Vanagon AAZ runaway

[chris24g]

Quite a few months ago, I was just cruising on the freeway when my vanagon started accelerating. I thought it was odd, maybe a sticky throttle cable. I soon figured out it was starting to runaway when I disengaged the clutch. I pulled over in 4th and stalled out the motor. The motor was making some crazy sounds and smoke until it died, so I towed it home and didn't even touch it because it was done.

Last week I finally mustered the emotional strength to check it out. Pulled the boost elbow from the turbo and turned the motor over to see how bad the motor was. The mother f##ker started 

Of course,  turbo was spewing oil but that was what I suspected.

I've got a spare turbo but now am thinking about changing the exhaust manifold. People say to use the quantum mani but I'm not sure if it would be any better. Did my AAZ manifold cause the problem somehow? The motor was running fine for over a year. Not sure how the other manifold would be any better...

[libbydiesel]

Assuming you are using the AAZ manifolds, then no, they aren't causing the problem.  The oil that your engine is burning is likely coming from either the turbo seals or crank vent from excessive blowby.  What's your turbo oil return like?  Stock AAZ?

[chris24g]

Yes basically a complete AAZ except for the oil pan, which came off of the old 1.6D motor. Is there any advantage to using the older JX style manifold?

[libbydiesel]

No. 

[burn_your_money]

You have the turbo clocked correctly?

[libbydiesel]

I doubt the turbo is clocked correctly if he is using the stock AAZ oil return line in a vanagon which is tipped over an extra 35°.  The braided stainless doesn't stretch that far.  I would guess he has either a T2 or a K03 as the K14 takes some significant mount bracket modification to fit.  If I had to hazard a guess I would say maybe the center section is all coked up from having the large puddle sitting there to simmer after shutdown.  Eventually the carbon mess got large enough to plug the oil return and now oil is blasting out the shaft seals enough to runaway.

[chris24g]

The oil drain tube was not perpendicular to the ground. I do try to let the turbo cool down if its been under load. I will check out the drain port on the turbo for evidence of colking. Rotating the turbo is a good idea. Does anyone make a longer oil return for this fitment or do I have to diy something with push loc hose AN fittings? What size are the fittings?

[libbydiesel]

Which turbo do you have?  I'm not aware of any ready-made oil return line options other than using the K14, JX turbo and accompanying manifolds and intake.  I would not recommend going that route.

[chris24g]

i'm using the stock k14 turbo. i don't see how i can easily reclock the turbo though, since it has a hard line from the compressor to the wastegate. i looked at the oil return ports and it is pretty clean, no carbonized oil to be seen. what to do...

[libbydiesel]

For the K14, you'll need to bend the supply which is easy.  The drain side is harder.  I wonder if Gizmoman still has his k14 setup.  You might check with him.

[Gizmoman]

First I tried this


I believe the abrupt 90 degree turn and forcing the oil to drain through the tiny holes of the banjo was too restrictive - still oil in compressor.
I finally fabricated a swept 90 degree bend from 1/2" hydraulic tubing - oddly I can't find any photos of it but it was a bear to make.

I ended up finding that I also had a restriction from the PO who had done a poor job getting into the oil pan gracefully. Whatever you do, keep ALL your ID's greater than 3/8". No droops in the line, and all bends as smooth as possible (not sharp). Never really did solve it but the K-14 didn't produce enough CFM for my application anyway so I went to the Holset HE200 - a  different story.

All said, I never ran the K-14 with the pan restriction removed. That fix may have been all I needed even with the banjo - I'll never really know.

[libbydiesel]

I do not try to drain to the block as the fitting is, IMO, too high to get proper drainage by the time the turbo is clocked properly.  The solution I have used on (4) K14s into vanagons is to remove the existing drain pipe, get an NPT to 1/2" barb 45° fitting that is the closest thread to the drain threads in the center section (I think the threads are 3/8 npt but I could be wrong), clamp the center section in a vice so that the hole is facing down, grease up the tap and make new NPT threads.  You need to remove the compressor scroll and the exhaust housing but the 45° fitting will thread into place without removing the compressor wheel and backing plate.  I angled the return fitting to go around the back of the wastegate and rotated the wategate 120° CW so the fitting on the wastegate was out of the way.  Remove the wastegate fitting standoff so that the wastegate bolt does not collide with the exhaust housing when assembled and then rebend the metal wastegate line so there is not any interference.  Done that way allows the drain to be at exactly 6:00 with decent ID.  The 45° brass fittings are available at mcmaster.  Best to have the fitting in hand to set the depth of the threads correctly.  For the return line itself, after using various AN and hydraulic fittings I have gotten fed up with them being weak, too small ID for their bulky OD, prone to leaking, etc... and now use either braided stainless or multipurpose hydraulic hose and clamp them onto barbs.  I also like to use a 1/2" OD straight metal tube for the majority of the return so that it is impervious to exhaust heat and maintains that nice straight downward slope.  The finished line does not look as 'slick' but is functionally far superior to the aluminum AN fittings that are commonly used.





 

[chris24g]

thanks libby and gizmo for the valuable advice! i was staring at the turbo and scratching my head for a long while. i didn't realize the extreme angle of aaz would be such an issue with turbo applications.

[chris24g]

i took the turbo apart. is there a way to test the turbo cartridge for leaks? if i plug the return and fill up the turbo will it show the leak or does it need pressure? the compressor housing had a puddle of oil in it, but could it have come from the breather?

seems to me the best solution would be to relocate the wastegate so the oil can have a straight shot down before bending. the wastegate bolt pattern doesn't look standard to me, though.

*edit*

scratch that idea. internal wastegate. the npt method looks to be the only reasonable one outside of going to a completely new turbo system.

[Gizmoman]

You'd be wise to take libby's advice 100%

Good luch Chris