Testing alignment on a PM1440HD

[Larry$]

xr650, interesting variations. In thinking about why, if there were variations in both the spindle taper and the test bar, could it be that at some rotations they were adding while at others they were cancelling? The two would likely be 180 Degrees apart. Was that the case?

 

[Christianstark]

Before I trusted any measurements, I’d try to get some kind of repeatability at different rotations, with some kind of assurances you aren’t adding errors with indicator placement and methodology.

 

[bill70j]

About .0035+ over 9" of travel with the indicator mounted on the cross slide. Only conclusion I come up with is the headstock is out of parallel with the bed.

Me too! I have the 1440GT.

FWIW:

Right after setting up the 1440GT, I carefully leveled it using the SHARS' machinist level. I called that good. That was 4 years ago.

But just recently I was fooling around with the Rollie's Dad's Method (RDM) and found your exact same 0.0035 deviation over 9".

My next step is to use the two collar method to align the headstock. - and then the tailstock.

Bill

 

[xr650rRider]

xr650, interesting variations. In thinking about why, if there were variations in both the spindle taper and the test bar, could it be that at some rotations they were adding while at others they were cancelling? The two would likely be 180 Degrees apart. Was that the case?

I probably only rotated thru about 45 degrees of rotation before I hit the 1.5 tenths spot, it's likely that it would repeat but not what I was after. It always bugged me that I'd put that bar in and have runout when rotated, runout would be nearly identical on both ends, so I'd continue on using it for alignment and just traverse in 1 position. You start thinking it could be caused by bearings etc. It's a really simple to test bearings, just put a dial test indicator on that spindle taper and measure. When I saw no wiggle of the needle along that taper, you realize the runout is elsewhere. Think about it, a joint that relies on 2 tapers for precision alignment is likely to be the source of error. Just a spec of a chip can cause it, which is what I was leaning toward but in reality the fitment can have a spot where runout is minimized, even if it's spotless.

And these precision test bars eliminate the need of making all those cuts using the Rollie Dad or 2 collar test. Can still make cuts as final check but not necessary while chasing your tail trying to align headstock.

 

[Larry$]

There is no such thing as perfection! Just accommodate the imperfections. Manual lathes have a pretty easy life. Nothing turns fast, bearing preloads are minimal, at least compared to machines designed for serious loads and speeds. Bearing fit to housing needs to have some clearance. The cup has to be able to slide even when run a full speed and load, maximum heating conditions. There has to be some slop, measurable?? One end of a shaft will be designed to accommodate the required clearances with some sort of preload system the other end has to have clearance to slide in the housing. I have no idea how consistent the design clearances are maintained especially on import lathes. Are the bearing seats in the casting exactly on center at each end? It's surprising they work as well as they do.