I felt like having another try at the valve, so followed Jo's excellent advice and made a profiling tool. Boy, what a difference!
The tool started off as a bit of oil-hardening, ground stock, here marked out with a 0.25" radius and some hand-drawn joins:
I cut the primary radius with an endmill, then ground the smaller radii on the grinding wheel by hand, with final smoothing on a stone, then attempted to harden and temper it (which I don't think really worked very well; not hot enough). But it made quick work of the brass!
and the result is so much nicer looking:
I also took the advice about using an end mill to form the valve seat, which worked nicely:
I used one side of the same profile cutter to do the neck of the side-arm, so that it matched the body of the valve:
Here you can see that I've got a shim under the back of the profile tool. I didn't put any relief on the cutting edge, since I was afraid to ding the edge by grinding up to it. So a slight angle on the tool made up for that.
Other things I did differently this time: I broke out the 6" rotary table, which is easier to bolt down firmly (and a pleasure to turn, unlike the 4"):
and used an end mill for the side pocket:
It's a bit hard to see here, but there's no flare on the side arm. It drops into a pocket made with an appropriately sized drill (no collar required):
This worked nicely because it allowed me to put the solder on the inside, resulting in a much neater soldering job.
The other operations were the same as before, though I did learn that I have to adjust my BSB dies on the holder to get a good fit to the corresponding threads.
Here's the result (new one on the right), complete with with the gland nut. I stole the spindle from the first one, and still have to fix the disc to the spindle.
Next: experiments with hand wheels! I have three different styles on the go
Simon