EJ non-turbo burned exhaust valves - consensus?

[GeneralDisorder]

*edit - pictures on page 4*

 

Just bought a '97 OBW with 169k on it. Pretty ugly misfire on #4. Pulled the plug and she's clean (so it's not oil fouling or ring wear - blow-by isn't indicated by the PCV, etc, nor is it burning any oil). Cylinder shows 60 psi. Looks like another burned exhaust valve....

 

Just about two months ago I had a customer's '97 Legacy L (2.2) in here with the same issue - except it had REALLY burned a valve and had zero compression on #4. 199k on that one.

 

I did a search of the board here and people mention seeing it. I haven't heard a ton of incidences of this and not much in the way of specific threads talking about it.

 

What's the consensus on the cause? I had figured a lean condition on the EJ22 I just did so as a precaution I installed new injectors, fuel filter, etc. Everything else checked out on it and it's running super smooth now.

 

My plan right now is to tear it down, rebuild the heads, and replace the injectors, fuel filter, and on this car both O2 sensors (the wire to the rear one has been cut off at the sensor - don't know what that's about yet).

 

Any other sensor's that people have found to cause a lean condition or is this just a faulty injector issue? I'm really starting to get paranoid about EJ injectors - this is Honda territory from what my head guy tells me

 

And slightly off-topic, but this EJ25D has near 170k on it and I bought it from the original owner. It's still got the factory head gaskets . It showed no signs of overheating and it was driven here from Bend most recently and before that it was somewhere in Vermont .

 

GD

Edited September 30, 2010 by GeneralDisorder

[daredevil1166]

The GD retirement plan at work here......

 

Have you checked the valve clearance? The DOHC is a shim and bucket adjustment. They never get adjusted and occasionally do get really tight.

 

Maybe it just wasn't allowed to close properly because it's too tight.

 

I dont know much about engines but over time couldn't that cause a burned valve?

[GeneralDisorder]

The GD retirement plan at work here......

 

Have you checked the valve clearance? The DOHC is a shim and bucket adjustment. They never get adjusted and occasionally do get really tight.

 

Interesting theory. I haven't got the engine out yet (brought it home today and verified their mechanic's claim of low comp. on #4) but I'll pay attention to that as a possibility when I pull the head off.

 

The EJ22 I worked on recently was a manual adjustment (no bucket/shim stuff) and baring a valve hanging open due to deposits..... well the head guys said it was from a lean condition and likely a bad injector so that's what I went with. The valve looked closed - well as closed as is possible for something with 1/4" hole burned into the side of it .

 

Maybe it just wasn't allowed to close properly because it's too tight.

 

I dont know much about engines but over time couldn't that cause a burned valve?

 

Definitely a possibility worth looking into. If the valve hangs open then it can't dissipate heat from the edge of the valve into the seat and surrounding aluminium of the head (which is primarily how valves are cooled). As soon as you create a gap at that point you generally experience a rapid enlargement of the "hole" location as that spot gets larger and dissipates less heat - hotter and hotter. Eventually it becomes like a cutting torch and just blows away material very quickly.

 

I'm not following how they get too tight as they age though - that's counter-intuitive to what I'm used to. What's the mechanism behind that?

 

GD

[Fairtax4me]

Could a bad exhaust gasket on the head cause an air leak and possibly cause the valve to burn? I know these cars have issues with exhaust leaks, not sure how commonly that particular gasket goes bad though.

 

I'm not following how they get too tight as they age though - that's counter-intuitive to what I'm used to. What's the mechanism behind that?

I've heard of this before and am also interested to know exactly how it happens. My suspicion is varnish/buildup under the bucket.

[GeneralDisorder]

Could a bad exhaust gasket on the head cause an air leak and possibly cause the valve to burn? I know these cars have issues with exhaust leaks, not sure how commonly that particular gasket goes bad though.

 

The manifold gaskets rarely fail..... at any rate that's largely a myth on modern engines. You won't burn a valve if you run without the exhaust in place let alone a small leak. That's not what causes burnt valves. As I see it there is only three possibilities:

 

1. Wear. It simply wore out to the point that there wasn't enough contact with the valve face and seat to properly cool it. Overheating and cracking/chipping/flaking of the valve face occurs till the hole opens up more and then the thing just becomes a torch and rapidly declines.

 

2. Lash adjustment too tight - valve never fully seats - see above.

 

3. Lean condition - basically running too hot. Retarded ignition timing, lean injector, no EGR on engines that need it's cooling effect (not specifically Subaru's), detonation, preignition, etc.

 

I'm just wondering what people's experience has been with burned valves and if they know for certain what, if anything, made it not return for them..... I've read at least one post where a member had the same valve burn again after only a couple months.

 

I'm leaning toward some kind of lean condition as I have seen plenty examples of EJ's that have run 300k without this happening. Thus it would seem not to be a "normal" wear effect. Tight valve tollerances might do it but I would assume that most of the EJ25D's on the road have never had the valves adjusted....

 

GD

[daredevil1166]

I'm not following how they get too tight as they age though - that's counter-intuitive to what I'm used to. What's the mechanism behind that?

 

GD

 

 

I believe it's because they wear the valve and valve seat together. Therefore the valve stem sticks further out.

 

IMO... Most valves get looser over time because of the adjustment mechanism backing out/wearing. So without an adjustment mechanism the DOHC, the valves don't back out. The valve and seat wear and make it tighter.

Edited September 28, 2010 by daredevil1166

[GeneralDisorder]

I believe it's because they wear the valve and valve seat together. Therefore the valve stem sticks further out.

 

Interesting. I see what you are saying.

 

IMO... Most valves get looser over time because of the adjustment mechanism backing out/wearing. So without an adjustment mechanism the DOHC, the valves don't back out. The valve and seat wear and make it tighter.

 

Most valve looseness is due to wear in the valve train - the adjustment's don't back off but the cam and it's follower do wear. That is minimal in a roller-rocker system though so it could be the case that the adjustment *doesn't change* appreciably and in doing so the tollerance is taken up by the valve seat wearing, the lash reaching a zero point and then further wear creates an actual gap between the valve and the seat.

 

This is all speculation though. I'll actually take measurements when I get it apart and see what, if any, changes are present in the valve lash.

 

GD

[wavekayaker]

Question along those lines. I just took my heads off for a timing belt job (decided to pull the engine for a re-seal, broke 4 of 6 exhaust studs, off to the machine shop. Origional head gaskets EJ25D, gaskets looked perfect!) One of the exhaust valves is light in color and has some flaking on the face. Is this the early stage of a burned valve? Should I replace it now?

 

Sorry about jumping into the thread, but it is on topic:)

 

Anthony

[GeneralDisorder]

Generally I've seen differences in exhaust valve color on cylinders that were exposed to some quantity of coolant - this also will often "steam clean" the carbon from the chamber and piston top. Typically the change in color is not an indication that the valve is actually bad or in need of replacement. Just that there was something wrong with the chemistry in that cylinder.

 

I would definitely lap the exhaust valves at the very least and reset the clearances. I've seen enough exhaust valve problems that this is just good sense IMO. The last set of heads I had redone the shop had to replace about 4 of the valves and he said they were all "significantly" worn on the face and seat. That engine had 200k on it.

 

GD

[WoodsWagon]

I've had hydro lifter 2.2l heads burn valves twice, different heads, same intake, same shortblock. That shortblock did burn oil though, but that was because it never got a chance to break in. The hood latch broke on it's maiden run, which was across the country. So no oil added between Buffalo NY and Lander WY, on a brand new engine.

 

My assumption was a lean burn in that cyl, #3, due to a clogged injector. I swapped engines and never got around to diag because another motor fell in my lap for free.

 

Next motor ran fine, and I beat the piss out of it just as hard as the first one. Same wiring harness and computer, so that's eliminated.

 

Seen one EJ25 with a burnt cyl #4 that we threw a 2.2l at to fix. Worked fine, got rid of the skip! 97 or 98 outback.

[Ricearu]

I know when working on air cooled engines, they say the exhaust valve difference is caused by stretch. Hot valve, plus spring pressure. makes sense. I know lack of valve adjustment on a VW will cause an exhaust valve to seperate at the head, and fall into the cylinder and make mash of the piston and head.

[Fuzpile]

Something which can be forgotten is that the valve is supposed to rotate slightly. This is necessary for all the reasons which flash in everyones head:grin: Offsets were supposed to occur to induce the rotation with rocker arms and that. Otherwise there is trouble in lube which becomes heat and stick and burn.

So then something is to induce the rotation. The actions of springs when they release contribute to that unless there is something different from the others.

[subaru360]

I've had a few burned valves. It's always been that the valve wears into the seat and the clearance gets too tight, hanging it open and burning it.

 

I've always been able to just lap in a new valve, reshim and it's good to go.

 

For whatever reason it always does seem to be #4

[GeneralDisorder]

For whatever reason it always does seem to be #4

 

Yeah I would really like to figure out what's up with that.....

 

You never replaced any injectors, etc? How long did you track these repairs afterwards - I'm curious to know if you are aware of the mileage on them since the repair.....

 

Would you replace the injectors?

 

Anyone else? I don't want to replace the injectors if I don't have to - I guess it comes down to checking the clearance and seeing if the lash has closed up and is holding them open. If that's the case then I can probably be safe to assume it's not the injectors....

 

GD

[subaru360]

No, no replaced injectors. The reason for failure was obvious. They had no valve clearance at all and the valve was hung open. I've seen this on EJ205 wrx, EJ25D outback and EJ22. Always #4.

 

I don't keep track of repaired cars. Only way I would know of a repeat failure is if it broke again and the owners contacted me. None have about burned valve repeat failure.

[Fairtax4me]

Of the ones you have seen do they have the same head/cam design (sohc vs dohc) or is any type susceptible to this?

I also wonder if there is any correlation between head gasket failure and if it may contribute to deposits forming/landing on the valve seat.

[GeneralDisorder]

The last one I did was an EJ22 ('97 single port) and all EJ22's are SOHC. This new one is an EJ25D which is DOHC. It doesn't seem to make a bit of difference as to DOHC vs. SOHC.

 

*something* is making this happen with #4 and not the other's though. I find it very strange that it's so consistent. Something about that cylinder runs hotter. It's not the heads because the same basic casting is used for both sides so if it were a coolant flow issue it should show up on #1 and #4..... but it doesn't.

 

Obviously when I say "consistent" I mean *when* this problem crops up. There are thousands..... nay - tens of thousands - of these engines on the road that this never occurs on. It's still an unusual failure in the scheme of things but when it *does* happen it seems to be always the #4.

 

Neither of the one's I've seen have had a single head gasket failure. Both have known history from the original owner and in the case of the EJ25D I have here now I can see they are original. The EJ22 also had it's original gaskets in place and they were not blown.

 

GD

Edited September 29, 2010 by GeneralDisorder

[Fairtax4me]

So coolant deposits can be ruled out as a major factor, and head design doesn't appear to make one or the other more likely to experience failure.

 

What about EGR vs non EGR?

I'm not at all familiar with how the piping for the EGR is routed on these. Never bothered to look it up since it's a gremlin that I don't have to deal with on my car.

 

Could there some attribute of the intake manifold that allows the #4 to lean out?

[GeneralDisorder]

Well - that's as good a guess as any. Indeed both of the one's I've seen had EGR and the EGR pipe is located on the #4 exhaust port as well as the EGR valve being located near the #4 intake runner

 

I'll have to think on that one and have a look at the manifold.....

 

GD

[987687]

The brake booster also comes off the cyl #4 intake runner. So if the MC were leaking it would cause a lean condition in that cyl. They put a lot of stuff on the #4 cyl.

[StructEngineer]

How are the valve guides? Heat is dissipated from the valve through a combination of the seat contact area and the valve stem/guide contact area.

 

As for why it's #4, I would be interested to see a schematic of coolant flow through the block....

[idosubaru]

the ethanol haters haven't commented yet?

 

is everyone not only pointing to the same cylinder but the same valve within that cylinder? subaru360, Theloyale, GD, et.al. - which valves are we talking within cylinder #4?

 

interesting indeed. i haven't seen it yet, wonder if davebugs has?

[GeneralDisorder]

Ok - little update.

 

As I suspected it's a burned valve.

 

It's the forward exhaust valve in the #4 cylinder (as it was the last time I found this). About a 10mm section along the edge of one face is missing and the valve is VERY deep in the head compared to the other's. It's obvious that it was quite hot. It has no deposits on it - they were all burned away.

 

Before I removed the head I checked all the valve clearances. They were as follows:

 

(front) .005"/.010/.000"/.010" (rear)

 

The burned valve had no clearance with the cam lobe. Thus it was not in contact with the seat.

 

I'll get a picture in a bit.

 

I noticed one thing right away - the burned valve is closer to the exhaust port exit (and at a more favoreable angle) than it's brother in that same cylinder and the #2 cylinder has both valves about equally distant from the port exit. That might indicate that the burned valve handles a higher volume of exhaust gasses.

 

Having given some thought to the EGR theory - that should actually cause a rich condition because exhaust gasses are inert. The EGR being open will not affect the pusle width of the injector so you would have less intake air for the same amount of fuel.

 

Same basic deal with the brake booster. It takes air from the intake runner but since the injector is down right next to the head the amount of *fuel* doesn't change. This would cause a rich condition not a lean one.

 

So far my theory boils down to two basic factors:

 

1. The modern DOHC valve train is very robust. Very little wear occurs because you only have two wear points - the cam lobe and the lifter shim. In the case of the EJ22 it had roller rockers and they are similarly light on wear. Throw in modern oil technology and you have a recipe for almost no wear to occur in the valve train.

 

2. If valve adjustments are not done in a timely manner, the only wear in the system is to the face of the valve and this has the complete opposite effect of wear to the valve train - it causes the lash to close up rather than to enlarge (which is what I'm used to with older engines - EA81's, etc). When it reaches zero the valve hangs open and rapid deterioration results as no heat can transfer to the head through face/seat contact.

 

GD

Edited September 30, 2010 by GeneralDisorder

[eppoh]

There is a third possibility for a burnt valve, and it is the most common reason on older BMW boxer engines from the early 80's. That is valve seat metallurgy. The seats are too soft. I had this happen to me, and found the burnt valve early by accident. Had to pull the head to have an stripped exhaust port nipple repaired.

 

I would guess it is a combination of slightly higher heat at that # 4 cyl valve and lack of proper maintenance.

 

It would be interesting to poll Subaru drivers and ask how many actually ever have valve clearances checked.

 

Loud valves save $$$$

 

Also on your high mileage engine, reduced flow from a slightly clogged cat could account for higher EGT's .

[GeneralDisorder]

There is a third possibility for a burnt valve, and it is the most common reason on older BMW boxer engines from the early 80's. That is valve seat metallurgy. The seats are too soft. I had this happen to me, and found the burnt valve early by accident. Had to pull the head to have an stripped exhaust port nipple repaired.

 

I think in this case that can be ruled out. The seat actually looks very nice and my head guys don't even replace them - just the valve itself and lap it in.

 

Loud valves save $$$$

 

Also on your high mileage engine, reduced flow from a slightly clogged cat could account for higher EGT's .

 

That's a possibility - this car has had both cat's replaced. In fact they fabricated a whole new J-pipe for them. I'm betting it was from a P0420 but it's hard to say at this point. The EJ22 I worked on did not have any cat problems that I could find. No codes to that effect and no unusual exhaust sounds or lack of power after I did the valve job.

 

GD