BluSpectre

Ok, so I've addressed my coils. Replaced the clutch. Dealt with the unfortunate mistake of cruising with 1/4 tank of fuel at midnight and overheating the pump. Now, it's time to figure out why my car is down on power and throwing lean codes and misfiring on every cold start! It seems to throw only a front rotor misfire and never a general misfire, so maybe it is actually just the front. I know that the #1 cylinder/front rotor is sort of the default for misfire and timing codes, but I don't get general codes; Only the front. Considering the engine chugs along and never stalls yet misfires for a solid 10 seconds sometimes goes along that at least the rear rotor is pulling all the work on startup. No smoke or coolant burning, just misfires and the smell of rich unburnt fuel sometimes. My coils are good, as I recently installed my BHR coils after replacing a set of faulty Mazda coils with just over an oil change on them. My fuel pressure seems to hold 58 PSI or so, meaning that my new fuel pump is working fine. I have my new OBDLink MX+ live scanner, and using my ECUs sensors I can see that my average AFR under load is indeed a bit low and my fuel trims are very high. My first thought was a vacuum leak, but I never heard anything unusual other than the air pump. I took apart my upper intake manifold, and everything was in good order. I cleaned it up (and damaged the black plastic color from some stupid chemistry) and reassembled, no changes made and no changes noticed. The car still idles rough, misfires on cold start, and is still occasionally down on power. The intermittent part is what gets me. Sometimes the car has perfect AFRs and the fuel trims go down, and I have full power. Then it goes lean, doesn't have as much power and idles rough again. I am suspecting maybe my front primary injector is bad or something. I am thoroughly confused here. Additionally, today when I was parking my RX-8 I noticed there was an excessive amount of visible fumes/heat waves from the exhaust. Strange. It never did that on warmups, and I've never seen that before. My current theory is that a bad injector may be not providing adequate fuel on startup to the front rotor leading to a bad fire, and the rear rotor is working fine. This seems reinforced in that I seem to have fine mid range power and the fuel trims seem to go down when cruising in the mid range. I genuinely don't know for sure, and the fuel injectors on the RX-8 are incredibly hard to work on.

Loki

iTrader: (1)

What's your MAF rate in grams/sec and fuel trims when the car is warmed up and idling? Should be around 5g/sec. Lower and you have a vacuum leak.

Can you provide actual numbers that you feel are a bit high?

What intake do you have?

It's not your injectors.

apex1

Seems like a fuel delivery issue, not an air leak. It would have to be a major leak. I was seeing pretty high ltft +16% and had not misfires or running issues. Misfires, once the ignition has been ruled out usually point to injectors.

BluSpectre

I have been logging data using Torque and OBDLink's proprietary app. Although the logs appear to be hard to read without opening the logs in the respective apps. I will upload a log and exact data tommorow.

BluSpectre

My car is completely stock, aside from the BHR coils and DeatchWerks 65C fuel pump. I have the stock airbox, no VFAD delete or any changes under the hood.
My long term fuel trims seem to run between 16-22%, and they lower as I cruise 4-5k RPM. That is why I suspected a primary injector, as the secondaries open up above 3k and add consistent fuel. I often notice that power is not instant and has a sort of jolt when I press the gas, even the lightest possible amount, at lower RPMs. Additionally, since I had worked on the fuel pump previously, it's possible that some dirt or something managed to get into the fuel lines despite my effort to keep them covered and clogged an injector. My SSV is still dirty but seems to actuate fine, but I'm going to remove it and clean it this week. When I do that, I will also attempt to cross the injectors as a method of elimination. I will swap the Front Primary with the Rear Secondary, and so forth. If I can reach them.

Loki

iTrader: (1)

Leave the injectors alone, that's pretty far down the troubleshooting tree for this. What is the airflow rate and trim at idle?

BluSpectre

It appears to hold 0.60 Lbs/Min on idle when warmed, and LTFT seems to be about 16-18% on idle. It shifts around every couple of minutes, as per how it should be when measuring LTFTs. My AFRs often go lean when I press the throttle, so that's not pog. It was really rough to start yesterday afternoon and today both in the afternoon and after work. In the AM, the ambient temps have been down to around 50 Fairenheight, and the afternoon around 80-90. Unfortunately, I set my logs to Lbs/minute instead of g/s so I will need to convert that. For future reference, I set my logs to record g/s instead of Lbs/min. It still misfires when starting cold, and it putters a lot until it settles the trims.
Edit: I converted 0.60 Lbs/min to get 4.53592 g/s MAF reading. I will take a more accurate log and get some fuel while the car is still warm. I will also scrub the GPS tags on my data logs so I can upload them in peace of not having the entire internet have access to where I live.

Loki

iTrader: (1)

That's a bit low on MAF rate. 16-18% low I would say. Bogging on throttle input is also a symptom. You're getting unmetered air in somewhere. Check the VFAD nipple under the throttle body out of sight.

BluSpectre

Ahh yes. I just went out to log some data using my little toy, the OBDLink MX+, and I was monitoring the MAF readings, manifold pressure, fuel trims and RPMs. The MAF readings on idle with the A/C off fluxuated between 3.82 g/s minimum and 5.12 g/s maximum, with an average reading being around 4.3 g/s. What's odd is that it often ran smoother and the O2 sensor AFR was more correct when it was around 4.8 g/s, with the fuel trims leaning off. Anyways, the verdict is back to being a vacuum leak again. IDK how, since I removed the entire upper intake manifold and secondary runners to inspect my solenoids, lines and tubing for cracks. The lines had vacuum when I removed the first one. I guess I'm going to have to try some creative search methods.
The VFAD is still installed on my RX-8, and that line holds vacuum perfectly fine.

BluSpectre

[QUOTE=BluSpectre;4955744I guess I'm going to have to try some creative search methods.[/QUOTE]
Since I couldn't seem to locate any cracks or leaks before, and I can't hear anything odd with the engine running, I am going to have to try and create artificial vacuum and measure the vacuum pressure. My current idea is probably the most redneck solution, but if it works then it works; Use a vacuum cleaner and meter the vacuum pressure it creates, feed that into a larger vacuum line behind the throttle body and listen that way. I could try creating positive pressure, but I'm not sure the check valves would appreciate that.

Loki

iTrader: (1)

You might also check the MAF sensor's o-ring. Common to scrunch it if the sensor had been removed and reinstalled.
Also if the straightening screens in the intake are missing you *could* get stuff like this but it's usually more erratic.

One other place to consider is the vacuum tank under the intake manifold where the solenoids are. It can crack.

I'm not sure what you mean with the vacuum cleaner, you'll need to find a way to block off the filter end of the intake. Going around with a soap water spray might help you find it faster.

BluSpectre

I will re-check those areas using "Soapy Wooder" on Wednesday, thank you for the search suggestions. I do recall that hollow hunk of plastic being loose and dirty when I was trifling around under the UIM last week. I will inspect that thing for a leak. Oh, and IDK how I can have the car create vacuum if I have the UIM removed. That's why I was trying to figure out how I can test vacuum safely without just sticking and oil pipe in my mouth and sucking it or something. I don't have access to a compressed air tank, so I can't use pressure over an opening to make vacuum that way. Maybe just using some sort of hand pump, but reversed for suction, could help. IDK, Maybe I'm too focused on trying to test pipes by creating a vacuum. My visual search and removing each of the vacuum feeds from each device on the vacuum diagrams didn't show any signs of damage or leaks with those. Maybe it is something like that plastic thingy being cracked.

BluSpectre

So I removed the manifold, removed the vacuum and oil fill assembly and it's all in good order. I broke the VDI Solenoid stem trying to remove the hose, so I guess I need to get a new one. For now I'll just swap it with the AIR solenoid until I get a new one shipped. The vacuum tank held vacuum even after I had removed it from the car, as when I removed the solenoid it hissed and sucked in air. So that part works fine with no leaks. I don't know where else a leak could be coming from.

BluSpectre

While everything was out, I figured it was a good idea to check around the EOMP system and double check my accordion tube for cracks. The little rubber accordion tube and intake muffler are well seated and have no cracks. I noticed the yellow hoses that attach the EOMP to the oil fill neck (with the yellow release tabs) are loose, so if that's a possible point then maybe those are my problem. They don't appear to be a part of the vacuum setup, and they don't have any airtight design to them. Additionally these lines have been like that since I got that car and I didn't have problems then, but I'm not going to entirely excuse them. The MAF is seated correctly, no cracks or damaged lines, vacuum tank still holds vacuum, VDI actuator is good but I haven't checked the SSV. This is quite the hassle for finding no results. I guess I'll go grab a new generic solenoid from the parts store for now and throw everything back together. I doubt anything will change. I guess I'm going to have to pay for a smoke test or something.

dannobre

iTrader: (3)

If you have a shop that can clean the injectors locally I would do that. Then replace the seals for the injectors in the intake manifolds and see what happens

The injectors are pretty solid.... but if you have issues in a vehicle this age it's worth the spend to have them cleaned and flow tested

BluSpectre

So an update after some time;
I Visited the local TRex Performance shop and talked to the guys there. I discussed my issues, talked about the condition of the car, and where to buy parts. They said that the accordion tube installed into my car is not the original Series 2 tube, that's it's the Series 1 accordion tube. There is an elbow that doesn't belong, and appears a bit loose but doesn't seem to be the source of the leak. I was instructed to try replacing all of the solenoids and remove the VFAD system. They also noted that my 8 was in surprisingly good condition for it's age and mileage, and seemed quite impressed with the body. This is the main reason for me buying this exact 8, even with the higher Mileage and I intend to keep it in good condition. I have purchased the ZBN Solenoids they told me to buy, and I plan on buying a new AEM/MazdaSpeed intake to replace the entire intake before the throttle body. If this doesn't yield any results, I will have to perform a pressurized smoke test followed by an injector test. I got the solenoids in the mail today, so I'll just install those and recheck everything. I will also remove the SSV and check for leaks, as I've noticed a new whistling noise right before the SSV opens up. Maybe the SSV gasket is leaking into the front rotor.

Loki

iTrader: (1)

Why not do the pressure test right away? An AEM is around $300, an expensive way to find out that wasn't the problem. Removing the VFAD sounds a bit arbitrary, it's open most the time you're running anyway, it only closes in low regimes. Not that you should or shouldn't remove it, it's just not really a solution to the problem?

BluSpectre

Well, I stopped by Harbor Freight to pick up a bleeder valve for when I do the brakes and instead found a vacuum tester hand pump and a whole kit for testing lines. So I bought that. I took apart the manifold again to replace the solenoids I just got, and my temporary hot glue fix did not hold. Unsurprisingly. I noticed that there is a line from the accordion tube down into the lower section of the oil fill neck, so that means the oil fill is metered vacuum. So that loose line on my oil fill neck may be a leak point. Is the oil fill neck meant to be metered vacuum? Can I block that section off or re-route it?

Loki

iTrader: (1)

That would be your PCV line. Any air mass that is coming from there was metered at some point so it shouldn't matter. Assuming the oil neck is sealed.

BluSpectre

I tried to use the little hand vacuum pump to test for vacuum on and behind this line, but the entire volume behind that point is either too large for a measly hand pump or it just doesn't seal at all.

BluSpectre

Well if there is a vacuum leak, mass air coming from anything behind the MAF is exactly what I'm trying to locate!
I also inadvertently ran the car without the MAF connected when removing my bumper after that little fiasco, and the car idled smoother but reported leaner conditions. This exact scenario here, combined with previous evidence of over fueling, keeps making me wonder if my wideband O2 sensor is reporting correctly to the ECU.
Edit: With the MAF disconnected, my OBD 2 scanner was reporting the ECU reading 13.4-13.6 AFR on idle, compared to 14.2 with the MAF connected and calibrated fuel trims. The idle was noticeably smoother, but still not as smooth as it should be.
At least since the car needs to be reviewed by a shop for insurance due to the minor accident I had recently, I will also get a proper pressure test done then. Replacing the solenoids did improve slightly, but since I had cracked one as mentioned prior I'm not certain this was the original leak I was after.

Loki

iTrader: (1)

You an cap the PCV line temporarily to see if it's that. Would be surprised but maybe. The MAF is clean right?

I would be very surprised if the O2 was reading wrong, they don't fail in that fashion, they become unresponsive or unstable but not steadily wrong.

What if you unplug the baro sensor?

wcs

iTrader: (2)

I believe when the Maf is disconnected the ECU is forced in to open loop.
At idle with the Maf plugged in (closed loop) you should see around 14.7 afr but this will vary slightly and move around especially if you have the air conditioning on.

I agree with Loki on the Front O2 sensor. When they fail they typically flat line. At least that's my experience. Having mine positioned immediately after the turbo charger shortens the life span greatly.

BluSpectre

The MAF was clean and new, only a few months old. I will try to cap the PCV and idle and see if that does anything.
I will try that, I forgot that thing can do this sort of stuff.

BluSpectre

Sorry that it has been some time, but the issues have all been solved. Everything, or at least found and parts on order. The false lean condition was my front O2 sensor reading incorrectly due to the coolant leak drying up on the sensor and foiling it's readings, fuel trims would draw up, draw back, reset zero, and wildly change all over. There was no vacuum leak, no solenoid failure, just a dirty O2 sensor.

The engine was replaced with a factory new Mazda crate engine by Rotary Performance of Garland to resolve the coolant leak. New radiator, solenoids, proper VFAD cap (I had a peice of plastics stuffed airtight into the hose prior), new manifold seals, new tune, oil cooler thermostats, etc. The whole kit and kaboodle.
The squeak in the rear was identified to be my wheel was slightly out of round. New bearings already arrived, new wheel hubs are on order. This will resolve that issue. New O2 will be ordered if cleaning the sensor fails to resolve the issue.