I ran my story through ChatGPT for easier, more concise reading (for full transparency). This has been a long and frustrating problem since getting my 2004 RX8 (6-port), so I figured I share what happened just in case someone goes through something similar in the future. I also know this is an existing/known problem.For a while, my RX-8 had this puzzling behavior: it ran fine in open-loop mode, but as soon as it reached operating temperature and switched to closed loop, the engine would die. The idle would surge, dip, or die as it transitioned, and I started getting a few OBD-II codes like P0171 (system too lean) and P2096/P2097 (post-cat O2 issues). I would get long term fuel trim readings of 22.656--mostly during idle or light cruise situations.

Over the next couple of weeks, I methodically went through nearly every common culprit. I replaced the MAF sensor, both O2 sensors, spark plugs, coils, plug wires, the air filter, the purge valve solenoid, intake solenoids, and resealed all the O-rings on the upper intake plenum. I even replaced all the fuel injectors. Throughout all this, the battery had been disconnected and and off for about two weeks, so the ECU’s fuel trims had been reset, which at least simplified some variables.

Despite all these replacements and careful checks, the car still died in closed loop. I monitored the O2 sensors live and saw the expected oscillations from the wideband front sensor, suggesting the sensors were working correctly. I thought maybe the fuel pump or wiring might be the problem, and after further inspection, I discovered that some of the injector wiring had been swapped. Correcting the wiring was the turning point — the engine ran and drove normally without dying.

Once the wiring issue was fixed, the next problem I noticed was a slight engine bog when spraying brake cleaner around the injectors. This indicated minor vacuum leaks, even though the injectors and O-rings were brand new. I removed the intake and injectors, cleaned all seating surfaces, lubricated the O-rings with a thin coat of engine oil, and carefully reseated them. After an idle relearn and a few city and highway drives, the P0171 code disappeared, and the idle and fuel trims stabilized.

The last piece of the puzzle was the post-cat O2 code, P2097, which is almost inevitable on a catless RX-8. I’m using a J-style Whoosh Motorsports defouler with swappable inserts on the rear O2. My approach was to start with the least restrictive insert, clear all codes, and drive while monitoring the live sensor data. From there, I could experiment with slightly more restrictive inserts or adjust the sensor angle to smooth out the exhaust pulses. The goal isn’t to perfectly replicate a catalytic converter, just to give the ECU a believable signal so it doesn’t throw the code constantly.

Key Takeaways

• Even new injectors and O-rings can leak if not seated properly; lubrication and careful installation are essential.
  
• Wiring mistakes, especially with injectors, can cause closed-loop stalling that no amount of parts replacement will fix.
  
• Open-loop operation doesn’t always reveal problems — many issues only appear once the ECU switches to closed loop.
  
• On catless RX-8s, post-cat O2 codes are mostly harmless; defouler tuning and sensor positioning can help minimize them.
  

 

I'm not sure defoulers work on cars this side of the year 2000. The ECU plays with the mixture to test sensor response, it's not just monitoring that the signal is in some expected range. You can try obviously but I don't see this being a permanent solution. More commonly people just mask cat codes in software via a VersaTune or similar.

Swapping injector wires is a fairly common mistake when replacing an engine, they're beige and yellow, and covered in grime, surely. Easy mistake to make.

 

I too, have heard that defoulers are a mixed bag on these cars. I have the same style defouler on my '17 Focus ST and it works great, but I do understand it's a different animal altogether. When I reconnected the injector wires I simply reconnected them the way the previous owner connected them--my mistake. The engine is a replacement engine with (allegedly) ~60k miles.

I've been looking into either Mazdaedit or Versatuner as the next likely step. I have heard though that Versatuner doesn't actually help you pass emissions. Either that, or that's just what they say publicly on their forums.

 

Depends on the emissions criteria. If it's just a visual check - sure. If they hook up a computer they'll find pending codes or system not in a ready state, both if which are fails.

 

It's an OBD only emissions test. I passed with my 2010 Challenger that had a big cam, headers, and no cats. lol If it's a post-2000 model year vehicle, it's allowed 1 readiness monitor to not be ready. I like the looks of Versatuner, but it sounds like it basically just turns the light off. However, if it does allow you to make it so a readiness monitor doesn't complete, I could just disable the downstream O2 sensor or catalytic monitor/codes and be able to pass with the remaining readiness monitors. I've also heard that Maricopa county doesn't check for pending codes, but I'm not positive on that.

 

Start with a check/inspect/replace of the following:

-Jet air line
-vacuum lines to each OMP nozzle and vacuum distribution block
-vacuum distributor for OMP system that acts as a spacer under the P1 fuel rail

 

I believe this is a post-cat O2 specific code. I previously had it threaded in directly to the mid pipe and got a P2097 code (too rich). I'm toeing the line between lean and rich right now lol. I did just no adjust the defouler to be about 1/4" further out of the exhaust flow, and turned to face the front of the car more. Next thing I might try doing is take the K&N air filter out and put a paper one back in. Not sure if it'll make a big difference, but it's worth a shot either way.

 

You need to remove that defouler. Then you need to find and resolve the problem. This is not a post-cat specific code. Like I said in your other thread, if you are throwing that code then you have an actual problem with the engine you need to resolve.

K&N filters cover your mass airflow sensor in oil leaning to bad readings. Remove it and reinstall a traditional air filter.

 

Like I said; I did remove the defouler. Then I got P2097 for post-cat O2 too rich—and that’s all I had. I resolved the P0171 already.

 

What I am saying is that even though your code at face value might seem to be due to the rear O2 sensor, its actually because of a out-of-range value between your front and rear sensors. The car knows what it should expect for different conditions. I would revisit the idea that there could possibly be a not completely failed O2 sensor front or rear even if they are new or recently replaced. Due to supply chain issues and thefts by replacing with counterfeit product during shipping to usually trustworthy local autoparts suppliers there are always concerns that any part could not be operating as intended but still well enough that its kind of under the radar giving you passive problems.

To help you get your car running you need to list and outline all mods and any services already done to your car, last time things were serviced etc. to help point you in the right direction, and all in one thread.

There are still helpful people on this forum that can assist if you do the legwork. I'd like to imagine myself and a couple of others who have replied are some of those people who could provide good insight. Don't dismiss anything we say right away, consider what we are saying and that we have been doing this a long time now and could help through the troubleshooting if you listen.

 

Definitely, I agree with what you're saying. I think a little more context (like you said) would be helpful.

Engine has 60-70k miles on it. Car has about 75k. In the last month it's had the following done:
New air filter
New OEM MAF sensor
All 3 intake solenoids replaced
All UIM o-rings replaced
Throttle body o-ring replaced
New EVAP solenoid
New vacuum lines on for just about everything
All fuel injectors replaced
Oil change
New plugs
New plug wires
BHR coils
OEM front O2 sensor
New rear O2 sensor (not OEM)
New header gasket
New header collector gasket

Existing mods:
Aftermarket (probably Chinese) header
HKS full exhaust

Previous owner had injector wires swapped and was causing the car to stall in closed loop. I eventually found that and fixed it after firing the parts cannon at it. J style defouler with all 3 inserts caused a lean code of P2096. Using no defouler at all caused a code of P2097 (as expected). I just put back in the defouler it came with after cleaning it. It's more of a "mini cat" style defouler. After the first drive with it back in after fixing the P0171 code, all monitors are set except the EVAP monitor and the Secondary Air monitor. No pending codes so far. If I can get just one more monitor to set, I can bring it in to get tested.

 

HKS is a nice exhaust, I've sometimes considered making a 3.5" version on my turbo setup.

I've run a chinese header before, and I can confidently say that it caused me headaches and going back to the OEM header was a great decision. I've never seen a chinese header that didnt have a leak of some kind that can be hard to know is there and find as its usually drowned out in the notably louder exhaust sound from the engine bay area. It likely is also getting most of the exhaust flow hitting it directly from only one of the rotors which likely doesnt help.

Speaking of, where are the new fuel injectors from? Pictures of the ones you got and the listing you got them from? I've purchased injectors before to later find out they flow WAY off from what they are supposed to. Heck, I have some myself that looked like and sold as genuine injectors for the car, and flowed WAY off from what they were supposed to and the car trims couldnt compensate.

 

Ooops--never responded to this! I did end up keeping all the existing exhaust components. I did have a small hole in the header that I found later, and I also deleted the air pump and switched back to a proper MLS gasket to eliminate any exhaust leaks I had. It's been running MUCH better. Been beating on it in the mountains, even. lol