Need some help diagnosing this. Fuel Cut?
 

Hi guys, this has recently started happening the past 3 days and has made my BRZ undrivable without risk of getting stranded on the side of the road.

The issue started as a slight hesitation, then became a feeling similar to hitting a fuel cutoff/rev limiter. It happens under all driving conditions (light & partial throttle, wot.) It very rarely happens at idle, or is masked well enough I cannot tell. Car has been running fine for the majority of the year otherwise. It seems to happen in closed loop (warmed up) way more frequently. In fact the first 10min of driving are problem free. After a good half hour daily commute the car violently bucks and jerks. After this has happened enough eventually the car will stall if not moving. After it gets to this point it becomes very difficult to restart until it has cooled or sat for a bit. Then the issue repeats. The exhaust smells pretty rich despite fuel trims not being too far off.

Here is a video of the issue. In the video the car is warm and being held around 3100rpms with a steady pedal. As you can see it cuts out, dips, stumbles and catches itself just as I let off the throttle. (trying to prevent it from stalling.)
https://www.facebook.com/847745593/v...7785091455594/



After the vehicle settles back down and idles in the video, I tried my best to quickly get a snap shot of what everything was reading.

I've been trying to log the even with ecutek but, the stalling and bucking keeps messing up the connection kicking me out of map access. :(

DANG, Kiske, you have been around long enough to know we need more information ........ :slap:

Miles on your car .. ??

Under warranty .. ??

Power mods .. ??

Do you track it hard .. ??

Aftermarket tune .. ??

Is it giving you a code(s) ..... if so, what .. ??


;)


humfrz

If it eventually stalls while idling I bet if you log engine parameters while that happens, we'll be able to see something.

Also, it's surprising that it hasn't thrown any codes.
@humfrz, he has it all in his build thread. FBM Turbo kit.

In the old days I would say clogged fuel filter in a heart beat. Hell I am still going to say that! I have zero clue about what all the numbers mean but your description is a clogged filter to a tee. Clogs up sucking fuel and then clears it self once suction is stopped. Rinse and repeat.
Ok at idle since the demand is low. Check
Can be unpredictable but usually happens after running a certain period of time. Check
Cough, sputters and stalls. Check
Trims look ok (will take your word on that one) Check
Doesn't trigger a CEL. Check
Smell rich (just before running out of fuel there is not the proper mix and ironically the car will run rich before stalling because not enough gas) Check

Hmmmm.... Except it stalls at idle.

"Very rarely"

I like it.

Miles on your car .. 30,xxx boosted since 2,000mi

Under warranty .. lol though I do sometimes fantisize about demoding and passing the car off as 'lightly driven' :P

Power mods .. yup, nothing changed recently though. Just reliability mods, maintance and suspension work this year

Do you track it hard .. not in a year or so. cars been needing everything under the sun before it will see track again.

Aftermarket tune .. yes, hri/jamesm. Has been running fantastic since turbo installation

Is it giving you a code(s) ..... if so, what .. zero dtc, nothing pending either. This is bothering me most because I'd think surely if the car is fucking up bad enough to stall and be undrivable I'd get something at least to narrorw down my search. I've tried to drive it safely as much as I can since the issue showed up trying to get a cel to pop on and nothing.

Like I said, I can't really log the event because when it the bucking happens and I stall I it boots me out of map access and kills the log file. Instead of saving it until the car dies.
I'll keep trying though.

I've already peeked at the filter in the tank. Best think about this car is how easy it is to pull the seat cushion to get to the pump. Sock still looks brand new.

Correct, the only time it will stall at idle is after it has been bucking for a while and I've been driving it a good half hour. When I first cold start it, it is perfect until changing over to closed loop after a good 5-10 min of operation. Because of this I started checking the maf, o2 and intake/exhaust for leaks. Even swapped the sensors with new ones.

Well that should rule the filter out then.

Maybe the crickets all died .......thus, the HPFP is not working too good ...... ??


humfrz

Finally managed some logs. Ecutek cord is crapped out on me.

Large file is me slowly taking the car to 3k then back to idle a few times. In this file everything seems happy except fuel trims at idle. The car isn't bucking or wanting to die in this log.

The small file is the car at it's worst. Me just trying to keep the car alive before it stalls. I'm holding the throttle open at around 30% the best I can and the cars bouncing around mad.

http://www.filehosting.org/file/deta.../kiskelogs.zip

Have you checked the fuel pressure at the fuel rail ......??


humfrz

That's logged.

I let the car sit a couple hours and restarted it, letting it idle 3-4min before this log. At the beginning of the log you can see the car idling, I then slowly rev the car up to 2500rpms and it does a cutout before returning to 2500 and then I let off the throttle ending the log.

http://s000.tinyupload.com/index.php...24056653087890

After looking at the STFT, the fuel trims are going nuts- from near normal to very high to very low. It happens quickly and sporadically. In file 2016 12-10-37.csv the LTFT is 0, I assume you just disconnected the battery- that may be why the LTFT aren't whacko yet. In the other file the LTFT starts to go negative (it takes a few miles for the LTFT to settle in). But your STFT are all over the place- if you reset the ECU by disconnecting the battery, watch the STFT not the LTFT until you have 100 miles or so on the car. Judging from those fuel trims, it won't go 100 miles though.

http://www.easterncatalytic.com/educ...agnostic-tool/ some info on fuel trims. I would not try to drive this car, if it is running as lean as the fuel trims make it look, you are very close to granading that engine.

Check thoroughly for leaks in the PCV or vacuum lines. Also inspect your exhaust around the header. Make sure the basics are OK, check all the wiring for the turbo set up (like the MAP sensor and whatnot). I could see a severely cracked vacuum line causing this as the engine flexes and the crack in the line opens

FYI- you can open those log files with Google Docs (google drive)

https://drive.google.com/file/d/0B6p...ew?usp=sharing

https://drive.google.com/file/d/0B6p...ew?usp=sharing

these are publicly shared- his data logs in a spreadsheet

Oh, thanks, I reckon I should either learn to read the logs ...... or keep ma hands off the keyboard ....... :bonk:

This story is for @Cole ; brings to mind the time that a divorced neighbor lady, back in Connecticut (a very pretty/sexy lady), had an MGB that wouldn't rev over 2,000 rpms. Since I had an MG, she asked me if I would take a look at it.

Well, of course I took a look at it .......one never knows what kind of "compensation" might be in store for me, if I fixed it.

I noticed that very little exhaust was coming out the tail pipe (you see, young-ens, back then, we could tell a lot about how an engine was running by looking at and listening to, the exhaust coming out of the tail pipe).

When I disconnected the exhaust pipe from the exhaust manifold, the engine ran fine. The insides of the catalytic converter had all come unglued and was clogging up the exhaust system.

Damn, she was a good looken woman ..... :D


humfrz

Do you have a separate AFR gauge? I'm looking at the AFR and it is scary high- actual AFR at 4030 RPM is at over 19, which is dangerously high. But if that primary o2 sensor is bad, that reading is false- that said, it's rare for an O2 sensor to go bad.

The question is, what's causing that AFR to be so crazy? That's certainly causing the fuel trims to be like they are.

Could only download that last log.

Something is up with the Direct injection in that log. Fortunately, the port duty cycle is 0 throughout, so I can look straight at the problem.

The thing that jumps right out at me is that Final injection time Literally drops to zero whe you hit your mysterious "Fuel Cut"

You said you were hitting a fuel cut, and that's exactly what's happening.

Fuel rail pressure takes an unexpected dip immediately before it happens, and is otherwise normal throughout the log.

Redo that log with fuel pressure target value logged.

Correct. I did do a hard battery reset. I did a smoke test, and boost leak test of the intake and exhaust. I found a tiny pinhole around the MAF letting air get by and fixed that (new O-ring.) I also found a leak at the exhaust flange just in front of the B1S1 oxygen sensor. I was sure that, that was the cause of my fuel trim swinging. Got it all sealed up with a new gasket last night and retested. No improvement. I then swapped out the B1S1 oxygen (a/f sensor) with new oem replacement to eliminate a fouled sensor being the cause. Same with the MAF and MAP sensors. With no changes there I did a hard reset overnight thinking the car just might need to relearn fuel trims.

If the car was constantly leaning one way or the other I might have a clue which direction I need to take but, it swings depending on when and how much if has been ran. Tomorrow morning I will start it and the first 10-15min of driving will be great minus some whacked stft at idle. Because of this it's really frustrating me. If it was a constant failure I'd be able to pinpoint it much faster.


I'm using a P3 gauge which is using the ecu to read afr, same as the log files.


Idling, then reved up and held at 2500ish, no fuel cut. Let it idle again and repeated. The second time I rev'ed it up I hit the fuel cut. When reducing the rpm afterwards it stalled. I logged the fpr target value.

http://s000.tinyupload.com/index.php...03595844939359

So that means that you are using the car's O2 sensor for AFR readings?

Correct, it does has a seperate analog input for a proper wideband but, I haven't done it yet because I would need to remove and tap my exhaust for another bung. Eventually if the car gets back to running and after the holidays when money allows the exhaust is getting redone with a reccirculating setup. A proper wideband/failsafe will be installed.

Thanks.

I attached the image where it happened, and highlighted the suspect line.

In both instances, you hit a complete DI fuel cut. The rev limit and the stall.

The log resolution isn't spectacular, but in those few frames you see a spike in fuel pump duty, a spike in commanded rail pressure, and a fuel cut. It looks like momentary fuel starvation, but I can only speculate as to why. In both instances, there is a spike in FPDC exactly 0.3 seconds before injection cut.

For some reason, the ECU sees fit to cut fuel completely, and there is a definite reason why.

Time to start troubleshooting, from tank to injector. The problem is in the fuel system, but I don't know if it's a physical or electrical problem from just the logs yet.

My initial suspicion is your Walbro in-tank pump is failing.

All the other parameters look fine.

God I hate direct injection.

Just kidding... Look at that Manifold pressure... Something very wrong there.

Since it's showing fuel rail pressure through out the cut, wouldn't that indicate the fuel pump is fine and the issue might be with the high pressure fuel pump? Just speculating. I can see the FPDC spike to 71% but, I don't really see the pressure spike until after the direct injection is getting cutoff.

HP fuel rail pressure is regulated. That last log was good. Take another like that but like @Spartarus said, log target rail pressure. Also, try climbing up on the throttle more gradually.

If rail pressure follows target pressure we can eliminate the regulator and your supply pressure as the cause.

We would then have to look elsewhere.

As a side note, Techstream has a diagnostic mode to vary rail pressure manually. Pretty slick.

edit: NVM. I'm waaay behind you guys. Thanks, Spartarus.

Problem is, throughout the log, rail pressure and commanded pressure are always close and following.

My suspicion with the CP and FPDC jumping at the same time was that the HPFP suddenly lost supply pressure, pointing to the in-tank pump cutting out.

Forgot to clarify, that's HP rail pressure. The LP rail doesnt have a sensor.

the problem with changing out components is that you are going to run up a big bill fast and it's really only random chance that you'll find the issue. It's not uncommon to replace a good working component with a component that is bad out of the box then you've compounded the issue. If you don't have the equipment to test the components that you are replacing you're really just costing yourself more money IMHO. You'll run out of money before you run out of parts.

Put the old components back on. Bring the car to a garage you trust and you'll save money.

Here. I uploaded his log to datazap.

http://datazap.me/u/ultramaroon/log-...g=0&data=15-40

Or we are back to my first suspicion and the filter is plugged even though a visual inspection says it looks fine.

I don't think the filter itself was inspected. @Kiske, you just mentioned a sock, right?

That makes sense. I guess next step is to print out the repair manual and start with testing the fuel pump. Will also have to check the fuel pump o-rings. I never had an issue with then nor had to use multiple o-rings in my install but, the issue does sound similar to what others experienced. Wounld be odd for it to not have manifested sooner though. Thanks

This car doesn't have an inline fuel filter, just a sock on the fuel pump inside the tank.

The diagrams don't help much but they call it a "fuel filter assembly." You don't still have your stock pump by chance.

You can keep your graphs.

Mine are all up here.

I don't even see the code anymore.

Yes, I still have it. Aswell as a spare engine and junk trans. :O I'd have to re-tune to use it though. The walbro pump reuses the stock housing, wiring, lines and filter.

Here is a picture of the sock.
http://www.ft86club.com/forums/attac...1&d=1397329474

Haha... Ok Neo. It helps the rest of us mere mortals.

Ah, I see. Ok, definitely not the filter.

As long as there is some positive pressure on the PI rail, the HPFP will supply and regulate the DI rail just fine.

30 seconds into your last log it appears to switch over to DI. The port injectors never fire after that. At 50 seconds you start giving it gas but the target and actual rail pressure never diverge.

I'm skeptical that it's a fuel supply issue. Still studying.

edit: The second time you give it gas, while you're holding it steady, the MAP shoots up and down like crazy. Is that real? It looks sketchy.

The ECU is cutting fuel. The injection time drops to zero.

All the speculation is about why... Something is triggering that fuel cut.

That much we know for a fact.

I was speculating that there is a momentary supply pressure drop to the DI fuel pump and the ECU cuts injection in some sort of protection scheme.

I seem to remember some issue with the pumps wiping cam lobes when the car was new... In this case the pump seems to be working fine.

Well unless I'm wrong, only the MAF and MAP can can tell the ecu to cut fuel if they are seeing too much air. The map sensor does seem to shoot up to 36.97 just before the direct injection is cut at 112.593 into the log

Oh my god, how did I miss that. Boost fuel cut...

You have a faulty MAP sensor, because you obviously didn't build 2.55 bar of boost in 0.1 seconds at part throttle and no load.