Well perfect is a very dangerous word in engineering.
In this case one only needs an error which is inside the thickness of the blue to guarantee a point contact, and which is not detectable.
The result will be something which is more prone to shaking loose under intermittent loads etc, or will permit runout under load. (Or worse will jam if made from soft material) As before, it all depends on the application. If its not critical – fine. Or if you use a drawbar with it, then a very small error be ok.
Still I’m glad that someone can get it perfect.
My method is as follows.
I do the maths for the taper (2 or3 morse), and that converts the solid angle in degrees into a deflection in thou. I’m using a base leg of 5″”. That’s 1/2 the length of the taper turning attachment guide, so I have a 2:1 advantage in sensitivity give or take. The taper attachment has a 40 tpi micrometer setting screw – like your M3 screw. When I have it close, I then measure my setting along the length of the attachment guide (10″), not with a clock, but with a DTI graduated in tenths of a thou, so you can read to a couple of hundredths over 10″. That is a microscopic error?
Then its cut with a new .8mm radius tip, under power feed, which you can do with a a taper tuning attachment.
Even like that, its still not perfect, and still not as good as el cheapo bought item.
So for the hassle involved in not getting it right, IMO, its easier to send off to Chronos/ Axminster/Warco etc
Edited By mgj on 23/03/2010 13:43:07
Edited By mgj on 23/03/2010 13:45:07