PDA

View Full Version : turning problem



gambler
07-10-2018, 01:40 AM
ok, so I am turning a shaft, 9 inches long on my old logan. between centers, the end closest to the head stock measures .515. the far end measures .505.
I think the bed is worn near the head, and the apron is moving away from the work, does that sound right? and is there anything I can do about it?

lakeside53
07-10-2018, 01:49 AM
Is your tail-stock perfectly in alignment? What do you measure if you take a light cut from a stiff piece of stock without tail-stock support?

gambler
07-10-2018, 02:41 AM
Is your tail-stock perfectly in alignment? What do you measure if you take a light cut from a stiff piece of stock without tail-stock support?

I have checked the tailstock and adjusted it recently, it was off. It is as close as I can get it now.

MattiJ
07-10-2018, 02:51 AM
ok, so I am turning a shaft, 9 inches long on my old logan. between centers, the end closest to the head stock measures .515. the far end measures .505.

and is there anything I can do about it?

Easy: re-adjust tailstock so that both ends come out same diameter. (You still might get slight barrel/ huor-glass shape but it will be a lot smaller error)

JoeLee
07-10-2018, 08:18 AM
How did you adjust your TS ??? There are a few different methods.

JL................

JoeLee
07-10-2018, 08:21 AM
Easy: re-adjust tailstock so that both ends come out same diameter. (You still might get slight barrel/ huor-glass shape but it will be a lot smaller error) If he has substantial wear in the bed he'll have to adjust his TS where ever he puts it. No real easy fix.

JL...............

old mart
07-10-2018, 08:46 AM
I bought a test bar on ebay, made in India which has a MT2 at one end, centres both ends and is 11" long overall. It runs better than 0.0001" tir and has been very useful in getting the alignment and height of the tailstock corrected. Don't make the same mistake as me in expecting live centres to be spot on, but use dead centres for testing. I put a piece of 3/8" diameter bar in a collet and turned a 60 degree tip (using a chuck would be ok too), which must be true to the axis of the spindle as long as it is undisturbed. Try the alignment test with the tailstock quill retracted and also with it fully extended to see if there is a difference.

smithdoor
07-10-2018, 09:16 AM
ok, so I am turning a shaft, 9 inches long on my old logan. between centers, the end closest to the head stock measures .515. the far end measures .505.
I think the bed is worn near the head, and the apron is moving away from the work, does that sound right? and is there anything I can do about it?Even if lathe was new this happens on long bars
This why the make follower rest so bar does not flex

Dave

Sent from my SAMSUNG-SM-J320A using Tapatalk

RichR
07-10-2018, 10:24 AM
Even if lathe was new this happens on long bars
This why the make Fowler rest so bar does not flex

Dave

Sent from my SAMSUNG-SM-J320A using Tapatalk

I didn't know that Fowler made them. I always thought they were called follower rests or traveling steadies.

smithdoor
07-10-2018, 11:00 AM
I didn't know that Fowler made them. I always thought they were called follower rests or traveling steadies.Phone struck again need to prof read before sending

Dave

Sent from my SAMSUNG-SM-J320A using Tapatalk

KiddZimaHater
07-10-2018, 11:04 AM
If it's tapering without the tailstock, then it's probably bed wear.
My old 1980's Jet lathe has significant bed wear, and it turns a .001" to .002" taper over 10".
I have to "Bump" my cross feed as I'm turning, to get shafts perfectly straight.
Just like the old timers did.

BCRider
07-10-2018, 11:30 AM
I'm surprised no one has asked the main question yet. Has the lathe bed ever been tested to ensure it is straight and parallel? A key part of any bench lathe setup is that it be trued up (often called "leveled") before you can trust it to do long parallel cuts.

Now some Logan lathes have a motor and primary drive cabinet. But that does not mean you can just plunk it onto a floor and KNOW that the bed is in flat from end to end. It still needs to be checked.

A good and surprisingly easy way to do this is chuck up a long and stout bar in the lathe that sticks out about 4 to 5 inches. Relieve the middle of the exposed bar by about .04" leaving two full diameter "collars" at the end and in close to the chuck jaws. Now skim cut the two collars and mic them. Using shims under the bed mounting bolts or the adjustable feet on the cabinet twist the bed until you measure the same size on both collars. Note that this does NOT involve using the tail stock. You don't want the TS in the equation at this point.

Once you remove any twist in the bed you can now use the identical size collars to check for any bowing by running a DTI over the tops of the collars. Any difference there is an indication of the bed sagging down or bowing upwards and that too can and should be corrected. Confirm you don't affect the zeroing of the twist by taking another skim cut and adjust as needed until you get same size test collars and a DTI reading the top of the collars shows no difference between them.

NOW bring the TS up to the end collar. If you chose your bar stock well it is roughly the same size as the TS ram. So with the mic'ed size of the collar and the TS ram you can use a dial gauge between them to set the ram to center side to side using just a little math so you know the step it should be. And while you are at it run the dial gauge over the tops and see if your TS ram is also centered for height. If needed put shims between the base and upper piece. And again while you are there extend the TS ram a few inches and check that it's both at the correct height and not angled up or down.

After all that the bed should be as true as it can be. If it still cuts a taper or a barrel or wasp waist long shaft then it truly is bed wear.

But unless you can see the wear quite easily I'm guessing it's just the bed has a bit of twist to it and you need to run this "collar skimming" test and true up the bed.

MrWhoopee
07-10-2018, 11:55 AM
While serving my apprenticeship, we had an old Sheldon lathe that was significantly worn at the headstock end. I got to use that machine, the real machinists got to use the real lathes.
More than once I had to gradually move the cross-feed in a few thousandths while turning to compensate for the wear. Not precise or desirable, but you work with what you have. Learning to make good parts on a bad machine is what makes a machinist.

JCByrd24
07-10-2018, 01:05 PM
I like MattiJ and BCRiders post best so far. If you're turning taper only then you need to adjust your TS again for the length of the project at hand. The reason as to why your previous adjustment of the TS doesn't work is likely wear or bed twist or both. A simple readjust at your current length should get you much closer at both ends. You may then get some barrel shape due to deflection without a follow rest, but it should be minor and easy to remedy with a file depending on the ratio of dia:length.

JoeLee
07-10-2018, 02:15 PM
On my lathe the follower can be mounted on either side of the saddle. If it's mounted between the tool post and the chuck and your moving towards the tail stock it's called a follower rest. If it mounted between the too post and the tail stock and you feeding towards the tail stock it's called a leader rest. Are you following me???

JL.................

BCRider
07-10-2018, 02:58 PM
On deflecting the work..... A good sharp cutter is a big plus. If you can cut a pass then go back without retracting the tool and not leave a spiral mark or cut any more material away then it's cutting with no bending pressure. And that's pretty important for doing long pieces.

PStechPaul
07-10-2018, 03:55 PM
I was thinking that one way to check for wear or misalignment of the bed with respect to the center line of the spindle and the tail stock, might be to stretch a wire between the centers. Then you can run a sharp pointed tool from one end of travel to the other and visually check the position of the point to the wire. You could also mount a continuity checker on the toolpost and align it so it just touches, and then back off 0.001". Run the tool from one end to the other so that it does not register any continuity over the travel, and then adjust so it touches, and continuity should be indicated the whole length. This can check for alignment in either axis.

JoeLee
07-10-2018, 04:26 PM
On deflecting the work..... A good sharp cutter is a big plus. If you can cut a pass then go back without retracting the tool and not leave a spiral mark or cut any more material away then it's cutting with no bending pressure. And that's pretty important for doing long pieces.You can still get some cut on the way back if you have any play in the saddle. And yes ... dead sharp on the cutter is a must.

JL...........

gambler
07-10-2018, 04:41 PM
How did you adjust your TS ??? There are a few different methods.

JL................

with two new dead centers. as close to dead on as I can tell.

gambler
07-10-2018, 04:46 PM
On deflecting the work..... A good sharp cutter is a big plus. If you can cut a pass then go back without retracting the tool and not leave a spiral mark or cut any more material away then it's cutting with no bending pressure. And that's pretty important for doing long pieces.

I get a spiral.

gambler
07-10-2018, 04:49 PM
I bought a test bar on ebay, made in India which has a MT2 at one end, centres both ends and is 11" long overall. It runs better than 0.0001" tir and has been very useful in getting the alignment and height of the tailstock corrected. Don't make the same mistake as me in expecting live centres to be spot on, but use dead centres for testing. I put a piece of 3/8" diameter bar in a collet and turned a 60 degree tip (using a chuck would be ok too), which must be true to the axis of the spindle as long as it is undisturbed. Try the alignment test with the tailstock quill retracted and also with it fully extended to see if there is a difference.

I didn't think to extend the tailstock ram while adjusting, I'll recheck with it all the way out too.

gambler
07-10-2018, 04:52 PM
I'm surprised no one has asked the main question yet. Has the lathe bed ever been tested to ensure it is straight and parallel? A key part of any bench lathe setup is that it be trued up (often called "leveled") before you can trust it to do long parallel cuts.

Now some Logan lathes have a motor and primary drive cabinet. But that does not mean you can just plunk it onto a floor and KNOW that the bed is in flat from end to end. It still needs to be checked.

A good and surprisingly easy way to do this is chuck up a long and stout bar in the lathe that sticks out about 4 to 5 inches. Relieve the middle of the exposed bar by about .04" leaving two full diameter "collars" at the end and in close to the chuck jaws. Now skim cut the two collars and mic them. Using shims under the bed mounting bolts or the adjustable feet on the cabinet twist the bed until you measure the same size on both collars. Note that this does NOT involve using the tail stock. You don't want the TS in the equation at this point.

Once you remove any twist in the bed you can now use the identical size collars to check for any bowing by running a DTI over the tops of the collars. Any difference there is an indication of the bed sagging down or bowing upwards and that too can and should be corrected. Confirm you don't affect the zeroing of the twist by taking another skim cut and adjust as needed until you get same size test collars and a DTI reading the top of the collars shows no difference between them.

NOW bring the TS up to the end collar. If you chose your bar stock well it is roughly the same size as the TS ram. So with the mic'ed size of the collar and the TS ram you can use a dial gauge between them to set the ram to center side to side using just a little math so you know the step it should be. And while you are at it run the dial gauge over the tops and see if your TS ram is also centered for height. If needed put shims between the base and upper piece. And again while you are there extend the TS ram a few inches and check that it's both at the correct height and not angled up or down.

After all that the bed should be as true as it can be. If it still cuts a taper or a barrel or wasp waist long shaft then it truly is bed wear.

But unless you can see the wear quite easily I'm guessing it's just the bed has a bit of twist to it and you need to run this "collar skimming" test and true up the bed.

I recently moved my shop so the lathe has been leveled and rechecked several times over a two month period. I can borrow a level and recheck. the collar thing will probably be next.

JoeLee
07-10-2018, 07:13 PM
with two new dead centers. as close to dead on as I can tell. I found the most accurate way was to mount a dial on the chuck. You can pick your mounting method, magnetic base etc. I made a fixture that I mount in the chuck so there is no dial rod sag when the dial is rotated under the TS quill. Only draw back is you have to chase the dial around the quill.
You can sweep the OD of the quill or the inside taper, compare them. You can mount a good dead center and sweep the taper of the center.
The TS can be OK front to back but can also be sitting high or low. This dial test will tell you where it is.

Others have mentioned putting a center in the spindle and bringing it up to the TS center point to point, some do the cigarette paper trick or even use a thin piece of flat shim stock.
But I like the dial method better.

JL...............

Robin R
07-10-2018, 11:49 PM
A Co-Ax indicator would be a convenient way to check the tail-stock quill, trying to follow a dial test indicator all the way round isn't the easiest thing. When checking the tail-stock it's probably worth testing with the quill retracted and extended, to see if the tail-stock has any sag.

Forestgnome
07-11-2018, 02:58 PM
Easy: re-adjust tailstock so that both ends come out same diameter. (You still might get slight barrel/ huor-glass shape but it will be a lot smaller error)

That's what I do. Just shift the tailstock .005" away from the tool and measure by watching an indicator on the tailstock end of the work.

gambler
07-20-2018, 01:10 PM
I found the most accurate way was to mount a dial on the chuck. You can pick your mounting method, magnetic base etc. I made a fixture that I mount in the chuck so there is no dial rod sag when the dial is rotated under the TS quill. Only draw back is you have to chase the dial around the quill.
You can sweep the OD of the quill or the inside taper, compare them. You can mount a good dead center and sweep the taper of the center.
The TS can be OK front to back but can also be sitting high or low. This dial test will tell you where it is.

Others have mentioned putting a center in the spindle and bringing it up to the TS center point to point, some do the cigarette paper trick or even use a thin piece of flat shim stock.
But I like the dial method better.

JL...............

I intend to do this soon.

gambler
07-20-2018, 01:12 PM
Easy: re-adjust tailstock so that both ends come out same diameter. (You still might get slight barrel/ huor-glass shape but it will be a lot smaller error)

ok so I did this, got the shaft to within .001, good enough for now. thanks.