FuZzBoM

Hi all,

Looking for a bit of help trying to figure out why my 2009 STi hatch keeps blowing turbo oil seals.

When I bought the car I was told the turbo had recently been replaced (VF48).. In less than six months the turbo started leaking oil into the intake/intercooler. When I swapped the turbo it looked a bit tatty so could have been swapped for a used turbo. I swapped the CHRA out for a new Mellett one as I was told they were of good quality. About 1500 miles later this started leaking as well..

Due to need to use the car I bought another Mellett CHRA and swapped it over. I sent the ‘faulty’ CHRA back to Mellett to see if they could indicate why it failed. They came back with a report saying everything was in tolerance but there were signs of oil contamination and overheating and refused the warranty claim. Talking to friends who have sent turbos back for warranty claim, they said that this seems to be the usual line most companies use.



A few days ago I noticed a puff of smoke from the exhaust on start up again… And when sat in traffic I could smell burning oil.

Have I just got unlucky with the Mellett CHRAs or do I possibly have another issue I should look into?



When changing the turbo I change the oil for my usual Shell Helix Ultra 5w40 A3/B4 that I’ve used on previous Subaru’s without issue.

Car is a daily driver running stock map. I never do track days and never drive it that hard. Previous owner did fit a sports cat so I have seen boost occasionally hit 17 Psi but normally around 15 Psi according to my gauge. Yes I probably should have the ecu remapped but under boost it will be nice and rich (I have an AFR gauge too) never had any running issues. I’ve also heard of people running VF48s at above 18psi without issue.

I run a Mishimoto radiator with a new thermostat and silicone hoses, I never see the engine temp go above 86c normal driving so I struggle to believe it is an overheating issue.

When replacing the turbo last time I also replaced the PCV and hoses in case that was an issue.

The only other things I can think of are I run a Grimmspeed air oil separator and the secondary air pump has been deleted. But I can’t imagine those would affect the oil seals?! I may remove the AOS for a while just to eliminate that.



I have just ordered one of these in the hope a race built core can cope with daily driving lol
https://www.turborebuild.co.uk/websh...bocharger.html



Sorry for the long post. I just want to cover all bases if it is something I’m doing wrong. Open to suggestions.
Thank you

NOSSY_89

I was going to suggest PCV valve. How have you got the AOS plumbed in thought that would have done away with the PCV.

FuZzBoM

Thanks for the reply. I have left the PCV system next to the turbo intact. I have both the breather hoses from the tops of the rocker covers going to the AOS on the oil filler tube then routing across to the intake before the turbo. Sort of a half AOS setup if you catch my drift?

I do have a catch can in the garage and was considering running that from the rocker covers and just capping the intake to the turbo. Although I’ll probably just revert it all to standard for now to see if that helps.

1509joe

iTrader: (2)

Check/change your oil return pipe turbo to sump and make sure your not over filled the sump Also make sure you have the correct banjo bolt/restrictor on the oil feed pipe

FuZzBoM

I forgot to say that. When I changed the turbo last time I also replaced the oil return hose because the original one was hard. I had a look down the return but couldn’t see any blockages and used an airline to move anything that might be blocking. Also running the standard oil feed restrictor as the turbo was a like for like swap. I believe these should be a 1.5mm restrictor but I’ll double check it.

Thanks for the input though.

FuZzBoM

I started to remove the turbo last night. Sure enough intercooler and throttle body full of oil.
Checked the banjo restrictor and found a 1.5mm drill bit fits snugly in there, so I’m convinced it is the correct bolt.

I’ll update if I find any other odd things when I split the turbo housing.

Henrik

iTrader: (5)

Do you have an oil pressure gauge?

gazney101

iTrader: (5)

It's a bit of a longshot but if the avcs Banjo filters are blocked could it be over pressuring the turbo feed?
​​​​​​

FuZzBoM

no, but it's not a bad shout to get one

FuZzBoM

Also a good suggestion. But I don't have any running issues I'm aware of. Also, the engine was replaced by Subaru around 15,000 miles ago, so I'd probably have to be unlucky for it to be he case. I'll see about checking them though

lockheed

iTrader: (1)

And sometimes the thimble filter goes in to far also blocking the oil feed, i took mine out.

FuZzBoM

Hi all, to resurrect an old thread. I fitted the enhanced core from turborebuild. It has lasted the longest at just over a year with 3500 miles and has started smoking…
https://www.turborebuild.co.uk/enhan...ridge-66-blade
I’m seriously getting peeved with this. That’s now 5 turbos in 3 years.

I have never seen the turbo pressure go over 1.1 bar when on boost.



As a precaution, I have done the following to see if I could reduce the chances of it happening again.

Replaced PCV and hoses

Oil has been replaced every 6 months (1500 miles) with Shell Helix ultra 5w-40.

New genuine Subaru restrictor banjo bolt on oil feed. I believe these are 1.5mm restrictor hole.

Removed the mesh filter on the oil feed banjo bolt on the block. (has been noted on many forums these fail causing oil starvation) The AVCS (VVT) is run on the same oil feed line as the turbo on the hatch. If it were a starvation issue, I would be receiving an EML as the AVCS would not operate as intended!?

Larger Mishimoto radiator with new genuine thermostat and silicone hoses (engine temp sits at a solid 86c)



I also now have an oil pressure gauge. On a cold start it will go to around 6 to 7 BAR then pressure reducing as the oil warms up. When warm idle shows about 2 BAR. When driving it will hover around 5 to 6 BAR at varying loads and revs.

I only use the JDM black oil filters as they are supposedly superior quality to the UK blue ones.


Any advice you can offer would be greatly appreciated. I'm seriously considering selling the car, because having to fit new turbos all the time is getting annoying and expensive.

Cheers,

Rob

1509joe

iTrader: (2)

Ask them to send the recommended restrictor for what you have ordered.

lockheed

iTrader: (1)

I happen to know theses turbos can come with the wrong restrictor i even had a turbo sent with out a air speed sensor tapping .......... But five turbos in a small amount of time ,there's something else wrong .

FuZzBoM

Thanks for the replies. it's clearly getting oil. So I'm convinced it isn't a starvation issue.
The website and the instructions that come with the CHRA do not mention using a different size restrictor. I’ve been in contact with them over this (they have been very helpful so far) I mentioned I run the OEM size 1.5mm restrictor to them but this didn’t prompt a response. I believe it is still a journal bearing turbo and standard size. The only difference is a billet compressor wheel and stronger internals.

I’m not chasing power, I just wanted reliability.



Anyway. They have asked I send this core and any others I may have to them for investigation. I’m going to see if I can get it off in the next few weeks.



I had another thought. Could coolant be an issue? I run Comma G30 in it as it is readily available and what they recommend on their website, I’ve had no issues with it in the past. Could it be boiling in the housing and therefore not cooling the turbo as it should?! Clutching at straws I know. But I’m running out of things to try.

Danjo

Not teaching you to suck eggs, but are you priming the turbo before installation?

FuZzBoM

Indeed I am.. These CHRA's come with a syringe of oil to prime them before putting them on. I also top up before putting the banjo bolt on.
I change the engine oil and prime by unplugging the crank sensor to get oil around everything before starting the engine. So I’m fairly sure it has been correctly primed. Cheers for the input though.

Danjo

To me it points towards excessive crankcase pressure allowing oil to be pushed past the seals, particularly with the amount of oil you are seeing in the intercooler pipework. Have you confirmed the PCV is good? And while I can’t see any major issues with the way you’ve routed the Grimmspeed AOS (other than it being fairly redundant as 90% of blowby will exit the crank case vent at the rear of the block rather than the heads, unless aggressive lateral g-loads are introduced) it’s worth putting it back to standard to tick it off the list of potential issues.

Excessive crankcase pressure limits the turbos ability to drain its oil and therefore it can only go one way, and that’s past the seals.

**edit** I’ve just re-read and seen that you have changed the PCV. Is all pipework connected as per factory (other than the AOS)? If you are introducing boost into the crankcase that would cause a similar issue.

JDM_Stig

iTrader: (3)

Utilising the dipstick tube can you not drop a meter on to test the actual crankcase pressure, also if it was excessively high wouldn't it be popping the dipstick out?

FuZzBoM

Yep I replaced the PCV too. And reverted all the oil breather system back to stock. I have had all the other breathers off and confirmed they are not blocked up. Maybe I should consider a catch can...

FuZzBoM

It's an idea.. No issues with dipstick popping out that I'm aware of.

**edit** I might buy another oil filler cap. I could drill a hole and make a fitting and attach a vacuum line. I can then either run that to a boost gauge or I have a motorcyle carb balancer and get a passenger to record what the gauge is doing. Worth a try!?

lockheed

iTrader: (1)

Id get a leak down test done by a professional company like tdr ,sc , subaru 4 u .........

JDM_Stig

iTrader: (3)

Has to be worth checking, 2.5 to 6 psi is the window Google suggests is correct.

FuZzBoM

I'm surprised it is as much as that. But yes, it won't cost me anything to try it. I have a spare boost gauge sat in the garage.

FuZzBoM

I have a leak down tester among other tools. I'll give checking the crank case pressure a go first I think. Cheers.

JDM_Stig

iTrader: (3)

Could be worth finding a local member with similar car for a side by side check, if you are drilling a oil filler cap and nice length of hose to reach your cabin then its not going to hurt their car.

bunsofsteeeel

iTrader: (1)

I had a similar problem years ago on a Jdm bugeye sti it kept blowing turbos but it only took about 20 miles to kill each of them, in the end powerstation did a type 25 oil modification where they drilled into the block and run a oil feed from there to the turbo, that solved my issue.

JDM_Stig

iTrader: (3)

Was the Bugeye standard?
Sounds a bit aggressive for high crankcase pressure when thousands run ok.

bunsofsteeeel

iTrader: (1)

no it had a hybrid turbo, injectors, induction kit, manifold, full 3” exhaust but one day just started eating turbos.

FuZzBoM

Today I finally got around to doing my janky mod to the oil filler cap. The fitting is sealed with silicone sealer and as airtight as I could make it. I plumbed into my boost gauge as it was already there. Engine off but ignition on gauge sat at zero as you'd expect. Start engine it went to vacuum just below zero. Driving normally gauge moves a few notches below zero. putting foot down and making boost the needle goes further into vacuum. I'd say it went no further than about -0.2 bar (at no time did the crank case go into positive pressure)
I personally feel this shows my PCV and general breather system is operating as intended?

I also think that’s a reasonably promising result to rule blowby out as a possible issue.



I suppose the next logical step is the oil pump? Could the pressure relief valve not be operating as intended? Or could it be worth dropping from 5w-40 to 5w-30?
Being a hatchback STi I believe these all came with 11mm oil pumps. The engine was replaced by Subaru under warranty, so I’ve no reason to assume they would use anything other than standard parts.