Strange R8 Collet Problem.

[rdsii64]

You could insert a 7/16 collet in the mill, then mount a 7/16"-20 rod in the collet. Screw a nut onto the threaded rod, then screw on the collet you want to work on using the nut to lock it in place. You could also use a drill chuck to hold the threaded rod but if you have a keyless chuck be sure to pay attention which direction you turn on the spindle so it doesn't unscrew.

This is promising. Since I'm going to order another draw bar anyway, I can cut off the top of my current draw bar and then put it in my drill chuck. Spin on my 7/16 collet and sand it untill it fits properly. Then
install the newly modified 7/16 collet, fit the modified draw bar (now just a threaded rod) and sand the rest of my current collet set until they fit properly. If I'm careful and only go half thousandths or so at a time, this should work.

 

[planeflyer21]

Used a similar solution on my Smithy Granite. Got it secondhand with one drawbar. The accessories that came with it had different threads.

Went up to the hardware store and got a couple of feet of all-thread, a big washer, and two hex nuts to serve as a drawbar. Worked great!

 

[rdsii64]

Well, its confirmed. My R8 collets measure .934 at the point where they bind. My R8 drill chuck measures .901 at the same spot on the shank. So that is a difference of 33 thousandths. Sanding 33 thousandths off is a lot of sanding. I'm going to go in search of better quality collets. But since these are basically useless to me, I'm going to try to sand them anyway just to learn something and to let the inner kid out and "play with my toys". If I make them worse, no harm no foul. While I'm here what are some good brands to look at.

 

[janvanruth]

why not take a file to it instead of sanding

 

[EmilioG]

All of my r8 collets are Hardinge, Japan and Germany. No problems. The latest one is an import but
well made Pro series r8 jt3. Great article on collets. I felt that collets were not all made the same which is
why I looked for the best ones I could afford. The keyway spec is: .156 wide x .094 deep.

 

[darkzero]

what are some good brands to look at.

If I was looking to buy new today for quality R8 collets, I'd probably buy Lyndex or Royal. Harding collets are nice but they're pretty expensive.

 

[RJSakowski]

Well, its confirmed. My R8 collets measure .934 at the point where they bind. My R8 drill chuck measures .901 at the same spot on the shank. So that is a difference of 33 thousandths. Sanding 33 thousandths off is a lot of sanding. I'm going to go in search of better quality collets. But since these are basically useless to me, I'm going to try to sand them anyway just to learn something and to let the inner kid out and "play with my toys". If I make them worse, no harm no foul. While I'm here what are some good brands to look at.

Before you sand/grind/mill, the specification for the cylindrical section at the top of the collet is .9495/.949". The central section of the collet is given as .94" or 15/16". and is not precision ground. Since the upper portion of the collet has to move tpast this position n the socket when you seat the collet, anything less than .949 is good.

More than likely, if you are experiencing binding it is either the keyway or you have galling in the upper portion of the spindle socket. I would check carefully with a known good R8 tool before I started to modify any others.

When I bought an R8 collet set, I had several that bound because of a narrow keyway. A bluing test will confirm that interference. I didn't have a Dremel at the time so I eliminated the interference with laborious hand honing. A Dremel will make short work of the problem

As I understand the situation, the R8 specification was an internal Bridgeport spec. and all other manufacturers reverse engineered the Bridgeport collets or clones of same to create their own spec. As a result, it is entirely possible that some latter day manufacturer has created a specification that has significant rounding errors. The same is true for the spindle socket. It is entirely possible that your spindle has a slightly larger key than typical, causing the binding.

 

[chips&more]

Well, its confirmed. My R8 collets measure .934 at the point where they bind. My R8 drill chuck measures .901 at the same spot on the shank. So that is a difference of 33 thousandths. Sanding 33 thousandths off is a lot of sanding. I'm going to go in search of better quality collets. But since these are basically useless to me, I'm going to try to sand them anyway just to learn something and to let the inner kid out and "play with my toys". If I make them worse, no harm no foul. While I'm here what are some good brands to look at.

Something is wrong, maybe it’s me, but I’m getting a big question mark? What does the .934 dimension reference to on the R8 collet? A pic please or drawing, or something of the problem…Thanks, Dave.

 

[Billh50]

Here's a picture:
I still say check the keyway first.

 

[rdsii64]

Before you sand/grind/mill, the specification for the cylindrical section at the top of the collet is .9495/.949". The central section of the collet is given as .94" or 15/16". and is not precision ground. Since the upper portion of the collet has to move tpast this position n the socket when you seat the collet, anything less than .949 is good.

More than likely, if you are experiencing binding it is either the keyway or you have galling in the upper portion of the spindle socket. I would check carefully with a known good R8 tool before I started to modify any others.

When I bought an R8 collet set, I had several that bound because of a narrow keyway. A bluing test will confirm that interference. I didn't have a Dremel at the time so I eliminated the interference with laborious hand honing. A Dremel will make short work of the problem

As I understand the situation, the R8 specification was an internal Bridgeport spec. and all other manufacturers reverse engineered the Bridgeport collets or clones of same to create their own spec. As a result, it is entirely possible that some latter day manufacturer has created a specification that has significant rounding errors. The same is true for the spindle socket. It is entirely possible that your spindle has a slightly larger key than typical, causing the binding.

This is something I failed to look at. After re-reading you post, this sounds more plausible since the top of the collet measures .949 (My caliper actualy measured it at .950) It looks like I will be chucking a drimmel attachment in my drill chuck and SLIGHTLY open up the keyway and see if it works.
first I'll measure the keyway in my R8 drill chuck since it goes in and comes out properly, and If I find a difference, I'll match then with my dremmel atatchment.