5C taper problem again

[GlennW]

I've pretty much ruled out deflection for a couple of reasons. The main reason is that I can put the workpiece in my 3 jaw chuck, make the same cuts, and there's no taper. Also, I've done this many times with different depths of cut.

I covered that on page 2.

When held with a chuck the bar has lots of support, but a collet has a very minimal contact length in comparison which would allow for more deflection with a cutter that isn't cutting properly.

[Russ Hanscom]

Are you saying that lack of parallelism between the workpiece and the spindle doesn't cause taper

YES - it really does not.

Put a bent bar in the three jaw or intentionally offset a bar in the four jaw. You will be able to turn a straight cylinder on the wobbling bit.

As an extreme case, you can make a small crank shaft by lots of offset in a four jaw.

[rexcsmith]

OK, now I get that the only thing that will cause a taper is the work piece flexing or moving away from the cutter, assuming the cutter is moving parallel to the spindle axis, which it is. That's progress on this end. hmm

Actually, when I put the precision bar in the 3 jaw Bison, I get less than .001 runout at 3", rather than the .002 I get with the collet. No taper on a cut.

This collet chuck is as rare as hen's teeth. My Monarch 10ee is 1979 vintage and the collet chuck/drawbar assy for it is about the same age. I know some places to ask around and will probably do so. Nothing like I'm looking for is on eBay or Craig's though.

Let me tell you about something else I tried. I put the caliper against the work piece right at the collet and reefed on the work piece with my hand. There was about .0004 (4 tenths) deflection. I then put the caliper against the nose of the chuck and reefed. I still got some deflection there, but more like .0002. This is uncharted ground for me, as I don't know what to expect. Should I see any noticeable deflection at all? rex

[spro]

In the first page I mentioned the tail of the drawtube. It didn't make sense because afterwards it was a collet chuck. I don't mean to get in the middle of this and confuse but it is something to think about. I made a drawtube and nose cone for 5C to fit my lathe. I was getting weird results from the collets. Under tension the tube was wrong or something, so a made a spider type cap where I could adjust the tail end. I failed in that. Something was happening in that tube under tension and pressure at one side. I don't have the answer, largely because my 3-jaw is so accurate.
I just relate this so you know that you aren't the only one to have this type problem.

[ctwo]

I was going to suggest clamping a longer test bar and reef on it with an indicator on it so you can see how much spring your machine has and if it returns. I can get a few tenths, close to a thou deflection on my regular chuck with indicator mounted to the tool post, but it comes back when I remove my pry bar

[spro]

Good one But we digress or do we? There is a tension somewhere which isn't seen by static measurements.
Okay. I screwed the thread to another place. If the tail center is correct horizontal And vertical, the tests should be right. This was asked before and relates to wear at the bottom front of the tailstock. The centers can be perfect at one point yet at different extension of the "barrel" will be off. This would place the centerline below the cutting tool and create that taper. This was mentioned earlier.

[GlennW]

This whole thing could amount to not enough pressure on the collet, but I don't know how to convey that using a handwheel closer.

I use a lever closer and it takes a pretty hefty pull on the lever to lock it, or I get push back of the stock when roughing using larger diameter stock.

[spro]

I would think the Monarch uses the lever pull. We can see how this could get out of whack, as there are centering shoes, fingers or something which wears more about the axis/ puts pressure at one side while spinning. Dang. Shouldn't have posted.
I almost deleted this except there is a point after closure and an actual lock adjustment.

[Harold_V]

It would be quite helpful to see a picture of the collet assembly. Some don't use a pull tube (the Hardinge-Sjogren, as one example). However, they are made such that it isn't beyond possibility that the cavity in which the collet locates couldn't be secure, moving at the rear, where it connects with the closer. That, or the register is oversized and permits deflection. I made mention of that possibility but have seen no response as to whether there is excessive clearance, or not.

Harold