Subaru engine sensor oil leaks and loom oil contamination.

[bigBADbenny]

Subaru engine sensor oil leaks and loom oil contamination.

Screen Shot 2020-09-15 at 09.35.20.jpg (51.66 KiB) Viewed 5530 times

https://www.clubsub.org.nz/forum/index. ... have-avcs/

This is a known issue and is referenced in my AVCS faq.

I had thought the issue was confined to preFL DAVCS cars because thats where most instances appear due to the extra exhaust cam and avcs sensors, below the heads.
However I recently saw a post on FB regarding an 04 WRX cam sensor issue which prompted me to do some more research that led to the ClubSub post linked above.

Definitely worth reading that forum post, however if you have a short attention span,
here’s a few observations to consider...

The loom can be contaminated all the way to the engine harness and ecu, but also other sensors on the engine.

So the focus here is the inlet cam position sensors on the upper rear of each head, but the issue also affects exhaust avcs cam position sensors, avcs oil control valves, oil pressure switches, eg also those related to ALVS on H6 etc.

The oil can harden loom wiring causing continuity issues.

Therefore its useful to check the extent of the contamination around all terminations on the loom.

It it’s not too bad, the affected sensors can be modified with sealant, remove plug seal, and or use bullet connectors etc to isolate the sensor from the loom.

This could save you a bunch doing the sensor mods as new oem replacement cam sensors can be quite expensive.


I did some research and some fixes include:
Pulling apart the sensor connectors and coating the individual wire crimped terminals with epoxy resin,

Using epoxy resin in the sensor plug (in place of the internal rubber seal),

Applying 3bond to seal the metal sensor transducer to the plastic sensor body, around the circumference of the join.

Also measure up for new oil and heat resistant oring:

This will likely be a metric oring and you’ll need to extrapolate the oring outer diameter and oring section width dimensions.

Causes and prevention.
As for causes of the issue, and prevention, the primary candidates are oil ingress due to leaks from faulty or worn sensors and their seals an orings...

The oil enters the loom via gravity, osmosis, heat or potentially excessive crank case pressure.

The sensor plugs and loom is water and potentially airtight meaning the oil has no choice but to penetrate the loom.

Therefore keeping the engine very clean along with regular inspections will help to spot leaks early before the loom and plugs become inundated with oil.

Equally important is checking crank case pressure as part of regular health checking, in paticular ensuring the PCV isnt stuck.

A quick way to check for PCV or compression related excessive crankcase pressure issues is to get a spare oil cap, tee a fitting for a hose to a boost gauge into the cap, and make a few pulls over 3k rpm. If you go positive pressure, youve confirmed the issue.

[bigBADbenny]

Quick update, the postfl single avcs inlet cam position sensor o-ring in my car was worn out and causing an inlet leak.

I pulled the sensor out, wipes it down and took it to my local engineering supply shop: Southern Bearings in Abbotsford.
We removed the hardened and squashed oring and measured the height and width, averaged them and extrapolated 15x1.5mm as the size for a replacement item in nitrile.

Popped it back in and the car, which has had most inlet gasket, seals and injector seat seals replaced during inlet pressure testing, ran superbly for a day.
Then unfortunately it got a no crank, no start issue with CEL/DTC’s...

DDD834EB-8383-469E-904C-FFE30AC5D05D.jpeg (79.07 KiB) Viewed 5461 times

So I fixed the o-ring on the bank1 sensor, but these CELs are for bank2 and crank pos.
I’m not worried about crank pos as that’s more or less due to the bank2 issue.

I’d hazard a guess that now bank1 sensor has been cleaned, bank2 sensor is reading differently than before and the ecu doesn’t like it.

But regardless, I’m going through the FSM diagnostic procedures.

[bigBADbenny]

I forgot to add that the car ran flawlessly for 2 heat cycles and about 50km, then sat all day before it wouldn’t start and threw the CELs.

I got cracking with the fix, foolishly I initially assumed it was bank1 issue since that’s what I’d removed and repaired.

At least that side is checking out ok as regards FSM diagnostics.

I’m repairing worn loom insulation on that side of the harness, then I’ll button it all up and clean, replace, test bank2.

Its pretty tight down there on bank2 so the inlet manifold might have to come off.

Here’s the diagram from the FSM.

F83C95D2-99C2-4560-8E71-DE9AE02AAD50.jpeg (61.67 KiB) Viewed 5443 times

[nvmylh]

Goodluck mate! Making me want to pull those sensors out and check them over before getting my spec.b wagon on the road!

[bigBADbenny]

I’d just check the area around the sensor/s and inside the plug for that sensor.

Oil in the plug and at the main connectors near the firewall, the latter two on a 2.5t: the red e2 plug under the tmic and the main plug e2.
Other telltales are oil pooling around the sensor, or simply dark dry goo around the sensor which is fallout from blow-by.

If there’s no evidence leave them be.
Apparently the cam sensors are very sensitive with age and just swapping out the oring on one side was enough to throw out the other side.

I’ll update my progress today: repairing the loom insulation on bank1 drivers side sensors along the upper cylinder head: tgv, injector1, avcs & cam pos.

DCEB7384-4D0F-48A8-B4CF-B44309D8958F.jpeg (218.92 KiB) Viewed 5388 times

Basically the outer insulation in this area just fell to bits whilst fixing all the seals and gaskets in the area.
Obviously with age every perishable item in the area gets cooked due to proximity to the uppipe and turbo.
Plus the loom is very tight as it passes around and down next to the head breather pipes and turbo coolant tank.

But as I found, the other bank is even tighter!

Then I’ll move onto testing the bank1 sensor.

Update: bank1 E35 #2 to E2 #2 passed continuity but failed the FSM resistance test.
I’m probing that wire on the last loom branch before the sensor, will replace and re-insulate the section with more support at the branch.

But zero oil contamination in the loom so far

[bigBADbenny]

Well, eventually I found the issue, here’s the story...

Ok after an absolute ton of misdirection and major headtruckery: its fixed! :love:

Over-fixed even.

Bank1 cam position sensor oring fixed (plus1)

I wish I’d spent an extra 10 minutes googling the exact scenario, which I did today, being at my wits end

So I didn’t research it thoroughly (strike 1), and I made a lot of assumptions (strike 2, make an ass of you and me, mainly me).

So onto the many assumptions I made. Yes I know better, but even didn’t follow my own advice.

Rather than google the three CEL codes and eg “no crank” “no start”, and because I’d been messing with the bank1 cam sensor, I went to the effort of testing the loom and swapping that out with a spare: no/same result.

This despite the FSM CEL codes mentioning bank2 cam sensor...

Then I repaired the rooted loom outer insulation (plus2).

After that I lifted the inlet manifold removed bank2 cam position sensor, tested the loom, replaced the oring (plus3), found a fault in the loom, fixed the loom (plus4?), swapped out the sensor, did it again.

Still no crank, no start and all three CELs

Meanwhile the battery had been put on trickle charge as I’d noticed the voltage was dropping quickly during diagnostic testing.

Right, well I distantly remembered this battery was no spring chicken, it came out of the same my07 STi that my gearbox came out of. Tightarse!

So the car did eventually fire up, and with all the fixes, especially holding crankcase vacuum, ran very well indeed.

However, it wouldn’t restart, and then I noticed the battery tray was wet.

Pulled it out to top up the cells, but too late, the indicator went black and the surface charge would drop alarmingly overnight after a recondition charge.

At this point I was thinking, well it’s just a rooted battery, can’t turn over the car. Yet another assumption!

Anyway I make it to work that day but the car won’t start at home time!
And the foreman has a sore back, won’t help, and is screaming at me.

I’m at wits end and bust an absolute gut pushing it out of that hellhole, down the ramp and onto the street where I magically jump start the beast. WTF?

Now I’m assuming the starter came off the low k STi donor. 60k km when it met its unfortunate end going over the edge on a touge run up Arthur’s Seat.

But more likely the crew that installed the box in my lib just left the original in place and probably binned the newer one.

And by this time I’m prepared to google it a little more as the beast is stuck in our driveway going nowhere fast.

Well, as you expect but never assume, there’s results online (Nasioc mainly or that’s how far I delved).

And they’re evenly divided between grounding issues (thanks above reply, all good in testing), and eurgggh the freaking starter!!!

So being ignorant of where’s what, I looked up the considerations in the FSM: tl/dr, couldn’t really see what goes on, so I ripped off the tmic one more time to take a peek, gave it the usual technical tap... BINGO!

Kind of a perfect storm of truckups and ASSumptions really.

Its amazing how sheer terror (missing work, no transport etc) really can cloud ones judgement.

Also how hard it is to do things without basic tools and enablers.
Eg with an engine crane to lift the inlet manifold a little and a very strong magnetic torch, the bank2 cam sensor is a 45 minute job, tops, and near impossible without.

I hope you enjoyed the story and maybe learned some whatnot to-dos.

I sure did. Yessir. Will not forget this in a hurry!

Now do I remember that @Corgiwerks recently began offering starter motor service kits or refurbs? Better take a peek

[bigBADbenny]

Now this exact thing I mention in one of my FAQ’s.

Multiple CEL’s can point to an issue upstream or related to the individual CEL’s.

This is a fairly extreme example.

The CELs point to a timing issue, but the actual issue is either a dead battery or worn starter.

Dead battery: not enough timing signal during attempted cranking?

No cranking on key on: no timing signals.

Its too obvious and a stretch at the same time.

And yet it had me fooled.

I even had the Tactrix & Romraider on the job peeping basics like neutral position switch and starter signal!

I really should take my advice more often