[Spenny_b]

Thinking about this while I'm typing it....I'm not convinced that the 930 part number won't work - what I think may be the problem is that the less-than-stellar headers I have (although they're spot-on dimensionally) may be the issue. When building the engine, I did carefully measure the protrusion of the primary "stub" that extends through the flange. Comparing that to the depth of the step in the heads, and adding the thickness of the sealing ring, should be sufficient, i.e., the step was shallower than the stub + ring. So, my theory is that the end of the stub is not parallel with the mating flange, in other words, I'm not getting a 360º seal against the ring.

The easiest way I can think to do this would be using a depth gauge in, say, 4 positions on each primary stub to verify (12, 3, 6, 9 o'clock). The other way would be to turn a jig on the lathe, which would hold a DTI close to the way of the stub...rotate the jig on the stub, and the DTI will give a clock reading around the circumference....hmmmm....yup, think I may make that once I've removed the headers.

Then I've got to go about correcting any variation. That'll be "fun" (tedious)