PM 5C Collet Chuck not attaching to spindle

[andrewgr]

Please excuse my embarrassing lack of vocabulary in describing these components. I've consulted the "manual" that came with the chuck, and it is in Engrish, and I can't really make heads nor tails of it.

I'm trying to use a Precision Matthews 5C-ADJ-D1-4, a 5C collet chuck that they sell to go with the 1340GT (and other models).

However, the cams won't lock. They turn maybe a 1/5th turn, not enough to properly lock the chuck in.

It looks to me like the problem is that the 5C collet chuck has a "lip" that prevents the cam from turning, though I'm not certain. I'm probably using all the wrong words, but hopefully two pictures will make it clear.

Here's a closeup of my 3-Jaw chuck, which works fine:

Notice the half circle, which the spindle is supposed to lock into? The curve starts right at the edge of the backing plate. When you stick it into the hole on the spindle, 100% that's inside the spindle is curved, there's no flat spot. When you turn the cam, it smoothly slides into the semi-circle depression, locking it in.

Here's the 5C collet chuck, which I can't get to work:

Notice that the half circle doesn't start up against the backing plate. There's a very perceptible flat that juts out, that goes into the matching hole in the spindle. When I turn the spindle cam to lock it in place, I believe it's catching on that flat part.

I've tried each of these pieces in each of the three holes on the backing plate, and they won't screw down far enough in any of them.

Is this a red herring? Is there something else I should be investigating? The cam will turn a little bit, which will keep it in place if no pressure is applied, but wiggling it even a small amount will make it come loose.

If this is the problem, is there any reasonable solution other than sending it back?

 

[Ken226]

Remove the lock screws, Loosen the cam pins out a full turn, put the lock screws back in, and try fitting it again.

I can clearly see in your pic that the pins of the non-fitting chuck are threaded further into the back plate than in the chuck that fits. The symptoms of that would manifest themselves exactly as you describe.

If it's better, but still not right after loosening out 1 turn, loosen out another turn.

 

[markba633csi]

I would call PM, if it is a manufacturing error you should plan on returning it
-Mark

 

[Ken226]

If the cam pins are threaded too far into the chuck, the spindle cams won't rotate far enough to lock between the locking index marks.

If they arent threaded in far enough, the cams will rotate 360° and not lock.

Thread them in far enough that the cams stop with the cam mark indexed between the two locking marks on the spindle. Do this for each pin.

 

[xr650rRider]

There is a groove around the circumference of those pins. Normally that is flush with surface of backplate. So like Ken226 said, they are screwed in at least 1 turn too much.

 

[Aukai]

I just messed with an old chuck on my new lathe yesterday, I actually had to screw mine in two full turns, one went 3 to lock.

 

[andrewgr]

There is a groove around the circumference of those pins. Normally that is flush with surface of backplate. So like Ken226 said, they are screwed in at least 1 turn too much.

Thanks, that did the trick! I never would have figured that out on my own, I needed to loosen them up like 3 or 4 whole turns. But now it works great.

 

[andrewgr]

So now I'm trying to adjust the runout. I've got the runout on the taper hole of the collet chuck down to about .025mm. Unfortunately, the manual says that it should be within .006mm. So I'm off by a factor of 4.

The manual says that the "hexagon socket set screw with cup point can adjust it". The screws in question are located on the outer diameter of the chuck, on the same face as the screws that take the chuck key (and connect it to the face plate via the cam). They're located closer to the headstock than the chuck key screws, and are smaller. However, I just spent 30 minutes screwing individual hexagon socket screws in as far as they would go, and out all the way, one at a time, in twos, in threes, every combination I could stand. I tried screwing them while the cams were locked and when they were unlocked. At no point in time did the dial test indicator move so much as a micron. As far as I can see, they simply do nothing; or at least, they don't seem to have any impact on runout.

Is there a particular method for leveraging these screws to fine-tune runout? The method that I used to get this close-- loosening one or two cam nuts at a time, and pushing/pulling/gently tapping the chuck into position-- is way too crude for me to have any hope of adjusting by .019mm.

 

[Christianstark]

Are their 2 hex screws at each location? Take one of the hex screws completely out, and see it there is another beneath it. If so I think the outermost hex screws are meant to lock everything in place once the chuck is adjusted using the innermost set of screws.

 

[xr650rRider]

Did you loosen the 3 screws in the face of the chuck? When those are loose, then you use the 4 around the perimeter to move the chuck relative to the backplate. To move it, you have to loosen the one 180 degrees from the one your trying to tighten, can't do 1 without the other. You need to tighten all the D1-4 cams as they have nothing to do with adjusting the runout of the chuck. Now having said that, there can be slight differences in runout depending on orientation on how you install the backplate. Every 120 degrees can be slightly different runout. So it's a good idea to match mark spindle and chuck, so that you re-install it next time in same relative position. My 3 jaw from PM has 1 position where runout is .0015 and the other 2 are greater than .005". With my 5C and ER-40 collet chuck, I can get the runout down to around .0001" on the taper, that doesn't mean when you install a collet that it will remain that low. You might get lucky but seems most are closer to .0005"-.0007", using ground gauge pins in collets.