mlampard

I have an 07 six speed standard, with about 110,000 km on the clock. Have owned it for a year. My son had it on the track last weekend and at the end of 2 days of hard driving he had a sudden and total very brief power loss. He was under full acceleration and thinks he was pushing somewhere around 7,000 rpm. It didn't last long - probably less than a second - and then full power and acceleration resumed. It happened twice in the one session. When he came in, I did a quick coil pack change (more on the why of that in a minute), but it didn't make any difference, as the next session out, it happened twice more.

Some background:
6 months ago I changed plugs, leads and coils. NGK 6700/6701 iridium plugs, NGK 4858 leads and Delphi GN10508 coils (maybe a mistake).
1 month ago I had a P0301 misfire code after a rather lively drive. I was suspicious of the coils, but left them in to see whether it would come back. Nothing.
During the track session after he experienced power loss, I ran a scan and found P0302 cylinder 2 misfire code. Hence switching out the coils for the old ones between sessions.
Next session out, after the power loss, I ran another scan and again came back with P0302.

So, I think the misfire CEL is a red herring for the power loss issue. I think I do have an ignition issue and will, in all likelihood, pick up some BHR or D585 coils, but the power loss has me stumped. There was an ex Mazda Tech at the track who said he thought it was probably O2 sensors, but before I start randomly replacing parts, I'd like to check in here and see what others think. SSV? VDI?....?

sonicsdaman

iTrader: (3)

Was it during a left hand turn with 1/4 tank of fuel?

mlampard

Hmm, I think it was under full acceleration on a straight, but have sent him a txt to verify that. Gas was probably at 1/4 tank first time it happened, but after that first time I threw 10 liters in, so probably closer to 1/3 tank.

You thinking fuel pickup in the tank?

Another thing I should mention is that the car has thrown a couple of p0420 codes since I've owned it. Previous owner put in a high flow cat, so I assumed that was the reason for that, but it may be possible that it's related.

Loki

iTrader: (1)

Yes, it's a known issue. You don't want to throw it around too much with 1/4 or less in the tank, it can momentarily starve. It might have 1/4 tank total, but if it's all on the passenger side of the saddle, the pickup isn't seeing all of it. There's a venturi driven tank transfer mechanism that works fine under normal condition, but if you've been throwing it around, hard to predict where that fuel ends up.

Agreed that the P0302 is unrelated or side effect. Also not O2 sensor, it's not used under WOT. That seems like a random guess.

mlampard

So, a little more info as I'm back and forth with my son on this. First time it happened he said he had between 3/8 and half a tank of gas. Also happened coming out of both left and right hand corners and on a straight. He suggested that he was more under hard acceleration than WOT, so my misinterpretation of what he originally said, sorry. And perhaps why the Mazda Tech would jump to O2 sensor?

Loki

iTrader: (1)

Hmm. Running through possible causes:
- fuel starvation as stated above
- loss of signal from eccentric shaft sensor (if true, the tachometer would drop or jump around)
- loss of spark: might explain the p0302, but there's no great reason to lose all 4 coils at the same time
- catalytic converter clogged: would be like hitting an rpm wall, not complete loss of power. Would show up consistently, not just sometimes.
- fuel pump overheating: typically once they stop, they stop for a while, not just a second or two.

So I don't have a better explanation than fuel starvation. The fuel level might read higher than actual during a race, again due to sloshing.

mlampard

Ok. Great. I think my next steps then are to try and replicate the problem. If I can consistently replicate it, then I can come back with an update. I live on a small island in BC, so emulating race track scenarios are not too easy. I see some "off island" trips in my future 😁

Thanks so much for weighing in on this for me, i super appreciate it!

mlampard

Just to update where this is at.... In chasing what looked like a vacuum leak (4.34 g/s MAF airflow rate and 16.41% Long Term Fuel Trim at idle), I did a smoke test on the vacuum system and found out a few things - no visible vacuum leak in or around the engine bay, but the vacuum system won't hold pressure at all - I push 10 psi into it and it immediately leaks out. The compressor runs almost continuously. After some time doing this, smoke starts coming out of the exhaust.

So, to summarize, smoke/air into the intake manifold (vacuum system) passes through the engine and comes out of the exhaust.

The only logical conclusion I can draw from this is that the engine has very little in the way of compression. Am I missing something?

It's a bit of a journey for me to get to a Mazda dealership to get a compression test done, but I'm assuming that's the logical next step?

I should add that there are no other low compression symptoms that I'm aware of. Never had an issue with hot or cold starts, no rough idle, seems to have lots of power (although I've never driven a different RX8 so I don't really have a baseline)....

Loki

iTrader: (1)

While the engine is running, even with low compression, the engine itself isn't a vacuum leak, so I don't think it's that. Also low compression doesn't produce sudden, intermittent power loss. Smoke tests are notoriously ineffective on these cars unless the leak is obvious (in which case you don't need the test). I'm not sure how it would get to the exhaust, maybe through the air pump or leakdown through the rotor seals after time. They don't seal particularly well when they're not oiled and spinning.

A comp test is never a bad idea, but given that you don't have any symptoms, I wouldn't rush into it.

For the vacuum leak, did you check under the throttle body for the VFAD nipple? Or the vacuum tank under the UIM? Those are common invisible failure points. There's definitely a leak based on the data, could be as silly as the MAF being poorly seated in its housing (which also wouldn't show smoke). Keep tracking it down. Is the intake stock?

Loki

iTrader: (1)

For the power loss issue, the best thing I can think of is to attach a fuel pressure gauge, lay it out on the windshield and go for a rip to see if you can recreate the problem and if fuel pressure drops when it happens. Going to be a bit tricky to hook everything up so it's not a fire hazard, but it would be pretty conclusive if you can pull it off.

mlampard

Intake is stock. I had the MAF sensor out to clean and the o-ring looked pretty good. In the attached photos, is the VFAD nipple the nipple that the line with the one way valve (with green colouring) is attached to? If so, that's good. And the vacuum tank - in the other photo there are 2 what look like vacuum canisters under the UIM on the passenger side (other attached photo). Is that the tank you're referring to? I can't find anything else that resembles a tank under the UIM, but it's not too easy to see under there!

I removed the brake vacuum hose and blew into it. There is absolutely no back pressure at all. I can expel all the air in my lungs into that line almost instantly. When I do that, it sounds to be escaping from under the UIM somewhere, probably towards the back (i.e. towards the passenger compartment) and maybe in the middle right to left. It's hard to exact pinpoint and sticking a wet hand back and around, I can't feel any air escaping.

The other point that may trigger something is that watching the live data I'm seeing LTFT stay high until around 2400/2500 rpm, then it drops to what I would consider normal. Is there some sort of solenoid or vacuum assisted valve that could be getting stuck open, causing a vacuum leak, at low/no revs....?

And I apologize if this has taken the thread off topic a bit, but I'm just not sure at this point whether or not it's all related.

Loki

iTrader: (1)

That's the VFAD nipple yep. Looks good assuming the other end of that line is attached. For the vacuum accumulator tank, follow the other vacuum line that comes off the UIM underside, and you should find the accumulator. It's next to the intake valve solenoids, quite out of sight but in the general area where you hear air movement.

mlampard

OK thanks. Found the accumulator tank. No issues there. I might need to remove the UIM and have a look underside, as that's where it sounds like the air is escaping. Weekend job.....

mlampard

Also, I've attached 3 screen shots of live data showing LTFT, STFT and MAF:
1st is at idle showing high LTFT and low MAF
2nd is at around 2000rpm, still high LTFT
3rd is at around 2200 - LTFT drops to 4.69% and STFT kicks up. This is consistent and replicable.

Not sure if it helps diagnose the vacuum leak though.... The sudden drop down of LTFT and rise of STFT from 2000 to 2200 rpm seems odd to me. I should mention that the engine was at normal operating temperature when I screen shot these (so theoretically in "closed loop" mode).

Loki

iTrader: (1)

This is normal. LTFT is not continuously variable, it has a few bands, I don't remember if it's based on engine load or airflow, but up to 2k is the "idle" band, followed by one of the cruise bands that goes up to some higher airflow or load value. What you're observing is the boundary between the two. A vacuum leak of a given size is bigger proportion of measured airflow when the overall airflow is small (idle). Once the throttle is open, a higher proprtion of air is coming in that way and is metered, while the leak stays the same size and its effect is diminished.

mlampard

Ah, OK, I see. Thanks for clarifying that, it had me a little stumped. So, as I understand it, the change I'm seeing in the different rev range is normal, but the high LTFT coupled with low MAF at idle along with the fact that the UIM won't hold vacuum (or pressure), still points to vacuum leak. How likely is it that it's something on the intake manifold versus the UIM?

mlampard

To circle back.... I took the RX8 in to a local mechanic, who has some experience with rotaries. Apparently it does NOT have a vacuum leak, even though LTFT and MAF would indicate otherwise. He suspects dirty MAF sensor, coupled with faulty O2 sensor(s). I had cleaned the MAF, but I got my wrist slapped by the mechanic for using rubbing alcohol (70% isopropyl) instead of MAF specific cleaner! Also, the fact that it won't hold pressure or vacuum when the engine is off and air intake is plugged, is a normal to-be-expected condition. I guess I've been watching too many youtube videos where vehicles hold pressure..... I have ordered a vacuum and fuel pump tester kit, so will double check that later, to ensure it holds vacuum at idle. O2 sensors have been ordered, but are genuine OEM and may take a couple of weeks. Mechanic suggested using only OEM genuine sensors, as apparently the aftermarket ones for the RX8 can be a bit flakey.

In terms of the sudden power loss. The mechanic has 2 theories. The first being O2 sensors, which is consistent with what the Mazda Tech at the track suggested. The second is interesting - The vehicle has the "Key" light on, on the dash. This has been on for at least a year (the time I've owned it for). I had it into Mazda nearly a year ago and they read a B1131 lift gate/trunk low frequency antenna circuit failure. This is a common failure on the RX8 and there's plenty of info on this forum about that. The mechanic is saying that his diagnostic resources say that this can occasionally cause sudden power loss, due to momentary immobilization of the engine. I'm less convinced, as it's curious that it has only happened under acceleration, during a heavy workout session. Having said that, the key fob battery was on it's way out (it failed 3 days after the track session), so I guess there could have been some momentary battery glitch on the fob that caused the computer to glitch because of the rear antenna issue..... you can see why I think that's a stretch (Venus wasn't aligned with Mars ;-).

I will see if I can re-create the issue after the new O2 sensors go in, and if so I'lll try the fuel pump pressure test that Loki suggested.

Loki

iTrader: (1)

Um. I think the mechanic is way off for the simple reason that the O2 sensors are not used at wide open throttle. You're in open loop. Also O2 sensors don't fail by reading wrong, they fail by fluctuating or not responding to mixture changes, which the ECU diagnoses every drive cycle.

The immobilizer theory might not be crazy, if inlikely. I'm not sure how one would troubleshoot it though, besides fixing the antenna issue and see if it never reoccurs? It does make me think of a similar problem with the clutch safety switch which could cause intermittent problems and may be aggravated by being thrown around on track. The fix is usually to make sure the switch is nicely engaged with the pedal and not slightly depressed with clutch out.

mlampard

My son talked to the mechanic directly about the power loss. Apparently he was more under "hard acceleration" (perhaps 80%) than WOT. The power loss happened again yesterday, 3500 rpm and just cruising. Car glitched and took a couple of seconds to come back.

I'll check the clutch safety switch - I did do a clutch play adjust, just prior to the track day, so perhaps.....

mlampard

My new vacuum gauge shows that the vacuum is quite low at idle - 14" Hg at around 750-800rpm. I tested the gauge on a known healthy piston engine and the gauge seems accurate (22" Hg). Aside from the low vacuum reading though, the RX8 is behaving as expected: no fluctuations in readings at steady idle; blip on the accelerator flips the needle back to zero, up to 25 and then back to 14.

I will get this in for a compression check, as soon as I can find a non-Mazda dealer who has a rotary compression tester, but don't really have any other failing compression symptoms, so I'm left a little stumped......

Loki

iTrader: (1)

14 isn't that alarming if it's a warm idle. Obviously not peak of your engine's life, but I wouldn't worry about it yet. What's your altitude above sea level? The gauge could induce a small loss as well, do you have a known good vacuum source to calibrate with?

You're not likely to get 22 from a renny.

mlampard

I'm at sea level, so no adjustment for that. I had read a "Vacuum Reading" thread on this forum (https://www.rx8club.com/series-i-tec...eading-207845/), but it left me a little unsure still. I believe factory spec is 17-20 at hot idle, sea level, so 14 seems low.

The 22" from a piston engine was just my way of "calibrating" - it's not really a known vacuum source, but I used my Nissan Frontier truck with 160k on the clock to "calibrate". It was kind of what I was expecting to see from that engine, which is still pretty tight. More art than science, I know.....

The other thing I should note, is that the RX8 has 110k on the clock, but the engine was replaced or rebuilt by a previous owner. Not sure on how many kms are on it, but if I recall correctly, I think it's probably in the 20-30k range.

mlampard

A month down the track.... I replaced both O2 sensors (and, as an aside, coils with BHR Globally compliant coils). The mechanic noted that the aftermarket high flow cat had the rear O2 sensor below the 9 o'clock position, which can apparently cause moisture issues due to condensation. So, he cut the pipe and turned it so the O2 sensor is now above 9 o'clock.

I have not been able to replicate the sudden power loss issue since this work was done, so fingers crossed it was a faulty O2 sensor that caused the original issue.

I am, however, now noticing a very strong unburnt fuel smell at idle and am seeing a consistent 19.53% LTFT at around 800rpm (goes down to a consistent 2.34% above 2000 rpm). MAF seems better than it was, now at a consistent 5.8 at idle. The high LTFT concerns me. Should I be doing an ECU reset after installing new O2 sensors? Or is high LTFT normal with a high flow Cat?

Loki

iTrader: (1)

Hmmmmm. Not sure if this was mentioned but aftermarket cat + rx8 is rarely long-term survivable for the cat. Do you have any idea what specific brand it is? But that aside, the LTFT is weird... 5.8 g/sec of metered air is very good for 800 rpm, but if it's metering all the air, it shouldn't need +20% of trim unless it's under delivering on fuel for soem reason. Normally I would say vacuum leak, but 5.8g/sec +20 % should give you 1000+ rpm. Never hurts to reset it and see if it relearns, but that's weird.

mlampard

Not sure on the brand of the Cat. Not surprisingly, I can't find any markings on it. I'll re-check the MAF and idle speed on the weekend, to make sure I didn't misread anything. BTW, out of interest, is the 1000rpm calculation that you did based on 20% of 800rpm, rounded up, or is there some other calculation that you used which takes into consideration air flow, fuel trim, rpm, stoich, etc (just trying to educate myself - the more I know the better chance I have of trouble shooting this, I think....)?