Is Taper Attachment worth it?

[PeterT]

Edit - didn't even get to sleep and realized this was wrong. A ball in a cone will never be an ellipse. It will always be a circle and will work just fine.

What's even more frustrating is I think I got to the bottom of this issue once before & subsequently forgot it.
This might not be incorrect so don't hold me to it, but seems like I did something like so. I tried to evaluate quick & dirty in CAD

make a cylinder with two 60-deg cone center drills

if you make an angle on the end, that is equivalent to setting the bar offset in TS & slicing the end in cross section. I know we are not turning the end, we are turning the OD, but it allows us to look at the bar end specifically what remains of the center drill but from TS oblique perspective.

project the remaining cone outline onto normal plane. By that I mean how the TS sees it staring at the HS

SW says the outline is an off-center ellipse

Yay, CAD confirms if we slice a cone at an angle we get an ellipse when viewed normal to cone axis
So I don't think a spherical ball will fit obliquely into an elliptical hole with full tangential contact?

 

[Dabbler]

Hey, Peter, something is taking you down the wrong path in your analysis. and I'm not sure what. If you use a ball centre with a cone and change the angle as for taper turning, you get a wear mark evenly along the taper. I have done this before and it works. that way. Since all the tapers I have done are Morse tapers at 3 degrees plus a bit, it is impossible to say if a more oblique angle will result in a non-perpendicular wear ring.

All I know is that it works just fine, with good contact along the ball. Once it is worn in a little (perhaps a thousandth of an inch), the whole thing feels quite solid once the centre is tightened up- and then no further intervention is needed for the turning. If the wear ring was an ellipsoid the X and Y axes must have been darn near identical - it looked like it was a perfect circle. Or too close to know better.

 

[Susquatch]

SW says the outline is an off-center ellipse

I did end up sleeping on it after all. Even in my sleep, it is a circle.

If you drop a ball into a cone and then rotate it any way you want, it stays in the same contact circle on the cone. So the trace of the ball on the cones surface must be a circle that is slightly smaller than the OD of the ball according to whatever the tangent of the cone angle is.

HOWEVER, the trace of the cone on the ball is elipsoidal! That is because the contact point on the ball changes from low to high as the ball rotates.

The trace of the ball on the cone is a circle, but the trace of the cone on the ball is elipsoidal.

Now picture my mind following that path! It's no wonder many people think I'm crazy!

Anyway, I think that is what your software is seeing. It all depends on your reference.

What really matters though, isn't the trace, it's the usability of the contact surfaces for positioning. In this case, I believe the ball and cone works just fine. The geometry allows for a solid contact that does not wobble and maintains both longitudinal and transverse position. So an adequately lubricated ball and cone will work.

I still maintain that a ball and socket would be better because the loads are spread out over a much larger area more like a regular center.

 

[Mcgyver]

For me, its one of those things that comes with the lathe, i.e. I usually just wait until a lathe comes up with everything because its all too bloody expensive to buy separately. They are not often needed, but when they are they really are. That is infrequent enough that I wouldn't buy one and doubt it would add a lot of resale value, but you should kick your ass once a day for a week for getting a lathe without one I'm surprised Murphy hasn't done it yet by presenting you with dozens of long precision taper cutting requirements.

 

[Susquatch]

you should kick your ass once a day for a week for getting a lathe without one

I can't even touch my own toes let alone kick my own ass......

My gearhead lathe was purchased new 12 years ago. A taper attachment was not even optional back then.

I think you and others are right about the resale value of a taper attachment. It's stupid money from that perspective.

So it totally boils down to how much use I could get out of it and what my alternatives are.

I'm surprised Murphy hasn't done it yet by presenting you with dozens of long precision taper cutting requirements.

You are likely referring to my smithing hobbies. I would agree. But barrels are normally sold with existing tapers that are pre-ground. A non-tapered barrel is usually more expensive than pre-tapered. It makes no sense to re-grind or re-cut them. The most I have ever had to do of that kind of work is to polish them. I use my old girl for that.

Murphy is alive and well in my shop though. I've done several farm repair jobs that required a long tapered shaft. I did them on my old girl (circa 1880s) by offsetting the tailstock.

 

[Johnwa]

Nobody has mentioned an electronic leadscrew?

 

[PeterT]

Yes ELS would be nice in its own right. I'm jealous of the threads people can make by simple programming buttons vs mechanical alterations. Maybe one day <sigh>. But for taper turning you would also require second ball screw assembly in the cross feed direction, no?

 

[Susquatch]

Nobody has mentioned an electronic leadscrew?

It's true, that's an option.

But it would have to control my cross-slide not my carriage.

And it would also send me perilously close to falling down that CNC Rabbit Hole that I'd prefer to stay well clear of. I lived my whole career in an electronic control world and in retirement I'm really enjoying the old-school mechanical as much as I reasonably can.

The best analogy I can think of is the pure raw joy of driving an old stickshift muscle car vs a modern CVT automatic. Different strokes for different folks. Ain't it awesome that we all have that choice.

 

[PeterT]

HOWEVER, the trace of the cone on the ball is elipsoidal! That is because the contact point on the ball changes from low to high as the ball rotates.
The trace of the ball on the cone is a circle, but the trace of the cone on the ball is elipsoidal.

Yes, I agree, sounding familiar again, same conclusion I eventually came to. A ball within a cone will always make a continuous tangent line within the cone at any angle. So that means perfect line contact between the 2 parts in TS setup. Now that same tangency projected to the ball or normal plane perpendicular to TS axis is an ellipse. But turns out that doesn't mean anything at all in terms of supporting the part with a ball ended center. Its just a geometric artifact.

I think where I was going with this at the time was center drilling with a ball end mill vs a conical center drill because it might be easier to control depth & therefore resultant angle/offset relationship. But it looks like more futzing & I'm not sure ball & socket (area contact) is any better than tangent (line contact). Maybe worse if more friction, more heat, more problems. I think the only reliable way to set up offset angle is start with a straight cylindrical bar, measure deflection with indicator, convert that to target angle, converge to target. But same issue, your part maybe be of somewhat random length between tangent centers and may or may not offer the benefit of suitable straight section to run indicator down. That's where a taper setting bar shines. Once set on a coupon bar, its done. Make one, make 100. Its limited to shallow deflection angles but really so is supporting within centers. Heavy cutting is probably not a good thing.

Turning a ball tool center is certainly doable. I was apprehensive about silver soldering a bearing ball on a shank. I was looking at what kind of premade options might be out there. I see these 'tooling balls' think were used in fixture setup or alignment / measuring. They were spendy so I didn't pursue it.

 

[PeterT]

Tooling ball example

 

[YotaBota]

I use the TTA every chance I get. The handles for my handwheels were beyond the limits of the compound and the TTA was easier to setup than resetting the TS so I'm glad my machine came with the TTA. Handle knobs, if not ball shaped, are much nicer with a taper and I've done a bunch of them with the TTA.


What the TTA is worth is strictly up to your wants/needs. I like having the QCGB as well but it's not a need just a lot more convenient than changing gears. As mentioned before the finish is better and more consistent than turning the compound by hand and much easier on the getting older wrists.

@Susquatch - the balls in your court, do think it's worth it for you? Might be an idea to see one and have a demo before making the final decision. There must be a member close that has the TTA and is willing to show it off.

 

[Susquatch]

Might be an idea to see one and have a demo before making the final decision. There must be a member close that has the TTA and is willing to show it off.

I love this idea. You are absolutely right. This will say more to me than any description ever could.

 

[jcdammeyer]

My two cents here. First of all my ELS does taper turning. A ball screw isn't needed as long as the taper is always turned so the screws remain loaded rather than moving across the backlash each time. I have turned MT2 and tried R8 on my Gingery with good fits given the quality of the lathe itself.

OK. So let's forget the ELS or a moment as my South Bend already has a taper attachment so I've never bothered adding the cross slide motor for the ELS.

Have I used it? You betcha. Under power feed the quality of the surface is better than I've ever gotten with the compound turned manually.

Where have I used it recently? My Mill spindle has a cone onto which the drive pulley sits then pressed down with a nut and a tabbed locking washer. When I cast a new two step pulley so I'd have room for a slotted encoder disk I needed to cut the same taper on the inside of pulley and I wanted the belt grooves to be 100% concentric with the internal taper.

So the first thing to do was to mount the original pulley in the 4 jaw and dial in until the internal hole had no runout.

Next although I had a rough idea it then mounted the dial indicator and enabled the taper attachment and tweaked it until it tracked the taper bore of the pulley.

Next I had to make an arbor to be able to hold the pulley blank between centers since the lathe isn't really large enough to easily cut the two pulley grooves in one setting. So I turned an arbor that was reamed and key cut to slide onto my between centers arbor.

A quick fit with the original pulley and some bluing showed it fit perfectly and no runout. I used the original mill locking nut to hold the blank in place after I'd mounted the cast blank and taper bored it to size.

After that cutting the pulley grooves was easy. Just tedious.

This would have been harder if the compound. Doable perhaps but harder.

 

[thestelster]

I love this idea. You are absolutely right. This will say more to me than any description ever could.

I think it's time to go visit your sister, and while you're at it come over to my place and play!!

 

[Susquatch]

I think it's time to go visit your sister, and while you're at it come over to my place and play!!

This I can do! I have a few goodies that I want to bring you anyway! I'll let you know when I'm coming well in advance!

 

[140mower]

I'll let you know when I'm coming well in advance!

........ You do want them to be home... Right

 

[eotrfish]

Here's what I use for TS offset tapers.

My TS live centre has interchangeable points. I just happens that one of them is a 60Deg cup point. Combined with a 1/4" bearing ball it works great.

 

[Dabbler]

That's brilliant @eotrfish

 

[Former Member]

In case you want to use your compound, here is how to set it accurately.

 

[PeterT]

My TS live centre has interchangeable points. I just happens that one of them is a 60Deg cup point. Combined with a 1/4" bearing ball it works great.

That's cool @eotrfish
So how do you drill your stock ends when using it - standard conical center drill or ball end mill profile...? Do you also use a ball support on the headstock end?