Brett, what would cause an engine to surge? It would seem to me that the total airflow through the venturie would have to change in relation to the needle setting in order for the engine to surge.
If the needle is set with a very small air leak around the spray bar as would be the case with an unseal spray bar, I don't see how the total airflow through the venturie would change unless the spray bar retaining nut is so loose that the spray bar could rattle around in the venturie.
I expect to get blasted again shortly, but it's a good question.
We are almost certain that it is air leaking into the spraybar pass-through. It's more complicated than you suggest above. The air doesn't leak straight into the venturi (that actually makes almost no difference on "straight-through" venturis). In the sprinkler arrangement, it's prone to two additional issues - one, fuel doesn't necessarily go from the spraybar to the venturi/spigot. It can also leak around the pass-through, down between the case and the venturi, and then dribble into the port as raw drops. This was a very serious issue on the ST46, but not nearly as serious on the PA/RO-Jett.
It does pay to choose the venturis that have the best fit in the fore/aft direction (and both the PA and RO-Jett venturis vary significanly from unit to unit, some are loose overall, some are tight overall, some are tight in one direction (like fore and aft) but loose in the other (side to side). On the PA the spraybar holds the venturi in at the rear, and it wants to spring up in the front from the tension of the o-ring you are compressing, which also tilts it so there is a gap right below the spraybar where this fuel leak from the spraybar to the crabkshaft through the boss on the case can occur. The RO-Jett is also held by the spraybar but there's little tension on it because instead of being compressed buy the insertion force, the o-ring goes into the bore in the boss. This is offset somewhat by the fact that boss is a little shorter and the spraybar diameter is larger, meaning the seal area is shorter along the bore.
But the issue appears to be *tiny* amounts of air getting around the spraybar (threads on one end and "head" on the other). The spraybars are different, the RO-Jett has the nut on the collet side, and the PA has the nut on the inlet side, but I don't think that makes any difference. On the Jett BSE this is a machined surface, on the PA its bead-blasted where the spraybar contacts the case. At least to start with, after a few remove/replace cycles it's worn shiny, too. What we think is happening is that air gets in this way (which is under suction when running) and mixes with the fuel just like a pinhole leak in the line. This then displaces some of the fuel and makes it go slightly lean, then clears up (since the flow rate is miniscule). I also think it's a bigger issue on the PA (which is where David first noticed it) because the fuel inlet on a stock PA spraybar is extremely small so any there is significant restriction on the fuel flow. The Jett is much larger, and counter-boring the inlet on the PA makes a huge difference with the 61 and 75. Don't drill it all the way through, because that will also remove the needle seat! Guess how we found that out....
We have found all sorts of interesting things on the venturi/spraybar/fuel flow/flow restriction side of the system, and almost the only fiddling we have done in the past 5-10 years is in this area with sometimes very surprising results. Merely being able to open the needle far enough to get the desired setting is not the only criteria, reducing the restrictions seems to change a lot of things that you didn't even realize were happening until you fixed it.
We started looking at this after the 2002 NATs and to give credit where credit is due Brad Walker and his good advice on fuel tanks and other fuel discussion for 4-strokes were the precipitating factor. I think Brad would hasten to point out that he got it from RC guys, Brian Eather, and Bill Wilson. We were never going to run a 4-stroke but physics is physics.
Brett