PDA

View Full Version : typical runout in 5C chucks?



mikem
12-29-2010, 03:37 PM
My friend came over last night and we got the backplate machined for my new CDCO 5C collet chuck. We checked the backplate first and it was right on the money when screwed on the spindle of my South Bend 10K for a test fit without the chuck. Less than one thousandth +/-.

He turned the backplate with a ledge to match the chuck's indentation and I drilled and tapped the backplate. When we got it mounted on the spindle and checked it for runout with a dial indicator, we got about .0015" of movement in a half inch diameter drill rod in the collet perpendicular to the spindle axis about 2 inches from the chuck.

This was the first try and am wondering if I should try to improve on that or be satisfied? I could try changing the torque on the mounting screws or using some shims.

Are most collet chucks dead on? This precision is better than my 3 jaw by a long way.

macona
12-29-2010, 03:44 PM
Does the runout get worse as you move from the chuck? If not then the chuck is not concentric. Turn that shoulder down a bit so you have some room for adjustment and just set the screws semi tight. Use a rubber or leather mallet to tap it true and tighten the bolts.

Shims are a bad idea.

Paul Alciatore
12-29-2010, 07:04 PM
Checking it with one collet is not conclusive. Also the drill rod may not be completely straight or completely round.

First you need to make comparative measurements both at the collet and then a couple of inches out from it. Mark the high spots for each one and see if they are on the same size and the same amount.

If they are the same, then mark the high spot on the chuck and rotate the drill rod 180 degrees and check again. If the high spots are at the same side of the chuck a second time and high by the same amount then it is probably the chuck or it's mounting. If they follow the drill rod's rotation, then it is the rod. Likely, you will get a bit of both. If so, then average the readings at the collet and average the readings at 2" out and these averages will eliminate any errors in the drill rod and reflect the error in the mount and collet.

But you still don't know if it is the collet or the chuck/mount that is off.

You could try several different sized collets. See if the errors are consistent for all of them. Again, you probably will see some variation. Once again, averages will probably help to eliminate the individual collet's errors and tell you about the chuck/mount.

If you get an indication that the problems are primarily in the chuck/mount, then you can remove the collet and use the indicator on the inside bore of the collet holder. Indicate on the tapered surface that the taper of the collet fits against. Ideally, this should be dead true. If possible, also indicate on the rear of the hole, where the rear, cylindrical section of the collet fits. This surface should also be dead true: if it is not, the collet will be at an angle to the axis. If these two, inside surfaces are running true, then the mount and the chuck are perfect. If not, then you can not expect even the best of collets to run true in it.

topct
12-29-2010, 07:36 PM
We must consider the fact that both the front and back of the 5c chuck was finished on another machine.

Accuracy costs money.

justanengineer
12-29-2010, 08:38 PM
My Sjorgen has around .0005

DKS
12-30-2010, 04:27 AM
No expert here but I had the same question the other day because I've noticed on serveral machines with collet chucks minor run out, I was led to beleive they were always dead on, minus all the variation when checking a piece,( work or rod, collet, chuck, spindle etc). So I went and asked my teacher, he told me all the collet chucks will have some runout and over the years and abuse etc they get worse only talking 2-3 thou tops over the years with good Hardinge and Sjorigen(spelling?) chucks and collets. So I would say .0015 isnt too bad but could maybe be improved with work, but .0015 could be all the variation in the drill rod, collet, chuck, backplate, etc. like I said no expert just had the same thoughts the other day.

tlfamm
12-30-2010, 08:23 AM
Last Fall my son thought he could visually detect wobble in his newly purchased Asian 5C collet chuck after the mounting plate had been trued on the lathe. He mounted a 3/4" precision ground round in the collet chuck, and then mounted the other end of the round in a 4-jaw chuck, itself mounted on the lathe, the 4-jaw adjusted to yield .0005+ runout on the round.

He found .017 wobble in the outside body of the chuck, and .008 and .006 wobble on the vertical mounting surfaces on the back of the chuck (the chuck mounting plate had been removed).


The vendor reluctantly authorized return of the chuck as he was quite dubious about our findings. Later, he confirmed that our chuck was indeed _seriously_ out of whack, as were the two or three others in his inventory. [I assume that these chucks are CNC-produced(?) and don't understand how they can be this inaccurately manufactured.]


We eventually upgraded to a Bison collet chuck, and once it was dialed in, no runout could be detected (at least with our instrumentation).

============================
Bob Warfield has a thread on truing an Asian collet chuck:

http://www.cnccookbook.com/CCColletChuck.htm

Tony Pratt
12-30-2010, 08:43 AM
To digress, being CNC produced does not guarantee anything, concentricity and accuracy can be compromised by all manner of things, swarf and incorrect loading are prime suspects. The only way to check these collets are all good to specification is 100% inspection which I doubt the manufacturers do in this case!
Tony

tlfamm
12-30-2010, 09:18 AM
re: incorrect loading:

a good point, and we might imagine that the loaders are under-paid, under-trained, and under-motivated.



re: 100% inspection rate:

I'm developing the impression that after the typical Chinese manufacturing line has had its initial debugging, it is allowed to run on auto-pilot, virtually inspection free. If there is an accumulation of consumer complaints and vendor annoyance, then a factory-manager might saunter on down and see what the fuss is all about.

jugs
12-30-2010, 10:27 AM
The vendor reluctantly authorized return of the chuck as he was quite dubious about our findings. Later, he confirmed that our chuck was indeed _seriously_ out of whack, as were the two or three others in his inventory. [I assume that these chucks are CNC-produced(?) and don't understand how they can be this inaccurately manufactured.]


We eventually upgraded to a Bison collet chuck, and once it was dialed in, no runout could be detected (at least with our instrumentation).

============================
Bob Warfield has a thread on truing an Asian collet chuck:

http://www.cnccookbook.com/CCColletChuck.htm


CNC = speed + consistency, NOT accuracy - if the first one is out then the whole batch will be the same or worse (due to tool wear).

john
:)

philbur
12-30-2010, 12:31 PM
Why use drill rod to check the holder run-out, why not clock the holder taper? That way you avoid two additional sources of error.

Phil:)

Carld
12-30-2010, 03:41 PM
Why even test it with a collet? Put a dial indicator in the taper area in the bore where the collet goes and see what the runout is.

I have been considering buying a collet chuck and if the Bison is that good I may go with one.