Tom Gardner wrote:

I have an idea for a new kind of full-auto weapon that I have been seeing in
my head for a long time. I'm no where near ready to start machining stuff
yet but I am wondering how to do legal research and developing of the idea.
So, do I call the ATF and say: "Hey I want to build a machine gun."?

Yes, basically. You'll need to jump through a bunch of hoops and pay
some money. Undoubtedly it will be unpleasant and difficult, but
Congress has given them unpleasant and difficult laws to administer.
There are some good guys at ATF, actually. You might start he
http://www.atf.gov/firearms/nlc/ffl/ffl_types.htm and he
http://www.atf.gov/firearms/nlc/ffl/faqs_nfa.htm The ATF site is pretty
comprehensive. State laws apply as well.

At
what point does it become a problem?

When you assemble one.

It looks like I'll only need a few
pounds of unobtainium and some disapearium. The idea eliminates one whole
operation and could increase cycle rate by as much as 30% while diminishing
recoil. It might do better if I could suspend some of the laws of physics.

Higher cyclic rate means hotter barrels. When you look at the GAU-8,
for example, you see lots of barrels. I think the thermodynamics are
tougher that the cycle rate.

Higher cyclic also means more total recoil energy/sec. On submachine
guns like the MP-5, they have mechanisms to slow the things down so they
don't end up pointing to the ceiling after a burst. My only tiny
experience was when I messed up the disconnector when adjusting a High
Standard .22 pistol and it emptied a 10-shot magazine in what sounded to
me like a single explosion. Only the first shot was on the target paper.

HTH,
Geo. Anderson