View Full Version : Ball turning WHY-CANT- I
01-22-2010, 07:10 AM
I have made a nice ball turner. Sturdy with a heap of adjustment.
But I just cant seem to turn a decent ball. I make odd cone shaped tips, and then when I feed the tool in it takes a bit off the tip but then the cutter ends up miles from the part as I crank the tool around with the ball cutter. I just dont get it...
What gets fed into what? Do I use the saddle or the cross slide?
Or do I go as far as moving the tool inward in the cutter?
I have been able to stand a cylindrical parallel on the center of the body of the ball turner and adjust the cutter depth so that it follows the shape.
But then when I move the cross slide or the saddle to feed the tool in, the geometry is out.
I have tried starting with the tool bit on centre with the stock bar, I have started with it on the outside of the bar but each time, at some point in the cut the tool bit ends up in the middle of no where.
I have found many instruction videos that have been awesome but.. cant seem to find much on this topic.
any ideas where to start..
thanks again, Michael
Your Old Dog
01-22-2010, 07:17 AM
Sounds to me like you are using the cross slide instead of the saddle in-feed. Actually it's not a bad trick for someone who wants something other then a ball. Should come out pear shaped right?
If you are using the saddle then your pivot point isn't concentric with your tool holder on the saddle.
Post some pictures of your setup.
01-22-2010, 08:27 AM
Ideally, your ball turning device will allow you to run the lathe and feed the tool radially inwards - ball turners made from boring heads are ideal for this.
(note - I'm assuming the ball turner has a horizontal spindle)
If your tool is bolted in place and infeed on the ball turning tool isn't easy, set the tool for the finished radius of the ball. Make sure the axis of rotation of the ball turner passes through the centreline of the lathe (ie, it's on centre height).
Start with the cross slide pulled back, run the lathe. Advance the cross slide until the tool starts cutting, swing the ball turning tool back & forth, let it cut whatever it touches.
Advance the cross slide, repeat until the tool is on the centreline; this'll be the last time it touches the work, and will result in a ball being turned.
01-22-2010, 10:24 AM
Can we get a picture of your ball turner? Or a google pictures stock photo of one that operates in the same manner as yours?
01-22-2010, 11:40 AM
With a Holdridge, or similar style radii cutter I will:
Set the tool at a radius about .010" greater than the finished part.
Touch off the "Z" axis on the end of the part.
Back off the cross slide.
Set the carriage as needed to put the radius in the proper location.
Take cuts, by moving the cross slide into the work.
Measure the ball diameter, repeat as needed until .020" over diameter is reached.
Zero cross feed.
Hone the tool, reset to final radius.
Turn on machine, set cross feed to zero, finish pass.
01-22-2010, 02:10 PM
Are you trying to turn a cone with a tip radius?
Getting the ball dimensions to line up with other shapes can get a little tricky, there are various ways to screw up the geometry and the execution thereof.
Are you trying to put a 180 degree hemisphere on the end of something that is already turned to a cone? It doesn't work. The cone has already undercut the hemisphere, as illustrated in the following diagram:
The steeper the cone is, the worse it will be. Subtract twice the taper angle (or the total included angle) from 180. That is how many degrees of arc you need to put a ball radius on the end of a cone. Or to put it less mathematically, setup your ball turning tool perpendicular to the surface of the cone, not perpendicular to the axis of the lathe, at the point where you want the radius to begin. Note the position of your handwheels (and take out backlash) or set a depth stop. Then back off the saddle for your roughing cuts and work your way back to that location but stop a finishing cut short. Stop the lathe, remove the depth stop and touch off again rather than relying on repeatability.
Since you are turning a cone, there is an easy way to touch off without using a square - and probably more accurate since it relies on the tip of the tool not the shank which might be misaligned. If your radius attachment is free to rotate, just advance the saddle into the work gently. The tool will hit the work and pivot to the right starting angle (tip tangent to surface). When it reaches the correct angle, there will be resistance to going any further. Don't push so hard that the tool bites in. Use this for both the initial and final touch off. A double check would be to note the Z position here, back off and repeat the process on the back side of the work; if you end up with a different Z then either your repeatability is off or your cross slide isn't properly set on axis. You can iterate using the error in Z to hone in the cross slide position until it is accurate. If the radius cutter isn't on axis, you will cut an arbitrary surface of revolution of an arc rather than a portion of a sphere.
Note that your cross slide must be set so the radius turning attachment is dead nuts on center - not just tool height but in/out.
Your radius attachment must be set to the same radius as the diameter of the part at the gage line where the sphere and the cone intersect. If you don't really care about the gage line and just want a smooth radius of approximate size, then the touch off procedure above will automatically find the gage line that matches the radius your radius attachment is currently set to.
If the total length of the part is a constrained dimension, rather than the imaginary point on an ideal (non-radiused) cone, then you have some math to do. Here is a diagram with the math:
If you use a calculator, make sure it is in degrees mode if you are using angles in degrees. If your taper is rise inches over run inches, then the trig functions can also be calculated as follows:
sin(theta) = rise/sqrt((rise^2)+(run^2))
cos(theta) = run/sqrt((rise^2)+(run^2))
tan(theta) = rise/run
The illustrated cone had a taper of 1"/4" or 3" per foot or an angle (theta) of 14.036243 degrees. The angle where the tool intersects the cone is also 15.036243 degrees. The cone (without radius) was 200 units long and 100 units high. The radius was 25.769410 units long. The center of the radius was 106.25 units from the tip of the cone. The gage line was 100 units from the tip of the cone. The radius at the gage line was 25 units. And the tip of the radius was 80.48059 units from the tip of the cone. And the distance between the gage line and the center of the radius was 6.25units. And the angle of the sphere was 151.92751 degrees. I didn't include the distance from the gage line to the tip of the radius in the diagram: radius*sin(theta) - radius = 19.519410 units.
01-22-2010, 09:36 PM
I have several experimental ball turners, but the common point is that they all have a pivot axis- this axis needs to intersect the spindle axis, at least for the final cuts. Also, it can't be moved either towards or away from the chuck once the cutting begins, unless you are carefully roughing out the ball. Once the roughing is done, by whatever means makes sense to you, then lock the carriage, then feed with the crosslide only. It would be best if you could put a stop on the slide such that the pivot axis of the ball turner can approach and meet with the spindle axis, but can't pass that point.
You can feed in with both the crosslide and the tool in the ball turner, but once the crosslide has reached that stop point, the tool is the only thing being adjusted. As you swing it back and forth, it will be cutting a perfect sphere, then the tool is adjusted in the ball turning device to get a deeper feed, and bring the ball to the desired diameter.
At no time should you pass the cutter past the 90 degree mark until the slide has reached the stop- otherwise you will end up with a dimple in the end of the ball. Only when the pivot point intersects the spindle axis can you go past 90 without undercutting the ball, but be careful because when the cutter reaches the spindle axis and passes it, the material will then be wiping up from the bottom of the cutter (unless you reversed the spindle). This final point, the pip, is probably best finished off with a file, since the cutting action will have ceased- just like when you're facing something and come to the center point. Here is where having the cutting edge perfectly on height will get you closer to removing the pip before you have to file.
01-22-2010, 09:53 PM
I measure the diameter of the unfinished ball,then the diameters at 45 degrees on either side of the 90 degree measurement. Then, adjust the compound( which should be parallel to the lathe's ways),until the ball measures the same all around. Finally,I adjust the cutter in the ball turner carefully,and make more cuts till the ball is the right diameter.
I'm probably saying something wrong here,as it's been some time since I made a ball. As said,get the cutting edge of the cutter exact at center height. My 8" Holdridge just fits perfectly on my 16" swing lathe,with just a bit of height adjustment on the yolk.
01-22-2010, 11:19 PM
thanks for your help everyone, being such a common problem it had probably been done to death so I was a lucky to get any help at all.
I will go and have another go at it with the advice I was given. I will take some photos while I am out there..
01-22-2010, 11:25 PM
It's not because you only have three fingers is it? (only joking , couldn't help myself) If your compound has a lock use it.
01-23-2010, 07:34 AM
I have a 4" rotary indexer I put on a plate.. remove the compound, bolt this on solid.. then set a HSM donated lantern tool post onto the indexer on a T-slot plate.
Cranking the indexer rotates the lantern toolpost..
I got blisters on my hands.. so put a stepper motor on it.. then used the stepper for something else now so no pictures..
01-23-2010, 08:26 AM
its ok mate, your fine words create the most vivid pictures in my mind.