Run out for collets

[Vern Smith]

I would like to get a general idea as to how much run out is considered normal for ER collets. I did some testing tonight with ER 32 and ER 16 collet holders. I used a .5" carbide blank in the ER 32's and a .25 in the 16's. The ER 32's were fairly consistent at .001 run out at the face of the collet and just a little more 1.25" from the face. The ER 16 were a little better, .005 at the face and once again a little worse 1" from the face.

I swapped around holders, collets, and nuts to get a representative sample. I torqued all the nuts to recommended numbers. I also checked the spindle and it was out .0002 so not a big factor in the evaluation.

Vern

[makingchips]

I would like to get a general idea as to how much run out is considered normal for ER collets. I did some testing tonight with ER 32 and ER 16 collet holders. I used a .5" carbide blank in the ER 32's and a .25 in the 16's. The ER 32's were fairly consistent at .001 run out at the face of the collet and just a little more 1.25" from the face. The ER 16 were a little better, .005 at the face and once again a little worse 1" from the face.

I swapped around holders, collets, and nuts to get a representative sample. I torqued all the nuts to recommended numbers. I also checked the spindle and it was out .0002 so not a big factor in the evaluation.

Vern

Vern,

I indicate all tools used in ER collets. I use a best test .0001 indicator on a Starrett inspection indicator base. I made a hard plastic tool to fit nicely around the end mills and drills etc. I then tap them ever so lightly with the plastic tool and a very small jewelers hammer. It usually does not take much to get the tools within .0005 or better.


MC

[Astonlee]

ER Collets Increase Carbide Cutter Tool life by up to 150%, check out this page to learn more: http://www.mctooling.com/index/listings/page972.htm

[Vern Smith]

MC

I need a little more information on your system for tapping tools into good run out alignment. Are you doing this before you torque the nut?

It does sound like you feel that .0005 is about as good as it will get with ER collets.

Vern

[makingchips]

MC

I need a little more information on your system for tapping tools into good run out alignment. Are you doing this before you torque the nut?

It does sound like you feel that .0005 is about as good as it will get with ER collets.

Vern

I'm doing it after I torque the nut while the tool is in the spindle. Very lightly I might add. When run out is real bad above .002 I take the tool out and clean the collet. That almost always fixes it.

That is as close as I try to get them, after all I’m not making triggers for an atom bomb lol.

MC

[Geof]

A tapping technique you can use that does help the tool settle into the collet is while you are tightening the nut.

You can probably imagine how the tool cants slightly in the clearance in the collet before you start tightening it. As you tighten the nut the hole gets smaller and the tool is forced to straighten up, but friction exists so even when the nut is fully tight and the collet is closed up tightly onto the tool it is still a tiny bit canted. To correct this before taking it fully tight just tap the tool all round with a little bit of plastic; gently tap, you are not trying to knock the tool sideways but just give a little shock so the spring in everything can close down on the tool as it is shocked away from the canted position. Then fully tighten.

You can do the same thing in the machine if you want to dial them in but all the tapping does is allow the tool to take up the most precise position within the collet. If it still runs out a bit there is nothing more you can do.

And it is all a bit redundant because as soon as the tool starts cutting it is probably going to jiggle into the most precise position anyway. But not totally redundant because as the tool lines up parallel to the collet it is not held quite so tight therefore there is some utility to doing it on the bench while tightening.

Regarding how good collets are I have seen specs from Lyndex that claim 0.0002" or something like that within an inch or so of the collet face and not more than 0.0005" two or three inches away.

[Andre' B]

Hopefully without completely hijacking the thread.
Anyone have anything to say about HPI's MX tool holders?
They worth the money?
http://www.gohpi.com/HPI2007Update0807MX.pdf

Got a job with some access problems and it looks like these or heat shrink holders.

[Geof]

I hope Vern forgives the hijack.

They look interesting but I think you need to read:"The MX system is designed for finish operations", and: "Note:
���� Cutting Tool Shank size must be h6 tolerance to work properly or cutter sticking or slipping may occur", and notice that it seems the pull stud has to be removed to tighten the collet drawbolt.

[Vern Smith]

My general knowledge of machining and CNC are so sketchy I consider the hijacks informative.

[xiondavis]

My technique is to tighten the collet, then loosen it and re-tighten... same idea, let the spring in everything relax a bit and re-tighten.

[Astonlee]

I found sometimes it's not the collet that the problem, often as not by just rotating holder by 180 degrees you cure it, but it's not so easy has that with a machine that using CAT-V (Din 69871) but at least you can test toolholder by removing the larger drive key from spindle and manually loading tool.

[Geof]

I found sometimes it's not the collet that the problem, often as not by just rotating holder by 180 degrees you cure it...

So you are using runout in the taper to compensate for runout in the collet. Nothing is perfect, all collets will have some error, all spindle tapers will have some error; this is a case of two wrongs being used, maybe not to make a right, but at least, to make things less wronger.

Dead simple to do on a Haas, the holders go in either way.