I have had company in the workshop this weekend; mostly Mr. Fudgeit and Mr. Bodgeit, but also on occasion Mr. S*dit. However, at the end of the weekend a helical gear has appeared:

In retrospect it might have been a bit ambitious cutting a 6DP gear in steel, in one pass, the first time round, but nothing ventured, nothing gained. The cutter was an el cheapo one from a model engineering supplier.
It was clearly cutting mostly on one tooth, more so than the commercial gear cutters I've previously used. Despite the mill weighing the best part of 2 tons you could feel the vibration in the concrete floor. However, there was a small plastic eye glass resting on top of the overarm throughout cutting the gear, and it didn't move, so it couldn't have been that bad. The finish on the teeth is reasonable, although visually slightly worse than I'd expect when cutting a spur gear:

Here's a general view of the cutting process:

And a close up:

I would hate to give the impression that I just knocked out a helical gear without breaking sweat, it was nowt of the sort. I started off by making a complete foul up of the mandrel to hold the gear; not enough room for the three SHCS screws seen in the picture above.
Despite being a bin job I wasn't overly upset, as it was crap steel that I bought years ago before I knew better and started buying from industrial stockholders.
I then knocked up some parts for the dividing head gear train. I needed a couple of locking collars to keep the gears in place and a keyed sleeve to hold an idler gear when I cut the second gear with the opposite hand. Generally these went well, even the internal and external screwcutting. The thread was non-standard, 7/8" OD and 16 tpi Whitworth. The auto trip on the Ainjest high speed threading unit makes it a doddle to screwcut into a blind hole at 260rpm. 
I started gear cutting proper this afternoon. The gear shown is the second gear I made; the less said about the first one the better. I learnt some important lessons; which is political speak for I ****ed up, but at least I didn't make the same mistake twice. The first lesson was – do the bl**dy nut up tight on the end of the milling arbor before cutting. I was also a bit ambitious on cutting parameters. Cutting at 108 rpm and 100mm/min feedrate should have been ok at about 4 thou per tooth, but with the eccentricity of the cutter I suspect the working tooth was cutting far more. I ended up at 117rpm and 50mm/min for about 1.6 thou per tooth. Throughout the cutting the mill didn't cough; nor was there any hint of the motor slowing as the cut came on. I don't think I'm ever going to have nerve to stall this mill by being over-ambitous.
The next problem was with the mandrel, again. I was relying on the grip of the chuck on the dividing head to hold against the cutting forces. It didn't, so I had to machine a collar so that the cutting forces were reacted directly into the dividing head thrust bearing. Having got into a tizzy I then proceeded to cut a few teeth one turn of the handle too deep. If there's a mistake to be made you can be sure I'll make it!
On the plus side, once I'd finished the gear, setting a gear vernier to the calculated addendum and normal chordal thickness showed that the final teeth were pretty much on the numbers.
Cutting the second gear went rather better, and I even managed to take a few pictures and a couple of videos.
Once I've finished off the parts for the idler gear, and made some collars, the acid test is to cut another gear of the opposite hand and see if they mesh.
Andrew