After mounting my 4 jaw chuck, I chucked the cylinder into it, and zeroed it with a dial indicator to something under .001 runout.   One end face was fairly smooth, and the other very rough from the sawing with an abrasive blade.   I decided to face both ends flat, and then work on the diameter.  Facing that first rough surface was a preview of things to come!

 As the cylinder seemed well supported, I decided to skip any tailstock support, and just face it directly.  The first few passes went well, starting with the bit pointing down the lathe axis, and drawing it out across the face on the Y axis.  It knocked down the high spots in .015 passes without a problem.  Once the face was smooth enough that the bit was cutting full time,  I had my first encounter with the Chatter Monster!  (I suppose he had always been hiding in my lathe, waiting for a big project to come out and make trouble!)  By feeding just a few thousandths per pass, I was able to make chatter-free cuts, but I assumed that once I finished the facing, the turning would go much smoother.  That proved to be wishful thinking.

    I reset the tool post, and made a very light pass just clean up the outer surface.  That was uneventful. I made my next pass at .005, and the chatter was very obvious.  A .010 pass shook the entire machine, all 700 pounds of it and the table it is bolted to!  If I had any doubts about the power of my Shoptask, they were instantly removed by that!

    I switched out the carbide bit to a brand new HSS one, shortened the overhang on the tool to the bare minimum, and center drilled the overhanging end. (The center drilling took more force than I expected.  More about this later....)   I added a center and extension to the tailstock, greased it and snugged it to the work.   This solved most of the chatter problem, but I could still only take light (.005 or so) cuts.  I resigned myself to many hours of machining, and that's when I had the bright idea of using the power feed to make the passes!

    Instead of standing, I pulled up a stool, and feed the crosslide in .005.  Then, under power, I'd make the cut.  Under manual feed, I'd stop every so often to break the long, thin strands of aluminum coming off.  With power feed, I ended up with one long, very hot and very sharp strand, which would make an acceptable substitute for razor wire!  They came of in such a manner as to miss my hand on the left/right switch, so everything went fine.  Then 'experience', the harshest of all teachers, decided I was in need of a lesson.

    I was reversing back to the starting position, and about to shift into neutral when one very hot strand landed on and burned my wrist.  I pulled my hand back instinctively, and the cross slide ran into the tailstock under power!  While the machine made an awful grinding noise, I reached up with my left hand and turned it off.
    After a few minutes of swearing, I got everything disengaged.  The machine, much to the credit of the folks at Shoptask, survived the abuse without damage.  I was sure I would have stripped a few gears, or broken something.  Nothing broke, but I did have to reset all the gibs.
 I got in touch with JT via the Delphi group, and he assured me that I was not the first and would not be the last to do this.  He suggested I check the lead screw bearings, which I had done, and the bracket that holds the shifter and gears, as these will sometimes break when abused.  Mine was fine, so I went back to work...

    When getting everything set up, I noticed that the center in the tailstock noticeably shifted "outwards", towards the front of the machine, when I locked down the tailstock.  The force of the cross slide apparently pushed the adjustable tailstock out of true.  I unmounted the 4 jaw chuck, and realigned my tailstock using two centers, a 12" piece of center-drilled stainless steel rod, and a dial indicator.  The tailstock was .030" out of true.  I hadn't checked it "as delivered", so I redrilled the centers on the aluminum cylinder.  It was noticeably easier, so I suspect my tailstock wasn't exactly aligned when I first received my machine, and smacking it with the cross slide probably didn't help any, either!
    After a few more cuts of .005, I decided to stop and find out what was causing the chatter on heavier cuts.  I checked the tool height, and it was right on (I had earlier added .005 of shim), and the tool was still sharp.  I reset the cylinder in the chuck, and worked with the jaws until it had well under .001 runout.  I reset the gibs yet again, even more carefully  than before, but nothing seemed wrong.  A heavy cut still shook the machine, and it took many light passes to remove the chatter marks from the metal.
 

 Finally, I checked the messages on the Shoptask Forum on Delphi, and saw one about the mill spindle preload being too loose, and giving a very rough finish.  I hadn't checked the preload on my lathe spindle, so I tightened it down until I could feel some noticeable drag.  I let it run for a few minutes, without a cut, and the motor didn't seem to be getting any hotter than it had been before.  I tightened it down just a bit more, repeated the test, and started to cut again.
 It went *much* easier!  I could take .015 cuts, but the belts would occasionally slip under such a load, so I backed down to .012.  Many many many passes later (without using the power feed!), I ended up with the basic shape below, and a two trashcans full of aluminum string...

    Some test-fitting of the cylinder into the bearing and its frame showed that moving the drive pulley up into the cylinder would simplify the remaining construction.  I did some "redesign on the fly", and rather than turn down the last 1.125 tail, I cut down a 3.5"x1.25" disk of aluminum into a flange with a 1.125" shaft.  I drilled a 15/32 hole into the cylinder and the flange, and then reamed with a .499 reamer for press-fitting of a .500 shaft to keep everything centered.  (I guess I could have bolted it up, and then turned the final diameter, but it seems to have worked out just fine as-is.)
    The redesign meant I had to move the 4.5" pulley, which was to have been bolted below the cylinder, into the cylinder itself.  Using a 6" zinc cast pulley as a pattern, I cut in the V-belt groove, using the compound tool post and a parting bit.  The compound tool post made quick work of this, and was much easier than rotating the entire cross slide.
 

 A few light finishing cuts, a .499 bore on the cylinder top for centering the table attachments, and the axle was complete.  It didn't go exactly as planned, but I certainly learned a *lot* about my machine in the process.
 
 

Any questions? I'll be glad to answer, just send me an email.