View Full Version : Help fitting a collet chuck

11-30-2010, 01:21 AM
After I get the bearings adjusted I want to fit my new ER32 collet chuck to my Colt.
I have a nice backplate
And shiny new collet chuck

I assume the drill is to turn a shoulder on the faceplate and drill the three mounting holes. I have never fitted a faceplate so bear with me.

Do I just put a small shoulder 0.100" to locate. Or slightly less that depth of recess?
How much clearance on the diameter? I dont think I want a tight fit as the reason I went with this style is to be able to dial it in.


11-30-2010, 01:43 AM
Woo Eee Dave, your up late :D
Never put on a Back plate in my life, so I should learn something here.

One thing doubtless is a must is that you face off the backing plate installed on your lathe as well as face off the OD.

The Collet chuck is obviously precision ground and must maintain center with the face plate, that is the issue that I see that must be addressed.
Looks like I'd turn the face plate to accept exactly the ID of the inner step on the chuck and then the three holes can be drilled/tapped to match, even a little error on the bolt pattern should not be a problem as center is maintained via the inner step of the chuck.

Still, as mentioned, I've never done this so just food for thought.

11-30-2010, 01:49 AM
I will wait for the experts to chime in but I think you want some play to adjust for true. I don't know. That's why I ask. ;-)


11-30-2010, 01:53 AM
While that is not how I would approach it, waiting for further opinions would be wise. :D
Once you face the backing plate and turn the step, installed on your lathe, it will be as perfect as can be.


11-30-2010, 02:00 AM
You're working toward something, what is it ?

11-30-2010, 02:12 AM
I want to get the Barker moving along. I need to make some leadscrews up. I have the Nook leadscrews stock but need to make some bearing mounts etc. A collet chuck should make it easier since indicating a 4 jaw on a leadscrew seems like a hassle. Other post is about getting the play out of headstock or figuring out whats up. I need to hold a fairly good tolerance for the bearings. My 0.010 taper of an inch won't do.I doubt its my bison chuck that is flexing but anything is possible. If I have to I will visit my friend and get him to do the ends on his Haas but would like to do it myself.


11-30-2010, 02:30 AM
I've found a very capable method to do just what you are requiring.
Unfortunately, Cant find it.
still looking though.

I saved the link somewhere.

11-30-2010, 02:34 AM
This is a great read, it was very unfortunate that he passed away, way too early in life.

irregardless whether you use it, still a good read.


11-30-2010, 03:02 AM
Thanks for that link! A good read indeed. Hopefully I can dial in the chuck. When I was at the workshop in Ann Arbor a neat trick for centering ball screws was shared. You wind a piece of wire in the grooves then chuck the helix in your collet chuck. Rick C was the originator of that gem!

John Stevenson
11-30-2010, 05:12 AM
Skim backplate in position.
Turn short shoulder 60 thou or so until chuck just fits on tight.
Bump mounting holes thru with transfer punch.

Remove and drill.

Then deepen the register and get the diameter to 10 thou under side.
Make certain you have a sharp corner or undercut to prevent interference.

Bolt chuck on lightly and dial in.

job done.

11-30-2010, 07:11 AM

This is interesting. Following on from John's posted method, I would like to ask why you can't just turn a register that is a tight fit in the chuck backplate as you would do with a conventional chuck? Is there something inherently different with theses ER backplate-mounted chucks. Not saying any one is wrong , just curious as to why.

Rgds. Ian.

John Stevenson
11-30-2010, 07:25 AM
Ian the idea is that the little bit of leeway allows you to tap the chuck so it's running true.

Sounds a bit of a mackle up but it's been around for years.
Some of the most precise machines made, cylindrical grinders, that have to work to microns, proper microns that is , not Canadian ones :), have no register and the tool of choice for the operator is a square block of lead about 3" cube that he taps the chuck with to get bang on.

In theory chucks and collets should hold dead on centre at any given size, truth is there are always tolerances.

Once Dave has got his chuck setup and running true, chances are he won't have to touch it again with a decent set of collets.


11-30-2010, 08:19 AM
Ian the idea is that the little bit of leeway allows you to tap the chuck so it's running true.

Basically, there are two ways to address the problem of runout:
1.) Make it as precise and repeatable(!) as possible, and only tackle runout if it is too high
2.) Have some slack and always be forced to make the chuck run true.

Ad 1):
If you mark the backplate/chuck how it was machined/ground initially you are good enough in many cases. If that is not true, like for precision work or generally when doing grinding there are other solutions that are commonly accepted:
1a) a Hi-Tru chuck
1b) in cylindrical grinding, a magnetic chuck (like a faceplate) that holds a 3 jaw chuck. It can be taped in place. This is the most flexible solution, because you can even make excenters. A friend of mine is a grindor at a major bearing manufacturer grinding gages that way. Re-clamping and aligning to 0.5 Ám is common practice for him. :eek:

Ad 2):
I personally don't like that method, because you always have to align the chuck, even for work that doesn't have to be highly accurate. It gets a waste of time whenever you have to change chucks, BUT NOT if that chuck sits on a backplate that precisely registers on the spindle. The later resembles the idea of an Hi-Tru chuck.


11-30-2010, 10:39 AM
Thank you everyone for the suggestions. Once I ensure that the spindle bearings etc are in order I will proceed. I am glad I asked. Thanks again.