EJ22 Idle and Intermittent misfire underload

[1980ea71Brat]

Hi all,

I am basically done with swapping a 1996 EJ22 into my 1987 Subaru RX. I wasn't having any problems until recently, after I changed the fuel filter. It doesn't seem to likely be the culprit but I will describe the symptoms:

Idle drops to 300-400 RPMs, catches itself before dying, jumps up to 700-825 and repeats

Increasing revs seems to be no problem when in neutral and not moving

Driving the car, it will occasionally stumble or misfire while giving it throttle in gear. It appears to be worse when at lower RPMs, but will still happen at 4000-5000 RPMs. Its like a sudden stumble/misfire and then it resumes.

I might be able to record the issue tomorrow and try and upload it to youtube if anyone would be interested. 

I have already checked the MAF readings using a OBD2 scanner, and those appeared normal ~4 g/s at idle and 12+ at higher RPMs. No stored codes - only one pending code: P1101

Could P1101 really be causing these types of issues? I hooked up the neutral wire off the ECU to what I thought was the neutral signal feeding into the EA82 ECU but truly, it was kind of guess based off of forum posts on here. The EJ22 was running fine when it was in the 96 legacy. I replaced the front O2 sensor and it used to have a stored IAC code. No codes are showing up, even having driven it 50 ish miles on long and short trips

I was really excited to finally start driving this project and this problem has really taken the wind out of my sails. Looking for any help and suggestions.

 

.

 

 

[forester2002s]

12 hours ago, 1980ea71Brat said:

I wasn't having any problems until recently, after I changed the fuel filter.

A brand new fuel filter shouldn't give you misfiring symptoms.

But if you want a quick and cheap test, try replacing the old fuel filter (if you still have it), and see what happens to the misfiring.

[1980ea71Brat]

So this morning, I bypassed the fuel filter to check to see if it was the cause of the problem - and it still persisted. Against my better judgement I replaced the spark plugs and plug wires with NGKs as well, to no affect. The old plugs where barely used iridium style replaced by the previous owner, wires had no wear either. At least I can safely say those have been eliminated.

I also rented a fuel pressure tested and it too came back steady. Max PSI was around 36 with the fuel pump priming correct, and held steady around 28-32 while running and revving. 

I also discovered that the misfire/stumble is happening while in neutral. I previously thought the car had to be moving for it be a problem.

First link is to the stumble while trying to hold the RPMs near 2k:

 Second link is showing the fuel pressure:

Third video is the stumble while at idle:

 

[1980ea71Brat]

Well I am at a total loss for the moment. I figured if it was throttle position sensor related, it would throw a code - but as you can see in the third video, the TP% does not fluctuate.

 

Removing the IAC valve looks like a pain in the rear. Could that really be what's causing the problems while under load like in video 1? 

Edited July 30, 2022 by 1980ea71Brat

[1980ea71Brat]

Okay. Update after a lot of back and forth to parts stores. I replaced both the MAF and the IAC. Spent the big bucks.

The MAF was 100% causing the problem with the mid acceleration stumble. I used my code reader to chart the MAF performance, it turns out it was fluctuating too fast for the code reader to display, but was logging huge spikes every time the throttle cut out. It was definitely acting badly, and replacing it with a refurbished unit in fact fixed that problem. It will now accelerate from 1000 RPMs to 5000 RPMs no problem, no hesitation.

HOWEVER. The idle is still completely messed up. It starts off running fine for about 10 seconds then starts jumping from 400-1000 RPMs nearly stalling out. The IAC was a $390 part brand new, so I don't think it's faulty, but I am going to run through the IAC tests tomorrow morning or later tonight (after the engine has cooled off).

Maybe the P1101 code could be causing the idle issue? Lack of idle control because it can't get a good neutral signal? Hoping someone out there can help me out on that one. The code went away and hasn't shown back up (it was pending before, and now its not). So I might try to take it for long drive tomorrow to see if I can get it to throw the code. 

Thanks for reading.

[lmdew]

|P1101 SUBARU code description Neutral Position Switch Circuit High Input (A/T) Neutral Position Switch Circuit Malfunction (M/T)
 

Yes it could be causing your idle issues.  I've done a few Auto Manual Swaps and the ECU needs to have a good signal.  You said this all started after the filter change.  Did you do anything else?  How long was it running well?  

[1980ea71Brat]

1 hour ago, lmdew said:

|P1101 SUBARU code description Neutral Position Switch Circuit High Input (A/T) Neutral Position Switch Circuit Malfunction (M/T)
 

Yes it could be causing your idle issues.  I've done a few Auto Manual Swaps and the ECU needs to have a good signal.  You said this all started after the filter change.  Did you do anything else?  How long was it running well?  

Maybe 30 miles before it started acting up. I was in the process of fixing a few other items on the car so I didn't really get too many miles on it after the swap. But I didn't replace or change anything else on the engine until now.

*Edit* To clarify, I put about 30 miles on it without issue. Decided to change the fuel filter since I swapped tanks etc, and it looked original. maybe 5 miles after that it started acting up. I think the fuel filter had nothing to do with the problem and ended up just being coincidence. The fuel pressure is consistent with or without the filter.

Edited July 31, 2022 by 1980ea71Brat

[1980ea71Brat]

I just did a full 1 hour of leaving the negative off of the battery post to 'reset' the ECU. No change. It starts up for about 5-10 seconds and starts into a loping idle. Seems to get worse after 20-30 seconds.

First lope is about 600-900 RPMs, second lope is about 300-1000 RPMs. I can film the whole sequence again, but it's similar to video 3 from above.

I did another test.... I unplugged the IAC. It bogged down, and held a rough idle at maybe 400-500 RPMs. As soon as I plugged it back in, it started to lope again.

Does this confirm that my $390 IAC is bad out of the box? Potentially a wiring issue? The voltage read out on the 3 pin connector isn't consistent with the Haynes manual, but sometimes those aren't as accurate as factory service manuals.

[el_freddo]

Potential wiring issue. 

Also check the alignment of your timing belt. Ensure the crank aligns with both cam wheels correctly. 

What’s the fuel return pressure looking like? 

Cheers 

Bennie

[1980ea71Brat]

Unfortunately I didn't get to check the return pressure line before I returned the rented tool to Orileys. I'm definitely going to consider adding one to my inventory in the future, because it was super helpful verifying I didn't need a new fuel pump. 

Timing looks good. No issues with it running past idle, and I can track the advance using the OBD2 scanner.

 

I took for a 30 miles drive today to get some data and try to get it to throw a code. The opposite happened - the OBD2 scanner gave me a green check for all systems, (before it was giving me the yellow 'something might be wrong' light.). 

When I push the clutch in, it slowly declines the RPMs (not instantly) so its very drivable. When I come to a stop sign, it will hold the RPM for a little bit before going back into a lope.

This makes me think... vacuum leak? Vacuum leak might not throw an engine code, but the IAC might be operating to try and steady the idle against the unmetered air? Just a thought. I'll try to drive it a bit more on some of the not as hot days this week to see if the scanner can pick anything up.

[1980ea71Brat]

Drove another 50 miles or so. No improvement, yellow light back on but no pending codes from the scanner. Not sure why it flags an issue without knowing what issue it's flagging.

 

I'm at a loss here. Other than coming to a stop and idling, the car runs exactly as it's supposed to. I couldn't find any vacuum leaks and I traced every line, sprayed every line, every intake gasket, no change. I tested the ECT, and it's also working correctly. 

 

Any other ideas out there in the internet? 

[1980ea71Brat]

Latest idea - Cut the Neutral Signal, and ground it/unground it. I tried this going back and forth to see if the neutral signal was causing the surging idle - no dice. Let it idle for about 15 minutes, sprayed the brand new IAC with throttle body cleaner. No codes, no change.

[1980ea71Brat]

Okay all - I have been reading. A lot. Probably 50-100 threads on this website and others. SO MANY idle problems die unsolved on the internet. I REALLY don't want that to happen here.

I may have a small break through and I am desperately looking for someone else with an OBD2 scanner and an EJ22 to compare with:

- Chilton's Manual section 4-9 Heated Oxygen Sensor testing section states that, after fully warmed up, hold the idle at ~1200 RPMs and measure the voltage. Voltage should fluctuate rapidly between 0.4 - 0.6 volts. If the the voltage stabilizes at the middle of the specified range, or if the voltage fluctuates very slowly between the specified range, the sensor may be faulty. If the 02 sensor stabilizes at either end of the specified range, the ECU is not able to compensate for a mechanical problem such as a vacuum leak, or a faulty fuel regulator. These types of problems will cause the 02 sensor to send a constant lean or constant rich mixture.

Chilton's has been wrong before. But my first bank 02 sensor is reading constant fluctuations between .88-.95 volts! This would imply a very rich condition. I just replaced this front o2 sensor, so its possible that it's bad out of the box. I held the Idle at 1200-2000 rpms for several minutes each and this was the reading, fully warmed up, radiator fans spinning.

 

Can anyone out there confirm that this repair manual is speaking truth? Is my car is running too rich? @Bennie, perhaps I should have measured that post regulator fuel pressure after all.

Likely causes for being too rich? Fuel pressure regulator? Fuel injectors leaking? I couldn't think of anything else.

[1980ea71Brat]

Also - As I am trying to learn more and more about this, I also checked the short term fuel trim. It's basically 0 to -30% 0 at idle and up to -30% while revving up towards 2k RPMs.

This also is indicating that it's running too rich, right?

How is it I still don't have a check engine light? Note: While the idle is doing its crazy idle fluctuations, the short term fuel trim stays at 0%.

[1980ea71Brat]

At this point I'm looking for any help for possible explanations for it running so rich. I wasn't able to check the resistance of the Coolant Temperature sender (its at an awkward angle for my ohm meter) - But it's sending correct signals to the ECU. If I unplug it, the fans run, and if its plugged in, the fans will turn on when they get to temperature.

I read that it could be the fuel pressure regulator - I was getting 28-30 PSI steady state while running and revving. Some sources on the forums say it should be higher. Could the missing PSI be the regulator letting extra fuel slip by?? The dreaded Chiltons says pressure is supposed to be at 20-24 PSI...

My understanding is that if it were injectors, the PSI from the fuel pressure test would slowly leak down. It held steady (when not running) at 36 psi.

I feel like I am close to the solution. Thoughts or suggestions?

[heartless]

ok, you need to ditch Chiltons, and Haynes..

go here and find/download the appropriate service manual(s) for your needs.

you may need to do a little digging around - the pages are not usually labeled well, but the information is there.

[1 Lucky Texan]

not sure it's been mentioned, but a sneak path for fuel can be the fuel pressure regulator, pull the vacuum line and see if it's wet with fuel.

[1980ea71Brat]

@heartless Thanks for the tip I will take a look at those and see if I can find some answers.

@Texan I checked for that today and did not have any drips or leaks

My research today lead me down the path or checking grounds - I checked all the ones I know of and replaced the ECU ground for a better location. No affect.

I checked to see if the problem was happening in open loop vs closed loop - it looks like it doesn't leave OL while idling, and immediately switches to CL when I give it throttle. No idea if that's supposed to happen normally.

I ordered an Ebay ecu to see if that will make any difference ($5 so probably not)

I am going to go rent the fuel pressure gauge and attempt to test the FPR by removing the vacuum line.

That's basically all I could come up with - the rear 02 sensor is reading consistently .2-.3 volts less than the front 02 sensor, which leads me to believe the front 02 sensor is working, and the cat is doing its job cleaning the air. 

[1980ea71Brat]

Unfortunately the factory manual doesn't give much for diagnostics when it comes to a rich condition. It does give me FPR standards so at least I can verify those, hopefully tonight or tomorrow.

*edit* Rented the testing unit, will check the FPR tomorrow.

Also - discovered that my car is constantly entering OL-Drive instead of staying in closed loop while driving around. This point to a possible ECU problem, or with the barometric measuring unit in the ECU. Not sure if this can be repaired or the ECU just has to be replaced if that is the case.

 

Edited August 4, 2022 by 1980ea71Brat

[heartless]

interesting.... is my understanding that it should only be in open loop when engine is cold/during warmup.... once it hits full operating temp it should go closed loop and stay there, even in cold weather, until shut down and cooled off again.

I dont think it has anything to do with the barometric unit (pressure), this is more a temperature related thing. temp sensor for ECU (or ECU itself), thermostat, etc - possibly O2 sensor(s)?

hopefully you can get it sorted out soon.

[el_freddo]

It’s an interesting one @1980ea71Brat! Dunno that I’ll be of much use from here on. 

The only good thing I can see is that the issue is consistent and not an intermittent issue that can be super painful to pinpoint. 

To add some confusion - @heartless- I thought closed loop was programmed into the ECU for cold running until warmed up. Open loop then reads the various sensors to run the fuel maps for the given parameters. I haven’t read up on this stuff in ages so I could have it wrong! Hence the confusion comment. 

See if you can test the temp sensor for the ECU @1980ea71Brat, it could be out of spec but not enough to throw a code. What’s your knock sensor look like too? If cracked, replace it! 

Double check those O2 sensors and ensure you’ve got them the right way around. 

When cutting down the loom, anything in there to do with the fuel tank in terms of pressure or temp sensors? If so, I’m wondering if these could be causing issues. 

Cheers 

Bennie

[heartless]

sorry Bennie... but open loop is the cold/warm up phase - the sensors do not affect much... closed loop is when it is using the sensors to adjust various things...

Via google search...

Quote

Open Loop vs Closed Loop in a nutshell

In a nutshell, closed loop is where the engine operates with a feedback loop [via the sensors] A sensor closes the loop.

In open loop mode, the engine doesn't listen to nobody, it just runs off a set of instructions based on where the throttle is at, where the revs are, and so on

 

 

[1980ea71Brat]

Alright fellas. Just like like heartless said - Closed Loop is supposed to be how the engine runs after its warmed up and reading off it's own sensors. If it is going into Open Loop, that sometimes can mean there is something wrong with the ECU or the Coolant Temperature Sensor (ECT in the manual).

A Subaru technician friend of mine said that they had problems with ECU's relearning the fuel maps for their altitude (he was in colorado) and it would take them some miles to relearn using the barometer in the ECU. I thought that could be my issue being that we are at 3500 ft in Billings - but probably not. My 'new' to me ECU from EBAY did not fix any of the problems.

 

I also did a full test on the FPR - everything is within spec according to the subaru manual, and it holds PSI (no leaking injectors). The return fuel pressure was 0 PSI. The interesting part of this test was when I unplugged the fuel pump relay to release the pressure from the system - it kept running for a solid minute. It wouldn't take throttle well, but it clearly was running okay just off of idling. Super strange. Even stranger, when I pulled the fuel line it exploded with pressure spraying fuel everywhere - why didnt the fuel pressure decline with the pump off??

I thought it could be a clogged return line, but I was able to blow air through that line, and like I mentioned earlier, the FPR passed all the test, and did have any PSI after the regulator.

Where does that leave me? Even though my ECT works for turning on the radiator fans - I bought a new one to replace tomorrow and see if it works.

But before I do that - I DID FINALLY get a code for the front bank 02 sensor after driving all around to return the rented tool and pick up the ECT.

The code is only pending, so its not 100% a fault. I'm going to try rewiring it tomorrow (EJ22 exhaust puts it at a weird spot for my cross member) and going for a long spirited drive.

 

If that doesn't work I am stuck replacing the ECT... and cleaning the injectors?? Out of ideas again - really thought the ECU was going to fix it.

 

BTW the new ECU still switch between CL and OL-DRIVE while under throttle and idling. 

*edit* I also replaced the front 02 sensor not 300 miles ago or so. Unfortunately I didn't save my paper work, because I'm wondering if it was bad out of the box.

 

 

Edited August 5, 2022 by 1980ea71Brat

[1980ea71Brat]

I also tested the Knock sensor and It was within spec - exactly 560 kohms, no knocks detected by the OBD2 

[lmdew]

I have sets of Red Injectors.  Pay shipping and a Donation to the USMB and they are yours.

Larry