Beautiful Flycutting

[Crusty]

I thought I was setting the head using a quicker method (using the tools I have available) but now I understand that I wasn't checking the spindle axis to table perpendicularity. Somewhere I have an indicator beam and arbor to mount it in the spindle - I need to find it. Thanks.

The difference I see in opposite directions is perplexing but it being due to slop somewhere seems like the right trail to follow to me.

[Toolmaker51]

Another part of tramming is to extend the quill; that doesn't change the angularity, just magnifies it. When you retract to mill, the error will be lessened.
When I tram, it's over well matched 1-2-3 blocks. It's then easier to swing a larger circle and X or Y for the difference. On Bridgeport pattern mills, do the 'nod' or Y axis first, then X, and recheck Y.
The only instrument I use; Indicol clamp, Interapid test indicator. If it needs to 'perfect', I extend the clamp/ indicator far as they will go.

[tonyfoale]

Another part of tramming is to extend the quill; that doesn't change the angularity, just magnifies it. When you retract to mill, the error will be lessened.

Could you elaborate on that, I think that I must be misreading it because extending the quill would have no effect.

[Toolmaker51]

With the joint of a mill to nod or swivel, that creates a point of rotation. Lengthening distance [hypotenuse] from indicator point to that center, like a sine bar, angle doesn't change but the side opposite 'expands' so indicator will detect out of perpendicularity.
I'm describing as if a right angle, but the condition is more truly scalene cone until quill is at a right angle, thereby an isosceles.