1 2
Keith Tanner
Keith Tanner MegaDork
3/14/21 7:17 p.m.

The 2.2 Subaru engine in my Vanagon committed suicide (as Subaru engines do) in the middle of Utah a few years back. I took the opportunity to build a "frankenmotor", a 2.5 bottom end with 2.2 heads. High compression, loads of tork, perfect for my baby motorhome. Because Subarus are Subarus, the donor engine I tore apart was basically junk and I ended up replacing everything but the block. I did put my existing 2.2 heads on after getting them cleaned up. Took me forever to do this partly because my local machine shop is the slowest machine shop in the world.

The new motor feels great. The van will cruise at 80 mph and has acceleration to spare, even with the power loss from 5000' elevation. Awesome. It went on its inaugural trip out of town last weekend and was a rock star. Better fuel economy and it was a rocket on the interstate. Except for an oil leak. Something didn't seal when I put the new engine together - and it got worse as the trip went on, so it wasn't something I could ignore.

Drove the van on to the lift yesterday and disassembled what I needed to disassemble to get to the front of the engine. Well, front in the traditional Subaru sense, it's the back of the engine when it's in the Vanagon. But you get the idea.

Found a cut front crank seal and a cam seal that looked suspicious. Okay, I will accept installer error on that seal. Sourced new ones (yay for high volume cars with very few changes over time!) at Autozone and put them in. I have the factory manual and I'm not that familiar with Subaru engines so I very carefully followed the instructions on how to time the thing. Marks on cams, mark on the crank. Checked myself over and over. Turned it over by hand a few times, felt decent although I was obviously fighting compression. Buttoned it up.

Went to start it and it had a bit of a stutter when cranking. I blamed that on a poorly installed ground cable and tried again. Nothing. It has a MAF signal and an RPM signal via OBD-II, so it's not a missing sensor. I couldn't figure it out.

I used the TDC indicator on the crank pulley, not the timing belt indicator. Yes, the cute little arrow on the front of the crank pulley is not the one you're supposed to use. And these are interference engines.

Sigh. I'll bet that stutter was my starter motor helpfully bending my valves for me. At the very least, I'll be digging back in to the front of the engine to time it right, which will be a few hours. Who wants to put money on what a compression test will say?

Slippery (Forum Supporter)
Slippery (Forum Supporter) UberDork
3/14/21 7:24 p.m.
Keith Tanner said:

...I'm not that familiar with Subaru engines so I very carefully followed the instructions ...

Look at the bright side ... now you are going to be very familiar with Subaru engines 

Mr_Asa
Mr_Asa UltraDork
3/14/21 7:30 p.m.

Ugh.  My condolences.

Still, its a mistake you (hopefully) only make once

barefootskater (Shaun)
barefootskater (Shaun) PowerDork
3/14/21 7:51 p.m.

Timing an engine is stressful. Hopefully the damage isn't too extensive. 

Aspen
Aspen HalfDork
3/14/21 7:54 p.m.

Reminds me of the time my Subaru rolled a few feet down my driveway engine off in gear.  The timing belt jumped 3 teeth, bent one valve. 

That was a 2l version 3, before they added a guide above the crank pulley to stop that exact thing from happening.  Fun times.

You might get lucky with the valves, I think the 2.2s are only interference between valves if the belt breaks, but the 2.5 bottom may bring the pistons into play.

Pete. (l33t FS)
Pete. (l33t FS) MegaDork
3/14/21 8:06 p.m.

In reply to Aspen :

Huh.  I knew the guide was only on manual transmission engines, but all this time I'd been assuming that it was there because manual transmission engines don't have a stonking big fluid filled damper on the end, and between that and the irregularity of the cam belt loads, the belt gets shocked and can jump time.

Never considered the now obvious possibility that there's nothing about an automatic that can rotate the engine backwards, but a manual trans in gear certainly can, and the belt tensioner would happily allow the belt to go loose.

Keith Tanner
Keith Tanner MegaDork
3/14/21 8:07 p.m.

It wouldn't bother me so much if I hadn't been so careful. The manual says nothing about there being a secret second timing point besides the arrow on the pulley - although I'd already figured that out on the cam gears, so maybe that should have been a clue. And I should have remembered from the last time I put this thing together, although I had it on the bench that time.

Thanks for the hope, Aspen. I really should be down in the shop right now tearing into this but I don't want to. I am taking the evening off. Let's hope I didn't screw up the valves, because we were going to go camping next weekend.

alfadriver (Forum Supporter)
alfadriver (Forum Supporter) MegaDork
3/14/21 8:09 p.m.

In reply to Keith Tanner :

Man, if I did that, I would not be able to sleep- thinking of what I did wrong followed by an entire night worrying about what I need to do.

Hope it goes well.

Robbie (Forum Supporter)
Robbie (Forum Supporter) MegaDork
3/14/21 8:53 p.m.
alfadriver (Forum Supporter) said:

In reply to Keith Tanner :

Man, if I did that, I would not be able to sleep- thinking of what I did wrong followed by an entire night worrying about what I need to do.

Hope it goes well.

Once or twice, while in this state, I fixed the car in my dream and turned out that is what I needed to do to fix it all along.

But man does it suck before you go to sleep.

Keith Tanner
Keith Tanner MegaDork
3/14/21 9:02 p.m.

Good news everyone!

It turns out that 2.2 SOHC engines with twin exhaust ports are the last of the non-interference Scooby engines. My heads are from 1996, the last year of that setup. Hopefully the 2.5 bottom end will not affect that. The fact that I was able to turn it over several times with nothing more than the bouncy feel of compression is also reassuring. I'd like to think I would have felt valve-piston interference.

So there is some hope.

I think the reason I missed the secret mark is because I was working from below. When the engine was on the bench, I could see the top of the crank pulley and I would have easily seen the secret mark. But not from underneath, which is a feature of the VW installation.

Keith Tanner
Keith Tanner MegaDork
3/14/21 9:08 p.m.
Slippery (Forum Supporter) said:
Keith Tanner said:

...I'm not that familiar with Subaru engines so I very carefully followed the instructions ...

Look at the bright side ... now you are going to be very familiar with Subaru engines 

This comment is karma for me being a smartass in the Land Rover thread.

Pete. (l33t FS)
Pete. (l33t FS) MegaDork
3/14/21 9:55 p.m.
Keith Tanner said:
Slippery (Forum Supporter) said:
Keith Tanner said:

...I'm not that familiar with Subaru engines so I very carefully followed the instructions ...

Look at the bright side ... now you are going to be very familiar with Subaru engines 

This comment is karma for me being a smartass in the Land Rover thread.

LOL.

I'd been thinking, if you did bend the valves, there's a good chance that the seats are still in good shape and you can just pop new ones in and be golden, no waiting for a machine shop.  You'd just be out the cost of eight valves, gaskets, and valve stem seals.  Here's to hoping that the valve reliefs are deep enough for the valves.

A word of caution, though.  I've had slightly bent valves seal okay in an engine due to spring tension straightening them against the seat.  I've had only one engine pop a head off from weakening (a 4l Chrysler), but it's always a possibility.  Can you get a borescope in the plug holes and look for evidence on the piston tops?  I mean, you'd have to get the timing roughly 90 degrees off at the crank for it to be a possibility if the engine is only slightly possibly an interference engine...

Kreb (Forum Supporter)
Kreb (Forum Supporter) UberDork
3/14/21 11:19 p.m.

One question: How come practically everyone puts 4 cylinder motors in the VW busses? My understanding is that the EZ30 is only a couple inches longer and puts out a lot more torque.

Keith Tanner
Keith Tanner MegaDork
3/14/21 11:30 p.m.
Kreb (Forum Supporter) said:

One question: How come practically everyone puts 4 cylinder motors in the VW busses? My understanding is that the EZ30 is only a couple inches longer and puts out a lot more torque.

Because that's what it had when I bought it :)

There are 6 cylinder ones out there, but I think sheer numbers make the 4 the first choice. Plus it can be legal in CA. 

Streetwiseguy
Streetwiseguy MegaDork
3/15/21 7:45 a.m.

I'm just going to state the bleedin obvious here for others, to help avoid this situation.

You should always roll an engine over to get everything in time, and make sure you understand the timing marks before you take  it apart.

I state this  from personal experience, including a Subaru with the fake timing marks once...

Keith Tanner
Keith Tanner MegaDork
3/15/21 9:45 a.m.
Streetwiseguy said:

I'm just going to state the bleedin obvious here for others, to help avoid this situation.

You should always roll an engine over to get everything in time, and make sure you understand the timing marks before you take  it apart.

I state this  from personal experience, including a Subaru with the fake timing marks once...

I did roll it over about three times. My mistake was not realizing that the cast-in arrow on the crank pulley was not a timing mark and that the factory manual does not provide this information. In fact, the manual (at least the portion I was using) seems more interested in making sure you can properly put Sharpie marks on the belt than anything else. I was using the timing belt R&R instructions, it's possible there's another set in the manual that is written for when you're building up an engine from scratch.

Again, I think the root cause was the fact that I was addressing the engine from below instead of above which is an artifact of the engine swap. From what I remember, the secret mark is much more obvious (ie, actually visible) from above. On the cam gears, the real timing marks are on the front of the gears but on the crank they're not.

I'll see how quickly I can rectify this tonight. 

Ian F (Forum Supporter)
Ian F (Forum Supporter) MegaDork
3/15/21 9:54 a.m.

I find these interesting...  but mostly because my only reference to working on interference engines is my ALH TDI which interfere quite a lot so there are a lot of tools for locking them down at TDC when doing service operations (timing belt changes). The general consensus is to run away from any TDI with paint-marks on the pulleys, indicating someone did a TB change using the "mark and pray" method, which sometimes works... and sometimes doesn't. 

Keith Tanner
Keith Tanner MegaDork
3/15/21 9:08 p.m.

Three cheers for happy endings! 90 minutes to tear everything down, retime the engine and fully reassemble. The manual compression check (aka turning over by hand) felt good so I just buttoned her up and hit the key. 

Purrs like a kitten. Nothin' wrong with that boxer. Whew!

Here is the offending pulley. That arrow on the top face (at about 8 o'clock) is what I used. The secret mark is at the lower right on the trigger wheel, so I was 90 degrees off. It is now blatantly obvious thanks to some sharpies. 

alfadriver (Forum Supporter)
alfadriver (Forum Supporter) MegaDork
3/16/21 7:37 a.m.

So if the valves didn't hit the pistons, what prevented the engine from spinning?  

And given the error, that lack of spinning was an incredibly good thing.  Easy fix, as I'd bet the path you ended up doing would have been later down the debug list.

Paul_VR6 (Forum Supporter)
Paul_VR6 (Forum Supporter) SuperDork
3/16/21 7:51 a.m.

Any mark on a pulley that isn't an obvious and correct timing mark can go berkeley right off. Glad this turned out well.

Keith Tanner
Keith Tanner MegaDork
3/16/21 8:05 a.m.

I wasn't clear in my first post - on re-reading, I see that. It never locked up. It had that little stutter the first time I cranked, which actually isn't unusual when I think about how this thing often starts. I reseated and tightened the negative terminal and it cranked easily and quickly. It just wouldn't fire (that's what I meant by "nothing"), it wasn't even trying to run.

It was acting like there was no ignition but my rule is to always go back to what you just did. Once I'd confirmed that there was an RPM signal, I knew it wasn't a disconnected/misaligned sensor so it had to be related to the timing belt. Took me some googling of pictures to discover the second mark.

Interestingly, the cams have an arrow on them that is incorrect - but looking at the geometry, it may be possible to time this thing correctly if you use the arrows on the cam and the TDC indicator. I'm not going to try it, but it's close.

The first time I went to start it after the retime it didn't even do a full rotation before it fired up.

Slippery (Forum Supporter)
Slippery (Forum Supporter) UberDork
3/16/21 8:18 a.m.

Awesome news!

Thats the good thing about spinning by hand, not only you check that the marks still line up, but you also can feel if a valve and piston are touching. 

I bet next time you can do it without the manual cheeky

Keith Tanner
Keith Tanner MegaDork
3/16/21 8:31 a.m.

I didn't need the manual to fix the problem, that's for sure!

But I am going to limit my future exposure to Subaru engines if at all possible. 

DeadSkunk  (Warren)
DeadSkunk (Warren) PowerDork
3/16/21 8:35 a.m.
Slippery (Forum Supporter) said:
Keith Tanner said:

...I'm not that familiar with Subaru engines so I very carefully followed the instructions ...

Look at the bright side ... now you are going to be very familiar with Subaru engines 

To quote that brilliant "philosopher" David Feherty (I find him hilarious) "Experience is what you get when you don't get what you want."

FMB42
FMB42 Reader
3/16/21 8:47 a.m.

Good news! We could all use more GN these days. Meanwhile, I (many years ago) learned to always remove the spark plug or plugs whenever turning the engine over by hand after cam timing, etc, work. However, Suba S plugs are not always the easiest to get at (took me ~ 4 hours to change the S plugs in my wife's 2012 Forester).

1 2

You'll need to log in to post.

Our Preferred Partners
9on4SPCLesLocnT2dl0RCK4yVUBLW7hSEp9s5DIQVxVCQDtSizp700UtQLjZQYb2