93 RS Legacy TT - intermittent stumble/hesitation at part throttle

[ris2t]

Hi

Looking for ideas on how to diagnose an annoying/intermittent problem.

1993 Subaru Legacy RS EJ20TT, 198k on the clock.

Rarely but atleast once in the weekend - At part throttle the car stumbles like an engine or fuel cut but comes back immediately. Usually only does it once or twice.

No issues under full throttle, e.g. clear on ramp powering up through the gears and boosting.

Can drive around all day and not have it happen. Seems more prone after the care has been in use for a number of hours. Or being reused later in the day.

Car idle is stable considering its age. Bit of fluctuation/hunting but all within about +/- 200rpm, aka normal for the car.

When:

#1 Most often occurs in 4/5 the gear driving along at constant speed around 2000 rpm. Hickup and continues on. Both motorway and clear city roads.

#2 Had it once going up the driveway in 2nd gear around 2-2500 rpm. In this instance rpm dropped away (suspect because of the hill) before coming alive and continuing up the drive.

#3 Happens now and again when gently pulling out of side roads around the 2500 to 3000 rpm.

From the gauges

- don't see any apparent drop in RPM

- don't see any manifold pressure (vacuum) changes

No check engine (CEL) light or codes when using the 1 wire plugs.

Things done to date:

- replaced spark plugs (over 100k km on them, but looking clean)

- compression tested - ok based upon cold cylinders

- changed crank and cam shaft sensors, just precaution

- cleaned MAF, multiple times with little effect

- intercooler and piping has been off and on multiple times with no impact or obvious pipe splits.

- swapped ECUs, back to factory from a MINEs one. Still happens although slightly less frequency.

- upper throttle and idle air control valve foam clean, no impact

- changed a number of vacuum lines that were brittle, no impact

- cleaned and flushed the vacuum assemble for the twin turbo circuit. No obvious leaks or switch issues.

Other factors

- been having issues with water rushing the in the heater core. Replaced the core as got blocked with stop leak (prevention).

- no detectable coolant leak and a pressure test (when cold) holds for a while. NO overheating ever.

- Even if there was a coolant, can't see how it would cause a intermittent stumble. Would expect poor idle.

Thoughts?

- fuel pump, fuel pump relay, injectors, regulator - I would expect these to be worse under load if an issue

- sensor, electrical - can't think of anything else that would impact at part throttle and not throw a fault code.

ANY Ideas appreciated

Thanks

Raymond

[b01z]

have you tried the 02 sensor? I had an issue with mine for awhile, tried everything BUT the 02 sensor, what you say is sort of consistent with the beginning of o2 sensor failure. Mine would normally stumble between 2000 - 3000rpm for about 2 seconds then just pickup and go.

[Leufkax]

Air flow meter could be bung and not yet throwing a code

[les666]

Could also pay to check coils for small cracks.

[McDoof]

I had both of the above issues with my 2000 GTB. Look for signs of arcing on the coil insulator. Mine was arcing to wall when the car got hot.

[ris2t]

I checked the coils when replacing the plugs. Replaced one of them as I cracked the insulator taking it out. No obvious marks of arching but probably worth me revisiting. I would have expected this to break down at higher revs though. And more feel like a misfire then cut out.

[ris2t]

 b01z said:

have you tried the 02 sensor? I had an issue with mine for awhile, tried everything BUT the 02 sensor, what you say is sort of consistent with the beginning of o2 sensor failure. Mine would normally stumble between 2000 - 3000rpm for about 2 seconds then just pickup and go.

This and the knock sensor are my next consideration. Did you use an OEM sensor or generic replacement? Been seeing notes that suggest the OEM O2 sensor is needed.

And fingers crossed it actually comes out

[b01z]

 ris2t said:

This and the knock sensor are my next consideration. Did you use an OEM sensor or generic replacement? Been seeing notes that suggest the OEM O2 sensor is needed.

And fingers crossed it actually comes out

I used a bosch replacement which i ordered through wingers.

I would also run carb cleaner through the AFM to give it a good clean, if either of the 2 are playing up you get false positives codes popping up. Knock sensor would throw a check engine light.

[Rosssub]

 b01z said:

I used a bosch replacement which i ordered through wingers.

I would also run carb cleaner through the AFM to give it a good clean, if either of the 2 are playing up you get false positives codes popping up. Knock sensor would throw a check engine light.

If you're going to clean the AFM use a dedicated MAF cleaner, or second best an electrical contact cleaner. Both safe for sensitive stuff, ok for plastics and evaporate completely.

Other auto cleaners are pretty harsh solvents. Carby cleaner for example is mostly acetone which eats plastic and will often leave an oily residue.

[ris2t]

UPDATE:

Replaced the knock sensor, that was an interesting exercise. Just enough clearance. Was relatively cheap sensor. The old one looked pretty good, no cracks, so doubt it was an issue.

3 days later I've fitted a second hand Oxygen sensor. 1/3 price of a new one so a gamble. No faults so far but its so intermittent. Will monitor and update if I don't get any reoccurrence.

[ris2t]

UPDATE #2:

So after swapping O2 and knock sensor and a good couple of weeks running the hesitation has happened again. Minor hits are part throttle when in car park and one clear issue when driving up an on ramp with gentle throttle. Still no issues when under full load.

I do wonder if I've got a minor coolant leak (intake perhaps) that impacting ignition some how under partial load/vacuum. The heater core gets the sloshing sound now and again but somes and goes. Never could find a leak (no over heating) even with a pressure test.

Next thoughts
- IAC, this did get a partial clean when I did the O2 and knock sensor. Might explain the couple of weeks of no issue.

- AFM - might have the bite the bullet and try a different one on the off chance

- some kind of fueling issue (fuel pump/regulator) - discount this as this would me more likely under load.

[ris2t]

Update#3: Replaced the AFM with a reconditioned unit. Been 3 weeks and no repeats of the problem, so I suspect this was the fault all along. Also smoothed out the power delivery which always had a flat the hiccup before go spot mid rpm.