Another lathe leveling question

Was wondering if I should follow up with using rollies dads method. I have a motorcycle fork tube that is straight and long enough. Since this Enco 13x40 lathe is basically a large benchtop lathe bolted to an Enco lathe bench, should I correct any mis alignment the with shims under lathe legs, adusting bench feet, or further corrections on headstock alignment?
 
Align the headstock to the crossslide as you have done. Then don’t touch it. You can then align the tail stock up close by lightly pinching a razor blade between centers—align the tail stock to make the razor pinch at right angles to the spindle.

Then do a two-collar test with a test bar to level the bed. The tail stock is already aligned so don’t mess with that, but bed twist will move it out of alignment as you slide it down the bed. You can use shims under the bed where it bolts to the bench at the tail end. Make sure it’s bolted tight to the bench.

If you align the tailstock with a test bar without first aligning it with the razor blade, you’ll miss bed twist and the alignment will change as you move the tailstock. If you align it by twisting the bed before the razor blade test, the same thing may happen. Take things in order. Check both alignments after doing this—it may take a couple of iterations.

Rick “just went through this” Denney
 
Do a two collar test using the tailstock for bed twist? What I've been reading on this site is that two collar test in chuck without tailstock is for bed twist and with tail stock between centers is for tail stock alignment. I'm not questioning you experience, just confused about all the different advice. It could also be I'm not understanding the different techniques.
What would you call close enough on the headstock on my 8 inch disc cut?
I guess your not a fan of rollies dad method?

Thanks for the help. Your concern helps with my frustration doing alignment.
 
Clunker1 said:
Do a two collar test using the tailstock for bed twist? What I've been reading on this site is that two collar test in chuck without tailstock is for bed twist and with tail stock between centers is for tail stock alignment. I'm not questioning you experience, just confused about all the different advice. It could also be I'm not understanding the different techniques.
What would you call close enough on the headstock on my 8 inch disc cut?
I guess your not a fan of rollies dad method?

Thanks for the help. Your concern helps with my frustration doing alignment.
Click to expand...
I appreciate everyone here that has been helping you with getting your lathe setup. I also have a Enco 13 x 40 that needs some attention
 
rwdenney said:
Align the headstock to the crossslide as you have done. Then don’t touch it. You can then align the tail stock up close by lightly pinching a razor blade between centers—align the tail stock to make the razor pinch at right angles to the spindle.

Then do a two-collar test with a test bar to level the bed. The tail stock is already aligned so don’t mess with that, but bed twist will move it out of alignment as you slide it down the bed. You can use shims under the bed where it bolts to the bench at the tail end. Make sure it’s bolted tight to the bench.

If you align the tailstock with a test bar without first aligning it with the razor blade, you’ll miss bed twist and the alignment will change as you move the tailstock. If you align it by twisting the bed before the razor blade test, the same thing may happen. Take things in order. Check both alignments after doing this—it may take a couple of iterations.

Rick “just went through this” Denney
Click to expand...
Is there a reason why you prefer the two collar test to the Rollie's Dad Method after alignment of the tailstock?

The latter seems a lot simpler to get right and cheaper to do to me?
 
SouthernChap said:
Is there a reason why you prefer the two collar test to the Rollie's Dad Method after alignment of the tailstock?

The latter seems a lot simpler to get right and cheaper to do to me?
Click to expand...
What is the rollies dads test?
 
Clunker1 said:
Do a two collar test using the tailstock for bed twist? What I've been reading on this site is that two collar test in chuck without tailstock is for bed twist and with tail stock between centers is for tail stock alignment. I'm not questioning you experience, just confused about all the different advice. It could also be I'm not understanding the different techniques.
What would you call close enough on the headstock on my 8 inch disc cut?
I guess your not a fan of rollies dad method?

Thanks for the help. Your concern helps with my frustration doing alignment.
Click to expand...
I'm not sure my lathe could do better than a thou on an 8" facing operation, but then it's not adjustable. :) It might be a bit less than a thou in 8", and a bit more in 12". Facing flat verifies that the headstock is parallel to the motion of the cutter on the cross slide.

For the rest:

Here's the thing: a twist in the bed creates exactly the same effect as a tailstock that is not centered on its adjustment pins. It moves the tailstock center laterally with respect to the spindle center. When you have two adjustments that might create the same effect, you have to isolate them to check them. That's why I start with pinching a razor blade between centers, with the quill retracted nearly all the way. That puts the tailstock as close as possible to the headstock, which minimizes (but maybe does not eliminate) the effect of any twist in the bed. Once the tailstock is adjusted using that test, you can be more confident that the tailstock is centered on its adjustment pins as well as possible for the moment.

(If you attempt a two-collar test without centering the tailstock first using the razor-blade test, you won't know whether the tailstock is uncentered, the bed twisted, or both. And there is even the possibility that for the specific length of your test bar they are both wrong but offset each other leading you to believe that everything is aligned.)

Then, back the tailstock away from the headstock and do the two-collar test. If it's cutting taper (the two collars are different, or the measurement of a two-collar measuring device such as the one from Edge Technologies using a dial indicator on the carriage is different), then it's less likely to be caused by tailstock adjustment and more likely to be caused by bed twist. If the tail end collar you just skimmed is smaller than the collar at the head end, then the tailstock center is rotating away from you as it follows the twist in the bed--shim the back right corner up a bit. If the tail collar is larger than the head collar, then it's rotating towards you as it moves down the bed--shim the front right corner up a bit. Once both are the same, go back and pinch that razor blade again. If the blade still stands at right angles to the spindle when it's pinched between centers, then your tailstock is centered and the bed is flat with respect to tailstock and carriage motion, or at least as well as one can do given the wear on the bed. You might have to back and forth a couple of time.

A tailstock can also be turned a bit on the bed if the v-groove is worn crooked. (This is in yaw, while a bed twist affects roll, if those aviation terms mean anything to you.) Ask me how I know! The only cure for that is scraping it in, or finding a better tailstock. I tried the latter and got lucky. If it's turned, the tailstock center will shift laterally as the quill is extended, which is not good.

A test bar with a morse taper on one end, inserted into the tailstock and run up to a center in the headstock, can expose a range of maladjustments, as long as the test bar is really accurate. I have one of those Indian test bars, but I'm not sure I really trust it. I'm more comfortable trusting the Edge Technologies alignment bar, or just making my own two-collar bar.

You can also measure bed flatness using something like a Kingway, but that's really for measuring it when scraping it in, and it still needs to be twist-free before attempting that.

Did that explain my reasoning? Is that what Rollie's dad also suggested? (I found it easier to just explain what I do and why than try to find faults in someone else's approach--who may be starting from a different point or with a different set of assumptions.)

Rick "been through this several times with a 76-year-old lathe that has been reground at least once in its long life" Denney
 
Last edited:
rwdenney said:
I'm not sure my lathe could do better than a thou on an 8" facing operation, but then it's not adjustable. :) It might be a bit less than a thou in 8", and a bit more in 12". Facing flat verifies that the headstock is parallel to the motion of the cutter on the cross slide.

For the rest:

Here's the thing: a twist in the bed creates exactly the same effect as a tailstock that is not centered on its adjustment pins. It moves the tailstock center laterally with respect to the spindle center. When you have two adjustments that might create the same effect, you have to isolate them to check them. That's why I start with pinching a razor blade between centers, with the quill retracted nearly all the way. That puts the tailstock as close as possible to the headstock, which minimizes (but maybe does not eliminate) the effect of any twist in the bed. Once the tailstock is adjusted using that test, you can be more confident that the tailstock is centered on its adjustment pins as well as possible for the moment.

(If you attempt a two-collar test without centering the tailstock first using the razor-blade test, you won't know whether the tailstock is uncentered, the bed twisted, or both. And there is even the possibility that for the specific length of your test bar they are both wrong but offset each other leading you to believe that everything is aligned.)

Then, back the tailstock away from the headstock and do the two-collar test. If it's cutting taper (the two collars are different, or the measurement of a two-collar measuring device such as the one from Edge Technologies using a dial indicator on the carriage is different), then it's less likely to be caused by tailstock adjustment and more likely to be caused by bed twist. If the tail end collar you just skimmed is smaller than the collar at the head end, then the tailstock center is rotating away from you as it follows the twist in the bed--shim the back right corner up a bit. If the tail collar is larger than the head collar, then it's rotating towards you as it moves down the bed--shim the front right corner up a bit. Once both are the same, go back and pinch that razor blade again. If the blade still stands at right angles to the spindle when it's pinched between centers, then your tailstock is centered and the bed is flat with respect to tailstock and carriage motion, or at least as well as one can do given the wear on the bed. You might have to back and forth a couple of time.

A tailstock can also be turned a bit on the bed if the v-groove is worn crooked. (This is in yaw, while a bed twist affects roll, if those aviation terms mean anything to you.) Ask me how I know! The only cure for that is scraping it in, or finding a better tailstock. I tried the latter and got lucky. If it's turned, the tailstock center will shift laterally as the quill is extended, which is not good.

A test bar with a morse taper on one end, inserted into the tailstock and run up to a center in the headstock, can expose a range of maladjustments, as long as the test bar is really accurate. I have one of those Indian test bars, but I'm not sure I really trust it. I'm more comfortable trusting the Edge Technologies alignment bar, or just making my own two-collar bar.

You can also measure bed flatness using something like a Kingway, but that's really for measuring it when scraping it in, and it still needs to be twist-free before attempting that.

Did that explain my reasoning? Is that what Rollie's dad also suggested? (I found it easier to just explain what I do and why than try to find faults in someone else's approach--who may be starting from a different point or with a different set of assumptions.)

Rick "been through this several times with a 76-year-old lathe that has been reground at least once in its long life" Denney
Click to expand...
I linked the Rollie's Dad method pdf (it's only 2 pages) just above. It's just an alternative to the two collar test that doesn't require the use of the tailstock (but obviously you'd have to do the headstock alignment first).
 
The RDM method is pretty quick to see where you are. You simply need a ground polished rod, like a shock absorber rod of reasonable diameter to get started. You are measuring horizontal and vertical deviation of the rod as it rotates about the spindle axis. Chuck errors are automatically subtracted out. I find the methodology easy to follow, others may prefer other techniques. It's just one (effective) tool in my toolbox.
 
You must log in or register to reply here.
Back
Top