Nasty engine stumble under load with highest premium 98 on top of 91

[Steptoe's photos]

The "cack your dacks"  sort of stumble/misfire in the EJ202 Forester MY1999, the other day up a mountain pass -100km/62.50 miles from home.

And it was only just going so well just before. .. actually pulling smoother up the same hills than the times before. It felt well worth the 98 RON (our highest grade, easily available pump fuel in Australia) I had been running factory spec 91 with no issue other than a bit gutless and a little ping/knock before the knock control circuit did it's job. I had to sit mid mountain for some roadworks and noticed the running cooler on the new fuel mix (78-82 deg C) had risen to 88 while sitting - no manual fan switch yet !

 

(The tank got 55% of 98, 45% of 91, so a rating about 94 - 95 is what you get according to the fuel supplier technical division.)

Next uphill bit became a battle. I got some harsh resistance to power, saw the tacho drop, despite speed being pretty constant. It seemed to prefer 2200 rpm instead of 3000, before playing up again. Traffic building up behind my Subaru is pretty unusual and AWD ignition stumble is no fun, so far from home. I had a spare unknown coil on board but was too scared to stop in case the EJ let me down.

I just cried out for my mumma and pointed the accelerator pedal towards home, through a few towns, past a mate's place in one of them (can't have him laugh at my old Subie playing up  It carried on from time to time even on smooth plains and jolted at pot holes, so was very sensitive.

 

The leads were original Packard from 1998 and were running fine before - always the way isn't it? 361,000km

I got home and there it has sat for a week while I think about things, I found some specs for the plug leads. Hmm, fitted a new battery in the DMM for good measure, found the DMM did not read well sometimes in the 20K scale, so upped it to 200k and got my readings. Three leads were duds according to two DMMs.

I pulled another set of originals from a 2000 EJ202 with only 140,000 km, found some bad, or rather, no readings at all on them either. I still could not make up a good set.

To up the octane rating of the fuel the car was made for - my experience mainly with LPG 109 AU-RON , has been to ensure the ignition is in top shape. That includes reduce the gap to make life easier for the coil to keep sparking. I am pretty certain if I was to return to 91 I would lose the carry-on of stumbling and scaring my sneakers of a walk back home or my wallet to pay for a tow truck  I have retained the 1.1mm gap of the plugs, but thinking if I was to tinker and drop gap to 0.60mm the problem may just go away. 

New plug leads coming up and retire old leads to the bin me thinks and hope she runs fine for ever after.

Has anyone had similar experience, running tired old ignition leads on upgraded fuel, above what car is specified for, to find horrid performance, then return to normal on normal lower grade fuel? I think I have found my problem ...and it begins with EJ

[el_freddo]

Geez mate fair shake of the sauce bottle pointing the bone at the EJ being the issue. 

It’s simple - don’t do your maintenance, don’t get a good ride! 

At that age/number of kms the fuel pump probably needs to be replaced too - I’d be doing it as preventative maintenance. 

The EJ engines are good engines - they crap all over the EA82 by a long shot and some of them do well in the longevity department to be up there with the EA81 - with more power and good economy  to boot. They’re not rocket science and before you go there they don’t scan your brain and report back to the government   

Stick with the proven formula - NKG spark plugs and leads, leave the gaps unless you’re running on LPG. 

I bet your fuel RON increase will have little to do with it. As for pinging, the older your engine is the more likely it is to ping from ring wear and increased oil consumption as a result. I’m sure you know about this already. 

Sorry, crap day while feeling like crap, not meaning to be an arsehole about it if it comes across that way. 

Bennie

[heartless]

correlation does not imply causation.

i doubt very much that the fuel change caused the problem (unless the fuel was bad for some reason), it just happened to be coincidental.

24 yr old wires are more likely the culprit. Yes, the OE wires do tend to last a long time, but good grief, even they will fail eventually, Replacements should be either new OE or NGK ONLY. Anything else is a waste of your time, effort and money.

 

[1 Lucky Texan]

sometimes, spraying water from a plant mister at the high voltage parts at night will show arcing - and maybe cause missing.

 

dropped/shifted valve guide?

Edited March 14, 2023 by 1 Lucky Texan

[el_freddo]

10 hours ago, 1 Lucky Texan said:

dropped/shifted valve guide?

Oddly I’ve not heard about this being an issue in Australia - but then again I’m not a mechanic that sees loads of these vehicles week in, week out. 

I doubt it has much to do with a dropped valve guide. 

Cheers 

Bennie

[Steptoe's photos]

The plugs in it look like they were replaced at the same time it's last tune service occurred.

Fuel and air filter had 2012 make dates on them, air filter spotlessly clean and pristine.

I went to a long, long time supplier who only supply the best and got some locally made and owned company HT leads - I know and trust Eagle Eliminator for my turbo LPG applications for at least 15 years - so went with them - in stock and no online BS

On they went late last night, then a quick test drive and STUMBLE and now the CEL decides to light up ...maybe needs good leads to report poo plugs

Wholly DENSO ignition including the plugs, and I like DENSO , but went for my long respect for NGK BKR5E-11 to stuff in - correct grade

The engine in it only has 250,000km on it. The plugs that came out of the original 360,000 rod knocker where heat range 6 which is colder and LPG spec. I will put it down to laziness of the previous installer.

I like the idea of preventative maintenance , as much as if it ain't broke , don't fix it theory. But roadside rescue missions are not desirable. I have a new pump on board, and discussing with a mate who has probably had more Subies than I care to remember, has never had a pump fail on him. They may have failed before or after his possession though - another mystery that we will never know.

I do like to keep an eye on things and will be installing a fuel pressure gauge in the dash somewhere, some time soon.

Next step is to fit plugs, then diagnostic unit by SnapOn in mate's shop.

[1 Lucky Texan]

exhaust blockage? maybe a piece of a cat. conv. is shifting around???

[Steptoe's photos]

11 hours ago, 1 Lucky Texan said:

exhaust blockage? maybe a piece of a cat. conv. is shifting around???

A thought, nasty one at that

 

[Steptoe's photos]

so, what did I write above ??

At no time did I have a CEL, until soon after start up with new leads fitted - tried and trusted for decades - Eagle Eliminators - only then did CEL come on.

New NGK BKR5E-11 went in, the CEL light went out.

It looks like if the HT leads are not within spec, the ECU CEL cannot report an inefficient spark plug.

First test drive, rather short, no problems.

Next test drive, within the mile, produced the same symptoms.

So, I have found bits not within spec, not the cause of my problems at all ! Preventative maintenance done then

Waiting until either I receive my fuel pressure gauge, or can find an oil pump gauge to test fuel delivery pressure T'd in after the fuel filter. Fuel filter is clean as inside.

Already stuffed an old fuel pump quick connect/disconnect trying to replicate a removal tool with pliers. Not a good look.

Have a post asking to see what others use to disconnect these babies

[el_freddo]

On 3/15/2023 at 1:14 PM, Steptoe's photos said:

I have a new pump on board, and discussing with a mate who has probably had more Subies than I care to remember, has never had a pump fail on him

I agree with that. The Subaru in-tank fuel pumps don’t just die - they will underperform causing issues before they outright die, they usually get swapped out before they completely die.

I had one completely stop working on my RS turbo - but I believe this was from sitting around with rubbish fuel in it.  That didn’t bother me as I was only moving it off a trailer, not 100s of kms from home. 

It’ll be ingesting to see what the fuel pressure gauge says once you get it installed. I didn’t know there were units available that has the gauge remote mounted from the sensor! I’ve only ever seen the ones that are plumbed inline and viewed from where it’s fitted. 

Cheers 

Bennie

[brus brother]

10 hours ago, Steptoe's photos said:

A thought, nasty one at that

 

Going uphill, gravity could allow bits of broken cat to fall backward and occlude the cat? Disconnect exhaust and take for a very loud test drive.

[Steptoe's photos]

On 3/17/2023 at 1:37 AM, brus brother said:

Going uphill, gravity could allow bits of broken cat to fall backward and occlude the cat? Disconnect exhaust and take for a very loud test drive.

"occlude" ? My goodness, I have only ever come across this word used in optometry text books ! I know what you mean   My same mate who has never seen a Subie fuel pump fail, thinks the higher octane burns hotter and canes the cat. I can offer that my engine coolant temp was actually 4 to 6 degrees cooler than normal for this run and an internet search suggests that higher does not burn hotter, just harder to ignite. And from experience, the higher octane allows for less downshifting scooting up the mountain pass.

I have had one blown cat on one car I bought with it's problem - it was a noisy rattle. I will look at cat after fuel pressure

[Steptoe's photos]

On 3/16/2023 at 11:30 PM, el_freddo said:

I agree with that. The Subaru in-tank fuel pumps don’t just die - they will underperform causing issues before they outright die, they usually get swapped out before they completely die.

I had one completely stop working on my RS turbo - but I believe this was from sitting around with rubbish fuel in it.  That didn’t bother me as I was only moving it off a trailer, not 100s of kms from home. 

It’ll be ingesting to see what the fuel pressure gauge says once you get it installed. I didn’t know there were units available that has the gauge remote mounted from the sensor! I’ve only ever seen the ones that are plumbed inline and viewed from where it’s fitted. 

Cheers 

Bennie

Experience across a few breeds is that the engine is at first, harder to start, before a pump fails in tank. My RXA came to me with good start but died as soon as some boost was asked of it. I whacked on the trusty old oil pressure gauge that has been adapted over the years to fit whatever - including old Lucas mechanical systems that needed 50 psi to properly crack open the injectors

T'eed into the fuel line after the under bonnet efi filter and ran the fuel pressure up a hose to the gauge taped to the outside of the windscreen. I could sit stationary and get 36 psi up and down the rev range, yet to take on the road and squeeze, the pressure dropped to 10 psi reading as power dropped away. This external pump was the original, 20+ years, 188,000 km, and had been in some suspect fuel supply regions over it's time I suspect. In the days of waiting for new pump to arrive, the old pump started to not even kick in on start up, until I knocked it with the old engineers hammer   We had a similar issue with a mate's Vortex turbo that sat in storage for 20 years - serviced just prior to storage. I used same original style pump up front from a fuel can supply to get it started - after that, the car's pump saw what it had to do, and go no trouble after. Next owner hopped in and drove it 1200 km home to Queensland - no dramas.

I have found a few efi  electric fuel pressure gauges on line over the years, some quote oil/fuel gauge. Other senders I have found quote oil/fuel sender. Happy to share my info finds Bennie .

 

I found my old never used oil pressure gauge bits (finally) amongst an unfinished project - I was making up an EA82 efi test bench for FPR and injectors, with gauge.

With the bits on hand back in the day, I had cobbled up something that I could straight swap to the post filter pressure line of the SF5 Foz. If magic works , I can share. I used a few tapered (I think) threaded barbs, and a manifold plug off - refreshed all the threads with gas tape.

The sender unit outer housing also needed to be grounded for good measure (maybe the carbon in the fuel hose and fuel inside itself may have been enough !?

Too hot in the outside parked thing to take for the first test drive at the moment. I wired up a cig lighter plug to direct power the gauge from just below where I will tape the gauge to dash top. Sender wire is just out the door and concealed amongst holes or gaps in panels into the engine bay. Initial start ups to test for leaks of fittings etc are all good, and pressure reads just under 40 psi - so, pump is delivering what I think is fine for stationary use. Looks like FPR not letting pressure go.

I renewed the brittle FPR vacuum line with all that I could get washer vacuum line in 4mm I think it was. It might be wise to make the length up from brake line and just use rubber hose to connect each end. EVERY Subaru FPR vacuum line that looks original from this era that I have probed - snapped ! It is so brittle

 

Edited March 31, 2023 by Steptoe's photos
remove big kb pic

[Steptoe's photos]

It is getting worse.....

the test drive watching fuel pressure gauge at just under 40 psi and the momentary yet repeating stumble / cut out happens without any change to the fuel pressure.

No need to pull fuel pump or FPR

I cleaned up both battery terminal clamps, radiator earth, #3 intake runner earth with wire brush and electrical contact cleaner spray.

ie ECU disconnected, so anything it learned in past 2000 km has been forgotten.

Start up results in CEL 40 seconds later, rough idle, stumbles, misses, almost undriveable on flat ground into shed

Next is to see if the CEL is gonna flash some codes when I join the diagnostic connectors.

EA82 conversion coming up !

 

[Link0demons]

I've done a lot of stupid things with the EJ platform, most notably and possibly applicable on the 22. Granted the EJ20 I had was amazing and never had any issues in the time I've owned it.

However, I did have a 22 blow a piece from exhaust valve #1 while accelerating up a steep incline, thankfully it blew out of the exhaust pipe and the car still got me home. It stumbled and ran like crap for months I couldn't afford to tear into the engine. I found out later that I blew out this valve chunk during a head gasket repair, which I thought was the cause. I had ran the machine for about a thousand miles with leaking valve guide seals and 0 compression on one cylinder unknowingly. Perhaps check compression?

 

The same engine, a few months later... I was running without a timing belt cover and was driving through snow deeper than the headlights. I caught a chunk of ice and spun the driver side cam 4 teeth out of time. The car still got me home, but was hard to start and also ran better than a garbage can. Perhaps verify timing?

 

I doubt this will help you much, but the symptoms sound like you need to dive deeper than any fuel issue. As far as I understand the injection system should be extremely happy with 35-40PSI. since you've checked the basics like plugs and ignition coils/wires.... I'd verify timing and compression next.

 

Please forgive me, maybe I missed it. What is your CEL reporting?

Edited March 21, 2023 by Link0demons

[Steptoe's photos]

LinkOdemons - demons are what I have I am sure, as there are no bright lights above in the sky , sucking life from beast.

I am yet to see what amazing info the CEL is going to offer. I suspect it will be one of those incorrect reports that diagnostic systems have tended to throw in the past on other vehicles. Comp test is my usual suggestion for others   I had done a few before and got good results.

Initially, the problem felt ..maybe like a full on injector cut, rather than an ignition cut as previous ignition faults usually have all those dash report lights come on while engine has died.

So, trying to find any reason an injector cut may occur.

2200 and 3000 rpm is not near over rev limit.

No turbo to create over boost

The manual suggests injector cut can occur if air intake between throttle body and head happens (allowing over rev I suspect)

This is an immobiliser edition. I note the manual lists that there are immobiliser and non immobiliser models.

Tis full on momentary engine cut out has evolved? to poor engine running, idle etc just with terminal cleaning - so weird

[Steptoe's photos]

Yes, thinking a timing light may now throw some light on things. I might start swapping components from a donor wreck to this thing , one at a time.

But first, I made this to allow self to be in upright position to observe what happens in the cockpit as the green and blacks are connected etc

I found what the manual calls diagnostic earth terminals in at the test plugs. On mine, they were exposed as male terminals. On the seemingly untouched donor, they were still wrapped in factory loom wrap vinyl tape.

[el_freddo]

@Link0demons- I like that suggestion about the cam timing being out. I heard of another situation like this but on an EJ22 from memory. 

Weird behaviour - full throttle it was fine, idle was poor/low, everything in between with load was jolty. 

The issue was the crank had skipped several teeth from alignment with the cam belts which were synchronised with each other. 

It’ll be worth the effort of removing the timing belt covers to check the alignment - and an easy fix if it’s out. 

It’s a cheap fix too if this is the issue. Typically it won’t show up as an engine code either, but may trigger other ones. 

Cheers 

Bennie

[Steptoe's photos]

Timing belt is only 22,ooo km old and 2,000 into latest inspection due to having to swap the crank sensor "wheel" to suit the crank sensor only, no cam sensor SF set up. I also installed the missing belt guide above crank with max 1.00 mm clearance....This is so weird, the better the electrical get - new leads, plugs, clean terminals, the worse things get ! 

The initial problem I was chasing feels like a random, intermittent , momentary injector cut.The more new components this baby gets, I expect better of it. It was parked out in the open for a few hours beforehand - just wondering if it was tampered with from beneath, so will be inspecting for foul play. It could be a CAS on it's way out .....like old EA ignition modules that first give a few warning hints that poo is about to happen 

Still to be code read - other stuff calling my attention......

[Steptoe's photos]

Progress ! With symptoms of a bad or failing CAS described online, sounding very much like my issue - no change after swapping the CAS for another.

So.... I read the code for the CEL

just got one code from the memory - joined black plugs and ign on - code 2-2 , unhappy knock sensor

just got one code from joined green plugs, with black disconnected, engine running - code 2-2 

the factory manual says ecu uses basic (limp mode?) fuel map and retards timing by 5 degrees

I discovered the two wires leaving knock sensor wire, turns into just one going to ecu - this akes the crush tube of the body of sensor , it's ground.

the second wire is grey, fat and is a shorting shielding wire that ends about 100mm from the sensor.

I also felt that  the injector lom for #s 2 and 4 injectors might be too close to the shielded wire section, so tied injector loom closer to #4's intake runner.

the sensor's light silver zinc coating was a bit brown, rusty looking and the block where it mated also a bit old and crusty.

replaced knock sensor with another used one, all metals cleaned up, joined both black and green connectors to clear the codes - this took a few minutes running before flashing codes stopped - i had the battery disconnected while doing the swap - following manual procedure

codes cleared and not returned.

nice idle returned, no stumbling at idle

BUT once at 65 degrees block temp , i carefully brought revs up in neutral then around 3,000 rpm the full on stumble began - this has been with me a month now.

Not happy Jan !

[Steptoe's photos]

OK. I now smell like petrol. I just swapped in a new fuel filter to be on the safe side when I can drive this demon again.

Autopsy on the old filter was not much to see, just a little black/brown sooty stuff washed out as I cut the can open.

Same issue of stumble, free revving around the 3,000 rpm

Swapped in a much lower km (140,000) MAP sensor by DENSO - same crap behaviour.

Ideas left are : to pull the exhaust to inspect the cat

                       : regap the plugs to 0.7 or 0.8mm to reduce the load on the coil , if the higher octane fuel is harder to ignite

                       : haul in a few jerry cans of low octane fuel to top the tank off, see if that has a change

[brus brother]

disconnect the exhaust before the cat(s) and do a very noisy test drive up hills and down.

[Steptoe's photos]

Cat. singular, as it is an AUDM   Pulled the exhaust (air intake plastics first to undo the O2 sensor plug) Cat is in one piece. Spied more earth, or grounding points while underneath. Took them back to bare metal, wire wheeled and emery clothed the components, sprayed with electrical cleaner, did same for ABS earthing bolt, assembled everything and went to start her up. Cranks but no fire. Tried the other ignition key. No change. I get one cod for the air temp sensor due to starting and reversing into shed sans air box. Previous start was an hour earlier, 15 degrees C air temp, TPS and MAP plugs disconnected, so it ran at a default rich, producing a lot of rich exhaust smoke - suspect even though it started again and gave no hint of being unhappy ignition- wise - it may be ( I am hoping for) wet spark plugs. Otherwise, this thing is possessed. 

[Steptoe's photos]

OK, now I can see it is a whole month since this event.

Today she fired up !

I found sooty spark plugs from the MAP being disconnected, but fitted another set of smaller gap 0.8mm plugs, still no start

I have been testing the relays on the bench, using an air horn compressor spinning freely, then partly blocking it's output to load things up - to the point of blowing 15A fuses on the 22A relays. 

I did the same for each side of the double throw main relay.

Today, I thought I would test a few outputs of the relays in place. I went to start, but nothing, just crank. Then I noticed the fuel pressure gauge was not registering any pressure. I forgot to install the main relay  Fitted the original main relay. Now she starts. Next will be to see if poling and prodding wires near underbonnet ECU causes any issues. Then see if 3,000 rpm stationary does anyhting bad. Next, a test drive up hill - in case I need to roll home.

[Steptoe's photos]

On 3/16/2023 at 3:37 AM, 1 Lucky Texan said:

exhaust blockage? maybe a piece of a cat. conv. is shifting around???

I observed your idea, pulled the air box off to get easy access to undo O2 plug, got the single cat out to inspect. Looked good. Cleaned up more grounds - the two front chassis rail ones, and the ABS bracket ground. Installed new air filter element as a matter of service while in there. Hopped in to start. Cranked, but no start. Pulled the new NGK plugs to find a bit black from the start up with MAP disconnected, tried 0.8mm gapped spare spark plugs. No change. That was a month ago.