Collet chuck

Can anyone help?

I have bought a collet chuck from a reputable supplier to fit my Myford ML7 lathe. Having machined the faceplate as per instructions, the closest I can get it to true is 0.08mm (0.003&rdquo.
Is this acceptable or is there something I can do to achieve higher accuracy?
Spent three hours trying but can’t do any better.

The lathe runs true.

Your register on the backplate can be turned a fraction undersize then with the fixings just nipped up the chuck can be tapped about to get it to run true at which point you can fully tighten the fixings

Where / how are you measuring runout?

Its something that always puzzled me that one would take care to machine a register accurately and then not use it as a precise and repeatable locating device.

The only function of a loose fitting spigot and recess is to roughly align the two before getting out the Manchester screwdriver!

I suppose its an acceptable method for something that is not intended to be taken apart.

Ian P

Apologies if this is a Granny & eggs situation, but have you tried rotating it around the three different hole positions on the backplate?

..there's usually a 'right' one..

There are several surfaces in a collet chuck which ALL need to be true if the end product is to work properly (ie first time, no messing). The holder taper (or thread) needs to be true in the lathe mainshaft, the inner collapsible cone true to the hole in it, etc. What you get can depend on what you ask for, and how much you pay (but a high price does not always mean real concentricity). And it is difficult, when you open the box and put everything togethe,r to work out which bits are most wrong. And even more difficult to convince the seller ...

Most of us struggle on with 'good enough I suppose' most of the time, and only need to fiddle with dial gauges and light taps with plastic hammers when something serious is in hand. Real professionals (ie people who work for firms which supply the kit) are beetter served (usually) as fiddling about costs real money.

Cheers, Tim

I have tried rotating using all holes in all positions.
Run out has been measured on the outer surface, at the back of the threaded section and inside the chuck. All are pretty much the same.

Tried tapping in with nylon hammer doesn’t move if tightened, works when slightly loose but runs out again as I tighten up.

Just noticed my original post looks strange. The (0.003&rdquo has morphed into something unrecognisable. I do sometimes post typos but it certainly was nothing like that.

Done it again...

(0.003"

Edited By Stephen Follows on 05/06/2023 18:02:38

Edited By Stephen Follows on 05/06/2023 18:03:09

The undersize register is often suggested here. It's the easier option than doing like I did with my 5C chuck which was to machine a stub of bar and then tighten the chuck onto that before skimming the back face and female register true before making it's backplate. As I got the chuck new for £20 I was happy to do the extra bit of work to it.

Most people only find the runout issue after machining their backplate so unless you want to make another with a larger register then the knock it into place is the common option.

Edited By JasonB on 05/06/2023 18:24:54

The horrible automatic smileys are a feature. When computers only had text screens, no graphics, it was the height of sophistication for folk to make emoticons by concatenating characters. Loads of possibilities such as:

• smile : - )
• sad : - (
• confused :-J

When everybody got graphics screens it became fashionable for software to show off by automatically converting text emoticons into their graphical equivalents. This forum does it, which is very unfortunate because engineers often use symbols like " that have been stolen by the emoticon crew. The text combination " ) is a wink. I stopped " ) being converted into a winky in this post by inserting an extra space. Type (0.003" )

Back to run-out, my guess is that the new collet chuck is a close fit to the register and either the register is worn, or the matching part of the chuck isn't quite true. So when the nuts are tightened the register pushes the chuck out of alignment. If a tight fit, I suggest opening up the chuck slightly to create room for it to move relative to the register. Then centring by tapping as per Jason's advice will work.

Registers wear a little every time they're used. How good or bad they are depends on the history of the lathe and the years are often unkind. A lathe that's spent its life constantly switching chucks is unlikely to have as good a register as one owned by a gentle hobbyist who did everything with the same 4-jaw.

Dave

Stephen,

Always type a space before typing a right-hand bracket - that leaves the smiley gremlins sleeping in peace!

Also, always type a right arrow after quoting from someone else's post and before typing your reply - that stops the left-hand quote line from following your post as well as theirs.

Best regards,

Swarf, Mostly!

The runout is not described properly, is it radial or something else? Assuming that the register is running perfectly radially and the face of the backplate has been skimmed, and the collet plate fits it snugly, where is the runout measured from? The OD of the plate or the bore? With a test bar held in a collet? Has more than one collet been tried? Is the collet bore concentric with the register in the back of the plate? With a test bar in a collet, if the only error is radial, then the error will be the same right up against the collet as it is 3" further out.

There is nothing wrong with having an undersize register, the bolts holding the collet plate and friction will never allow unwanted movement. In the real world, you have to make these things work, not just say it is not allowed.

Stephen,

I had exactly the same experience. I won't quote the suppliers' name. It was an ER32 chuck.

Such investigation as I was able to perform suggested that the front and rear ends of the actual collet chuck were not concentric. (Or do I mean 'coaxial'?? ) As I had bigger fish to fry at the time, I proceeded no further.

Best regards,

Swarf, Mostly!

Dave, we are talking the freshly cut register on a back plate not the spindle that may wear or get damaged over time

Chucks are usually hardened to some degree so easier to just take a bit more off the backplate register

Edited By JasonB on 05/06/2023 19:29:16

UPDATE ON CHUCK

Spent some more time time today trying to get it true. Although the backplate had been machined and was running true the whole assembly wasn't. Did some measurements on the chuck part and found it to be out by about 0.1 mm in depth. Took a light skim off it and refitted it to the backplate. Ten minutes later everything running true.

I thought the chuck would be perfectly machined, the instructions only mentioned machining the backplate.

Glad you got things running well. I have an R8 er25 collet holder that has problems with similar runout which would probably need internal grinding to rectify. The other two that we have are fine. You also have to be cautious with cheap collets, they can be poor if you are unlucky. I usually buy the sizes to match milling cutters from APT, they have three grades that you can choose from, and good prices.