety zone when pushing for more power. Do not blindly jump into this if you are getting your engine rebuild. Do your research and homework first (as always)
  • FD side seal springs: This has been suggested as a "solution" for the side seal spring overheating problem seen by track and race cars. The incidental information is that this will improve this point of failure, though it's a bit shy of being "hard proven fact".
  • E-shaft and rotor balancing: Balancing is another point of debate and contention. All rebuilders seem to agree that picking rotors that are approximately the same weight is important, however they differ on the recommendations of how far to go to balance the assembly. There is also known fact that at least SOME factory and reman engines are out of balance enough that the rotor tips can contact the housing as low was 8,500rpm (below our rev limit), however it's also fairly clear that not all factory and reman engines have this problem. There is largely a void of information on the balance requirements for a given RPM to be safe, and usually discussion about balancing so as to allow a higher RPM is met with scorn and ridicule about why you would want to be going higher anyway, rather than actually discussing the balance requirements. A better balanced assembly should be easier on the e-shaft and certainly smoother/freer reving as well as a benefit to engine lifespan. Many rebuilders find the basic balance "sufficient" while others disagree and go much farther.