mr. GrandGame

Hello everyone!
Sorry to just have registered on this nice forum and be posting here about my problems straight away) But I'm (and my mechanic too) already kinda desperate.
The story so far. I have a 05 JDM 6-port HP rx8. It's previous engine was blown, most probably due to an oil shortage, though I was rather well informed about how important was to keep oil level high. That was a very sad story, occured right in my birthday, and I could tell much things which would try to excuse me, but this post is not about this at all. It's about that a couple of months ago I've bought a new factory engine (not even rebuild) to resurrect my car. It wasn't bought from the dealer or directly Mazda itself. There was an intemediary seller. But the engine came in the box, with all theese Mazda logo stickers, serial number etc. Along with it, I bought: new coils (seller listed them as last revision C OEM), new Desno plugs, new OEM plug wires, new OEM oil injectors, and a bunch of irrelevant stuff (clutch disk with bearing, polybushings etc.). What was left old - MAF sensor, throttle body, intake manifolds, fuel injectors - were thoroughly cleaned by mechanics. When they were installing the engine, they noticed that the rear O2 sensor was broken (bent) and I've bought a new one.
After all the work was done and we started the car, it ran normally for some time, but then THE problem appeared - unstable idle and stall. No CEL with any codes. First we thought that we should just reset the memory. What we did, including famous 20 pedal stomp voodoo magic to make it forget about old engine's e-shaft. That wouldn't help. I've driven a car for almost 500 km (sorry for my metric numbers, it's something about 300 ml) and I beleive that was more than enough for ECU to get used to new engine, but the problem persisted. Then it came time for first oil and plugs replacement, along with wich we tried to install different used MAF. Finally I've bought entirely new MAF and used throttle body (and ofcourse it was cleaned) - no effect. ECU was again reset, I've came driving and had a feeling that the problem got even worse. Oh yes, did I mention that mechanic convinced me that he checked every square cantimeter of intake system to detect leaks? He filled with glue every place that were suspected. As I already said it's like the problem worsened, after we installed new MAF. Before that, the car would never stall not being warmed up enough for the temp indicator start travelling up from leftmost position. And even then, it would first start to hunt for idle for some time and then stall. Now it may stall significantly earlier during warm up and without any idle hunting. Mechanics say that they checked fuel pump, air pump, all theese air valves (SSV, etc), and all works. They activated them through the diagnostics device and saw them switching. I've read somewhere about the following procedure to check proper ECU idling behaviour: rev the engine to something about 5000 and then drop the pedal rapidly, the rpm should drop fast but should noticeably slow passing 1250. This was the perfectly what I saw beforce MAF and throttle replacement, though problem still existed. After MAF and throttle replacement I don't even see this! RPM drops like a rocket-hit aircraft all the way down to stall without even trying to slow down near 1250.
I guess I've told everything or so. I have no idea what to do next, as well as my mechanics. I think it's unlikely fuel injectors failure, though in one moment I was almost getting myselft (and my pocket) ready to buying a new pack. As far as I know injectors would cause problems in all rpm range, which is something that I don't observe. Compeletely new throttle body? Well since it's also not very much probable I'd like to be more certain that it is the real cause, before throwing out some 300$ on it.
Any help, suggestions, tips. I would appreciate that very much. This seems to be a hard problem, and even harder to diagnose remotely, but if *anyone* can tell *anything* useful I would be happy and grateful.

9krpmrx8

iTrader: (46)

Does it only stall when hot? No compression check with a rotary engine compression tester? Where are you located? There are no NEW mazda engines so unless yours was sitting in a warehouse somewhere, it is probably a Mazda reman. But anyway, when you say it ran for "some time" without problems how long was that? Was the purge valve checked? Fuel pressure checked? Fuel pump resistor checked?

mr. GrandGame

Yes, it stalls only when hot.
We did not measure compression because for a couple of various reasons it's troublesome and I thought it IS the new engine, and it should not require one. Ofcourse I could have done it if I would suspect it might be needed on a new engine, but without a good reason I did not want to mess with a gentleman who has a rotary compression gauge and who several times gave me the same tip before: not even look under the hood and into the oil level *before* the oil sign flashes on dashboard (which is ABSOLUTELY normal as he said to me). After my previous engine died I kinda sweared to myself not to deal with this man.
I'm in Russia. In principle ofcourse this could be a sellers fraud. But he also sells another engines which are specified as USA Mazda factory rebuilds. This one was listed as a brand new, unused, from Mazda Hiroshima factory. As far as I know this man has rather decent reputation in Moscow, contrast to my local Vladivostok's turbo-mega-famous-dedicated-rotary-specialist which every experienced RX owner in our city tries to avoid.
When I said ran for "some time" I meant probably until engine's complete warm up. The car was not driven before the problem appeared. Thus you can say it appeared immediately after engine swap.
As my mechanic told me, all secondary air lines along with corresponding valves were cleaned. Air pump checked. Valves are working when activated from diagnostic device. Fuel pump was inspected but I'm not sure about resistor.
IMPORTANT UPDATE: just an hour or two ago, I've bought bluetooth OBD scanner, plugged it into the car and connected it to Android phone. A whole pack of a new information to me to think about! Altough mechanics had scanned the car, they told me they did not find any errors. But it took around 5 minutes for me to extract P0172 "System too rich (Bank 1)"! Furthermore, now I can observe what's happening with the car. When I start it up, Torque App tells me that the ECU is in open loop. After less than a minute it switches to closed loop (using O2 sensor) and problems immediately appear: the car stalls or starts to hunt idle (and then stalls too ). After it stalls I can see it complains about 1 02 sensor failure. Could it be THAT simple? We replaced rear O2 sensor during the swap but we did not touch the front/top. But could it be that important? As far as I know there are many cars completely withouts cats. And even if there is one, I've spoken with people who had been driving around with broken rear O2s for a long time and had not any problems, at least problems with idling. They even told me that I've wasted money when bought new rear O2 sensor. The question is: is front/top O2 sensor so much more important for proper engine working than rear O2? And could it be a reason? And that's strange, there is a reading from "O2 1x2" in Torque App, which fluctuates between maybe 0.1 and 1.0 V when car hunts for idle. Which of the two sensors is this reading belongs to?

Nicon

Look at the fuel pressure.

mr. GrandGame

Would low fuel pressure realy produce "too rich" error code? And why would problems start immediately after ECU switching to closed loop?

stickyRice

I had a problem sort of like this once. For me, I had to 20-pedal-stomp it, but then start it and keep it "idling" (with my foot on the gas) for about 10 minutes. It never made sense to me, but my mixtures would never completely reset until I did this.

mr. GrandGame

Yeah, tried this (as well as all sort of other ninja techniques). It doesn't matter how long do I hold the pedal - rpm just rushes to stall when I release it.
Anyway, just ordered Denso DOX-1449 front o2 from the default russian city with no chance of getting it in our noble town Should come in a week or two. I guess it must be the cause, otherwise I'll just drown in frustration and confusion

mr. GrandGame

Hello everyone, it's me and my bad rx-8 again!
New denso primary o2 sensor finally came to me, and we installed it. What do you think? NO effect! I was pretty sure it was the cause, but it seems that this pain will last forever. I start to think that the ECU hates this new engine. Could this be the cause? The car is of 2005 model year, and the engine is new (as the seller sweared). May they be incompatible? Probably I should replace ECU or reflash it with newer software?

mr. GrandGame

Just caught P0410 by android app. Looks like it doesn't appear all the time, otherwise I would note it before

hoosier1104

iTrader: (2)

I am going to ask, well because no else has, what state is your cat in? If it is clogged, it would cause idle issues.

mr. GrandGame

Well, it appears that mechanics did not tell me 100% of truth before. First of all, they still have not checked fuel pump, because it's so hard to unwind that plastic holding ring without damaging it. Now, as the set of suspectives continues to shrink, we will probably have no other options but to keep trying until plastic eventually cracks. Not really sure what to do then, because while I'm trying to find this ring somewhere in junkyard, it seems to be at least as much valuable as the pump itself, if not more)
Clogged cat? May be. I just thought we've excluded this from the list, as well as the fuel pump.

tapeman

Are you talking about the fuel line connectors? Those are pretty easy to pop off without damaging them, there's some videos on youtube on how to do it.

tapeman

https://www.youtube.com/watch?featur...3nNy8BsE#t=320

mr. GrandGame

No, actually I was talking about this white ring with ribs, which a man on the video manages to unscrew near 11 minute. But thanks for the video anyway - it may still be useful!
My mechanic says it's stuck very hard. He welded a tool somewhat like a one on the video. But even with its help it does not turn with a really threatening amount of force. He afraids plastic will simply crack if one pushes even harder. I guess I better find that ring for replacement first, or if that happens that won't be funny at all.

hoosier1104

iTrader: (2)

What about the catalytic converter? Has that been checked.

04Green

iTrader: (1)

For the fuel pump, look half way through this thread.
https://www.rx8club.com/series-i-tro...change-234314/

The key is slow steady pressure.

Have you, on a cold engine, reset the ECU (20 stomp), started the car, and left it running, without touching the gas, for 15 minutes. Until it is completely warmed up, and idling smoothly? This is a learn process for the ECU. If you stop it early, it will not learn how to idle. Ideally, do this twice from a cold engine. If it has problems, turn on the AC. That will bump the idle speed a bit.

Did you add any premix with the new engine? It can take quite a while for the oil to get through the OMP lines and into the engine.

mr. GrandGame

Just called my mechanic. He confirmed that he already disassembled and inspected catalytic converter before, and it was very fine and clean. So it's hardly the reason, but if nothing else helps, I guess we will disembowel it. Probably we will disembowel it anyway, to avoid future problems.
Just bought used fuel pump but still have no luck in finding that plastic holder ring. Looks like I did not understand the true problem with it. Now after talking with mechanic I think the major trouble is breaking away theese plastic ribs when trying to unscrew it. He also says that he tried to spill wd40 there as well as other chemical stuff with no help. I already think whether I should order one from Japan, but it will come here soooo loooong - about 2 months :[

mr. GrandGame

Are you really sure I should 20 stomp it now? I just mean after reading this:

Real Way to Clear the Memory - RX8Club.com

I thought there is no any sense in resetting NVRAM because it only holds e-shaft plate profile. All manipulations we did after finding that problem (replacement of MAF, throttle and o2 sensor, spilling barrels of sealant on intake manifold) did not affect e-shaft plate. And aforementioned post says that relearning its profile is not that quick. So I did not want to make the problem of learning even harder for ECU by resetting this thing, which should have been learned properly. So to put it short: no I did not 20 stomp it AFTER replacement of o2 sensor, BUT I did it earlier. With and without this procedure we tried to leave car running. Sometimes it holds longer, sometimes shorter. But after some time rpm inevitably starts to wobble and finally it stalls. So the question is: is it really neccessary to 20 stomp it, or was it enough to just disconnect battery to reset KAM?

04Green

iTrader: (1)

MR.

The goal is to reset everything and have it learn everything. From my experience, if it gets confused, or does not learn all the way, life sucks. You are right, about only doing ESS. I should have said do KAM too, disconnect battery for a bit, or room fuse.

The key is to get a blank slate, then follow the learn process (cold start, warm up completely then a few more minutes). I sat in mine once through the process. At the end, the idle searched around a few times then the throttle blipped once and it settled down to a very smooth idle. If you touch the accelerator at all, it will assume you want to drive and stop the process.

If it works, Grand. If not, well at least you know it is not a confused ECU.

Best I got.

mr. GrandGame

Ok, I could not disagree that it's at least worth a try and yesterday I did it.
First, as it's a faster procedure, I just 20-stomped it without touching the battery. Oil indicator jumped showing success. Just out of curiosity I connected a phone to OBD port and started the engine. Fuel trims and test statuses remained the same as before the procedure - something that I actually expected, as it should not have affected KAM (to my knoledge). After warm up (which was followed by usual rpm wobbling) I've driven around a bit. To be honest - I did not notice any differences in car's behavior.
To let the car get completely cold I've continued an experiment in today's morning, leaving it for the past night to rest. I disconnected the battery for a half-hour and was holding brake pedal too. This resetted odo, climate control settings and what's more important - fuel trims and test statuses. I've started the engine. For the long time it showed 0 fuel trims. After some minutes it switched to closed loop and somewhere then STFT dropped to -25%, while LTFT remained 0%. Eventually the car managed to come around 850 rpms, but worked rather rough and never settled completely. Throttle was around 1.6%. After even more time rpms started to periodicaly drop under 800 and finally the car stalled. I've caught 'system too rich' code as already did before. Started it several times after but it stalled anyway. Then I went driving. Looks like the problem lessened, but I don't think that it's connected to prior 20-stomping. The same thing I observed before - the problem is minimal after resetting KAM and gets worse with more driving, like ECU learning makes it worse.
You may say that I should have reset both things absolutely simultaneously. But I was curious would only 20-stomping affect something. It's effect appeared to be hardly noticeable.
Now I beleive it must be a pump. I never saw this car working smoothly at rpms around 850. Even if it did - it was rough and vibrating. After all, it waited for a new engine about 2 years. Just think of it. How would that old fuel (not to mention it's poor Russian quality) affect the pump if it was in contact with it all the time.

mr. GrandGame

After some time of not-bothering community I'm back and necroposting here because my epic battle for idle continues.
At last I've received brand new plastic holder rings. We've cut away the old one and replaced a fuel pump. First sign of trouble was rather nice and clean filter in old pump. And after we thoroughly cleared what we could (20-stomp - oil arrow jumped, disconnect battery for 30 min - odo trip reset to zero) scenario repeated precisely as before. Car starts, warms up, rpms gradually drop to 850 and further to 800, then start to oscillate, then oscillations amplify, then car stalls. STFT = -25.0% just after starting. I understand that it's just began learing blah-blah-blah. But I still suspect it's not normal.
Probably the next stop is pulling the engine out of car. Will try to find vacuum leaks one more time after that - it should be more reliable. Will check electronic connections and wiring for broken contacts.
Any help and suggestions are still welcome. I've been thinking about ECU program. Does rx-8 have several generations of ECU firmware? And if so, can anyone tell if reflashing it to a newer version would give something?

9krpmrx8

iTrader: (46)

9krpmrx8

iTrader: (46)

tapeman

Whooa, where'd you get those color diagrams? That would have made my life so much easier

9krpmrx8

iTrader: (46)

At the top of the page there is this big white box that says, "Search..." in it with a big red button next to it that you click that says, "Search" on it.