inlet location questions

[Silverback]

2 different inlet location questions:
- the original design specifies an inlet through the lid, with it trimmed to make a roughly .75" gap between it hand the side wall of the container, presumably to allow the "cyclone" to make it past the inlet.  A number of the posts on this board tell people to move their inlet over so that they're closer to the side wall of the container.  I would guess that this is a better "injection " location, but would disrupt the "cyclonic action."  Which works better/which is right?
- people have posted a number of versions with the inlet coming into the side.  What is the best (best separation performance) geometry for this configuration?

[dbhost]

They both do the same thing, get the air in, spinning around the outside of the vessel near the drop slot. The difference is...

A. Top inlet is a LOT easier to build.
B. Side inlet eliminates 2 90 degree bends, and thus the restrictions those bends put in the system.

While the side inlet is harder to build, and somewhat nerve racking cutting into a barrel for example, the results are well worth it.

[Sub-Dooood]

My impression is that a side inlet configuration, by removing the fitting projecting into the top chamber, would have less turbulence.  Could this lead to increased separation efficiency, especially for the finest dust particles?  I only have anecdotal evidence to support this theory, as neither of my Thien separator builds utilize an inlet protruding through the top lid.  Both systems are extremely efficient at removing everything from the debris stream.