View Full Version : Collet closer concentricity: I need some advice

01-17-2007, 11:44 AM
I'm almost ready make a collet closer for my new Grizzly 12x24 (G4002) lathe. I've received some good advice on this forum, and I've been researching articles and plans.

I plan on building the closer in the HSM (May/June 2000). I have purchased (from Grizzly) a 5c female-to-#5MT male adapter, so that's one part I don't have to manufacture.

Here's the question:
How important is concentricity between the closer rod and the collet? This is mentioned in the article:

. . . insure good axial alignment of critical parts so the collets are drawn perfectly straight back into the spindle to minimize conical wobble.
OK, I understand that, and I know that concentricity is important. But it seems to me that a slight amount of off-axis pulling would not keep the two cones from mating correctly (Please, no mating-cone jokes.).

I'm still a beginner, and this collet closer will be one of the more complex things I've attempted (Don't laugh--it's true.). I guess I need to know just how perfect my machining needs to be.

I'd appreciate your thoughts.

James (seatlanta)

Spin Doctor
01-17-2007, 12:05 PM
In an ideal world the total runout should be zero throughout the whole assembly. At low RPMs .005" here or there shouldn't matter too much. But as the RPMs increase the load on the system from imbalances (and that is what runout is going to introduce due to the mass of the closer being offset from the centerline of the spindle) will climb. This sets in vibrations which will eventually cause problems with finishes etc. This is where IMO the after market lever operated closers fail the lathe owner. They where never designed as an intregral part of the lathe's Spindle Assembly from the get go. As to the tail end of the closer influencing the closing of the collet on the mating tapers. I really don't think it will effect the accuracy all that much. Most collet work (3 or 4 xs the diameter max)is fairly short unless centers are involved. figure out the length from front to back and the offset to get the angle. Just how many arc minutes or seconds is it going to be. Unless you are using Hardinge Super Precisison 5C Collets a couple of thou is not going to matter too much. Plus if you are using a lever type there has to some clearance in the system to work. The only way I can think of to get a hand wheel closer to run with no clearance is either in a lapped bushing or in preloaded bearing setup on the handwheel.

J Tiers
01-17-2007, 12:59 PM
It is probably more important to ensure that the threads are accurately cut at right angles to the axis. The long lever arm to the back is less likely to make trouble than badly cut threads, I would think.

of course they might have to be fairly bad, but still.

In the end, it is worth doing all parts of the closer as accurately as you can.

Oh, and the "hard part", the collet to spindle taper adapter, is probably the most important part..... Worth doing yourself so you know it is made for your machine.

If you make it, and always replace it in the same orientation as it was made, it will tend to compensate for any runout in your spindle. This is because it was made to as true a cylinder (taper) as possible inside, while inserted in the spindle.

So if the spindle was off-center, the hole in the closer might be off-center with respect to its OD. But it will be on-center with respect to the center of rotation of the spindle, when made.

If replaced in same position as when made, it will therefore be as close as it is possible to get to on-center. A store-boughten part is only as good as the maker made it, and does no compensation for YOUR machine.

Spin Doctor
01-17-2007, 02:07 PM
Good point in that making your own and always installing the adaptor in the sam eorientation will ensure that closest possible running condition. But one thin I always wonder about. Why is it that every time we see a shop made collet adaptor no body ever makes the OD with the Hardinge 4D taper. While I won't go into the 5C vs ER series debate (again, ad infanaseum) to me the major point of the 5C collet system is the whole bag of tricks it employs. Including the step chucks that require the use of a closer ring. Granted a lot of HSMs would never consider the purchase of said items but there is another aspect of having the collet just a little bit further out from the spindle face. The ability to position the tool post more solidly on the compound and not having it hang out at the edge because of the need to get close to the spindle face. Just my .02

John R
01-17-2007, 02:22 PM
I just made one. I used the 4 jaw chuch for all set ups and always set zero-zero. I'm sure is result is not exactly zero-zero, but I figure that's as close as you can get it. I worked fine and that's what matters.
Good luck
John R

01-17-2007, 02:41 PM
There are two somewhat critical aspects to the premade 5MT/5C adapter you bought: concentricity of the 5C part to the lathe spindle bore and how closely the inside of the adapter fits the cylindrical part of the 5C collet, clearance there should be 0.002-3 at most. You may find your collets vary a bit in this area however, and with my set the cylindrical portion ranged over about 0.002 range with possibly one outlier. Since this is premade your only option is to mark the spindle end and try the adapter in various positions with a collet and see if there is any significant variance with the adapter rotated at 90D intervals around the spindle. If you find a 'least wobble' site you should mark the adapter so that it is always inserted in the same way into the spindle. It is assumed that the taper is ok here. Checking that would mean rotating the collet around the taper and checking for wobble when clamped. The question arises of what you should use for a standard collet,
and the best answer might be a new Hardinge or equivalent.

01-17-2007, 05:26 PM
I'm working on a 5C closer tube right now too. I turned the OD to clear my spindle ID by some 0.004 or so, slipped it through the spindle and 4 jaw, indicated it in within less than 0.0005 on the turned OD, then turned the ID to the minor diameter. Should be well good enough I think...