WRX idle + runs for 5-10 mins then hesitates worse over time until cuts out

Update 2:

misfire/break down related:

Trunk monkey (Phil the mechanic) climbed inside the engine bay today to pin point the misfire and diagnose the issue.

peeled back my footwell and discovered a stock A/T "mines" ecu. there is only a red LED in the fault LED window. I thought there was supposed to be a green one and a red one?

LH (passenger) O2 sensor was disconnected (solid 0.32v shown in ECUtalk) - reconnected it and it looks like the sensor is bad, stays at around 0.4v under most conditions and then under heavy load will report 2.55v - where as the RH (driver) will flip flop between 0.1v and 1.0v while holding rpm (like I expect a narrow band should) - while driving the pulse width between banks was as much as 10-15% out

Fuel pressure gauge connected, solid 40-45 PSI under all conditions - even when cutting out on right hand turns there was no loss of fuel pressure

consult connected. it reported knock sensor circuit failure (1 time historically) and cas failure (30 times historically). cleared codes. still reporting knock sensor circuit failure.

CAS plug still not clipped however takes a very intentional "pull" to cut the engine - I doubt this is coming undone on right hand turns - Consult sensor readouts confirm this.

misfire pinned to front most cyl on LH (passenger) bank. clean up of injector plug and the misfire is no longer WRX in nature, however it is a rolling stumble and not smooth at all - it hunts for idle rpm - trunk monkey thinks that this is due to the bad o2 sensor

Took the car for a drive, after idling for 20 mins diagnosing the missing cyl, car was fine for 2 or 3 mins until we took a second right hand turn at very slow speed then it begun to cut in /out - observing sensor readouts during this showed CAS/TPS/O2/MAF/inj. pulse/RPM etc. all normal and exactly as as they are when its not cutting out - the exception being I once noticed injector pulse width going from 3ms to up to 10ms while it was cut out - perhaps in response to additional throttle by trunk monkey while car power cut out entirely? Trunk monkey noticed "dash lights" when it cut out (tacho was still reporting engine RPM) - (I have not reviewed any log files to see if there was a single bad sensor reading that thew the ECU into a coma)

massive backfire on recommencement of power

upon returning to the workshop (in this "turn right and you die" condition) car would hardly idle and was difficult to start after stalling, engine would stumble excessively on a jab of throttle from idle, using an injector pule light confirmed that the injectors are pulsing in response to the load. ECUtalk showed ignition timing was reacting to the throttle however the engine was struggling so much it was almost not catching it back to idle after a quick rev.

Once it has sat for an hour I am sure it will be ok untill the next right-hand turn.

at one point we drove down a 60kmh street, and did some left-right-left-right at speed and every time the car turns right the engine cut out...

---

Trunk monkey expects that the ECU is kaput - also wants to test another R33 PTU to be sure that one is ok / wiring is good.

just waiting on an ECU to be sent to me from QLD and plan is to drive it untill it shits the bed again and then throw the replacment ECU in it. see if it drives ok, if not then replace PTU (guy was too busy to send during the week - said he got it there on Friday though)

Other discoveries:

my brake master cyl is spitting brake fluid / letting air in from between the two hard lines - looks like ill be buying one of those.

Trunk monkey detected no wheel bearing problems on test drive, suspected I was feeling 5-cyl running.

 

This might be painfully obvious but check your battery terminal connections are solid. Cutting out under cornering definitely sounds like PTU wiring to me, cut back the loom from the PTU to check for split wires around the front of the engine then swap the PTU out. I'd do this before swapping out the ECU as that doesn't seem very likely given the problem only occurs under cornering. I had these symptoms years back and it was just faulty PTU wiring so swapped out for a series 2 at the same time.

 

Thanks, the battery terminals have been wire-brushed and a new battery fitted.

The PTU was recently wired from a S1 to a S2 with no change to the behavior of the car.

I will strip it back and check it though - and the connectors, and also check the grounds while I am at it.

Just sitting waiting for parts at this stage...

Once it has begun to cut out on corners it will continue to be rubbish until its sat for a little while.

 

thinking out loud here, could it be a header or manifold leak? would explain the wrx burbling sound, and sometimes where a header leak is present it becomes worse as the joint expands with heat.........

certainly seems to match your symptoms anyway

 

Certainly open to the suggestions.

I will take some videos this weekend of it running / not running - I am fitting brake master, PTU, ECU and injector wiring... (postman pending)

There is the slightest 'crack' for a few rotations on suddenly jabbing the throttle from idle. coming from rear driver side cyl. Trunk monkey suspected it was a coil pack shorting out to the plenum with the increased load - it so subtle I almost feel its just valve train noise - I am not so sure

Perhaps the crack is a leaking exhaust manifold? Even so I wouldn't expect it to cause such a dramatic idle problem.

--

A plenum pull was previously done and the plenum gasket was replaced.

I have not yet followed every vac. line to make sure there are no leaks - nor have I measured the intake manifold pressure.

--

I should add that the misfire at idle is enough to shake the bonnet holder quite noticeably - I would expect a smooth idling engine to not vibrate so much - but this is the only Z I have spent more than 20 mins with (aside from wrecking a crashed one YEARS ago)

Remember that we idled it for 20 mins with no problem, went for a drive and 5 mins later it would hardly idle at all.

The more my lizard brain spends obsessed with this car the more it begins to feel that the PTU/PTU wiring is causing the cut out issue - and the injector harness/o2/knock sensor is causing my shitty idle.

I also note that the S2 PTU has been mounted in the original location with some spacers causing it to be directly contacting the radiator hose. Maybe its just getting too hot?

 

Well. my replacement ECU, CAS and BM57 have arrived. pity I am too busy to work on the car this weekend :bash:



I'm not sure if this ECU has previously been used as a murder weapon...

I just hope there is no rot on the inside

 

You mentioned a "massive backfire".
Is the backfire in the induction manifold, or exhaust system?
A backfire normally occurs when unburned fuel is exposed to an ignition source.
Have you done a compression check of this engine yet?
If not why not?
If you have, what are the compression figures?
What do the spark plugs indicate?

 

Exhaust. gunshot loud. the system has no cats. It only happens when ignition returns after coasting in gear while the engine is not producing its own power after the onset of the issue. I would say these backfires have happend a total of 6 times in my presence. It will do this only when it is seconds from not running at all - I am not sure its mechanical as leave the car for as little as 60 seconds and it start again, idle poorly and on light load run reasonably well for 10-90 seconds provided you are gentle with it.

Consult&laptop + fuel pressure gauge + injector pulse light indicates fuel is still being injected during the cut outs - once ignition kicks back in is when the backfire happens (dash lights come on when the ignition cuts i.e engine light, yet consult keeps reporting sensor values during this time). I hope this is because its throwing a spark during the exhaust stroke - could be a sad valve but its prompt return after a short rest makes me feel confident

The way its running I really don't know how it hasn't throw a rod out the side of the block - hence its only been driven to assist diagnose the issue. I have parts on hand now.

compression test? no. simply havn't gotten around to it yet (my mechanic was quite sure it was a fuel delivery issue to begin with and we only had as much time as it took to confirm that this is not the case) A compression test will be done. I will check and replace the plugs at the same time.

I cant get my hands on the car again until the 28th unless I get some time off work (or I stick a rocket up trunk monkey)

 

As from above sound like you have leaking injectors...
+ Potential problem with harness/plugs

 

Just remembered I had a similar problem with very inconsistent running / idling etc. Took me a while to find it and was an accordion pipe had popped off down under the head light.
Couldn't believe that was the cause of what seemed like a variety of problems.

 

Today:

• replaced ECU
• bypassed knock sensor
• fixed kinked air intake snorkel
• blocked off open vaccuum lines from EGR/AIS delete
• replaced one o2 sensor
• modified PTU mount (no longer touching rad hose)
• found that the passenger side VCT was disconnected... reconnected.
• replaced missing throttle body screw

car. still. cuts. out... :bash:

BUT - the cutting out is probably fixed!! :rofl:

After doing all that and it still cutting out and the ECU giving me a code 55 (all clear) but still dying and carrying on - just about to get the rope out to hang myself - started wriggling wires / tapping on things with the engine running. took the relay cover off, tapped on the relays, boom... car cut out. removed the ignition relay and what has happened is the small screw in the bottom of it had come loose and the insides were moving around, so as I would turn a corner the relay would turn off and cut all the power to the ignition system... moved the AC relay to the ignition port...

I cant call it 100% certainly solved as I didn't get a chance to test drive it though... we moved straight on to the brake master and the HFM Dot BM57 is faulty -.- so I will get my stock one rebuilt this week and see if I can get my cash back from HFM (or a replacment unit)

Given it was so shagged that even tapping on the relay would cause the car to stall - I am confident that this is what was causing it to act up while driving.

TLDR; ignition relay failed

still to do:

• bleed clutch
• screw in altenator bolts (found them undone)
• fit a working brake master
• bleed brakes
• replace inj. wiring
• fix leaking p/s lines
• tighten exhaust cam covers and deal with other 4,000 other oil leaks
• clean up speed sensor wiring - pray to Nissan gods that this fixes the bouncing speedo
• source a clutch fan shroud for the radiator
• throw vg30de in the bin, buy an ls1, sticky tape it in, do a skid.

 

With respect to the "bouncing Speedo",
do yourself a favour and re-solder ALL the components on the small Circuit board inside the Speedo unit(this will be easily identified when you investigate inside the instrument)as they are notorious for developing 'Dry Joints' which lead to numerous annoying symptoms. Don't ask me how I know this!!!:bash::br:

 

I was planning to go for a drive down the highway with consult connected to see if its speed signal or the gauge itself. I have seen a guide floating around on the internet confirming the solder joints to resolder.

Thanks

 

Last Update:

Since relay has been replaced - car is no longer cutting out :br:

car still idles rough, have replaced injector connectors.

Given the cutting out and rough idle are not at all related this thread has run its course

thanks for the all the suggestions - I will soon be making more threads specific to other issues as I need help.

:rolleyes2: