New chuck and making a backing plate.

[vederstein]

My lathe is old. Quite old. 76 years old. As such it also has a 76 year old 8" three jaw chuck.

The chuck is worn out. Runout is awful. At the lathe's highest spindle speed (1100 rpm), the whole lathe vibrates and therefore that spindle speed is unuseable. I know it's from the chuck because there is no vibration when I have the faceplate attached.

It's time for a new chuck, but what kind?

I still wanted a three jaw for most work. I have an 8 inch four jaw when needed. So what size of 3 jaw? Usually people always want to go bigger, but 8" (and larger) chucks have an issue, the smallest stock they can clamp on is limited. In my case it's 5/16" diameter.

I wanted smaller clamping, so I need a smaller chuck. After doing some research, a 6" seems like a good option, but what kind?

There are various types of three jaw chucks:
Front Mount
Rear Mounting
Solid Jaws
Two piece jaws

As for the mounting, I was kind of stuck with a plain back with a backing plate. My lathe spindle is 2 1/8" - 8 thread. It's a bit of an oddball.

In the end I decided on front mounting, plain back 6 inch with two piece jaws:

http://www.shars.com/products/view/...lf_Centering_Scroll_Lathe_Chucks_2_Piece_Jaws

I wanted two piece jaws so I could make soft jaws if necessary and front mounting is just easier than back mounting. I went with Shars tool for both price, availability, and that the specifications are about the same.

I also needed to order a faceplate. I purchased a finished metric threaded plate knowing I'd need to completely re-machine it. No big deal though...

http://www.shars.com/products/view/8468/63_M39x4_Threaded_Back_Plates_

Decision made I purchased the parts. When they arrived, the tricky bit of machining the back plate can begin...

 

[vederstein]

This procedure is just how I managed to do this operation. I know there's a lot of old hats out there that will tell me I did all sorts of stuff wrong. But this worked for me...

I chucked the faceplate in the four jaw backwards. The I faced off the two vertical surfaces and turned the OD of the rear boss. This is more for looks and to get everything concentric and balanced. The dimensions aren't critical at all here. Just clean up the part.

 

[vederstein]

Then bore out the center to the required minor diameter. After that thread the part.

To get the boring and threading bars perpendicular to the back plate, I used a 123 block pressed against the earlier machined surface then tightened down the QCTP.

When boring 60 degree internal threads, it's critical the compound be 29 degrees to the left not the right.

Last year I made a back plate for my four jaw. When I did, I made a replica of the spindle. I used this part to check my thread fit.

 

[vederstein]

If you look at the spindle replica from the earlier post, you will see the spindle is threaded, then has a slip fit bore.

So after the threads were made, the slip fit bore is bored out. This is why the faceplate is mounted backwards, so I can make the bore, which is just larger than the major thread diameter.

Then I removed the four jaw and mounted the new back plate on the spindle. If it doesn't fit, you're boned and have to start over with a new backing plate blank.

From here I faced off the front and turned down the OD to the same diameter as the new chuck. These two cuts are not critical.

 

[vederstein]

This is the absolute critical step!!!

The boss on the back plate must be machined for a tight fit to engage the recess on the rear of the new chuck. When getting close, I was only cutting "dust". Many many trial fits were required. Because if I made the boss too small, I cannot guarantee the concentricity of the chuck.

I've read of some people heating the chuck and freezing the back plate to get an interference it. I've never gone to that extreme.

What happens if it's screwed up? Face off the front and try again. This works until there isn't enough material left. So realistically, there are about 3 tries before the back plate is scrapped.

 

[vederstein]

Next is to remove the back plate and chuck. I used a transfer punch to mark the mounting hole locations. This is why I prefer a front mounting chuck.

I drilled out and tapped the back plate and after mounting the chuck I tested for improvements.

As they say, the proof is in the pudding. With my old 8" chuck, I had a runout of about 0.010". With the new chuck, I had a runout of 0.003". That's a hell of an improvement. I had hoped it would be a little better, but the 0.003" is within the specifications of the chuck.

Also I used a rod from a magnetic stand because I don't have a ground rod to truly check for runout. And again, if I really need super accuracy, I can make soft jaws or use my four jaw.

Also, the high speed vibration is gone.

Overall, a successful project!

http://youtu.be/PyHcQzZ-gcI

 

[petertha]

Timely & informative, thanks. I just posted this question & seeng your similar operation helped me visualize it better.

http://www.homemodelenginemachinist.com/f13/chuck-back-plate-adapters-19937/

- maybe I missed it, but did you pre-machine the matching threaded arbour specifically to check the thread fit on the adapter plate as you were machining it?

- those blank plates from Shars are quite a bit less expensive that what I was looking at locally (Canada). I asume they are cast iron? Did they machine reasonably well? I was also thinking of getting a larger OD blank to make a mounting plate for rotary table.

 

[xpylonracer]

Another advantage of the front mounted chuck is it's simple to remove from the lathe and fit on a rotary table or dividing head for milling operations with the part still in the chuck thus ensuring concentricity.

 

[vederstein]

petertha:

Concerning the test arbor: I made it when I was making a back plate for my four jaw. At that time I could mount my three jaw and my faceplate. I took very accurate measurements of the spindle and recreated it on the arbor.

I used the faceplate to check for fit of the arbor.

For this job, when I got close to my final thread dimensions, after every cut I would test the thread with the arbor and check for screw fit. Once the threads were to my liking, I did the same thing when boring out the slip fit.

The main thing when doing this is not to rush. You're making tooling to make parts. Take light cuts on the important details. Check for fits frequently. To do this complete job took me about four hours. But when you look at it, there really aren't many steps involved. Just lots and lots of checking.

I believe the Shars back plate was cast iron. It certainly cut like cast iron. I didn't have any issues cutting them. I purchased a plate already machined then re-machined for my uses. This way I wouldn't have to cut through the crust on totally unmachined plates.

As for a rotary table, I don't know your application, but Shars has a several rotary tables already mounted with chucks. Below is a link for one of them...

http://www.shars.com/products/view/8011/8quot_HV_Rotary_Table_with_6quot_3_jaw_Chuck

 

[powderhorn01]

vederstein
Welcome to the fascinating world of the Sebastian Lathe's. As soon as you said the thread size, I knew what your lathe was. D.C. Morrison Co. still has some parts for these old critters. Also you can find some info over on the yahoo Sheldon site.

 

[vederstein]

Powderhorn:

Thanks, I didn't know anyone had parts for these things still.

 

[robcas631]

Excellent machine! I can see how the four jaw chuck makes a difference!