Chris,
I used the technique as described in Wayne's recent book. Install the install the heads, hand tighten barrel nuts, install cam tower, torque down, install camshaft in tower, torque head studs(two step), check camshaft for free rotation.
Again, if it were a clamping issue, wouldn't one of the cylinder spigots be dry and the others on that bank be wet? All the spigots are leaking on the left bank. That's what makes this so confusing.

Dave,

Like Sherwood, I didn't realize the nature of your leak (all cylinders) so my post was a off target.

The pros have everything machined so it is flat. I know decking the case is standard operating procedure for at least one well known engine builder. I've been burned by this before so now I check or deck the case (or both). I think there are other advantages to doing the heads and cam carrier separately but I think I'm probably in the minority.

Did you save the old base gaskets? I wonder if they were coated in sealant by the previous engine builder? I wonder if there is some sort of difference between the C2 cylinders and the stock ones? I've got lots of Curil-T if you need it...

I feel for you, this must be pretty frustrating. Believe me, BTDT.
-Chris

Chris,
I am beginning to think that the tower could be masking the issue. One cam tower was machined because of a very small discrepancy on the mating surface. I can't say that this is the cam tower associated with the left bank, but.... if the machining was done in such a manner that the mating surface is no longer square with the heads, it would cause a clamping force issue where there is more force on the top end than the bottom or vice-versa. When the heads are torqued down after the towers, this would exacerbate the issue, giving the cylinders unequal clamping force.
I am sorry to say I did not save the old base gaskets or even to think measuring their respective thicknesses....rookie mistake!:confused:
There are differances in the early C2 turbo cylinders versus the standard 3.3l 930 cylinders. The C2 turbo cases stud bosses were built up to mate with the stud boss area of the cylinder. This added more rigidity to the surface area. The cylinders are also fully finned, top and bottom for additional cooling. I spoke to EBS Racing earlier today and they stated that the C2 turbo cylinder upgrade is a common one. They stated to use the 930 base gaskets in this upgrade, which is what I had done as well as the previous owner had done. EBS had stated they have had no issues with leakage using this technique.

Dave,
I don't think the cam tower(s) could cause this. Long before the tower would cause clamping problems you would experience cam binding I think.
Believe it or not, things could be worse. Today I talked with a guy who heard his 930 motor go boom and then it dumped a lot of oil. He parked it and hasn't had the heart to looked at it yet.
-Chris

It is possible that there were thicker than stock base gaskets on the cylinders. E.g. 0.020" thick vs stock 0.010" or even thicker ones or possibly a pair of thicker ones. Check of piston clearence, cylinder height or compression ratio should answer this question. If the piston were to get to close or make contact it could cause a leak. Not likely for any of this to happen in my opinion but may be worth checking. What else is there???

Did you put some 574 on both sides of the base gaskets like I recommend?

A problem with clamping force would result in at least one cylinder clamping correctly - having all three leak is suspect.

Perhaps the case was disturbed when the studs were replaced? It's not uncommon for the aluminum/magnesium to get slighly pushed out when installing studs like these. Perhaps the cylinder spigots are no longer flat (not uncommon)...

-Wayne

Good point Wayne, Dave, was the case decked? IE the cylinder surfaces redone? Between normal case gyrations and replacing the studs its quite likely the problem is that the cylinder bases are no longer flat and perpendicular to the crank and all on the same plane and orentation.

Personally I do not like using anything under Copper gaskets. To me its sort of like silicone is to most everyone else, something you shouldn't do. And if you need more than the copper can do, then you need to machine it.

Update:
Just measured deck height on cylinder #1. Removed cylinder heads of left bank and found true TDC by mounting dial indicator on center of piston. Z1 is very, very close to true TDC, btw. Measured deck height on number 1 is .040(1.016mm). I suspect this is too low because the head was not torqued down to the cylinder which would have decreased deck height even more by a few thousandths. It is pushing the limits to begin with so I'm going to swap out the .25mm with .50mm shims. Or should I stack .25mm AND .50mm for .75mm total shim space? What do you think?

I measured 1mm (.04") of deck height on my 3.2L. A couple of years back I posted back and forth with Steve Weiner on Rennlist and he said this about deck height:
"Get that CR down to something reasonable without losing deck height. Anything more than .040 is nfg. I like .035 as this leaves less pockets of end gasses out there at the piston edge. This is always where it begins."
-Chris

Interesting Chris.
I went ahead and measured the other two cylinders and found the following;
Deck Height
#1 - .040 (1.016mm)
#2 - .031 (0.787mm)
#3 - .034 (0.864mm)
Nearly a ten thou variance! Can it be that the .25mm shim does not have enough "crush" variance that this type of deck height variance can't be compensated for?

Deck height (top edge of piston to top edge of cylinder) is a tricky thing to measure in my limiited experience. It could mean some cylinders are higher than others or some rods are shorter or a combination of both.
I'd put a straight edge across the tops of the cylinders or (better) the bolted down heads to check. I'm working at home this afternoon if you need to borrow one.
-Chris

you really need to precision measure the cylinder heights. just straightedging them, untorqued is a waste of time. also measure the thickness of all the heads, with a large micrometer, between the cam tower surface and the cylinder contact surface. they may have been flycut inaccurately in the past.

Copper gaskets can soak up maybe 0.001 to 0.005" and this is stretching it a lot. THe cylinders must be within about 0.001"max. this is very difficult to measure accurately. If you are seeing 0.010" differences it is probably real to some extent, way way out of spec.

The cylinders can be measured directly with micrometers, or with a drop guage on some kind of marble table. The case must be measured as well. This is not likely within the scope of any home measurement capability. I would suggest taking the whole thing to someone like CE or Ollies and have them check it and fix it. This is one that really requires a true machine shop to do, a Porsche machine shop.

I put a motor together that should have beene decked and it didn't leak. It wasn't aware of how bad it was until I had that motor apart again and checked it.
Like I said You can tell if they are different heights and if they are tilted that way. Perhaps I should have said "torqued down" instead of "bolted down" but that's what I meant.
-Chris

Update:
1.) Went back and measured all deck heights 1-6. Variances between all cylinders; .031-.034". I suspect that the reason why cylinder 1 was off reading a higher value before(.040") was due to the dial indicator may have not been mounted securely enough.
2.) Tore engine down to the case. Diassembled heads to cam towers. Inspected cam tower sealing area. Found some Locktite 574 had seeped into one of the dowel pin bores for head #2. This created a hard locktite plug which may have had some influence in the alignment of head #2 to the cylinder bore.
3.) Replaced the blue silicone thru-bolt seals with the green viton seals. These are a tougher material seal and I would strongly suggest to anyone doing a rebuild to use these seals. Also, as mentioned in Waynes 911 Rebuild book, the thru-bolt/nut orientation may not be that important. I would strongly advise that the nuts be located on the right side of the case. I installed mine on the left. I inspected the thru-bolt bores and found the right side of the case has the bores chamfered while the left does not. When torqueing a nut and bolt assembly, I have always tightened from the nut side, which in this case, may have been pushing my luck in terms of the seal staying in its seated position. If the washer turns ever so slightly, the likelyhood of that seal kinking or tearing is very high because of the sharp edges on the left side of the case. If done from the right side, the bore is chamfered reducing the possibility of that happening. I don't know if other cases are like this. I am only referring to the 930 case.
4.) Installed .5mm base gaskets instead of the .25mm. This will bring the deck height to a more thermally acceptable value of .04x". I did not use Locktite 574 to seal these gaskets. Instead, I used a Copper based adheasive as a sealer. I spoke to three engine builders and they informed me the copper base gasket acts as a heat sink between the case and the bore. This may not be as important on a NA 911 engine, but with a turbo, I wanted to get as much heat transferred out of the cylinders as possible.
After all of the mods, the bottom line is NO MORE LEAKS!!!http://www.pelicanparts.com/support/smileys/clap.gif

I suspect the most likely scenarios with this leak fiasco are items 3 and 4, with 3 the most likely.
Thank you all for the posts. They gave me a lot to think about. http://www.pelicanparts.com/support/...s/beerchug.gif

High and Dry!!
http://forums.pelicanparts.com/uploads/930underside.jpg

I don't think that those chamfers you are talking about are stock - it's common to spot-face the through-bolt holes flat...

-Wayne