Logan crossfeed dial acting...weird

[daveog]

So I finally started actually making chips with my project lathe this past week as I've "found some time" as of late. Last night I was turning a part down to diameter and was really perplexed at what was happening. I always try to sneak up on final dimensions to get them right on. I was coming up on my final diameter and was within 0.080 or so. I was dialing in 0.010 per pass but it was consistently taking 0.003. So, I backed off and checked on the graduations to make sure I understood how the dial was supposed to work. As I figured, the indicated depth on the crossed dial is designed to indicate the diametric change per cut, so a 0.010 indication is taking 0.005 in depth, but 0.010 in diameter. I confirmed it with a precision ruler and the dial was accurate on large scale movements. So, my 0.010 cuts should have taken 0.010 in diameter. What I noticed was this. I always make sure to take out any backlash in the gears before I start to cut. I would start my cut by touching off on the part and setting the dial to 0. Then, without backing off, dial in 0.010 and take a pass. I stopped the machine, measured, and the cut was 0.003 in diameter change. Without touching the dial setting, I touched off on the part again and instead of the dial indicating 0.010 where I set it, it indicated 0.003. So, something is happening to affect the amount being fed. I hope this makes sense. Does anyone have any ideas as to what is happening and how I can fix it? DRO is not an option, I tried installing one but the hardened steel on the saddle broke 5 bits trying to drill for taps and I never even got to tapping. It's hard to make precision parts when you don't trust the dials though.

Thanks!
Dave

 

[ErichKeane]

How sharp is your bit, and how rigid is your setup? It is not at all uncommon to have a cutter 'flex' away from the material and take a much smaller cut as a result. You're supposed to do a 'spring pass' (where you don't change the depth of cut) as you near your target. You'll find that you end up getting (depending on your initial DOC) 2-3 spring-passes worth of cut.

To visualize, put an indicator on your carriage up against the tool/toolholder and watch it as you start your cut. I'm sure you'll see quite a bit of movement that goes away after the cutter is unloaded.

I have a massive, very rigid setup and still get significant cuts on my spring-pass unless I've been taking tiny passes.

 

[Alexander McGilton]

Im confused as to how to interpret your diagnostic process. Lets start from the beginning with some basic questions. I have solved similar problems at Dawson and Mcgill. One where the DRO scale was loose making for a double back lash and faulty clutch that was self feeding the cross slide. both problems diagnosed by a dial indicator on a mag base, and solved by adjusting a few screws.

What Model of Logan do you have?
Does it have cross slide power feed and/or tapper attachment?
does the handle de-clutch with the power feed?
How much do you progress per revolution?
how many increments on the dial, and how are they displayed.?

A shot in the dark as a possible cause, there is rotational play on the dial between it and the screw.
You are approaching the tool to the work piece, all backlash is taken up. you touch then zero the dial, inadvertenly adding .007" of back lash to the dial. you take ten thou of cut but only proceeds .003".

Try placing the dial indicator on the carriage, then measure the back lash and feed increments. see if these all add up.

 

[daveog]

Im confused as to how to interpret your diagnostic process. Lets start from the beginning with some basic questions. I have solved similar problems at Dawson and Mcgill. One where the DRO scale was loose making for a double back lash and faulty clutch that was self feeding the cross slide. both problems diagnosed by a dial indicator on a mag base, and solved by adjusting a few screws.

What Model of Logan do you have?
It is an 80's model Powermatic, I can't tell you much more beyond that. It is the variable speed model.

Does it have cross slide power feed and/or tapper attachment?
The cross slide has power feed, but not taper attachment

does the handle de-clutch with the power feed?
I'm not sure what you mean by this.

How much do you progress per revolution?
I have tried many different speeds, but generally slow.

how many increments on the dial, and how are they displayed.?
The dial has 0.200 per rotation, marked in 0.001 increments displaying diametric depth

A shot in the dark as a possible cause, there is rotational play on the dial between it and the screw.
You are approaching the tool to the work piece, all backlash is taken up. you touch then zero the dial, inadvertenly adding .007" of back lash to the dial. you take ten thou of cut but only proceeds .003".
I don't see how I could be inducing backlash. I am taking backlash out by feeding toward the work and holding the handle while zeroing. I think I would notice a 0.007 turn on the handle.

Try placing the dial indicator on the carriage, then measure the back lash and feed increments. see if these all add up.

 

[daveog]

How sharp is your bit, and how rigid is your setup?
I am using a new AR Warner HSS bit. I have also tried new carbide. The setup is as rigid as a 12x36" lathe can be. I'm using an Aloris tool post.

It is not at all uncommon to have a cutter 'flex' away from the material and take a much smaller cut as a result. You're supposed to do a 'spring pass' (where you don't change the depth of cut) as you near your target. You'll find that you end up getting (depending on your initial DOC) 2-3 spring-passes worth of cut.
I can't imagine that cutting 6061 Aluminum (should have mentioned that originally) with new cutters at a 0.005 depth would spring that much. I end up doing a dozen small cuts to get to my final dimension because a 0.010 pass only gets me 0.003 change.

To visualize, put an indicator on your carriage up against the tool/toolholder and watch it as you start your cut. I'm sure you'll see quite a bit of movement that goes away after the cutter is unloaded.
I will try this, but again, I have a hard time believing it is a flex issue. It is a very consistent number. No matter the speed and feed, a 0.010 dial gives me 0.003 diameter change. I have seen this through a dozen or more passes.

I have a massive, very rigid setup and still get significant cuts on my spring-pass unless I've been taking tiny passes.
Wouldn't you consider a 0.010 pass a tiny pass? It's only a 0.005" cut?

 

[ErichKeane]

Factory Carbide tends to not be particularly sharp, they keep them somewhat dull I've seen for the purposes of industrial situation. I don't think that pass is too deep (TBH, I have no idea how rigid a 12"x36" lathe is).

Make sure you're cutting on center as well, I've seen this in the past with a tool that is either too high or too low. You end up cutting less than you want, and get a bunch of material deflection.

Since you can 're touch' off and see only the .003 cut, I have to imagine that it isn't the dials at play, but some sort of deflection causing the problem.

 

[daveog]

Factory Carbide tends to not be particularly sharp, they keep them somewhat dull I've seen for the purposes of industrial situation. I don't think that pass is too deep (TBH, I have no idea how rigid a 12"x36" lathe is).

Make sure you're cutting on center as well, I've seen this in the past with a tool that is either too high or too low. You end up cutting less than you want, and get a bunch of material deflection.

Since you can 're touch' off and see only the .003 cut, I have to imagine that it isn't the dials at play, but some sort of deflection causing the problem.

It certainly isn't the most rigid setup because of it's relatively small size, but it is fairly solid. I've checked the centering of the tool and it is dead on center. You make a good point about the dials accurately showing the actual cut. I will take another look at everything and make sure nothing is moving. Perhaps I will tighten the gibs on the compound to make sure it can't move.

 

[Nogoingback]

Since you're getting a smaller cut than what you dial in, the tool must be "backing out" ( not moving at a linear rate) of the work, possibly
as a result of cutting force.
On my (older) Logan, I can get "backlash" from my cross slide as a result of two factors. The first, of course is
slop between the crossfeed nut and screw. The second factor results from how I install the crossfeed handle on
the screw itself. My machine has two nuts that secure the handle: one on each side. The inner nut's adjustment
on the shaft affects how much slop there is between the crossfeed screw and the crossfeed casting. When I
install the handle, I carefully adjust the inner nut to minimize slop and then tighten the outer nut to lock the handle to
the screw.
Without knowing which machine you have it's hard to tell if your machine is assembled the same way mine is,
but it's worth a look. After you first loosen the set screw on the dial, try to push/pull the crossfeed handle in and out without turning it. If you have any slop there, then that may a source of movement as you feed the tool into
the work. You could probably test that with a dial indicator while taking a cut: set up the indicator against the back
(near) side of the crossslide or tool post, set up for a cut the way you always do and take the cut while watching
the indicator.

 

[daveog]

Since you're getting a smaller cut than what you dial in, the tool must be "backing out" ( not moving at a linear rate) of the work, possibly
as a result of cutting force.
On my (older) Logan, I can get "backlash" from my cross slide as a result of two factors. The first, of course is
slop between the crossfeed nut and screw. The second factor results from how I install the crossfeed handle on
the screw itself. My machine has two nuts that secure the handle: one on each side. The inner nut's adjustment
on the shaft affects how much slop there is between the crossfeed screw and the crossfeed casting. When I
install the handle, I carefully adjust the inner nut to minimize slop and then tighten the outer nut to lock the handle to
the screw.
Without knowing which machine you have it's hard to tell if your machine is assembled the same way mine is,
but it's worth a look. After you first loosen the set screw on the dial, try to push/pull the crossfeed handle in and out without turning it. If you have any slop there, then that may a source of movement as you feed the tool into
the work. You could probably test that with a dial indicator while taking a cut: set up the indicator against the back
(near) side of the crossslide or tool post, set up for a cut the way you always do and take the cut while watching
the indicator.

Yes, I will definitely look at the crossfeed handle screws. Here is a picture of the lathe, it's better than a description I could give.

 

[eeler1]

Dang! That sure is pretty for it's age.

Not familiar with the Powermatic version, but most Logan lathes have an indirect (or maybe direct, I get them mixed up) reading on the dial, i.e., you dial in 0.010" and you actually remove double that, or 0.020", since you are taking 10 thou off each side of the work. I don't recall any lathe where you dial in 10 thou and only take off 5, but I could be mistaken.

Only suggestion I can make is to check the gibs on both cross slide and compound, tighten them to where you get resistance on the cross slide and can barely move the compound. See if it makes any difference.