View Full Version : Help double conical thingamadoodle from last HSM issue

06-20-2010, 02:00 PM
I picked up the last issue of HSM magazine wherein they showed a desk toy consisting of a mildly inclined "base" consisting of two rails, and a dood-dad that runs on it. The geegaw looks like two cones glued together.

I looked at that and said, "gee! 1 part, simple geometry, I can do that one!" Yeah...not so much. He does the thingamadoodle by turning in steps, then shaving off the steps to make the first 40 degree cone, then flips it over in the chuck by gripping the cone (presumably not hard enough to mar it) does the second cone. Basically, this is what I'm trying to do but it's not working for me.

I'm basically what's happening is I get done with one cone, flip it over, and at some point the tools just rips the work right out of the jaws. Then of course it's "start over" because now the work has big divots in it both from the tool and the jaws.

I'm on about round 7 of "start over" and I thought I'd stop and ask your help. I'm using a sharp tool, speed is within the range recommended by Mach. hndbook. Of course, on a cone it doesn't matter where you set the speed it will also be too slow near the tip OR too fast at the OD. "feed" is manual (compound) From the picture below, you can see what's doing, including the divot from the last tear out.

I seem to have fewer problems (or at least the horror is spaced farther apart) if I cut from the headstock towards the tailstock (base towards terminus). But it looks like when I do this there is some strange back and forth wobble in the work. This may be an optical delusion because when I grab the chuck and pull back and forth I cannot feel any motion. I can't discount the idea that this is the transaxial component of the cutter force causing the crosslide to wobble back and forth by the range of the backlash. But my cross slide has no gibs, and I'm not seeing it when I look. I do know my crossfeed not has a lot of slop in it an probably needs changing out. This is a 12" 1944 Atlas Craftsman. Note the author says to use steel. I don't have any in the proper size so I cut the stepped version in wax and cast it in scrap brass. What your seeing is after the steps having been turned off of both sides, and the OD reduced multiple times as I've had to machine off divots and get a new grip.

The point to the left you can't see, but it has to be redone because the tip is bent from smacking the cement after being thrown out. The tip to the right I'm happy with. So ANY help you could offer that would allow me to remachine the tip to the left would be deeply appreciated. Also, as much as I would like to have the full double cone, I'm not wedded to it. So, some "flat" in the middle is fine. But that also raises the issue of how to machine off the divots IN the chuck in the picture without then losing concentricity after the flip or having a "step" in the flat section.

Also, as you can probably tell I'm am a hair raising sliver of light away from crashing the tool into the jaws when trying to machine where I am. I would deeply appreciate any guidance you could give me. I was hoping to take this on a trip next week to a family reunion where I can "WOW!" all 13 of my nieces with Unka Joe's gravity defying doodad.


06-20-2010, 02:10 PM
For starters, I would use a collet if possible, or at least a 4-jaw. They hold much better than a 3 jaw.

06-20-2010, 02:16 PM
Why not turn the first cone then remove the part and fit a slug of ally in the chuck and turn an inside cone in that at the same angle, then glue the brass part into the ally one to machine the other side?

06-20-2010, 02:23 PM
Well if you have the setup to cast brass, I assume you can also silver solder. How would you feel about machining two cones and then soldering them together. You could then use a file to clean up the joint if needed. You should be able to eyeball it and get the two cones very nearly concentric. A couple thousandths shouldn't matter.

Alternatively, you could make some soft jaws for you chuck and cut a taper in them so that as you clamp down on the piece, it engages the whole tapered section. If the angle is too steep (i.e. it still ejects the work piece as you tighten the jaws) you could use the tailstock to keep the part in the chuck.

The flat section would work well with a collet - you are less likely to run into big concentricity errors with a collet than a chuck.

Or how about some kind of fixture? Cut a matching taper on a piece of scrap aluminum or steel and then super glue the part in?

<EDIT> I see Peter types faster than I do...

06-20-2010, 04:06 PM

Very interested in the soft jaw/internal cone idea. Can you all walk me through/point me to a resource for learning this? It may be straightforward to you all, but for me issues like matching the angles, etc, often prove difficult.

06-20-2010, 04:12 PM
Matching the angles is easy - just don't move the compound! You turn the brass cone on the front with the lathe turning forwards and turn the matching internal cone on the back with the lathe running backwards, or if you are worried about your chuck un-screwing turn the inside cone on the back with the tool inverted.

06-20-2010, 04:18 PM
How about machine the tips to the proper angle first, then make two recessed parts, one held in the chuck, the other with a center hole drilled in the back to fit over your tailstock. clamp the entire deal beween tailstock and chuck, and then its just a matter of moving a dog around (to keep it outta your way) as you turn the rest of the taper.

Also brass is natorious for grabing tools. Are you using a 0 or negative rake tool? If not, you should be.

06-20-2010, 04:25 PM
On brass you don't want any top rake on the tool and you need a radius on the cutting edge. Sharp pointed tools will do what you are experiencing. Also set the tool slightly above center.

Apparently I don't understand what you are trying to do as I don't see a problem based on the info provided and the picture.

06-20-2010, 06:57 PM
The sharp point isn't really necessary for the thingie to function properly, and a flat end with a center hole would make it easy to stabilize the piece for machining. It would be more attractive and much safer, too.

IIRC, the two rails aren't level or parallel. They slope downward and get closer together from left to right. The double cone will roll from right to left, appearing to defy gravity.

06-20-2010, 08:20 PM
Yes, a flat bit on the end would certainly help the machining process but I thought it might take away from the aesthetics.

jgourlay - Do you have a four jaw chuck or a collet chuck?

Ideally, you would chuck up the blank and cut one taper. Then chuck up a piece of scrap and, without moving the compound, cut the taper on the inside and also take a skim pass on the outside to make sure that the outside is concentric with the inside. This will allow you to indicate off of the outside when you dial it in.

In more detail:

For a concrete example, I am going to pretend that your cone is 1" in diameter at the largest point.

Chuck a piece of 1.25" scrap metal in the lathe and drill a 3/8" hole through it. (This size is unimportant. Basically it is just clearance for a boring bar which will cut the taper. Obviously it needs to be smaller than the base diameter of the cone or you'll only have "line contact". You could go all the way up to 7/8" if you wanted, but then you'd have a pretty tiny surface area contact).

You may find it easier to step drill the bore to rough out the taper - similar to the steps that are cast in the brass. This isn't neccessary but will probably speed things up. Once you have your hole made and without ever moving the compound from its original angle setting, you will use the compound to feed a boring bar into the hole and cut the taper. Now it will match the taper on the brass but it will be going "the wrong way".

I suggest that you now also take the time to take a skim pass on the exterior of the scrap. This will give you a surface that is concentric to the inside and is necessary if you plan on using your three jaw chuck or a collet. It's not needed for a four jaw chuck but may be handy anyway.

Once you have done that, you can remove the piece. Turn it around and chuck it up again. If you have a four jaw, you can indicate from a place in the bore or from the area that you cut on the exterior of the part, which may be easier and less sensitive to longitudinal motion (old four jaws sometimes push the work outwards as you tighten them. This means that the indicator will move off of your reference point as you tighten since the indicator point is moving along a cone - not a big deal as long as you are aware of whats going on. If you're not, you may spend the next 20 years trying to dial it in just right ;) )

If you are using a collet or three jaw there isn't much you can do to improve concentricity. There are a few tricks for sorry three-jaw chucks. You can try rotating the piece until you find a position that has less runout or shim one jaw slightly.

Once you've done that, put some super glue on the brass part and "clamp" it in place using the tailstock to provide pressure. Take a cookie break and come back in twenty. Now you are ready to machine. If you can manage it, leave the tailstock there to apply pressure and keep it from popping out (which is unlikely but possible if your tool geometry is wrong and "grabs" it).

The soft jaw idea is probably out. Your three jaw chuck pictured doesn't have a provision for bolting soft jaws on so to make a set of soft jaws you would have to cut the "teeth" on the back of the jaws that engage the scroll. That is certainly possible, but without a k&T 2D (:D) it would take some ingenuity. On the other hand, they probably don't need to be very accurate - just repeatable so you might be able to widdle them out in short order. Never tried it myself.

06-20-2010, 09:19 PM
When I made my version of this toy I drilled and tapped a 1/4-20 hole in the end of two pieces of 1 inch brass. Then I put a peice of 1 inch steel in the collet and drilled and tapped a 1/4-20 hole. Then it is easy using a short piece of 1/4-20 ready bolt to fasten one of the peices of brass to the steel and turn a cone. Then do the same for the second cone. I actually placed a steel disc 1/4 inch thick by 2 inch diameter between the two cones to get a better flywheel action.
My rails are 1/2 x 1/2 x 1/16 inch Aluminum angle which makes it real easy to fasten feet and spreader bars as needed. I also have a polished brass bar that I use first in the demo. The brass bar rolls the opposite direction.