Jump to content
Ultimate Subaru Message Board

Reman turbo losing shaft nut within first minutes of running


Recommended Posts

So I cut up the oil filters.

 

First one (from the OE turbo failure)was dark and sludgey, but there was no sign of debris

 

Second one (from the 2 replacement turbos) had nothing but clean fresh oil in it.

 

the oil cooler had been leaking pretty badlyat it's upper o-ring, but the Oil cooler itself showed no signs of blockage or debris.  

 

Same for oil pickup tube.  Pickup is not cracked either.  it's O-ring seamed fine still.

 

Debating tearing the timing belt off and inspecting the oil pump.  Don't really want to, but it may be called for.

  • Like 4
Link to comment
Share on other sites

Who are they getting these CHRA's from? We only use Melett CHRA's when we do rebuilds. And I have not had a single one come back. 

 

It is possible that one of the housings is impinging upon the turbine or compressor wheel at speed? That is - the natural radial shaft play inherent to journal bearing turbochargers is allowing one of the wheels to graze the housing once it spins up.

 

I would get a CHRA and do the "rebuild" yourself. It's easy - takes about an hour. I do about 2-3 a month. Depending on the condition of the housings, you may want to swap those out for a different used set at the same time. I have a whole shelf full of dead TD04's if you need housings. I probably have 20 or more. We almost always just upgrade those to VF's or 16G's.

 

ALL turbo replacement's due to failure get an oil pan pull. Typically we just replace the pan with a new STi pan, Killer-B pickup, and matching dipstick. That 04 is going to have the inferior flat-bottom pan. 

 

Also ALL turbo engine have ALL the banjo screens removed if we are anywhere near them or during a turbo swap. 

 

GD

Edited by GeneralDisorder
Link to comment
Share on other sites

Who are they getting these CHRA's from? We only use Melett CHRA's when we do rebuilds. And I have not had a single one come back. 

 

It is possible that one of the housings is impinging upon the turbine or compressor wheel at speed? That is - the natural radial shaft play inherent to journal bearing turbochargers is allowing one of the wheels to graze the housing once it spins up.

 

I would get a CHRA and do the "rebuild" yourself. It's easy - takes about an hour. I do about 2-3 a month. Depending on the condition of the housings, you may want to swap those out for a different used set at the same time. I have a whole shelf full of dead TD04's if you need housings. I probably have 20 or more. We almost always just upgrade those to VF's or 16G's.

 

ALL turbo replacement's due to failure get an oil pan pull. Typically we just replace the pan with a new STi pan, Killer-B pickup, and matching dipstick. That 04 is going to have the inferior flat-bottom pan. 

 

Also ALL turbo engine have ALL the banjo screens removed if we are anywhere near them or during a turbo swap. 

 

GD

 

good to know

Link to comment
Share on other sites

  • 2 weeks later...

Update on the engine.

 

        Pulled the 2, 12mm hex plugs from bottom of oil pump, inspected for obsructions, there were none visible, clear up to the inlet and outlet of the rotor pump rotor.   After that, and reassembling the oil pan and cooler, we again ran the engine (with new oil an filter) without the turbo (yes it's loud with no exhaust)  and tested pressure in the oil sender port and again at the oil feed line to the turbo.  both gave good readings (85+ psi at idle cold, 50+ at idle warm, raising with RPMS)

 

        Every test I have done, and every condition observed point to a good running, well oiled engine.

 

SOOOO.............

So the rebuild shop, Bell Turbo, in Albany Oregon, has sent the 2 failed turbo assemblies back to the supplier they bought them from.  That supplier is refusing to warranty either.  They state "lack of oil"

 

       This is hogwash.  The original turbo had wear and was leaking oil into the exhaust causeing a huge cloud.  That was the whole reason we took it out to replace.  The car actually ran fairly OK still.  So it WAS getting oil.....lots of it. And that turbo had tons of nasty ground up wear on it's bearings, but had not lost it's shaft nut.

 

        Both of the replacement turbos had in them when removed.  And the one that I personally saw disassembled (2nd replacement) had virtually no wear on the shaft were the radial bearings ride.  The owner of the turbo shop rubbed his fingers together feeling the oil and said "o yeah, that's good synthetic oil and it's not coked up"

 

 

       Here is my theory.  The shop buys a balanced assembly.  one that has been torqued down, and balanced by grinding the nut/shaft in one spot.  The shop then has to disassemble the shaft, and install it through the housings.  Then when they reassemble, they don't go by torque, they just turn the nut back to where the balance "notch" lines up with where it had been to maintain balance.  

 

      Additionally, I've been reading that many small turbos need to be heated to 275 degrees before assembly, torqued within 5 mins to spec, and then allowed to cool.  this way it shrinks up even tighter, and won't be likely to loosen when heated during use.  I do not believe the shop did that procedure.  They are kinda done giving us answers or further information (probably trying to limit evidence in small claims court).  BUT......in the end......

 

 

 

The shop owner has agreed to refund our money, and return our OE core parts.  While I am upset at the wasted time and failed parts, I must commend Bell Turbo for at least "making it right" and returning our money so we can move forward with a turbo from a different source.

 

We are purchasing a genuine Subaru reman from the local dealership.

 

       I will be pulling the AVCS control solenoids and associated parts to inspect and reseal, but I can't imagine that would have any part of the failures.  

 

       I will also be verifying the bypass valve opening, and will have a boost gauge on the turbo to watch for overboosting (or boost leaks) once we get it up and ready to run again.  Again, everything in that regard seemed to be okay, no codes, no driveability issues or backfires.....but I want to be able to check every possibility.

 

Stay tune for the next episode of "as the compressor wheel turns"

  • Like 2
Link to comment
Share on other sites

At least they are willing to refund. Kudos to them for that.

 

TBH, taking a shop to court is a serious headache, and their attorney will drag matters out. I was out $1k after a shop spent months trying to track an electrical issue, and kept replacing unneeded parts w/o actually fixing it. Spent another $400 at Ford and they discovered the issue in a couple hours as being a crank sensor loom pinched in a timing cover, and rubbing on a pulley (hard to see as the cover hid everything; 92' SHO; was working 90 hours a week at the time so unable figure it out on my own).

 

Took the 1st shop to court, claiming they not only didn't fix the initial issue, they replaced parts that were unneeded. Unfortunately, I got this irritable female judge that knew ZERO about cars, and ruled against me. Was beyond irate as $1k was wasted.

 

GL with the new turbo. Be sure to fill it and hand spin the shaft beforehand to be 100% certain it's lubricated. Have a feeling your shop overlooked something and not your fault.

Link to comment
Share on other sites

Here is my theory.  The shop buys a balanced assembly.  one that has been torqued down, and balanced by grinding the nut/shaft in one spot.  The shop then has to disassemble the shaft, and install it through the housings.  Then when they reassemble, they don't go by torque, they just turn the nut back to where the balance "notch" lines up with where it had been to maintain balance.

 

The CHRA does not require ANY disassembly to be installed into the housings. That shaft nut is torqued at the factory and is never touched till end of turbo life. 

 

That is 100% definitely not the correct theory.

 

Take one apart - then you will understand.

 

I have been through HELL with turbochargers. And there are literally 1000's of reasons I have the policies I have. As in I've lost 1000's of $ down this hole. Turbo replacements, engine rebuilds, etc. 

 

We install Melett CHRA's, OEM Subaru turbo complete assemblies (new or at times "good used"), and for aftermarket turbo's we buy Blouch or Garret ONLY. If you want something else - buy it yourself and I will happily install it. Then when it fails I will shrug and say "Your problem - told you not to buy that!". 

 

Everything else we have ever installed has a 90%+ failure rate. 

 

With my policies..... I  can't remember the last time I have had to warranty a turbo related failure. Been several years. 

 

I would say there is a 99% chance they are using junk CHRA's. You can do EVERYTHING right and if the turbo is $hit, then the results will be $hit.

 

GD

Edited by GeneralDisorder
Link to comment
Share on other sites

The CHRA does not require ANY disassembly to be installed into the housings. That shaft nut is torqued at the factory and is never touched till end of turbo life. 

 

That is 100% definitely not the correct theory.

 

Take one apart - then you will understand.

 

Some shops can balance the shaft w/o the actual housing. They'll then ship the hot shaft with the cold wheel bolted down and marked at balance. Then, you have to remove the nut, reinsert through the actual central housing, and try and realign, though there might be an issue of getting the nut exactly back where it was while remaining tight, w/o over-tightening.

 

Think Gloyale was referring to them getting a balanced shaft and reusing an on-hand central housing, and not being 100% correct with the torque value?

Edited by Bushwick
Link to comment
Share on other sites

idosubaru, on 23 Feb 2018 - 04:54, said:

idosubaru, on 23 Feb 2018 - 04:54, said:

Wow. crazy series of events,

Great troubleshooting. frustrating when there’s a cause and people are just blindly regurgitating some idea that’s not it.

 

Are you nervous with the third attempt?!

 

 I don`t know how you can be so dimissive of the opinions of turbo builders in general and this one specificly

without any supporting evidence.

They see 1000s of turbos and are well aquainted with failure modes,after all.

Not everything is someone elses fault(or anybodys).

 

Good oil pressure with the turbo fitting deadheaded does not mean you have good oil pressure with the

turbo installed and flow is required.

 

A restricted (sludge was mentioned) oil feed line with only a pinhole passage would show good pressure

with the fitting deadheaded,but,poor pressure with the turbo installed and flow required.

To make an electrical analogy,it is like measuring good voltage on a low capacity battery with no load,but,

poor voltage with a load applied.

 

At least one turbo builder wants you to replace the feed lines,never mind clean them.

 

I would modify a banjo bolt so that I could monitor oil pressure at the turbo with everything installed.

Small price to pay considering.

 

Unless i missed something,overboost or a intake backfire would tighten the nut of a reverse threaded shaft.

They only loosen the nut on a straight threaded shaft.

I suppose an exhaust backfire might loosen a reverse threaded nut.

Edited by naru
Link to comment
Share on other sites

Some shops can balance the shaft w/o the actual housing. They'll then ship the hot shaft with the cold wheel bolted down and marked at balance. Then, you have to remove the nut, reinsert through the actual central housing, and try and realign, though there might be an issue of getting the nut exactly back where it was while remaining tight, w/o over-tightening.

 

Think Gloyale was referring to them getting a balanced shaft and reusing an on-hand central housing, and not being 100% correct with the torque value?

 

None of that applies to Subaru style turbo's. CHRA's are a bolt-in with the housings. Using bolted tabs/washers or snap-ring on the cold side, and v-band clamp on the hot side. ALL stock location Subaru turbo's are built this way. 

 

GD

Link to comment
Share on other sites

 I don`t know how you can be so dimissive of the opinions of turbo builders in general and this one specificly

without any supporting evidence.

They see 1000s of turbos and are well aquainted with failure modes,after all.

Not everything is someone elses fault(or anybodys).

 

Good oil pressure with the turbo fitting deadheaded does not mean you have good oil pressure with the

turbo installed and flow is required.

 

A restricted (sludge was mentioned) oil feed line with only a pinhole passage would show good pressure

with the fitting deadheaded,but,poor pressure with the turbo installed and flow required.

To make an electrical analogy,it is like measuring good voltage on a low capacity battery with no load,but,

poor voltage with a load applied.

 

At least one turbo builder wants you to relace the feed lines,never mind clean them.

 

I would modify a banjo bolt so that I could monitor oil pressure at the turbo with everything installed.

Small price to pay considering.

 

While that is all correct, I have never seen a clogged feed line. The banjo bolt screens typically hold all the filth. Not that it couldn't happen, and we do regularly swap to better feed line systems on certain models (05-07 LGT platform, or larger turbo installations), but we don't typically see that. We do wash out the lines with solvent if being reused and insure that the orifices in the banjo fittings and bolts are clear. 

 

I don't even measure pressure at the turbo feed. We pre-lube the turbo, fire up the engine, and crack open the banjo fitting to insure we are getting good flow. After you do a few of these and know what good flow should look like at idle - that's really all the more testing we find necessary.

 

We replace or install upgraded turbochargers on Subaru's at the rate of 3-4 a week and haven't had a failure in years. 

 

GD

Link to comment
Share on other sites

 I don`t know how you can be so dimissive of the opinions of turbo builders in general and this one specificly

without any supporting evidence.

They see 1000s of turbos and are well aquainted with failure modes,after all.

Not everything is someone elses fault(or anybodys).

.

 

true - I shouldn't have added "that’s not it ", i should have said "just blindly regurgitating some ideas".

 

i'm dismissive of "regurgitating" *some* failure modes over a more complete list of possibilities and "blindly" not explaining or giving data to chew on when someone is doing due diligence.

 

Maybe they said, "Ah man, we've seen this 100 times, every backyard guy slapping these on with no testing has this happen, seriously replace the feed line, it happens 50 percent of the time"....but i'm guessing they didn't and just fired back some guesses.  i could be wrong, maybe Gloyale didn't tell us the whole story. 

 

granted some things are hard to quantify.  and as well, i do get it - it's tough trusting someone else's diagnosis and two consecutive failures is a serious red flag that no one should ignore.  it seems they would give more than guesses, and maybe they did...

Link to comment
Share on other sites

 i could be wrong, maybe Gloyale didn't tell us the whole story. 

 

I have given you absolutely every detail I can without describing the hefty spoob I took while we warmed it up before the second turbos test run.  

 

For the record, the feed lines are clear and flowing.  Sadly. They are just tiny and restricted to begin with.

 

Here's a video (hopefully) 

 

This is the hose I had attached to the turbo line with a doubled up banjo fitting to put the gauge on.  This is without the gauge just to show the flow of oil at idle.  No exhaust.  It's loud.  make sure your sound isn't too far up.

 

39551738985_569e93f5c3_b.jpgIMG_2864 by Dans Subaru, on Flickr

Link to comment
Share on other sites

Yeah - as I have said - the car, and your work isn't the problem. 

 

It's a garbage CHRA (probably from China) problem.

 

I see this frequently. If it's not one of the turbo's or CHRA's I listed above, there's about a 90% chance it will f*ck off within a short period if not immediately. Been there, done that. BNR, Forced Performance, Kinugawa, random eBay junk.... seen them all fail in short order. And then had the supplier argue with me that it's not the turbo's fault. In EVERY case we swap it with stock or Blouch and the problems are gone.

 

As I said there's reasons for the policies I have. Good reasons. Reasons backed with painful experience that cost a lot to acquire. 

 

GD

Link to comment
Share on other sites

Yeah - as I have said - the car, and your work isn't the problem. 

 

It's a garbage CHRA (probably from China) problem.

 

I see this frequently. If it's not one of the turbo's or CHRA's I listed above, there's about a 90% chance it will f*ck off within a short period if not immediately. Been there, done that. BNR, Forced Performance, Kinugawa, random eBay junk.... seen them all fail in short order. And then had the supplier argue with me that it's not the turbo's fault. In EVERY case we swap it with stock or Blouch and the problems are gone.

 

Yeah, First they said that the shaft assembly was from Europe.  But then later they were talking about "we shippped both failed units back to Ohio" so who the F%#$ knows where they were actually getting their parts.  They claimed "their cost" on the rotating assembly was $430.....?  And I am really not sure they were getting CHRA, maybe just the shaft and compressor wheel.  The core of each turbo they provided looked like our original rusty cast core, just freshly sandblasted.  Not a fresh cast gray.

 

 

 

SO again.......I would ask for all to brainstorm.  I am particularly wondering about testing Bypass valve, and TGV (tumble generator valve) and AVCS (Active Valve Control System) operation.

 

As I said, no codes, no audible backfire, no signs of intake air leak, or of faults in the TGV or AVCS (no codes)

 

Although there seems to be no problems with those systems, is there any wild chance an overlooked fault with these could contribute here to the turbo failures?  

Edited by Gloyale
Link to comment
Share on other sites

I get my Melett CHRA's directly from a wholesaler in the UK. The TD04 units are $225 shipped to me from the UK. About 5 days to my door. Takes a couple hours, I would say, to swap it into the housings if you are being careful. Sometimes the exhaust housing likes to be rude about it. Requires persuasion with my induction heater. I charge $600 retail for full rebuild including all parts, labor, and usually I'll swap out the housings if they are wrecked and I have better used ones on hand.

 

They are not remanufactured, they are brand new. They come with copper crush washers, oil return pipe gasket, and a syringe filled with oil to pre-lube them.

 

If you had a problem with the TGV system it would throw a code. What's an intake runner control flap going to do to a turbo anyway?

 

Likewise, the AVCS system would throw codes if it had insufficient oil pressure, etc.

 

The bypass valve.... well that could be rough on a turbo if it malfunctioned. But its unlikely to kill a TD04 in the short term. In any case I have never seen a factory BPV fail.

 

This will end up being a junk turbo rebuild situation. Mark my words. Happens every time.

 

GD

Link to comment
Share on other sites

I get my Melett CHRA's directly from a wholesaler in the UK. The TD04 units are $225 shipped to me from the UK. About 5 days to my door. Takes a couple hours, I would say, to swap it into the housings if you are being careful. Sometimes the exhaust housing likes to be rude about it. Requires persuasion with my induction heater. I charge $600 retail for full rebuild including all parts, labor, and usually I'll swap out the housings if they are wrecked and I have better used ones on hand.

 

They are not remanufactured, they are brand new. They come with copper crush washers, oil return pipe gasket, and a syringe filled with oil to pre-lube them.

 

If you had a problem with the TGV system it would throw a code. What's an intake runner control flap going to do to a turbo anyway?

 

Likewise, the AVCS system would throw codes if it had insufficient oil pressure, etc.

 

The bypass valve.... well that could be rough on a turbo if it malfunctioned. But its unlikely to kill a TD04 in the short term. In any case I have never seen a factory BPV fail.

 

This will end up being a junk turbo rebuild situation. Mark my words. Happens every time.

 

GD

 

CHRA,:  Yeah...looking back I should have just ordered a CHRA and swapped myself.  But I am busy and turbos aren't my specialty so I thought it better practice to go to a professional shop.

 

TGV,:  My thoughts too.  It's a stretch, but I was pondering if ALL 4 valves closed at the wrong times, during full boost......could it have similar effect to failed a backfire or restriction?  But then Bypass would blow so it would have to be failed too for in issue to occur.  And tehy n't fully block the intake, and I would see a code for 4 failed TGVs

 

AVLS....  This is my only real dubious system still so far.  It is directly connected to the oil feed system.  Something in the system opening up so much that all the oil dumps and the feed line loses too much flow???

 

I think it's pretty dumb that the turbo and the pass side AVCS are fed out of the same banjo bolt.  Maybe I should build a new feed line for the AVCS that taps the port on top of the block or one of the plugs on top of the head?...I think there is a plug on top of the 04 heads?  IDK, It's the weekend and I am at home not at car.  I will find out soon enough.  The AVCS system is getting opened and cleaned.

 

For the bypass...., It's an automatic car, so there isn't the sudden closing of throttle between shifts.  I guess doing full throttle runs then suddenly slowing as you reach the limits of public roads?  If the bypass didn't open, then that might be an issue.  Seems unlikely but I will look into it for sure

Link to comment
Share on other sites

We routinely re-route the oil feed for the turbo from the rear gallery plug on the engine block. Leaving the head to supply the AVCS only. We use braided stainless -3 AN for both the turbo and the AVCS feed. We always leave the original banjo bolt in the turbo CHRA for the feed restriction - it is true that too large of a feed orifice can cause oil to blow by the shaft seals - mostly this is a problem for the IHI turbo's. The Mitsubishi based turbo's don't seem to have this problem - we often feed the journal bearing Blouch turbo's with completely unrestricted oil flow. Ball bearing turbo's use feed restrictors (usually .030") due to not requiring much oil and too much will actually create bearing drag and again with the excessive shaft seal pressure which the oil now has unrestricted access to through the voids in the ball bearings.

 

After this mess - I would re-route the feed lines if it were in my shop. 

 

Also - has the pickup tube been checked for cracks? Aeration of the oil at higher pump volume is something I wouldn't discount. Broken 255/257 pickup tubes start with just a crack. They progress from there till ultimately they fall off. 

 

GD

Link to comment
Share on other sites

We routinely re-route the oil feed for the turbo from the rear gallery plug on the engine block. Leaving the head to supply the AVCS only. We use braided stainless -3 AN for both the turbo and the AVCS feed. We always leave the original banjo bolt in the turbo CHRA for the feed restriction - it is true that too large of a feed orifice can cause oil to blow by the shaft seals - mostly this is a problem for the IHI turbo's. The Mitsubishi based turbo's don't seem to have this problem - we often feed the journal bearing Blouch turbo's with completely unrestricted oil flow. Ball bearing turbo's use feed restrictors (usually .030") due to not requiring much oil and too much will actually create bearing drag and again with the excessive shaft seal pressure which the oil now has unrestricted access to through the voids in the ball bearings.

 

After this mess - I would re-route the feed lines if it were in my shop. 

 

Also - has the pickup tube been checked for cracks? Aeration of the oil at higher pump volume is something I wouldn't discount. Broken 255/257 pickup tubes start with just a crack. They progress from there till ultimately they fall off. 

 

GD

 

I am strongly considering re-routing the oil feed.

 

I inspected the pickup tube for cracks. Didn't see any, although I did not put it under pressure to check for hairlines, but it sure seemed OK.  Could maybe have been the o-ring?  it wasn't very soft anymore, but wasn't awful or cracked either.  Meh?

Link to comment
Share on other sites

Since a new pickup tube from Subaru is like $35 or something silly, we install a new one when the pan is off even if the customer declines a Killer-B. I'll eat the cost if they complain too loudly. Basically I don't want anything in there being blamed on me cause I touched it last. I lay down the law with customers when it comes to blown turbo and pickup tube issues - either do it my way or take it elsewhere. They usually either go with my suggestions or end up back here with another blown/cracked/messed up engine after they took it elsewhere or attempted it themselves under the shade tree. No one ever wants to pull the pan - they always grumble about it. It's not optional with turbo failures. You just don't know - the turbo could be the canary that's warning you about impending pickup failure. 

 

GD

Edited by GeneralDisorder
  • Like 1
Link to comment
Share on other sites

  • 3 weeks later...

Update:

 

Subaru Genuine OE reman turbo installed.

 

Car running perfectly.  No smoke.  Quick, smooth spool.  Pulls hard to redline with no issues from turbo.

 

Now, this is just after a few short miles of testing.  Need to get tags on it to take it for a longer highway drive.

 

Don't buy cheap turbo parts.  Just go OE from the start.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

 Share

×
×
  • Create New...