Help wanted: How to zero runout on Hardinge collet chuck

Long story, please be patient.

I bought a No. 1 Sjogren Speed Chuck by Hardinge Bros. at a local yard sale. Great deal. It is well used, but in excellent condition.

I mounted it on my lathe (a Chinese model, NOT THE PROBLEM here). I bought a D1-6 adapter plate and machined the alignment boss to match the collet chuck and drilled and tapped the mounting holes. My runout on the adapter plate is about 0.2 - 0.3 thou, which is good for my lathe.

For final collet chuck testing, I used a 1/2" hardened dowel pin in the collet chuck. My runout on the dowel pin is 0.010. Yikes. I cannot tap it into alignment. There is no way to adjust the runout of the collet chuck that I can see.

I have only three options that I can see:

1. Enlarge the four mounting holes on the collet chuck so I can move it a little when mounting. I Do not like this method. I don't want to modify the collet chuck.

2. Turn down the heads and upper shanks of the mounting screws (3/8" SHCS) to allow the chuck to move a little for alignment, then tighten it down.

3. Turn down the alignment boss so the chuck is held only by the four screws. Hope there is enough room in the chuck screw holes to tap into alignment before tightening.

Is there a better approach?
Any help is appreciated.

Thanks much,

LM

LiquidMetal said:

Long story, please be patient.

I bought a No. 1 Sjogren Speed Chuck by Hardinge Bros. at a local yard sale. Great deal. It is well used, but in excellent condition.

I mounted it on my lathe (a Chinese model, NOT THE PROBLEM here). I bought a D1-6 adapter plate and machined the alignment boss to match the collet chuck and drilled and tapped the mounting holes. My runout on the adapter plate is about 0.2 - 0.3 thou, which is good for my lathe.

For final collet chuck testing, I used a 1/2" hardened dowel pin in the collet chuck. My runout on the dowel pin is 0.010. Yikes. I cannot tap it into alignment. There is no way to adjust the runout of the collet chuck that I can see.

Click to expand...

Correct. No adjustment was provided for. Hardinge just shipped them dead-nuts accurate.

OTOH? They aren't "supposed to" be able to even develop excessive TIR, and generally do not do ... no more than about ten or fifteen percent of the runout you measured, anyway. So a touch-up regrind to correct "lobing" the collet generated from wear usually covered the need.

Do try another collet, pin or drill-blank FIRST!

Unless it has been crashed, your one sounds as if it has a bump, burr, or bad collet. Look for evidence of such before doing anything further.

Does this chuck have a flat back? If so, you can probably turn your adapter plate into something like an Adjust-Tru backplate, using long setscrews to center the chuck before tightening it down. This will require some more-or-less minor mods to the chuck, not just the backplate. For example, you could sink the setscrews entirely in the backplate, with slots to accept short dowels added to the back of the chuck. Setscrews would bear on the dowels for positioning. The advantage of this over "tapping into position" is that you get positive control that won't be messed up when you torque down the fastening bolts.

Personally, I don't see any problem with modifying a chuck. But you will have to gain 0.010" of wiggle room one way or the other.

However, if your chuck is bent out of whack due to some past crash, centering alone won't fix your problem.

Take your test dowel (or a longer one) and try to determine if the angular alignment of the chuck is correct. I.e., with the spindle in a fixed rotation, mount an indicator on the carriage and run it up and down the length of the test dowel, looking for changes as you change the carriage position.

BTW, don't neglect to clean the chuck and the collet you're using scrupulously. Properly, you should test the collet for runout on a different device (different chuck, spin indexer, etc) before blaming the new chuck. [Added in edit] Thermite beat me to this point. Consider this a strong second.

I have 3 Sjogren chucks, all with D1-6 backs, and one has a related problem due to "work" a previous owner performed. He turned away most of the internal taper on the chuck back (presumably to make it fit better on his D1-6 spindle nose), and the result only bears on a ring about 0.050" wide right at the chamfered tip of my lathe's spindle nose. Every time I mount this chuck, I have to spend 5 minutes playing with the D1 cam tensions to get it pointed straight. Every time I swear I'm going to fill it in with bronze and turn a new internal taper, but haven't done that yet.

sfriedberg said:

...test the collet for runout on a different device (different chuck, spin indexer, etc)

Click to expand...

+1 on the rest of that post...

As to not having "several" other collet options..

.. one can "get there" simply by applying a coupla sharpie marks, one on the closer nose, another on the collet nose - then "clocking" same-same one-each to different relative positions each go, and recording where the high/low points are, relatively, and if the TIR changes, etc.

The long-rod test may tell a sad tale even sooner if it was pranged when a long and unsupported workpiece whipped and tried the classical leap over top the compound rest.

Ten-thou out is a LOT for a(ny) collet system better than an old-skewl two-prong carpenter's hand-brace.

I fixed my D1-3 Sjogren by chucking the a piece of stock in a four-jaw and turning a stub to the size of my biggest, bestest 2J collet. Then, without disturbing the stock, I clamped the Sjogren on it backwards and cleaned up the back registers. I did a little trig to figure out how much to take off the flat and taper and maintain the fit. It dialed in within 0.0003 when all said and done.

Thanks much, guys.

I determined, based on the comments, that the collet chuck probably did crash. I put a 4 thou shim on one side of the adapter plate and the runout went down by more than half. Also, there was precession (aka "wobble", a sign I think of crashing) which disappeared with the shim.

So, I still have to work out 0.004 runout which tapping will not achieve. My plan is to turn the D1-6 adapter plate to shrink the boss a few thou to give me more room to tap.

Thanks rklopp -- I was going to do the same thing, but I am afraid my lathe is not stiff enough to hold the collet chuck and do it right.

LM

your #3 option is where I would start...nothing to damage there and it's nice to be able to tap true when needed...should be enough room to allow .010"


.002-.003" on your plate is not good IMO...would look into that too....even on a "cheap" lathe it should do better, and with D1-6 it's not exactly a lightweight.