D1-4 on a PM1340

[Ken226]

Have not messed with dx tapers so just guessing, assuming they do not fully seat on anything flat as the taper is the interface.

Sent from my SAMSUNG-SM-G930A using Tapatalk

On D1 spindles, the chucks do indeed seat against both the taper, and the flat face of the spindle. As @mikey said, it's a secondary consideration. It doesn't need a ton of force holding the faces together, just contact. The taper/register controls radial runout. The face contact just prevents axial deflection under cutting forces.


The tapered nose itself is too short to provide significant resistance axial deflection due to either cutting forces, or one cam being tightened to a higher torque value than another.

There is less than a half inch of taper, so under significant cutting forces, a gap at the rear of the chuck would cause tension/compression forces be transferred into the cam pins. The amount of deflection would be inconsistent, with 3 spots 120° apart per revolution having more deflection (the gap between the pins. The pins are (3, on a d1-4) 120° apart. This would likely contribute to surface finish issues and possibly chatter.

The typical procedure for fitting a chuck that doesn't fully seat against the spindle face, is to mount up a good chuck/collet, then mount a bar into the properly seated collet/chuck. Then flip the offending chuck backwards, tightened onto the bar, and polish the inside of the female taper with emery cloth untill it seats with contact on both the taper and face. A little dyechem helps. It also helps to remove the camlock pins from the chuck, for safety.

Most of the backplates I've purchased for custom fixtures have needed this done. Some more than others. It usually takes be about 20 minutes to get a new d1-4 backplate properly fitted to my spindle.

I used a Shars backplate for the receiver truing chuck I recently made. It had a couple thousandths gap when new. A few minutes with some 220 grit sandpaper got it sorted out.



This spider chuck used a chinese backplate as well. It had a gap, and took about a half hour to get properly fitted.


It would seem that most of the new backplates have the taper made a little undersized, to allow for final fitting.

Several of my chucks were fitted to other camlock lathe's I've owned in the past, and still fit my current lathe quite well with no additional fitting required. That implies that the spindles were all the same, and the backplates tapers were undersized. I've never found a new backplate with an oversized taper. Always undersized.

Only 2 chucks I've purchased ever fit perfectly out of the box. One was a Kalamazoo 3 jaw, the other was a Chinese 6 jaw.

From Rohm's instruction manual:

Mount the chuck and without tightening the cams there should be a gap of .0008 - .002 inch. The manual states to evenly tighten the cams. The manual then states "Both the short taper and the face of the chuck must be in full contact with the spindle nose after mounting

So, with the chuck in place on the spindle taper, a gap of up to .002" should allow the cam pins to pull the chuck onto the taper, sufficiently to close the gap and bring the chuck and spindle faces in contact, providing a firm axial register.

 

[Dabbler]

It would seem that most of the new backplates have the taper made a little undersized, to allow for final fitting.

This could be the entire problem. When I bout my Pratt-Bernerd chucks back in the 80s, they fit perfectly, and I hadn't thought of that one!

 

[Ken226]

This could be the entire problem. When I bout my Pratt-Bernerd chucks back in the 80s, they fit perfectly, and I hadn't thought of that one!

I just want to clarify, that was a guess on my part. Based on my experience with chinese back plates always being a little undersized. I don't know that for certain.

When I've bought nice, expensive chucks, they fit.

According to Rhom, when the cams are loose, a gap of .0008" up to .002" between the backplate and chuck face is normal. Tightening the cams should pull the faces together. Over .002" might require more leverage than just a chuck key to pull the faces together.

I don't want to have to use a cheater bar to change chucks, so I like them to have about .001" gap when loose. After loosening the cams, a gentle smack with a rubber mallet pops them off.

 

[Tired&Retired]

Nope, you all didn't lose me. Just have other fish to fry and only get back to this when I can.

Yes, I did measure the diameter of the spindle taper at the back edge, as best I could and came up with 2.4800 more often than not. I haven't tried to measure the front of the nose taper as I doubted I would be able to get a good read there that would mean anything. I did try to measure the female taper on one of the chucks and came up with 2.4725, but I honestly don't remember which chuck I measured.

As I mentioned earlier, I used a 3000 grit stone rod mounted in a drill fixture which is in an MT3 adapter on my compound. I have that set to 7 degrees to match up as best I can with the angle of the taper. I only used slow speed on the lathe, and didn't apply all that much pressure, nor spend a lot of time with it. The idea was just to deburr and possibly polish the spindle nose taper a bit.

So anyway, when I took the earlier mentioned video, I mounted four chucks on the spindle while taking the video. The other day I was looking at the spindle nose and noticed the marks made on the taper during the process of mounting those chucks and took the picture below with two crops to get a closer look at the metal surface.

Got to run....

 

[mikey]

The spindle taper is 7 degrees, 7 minutes, 30 seconds.

7 degrees will not fit.

 

[ddickey]

ISO standards on the diameter of the taper should be 2.5005" +.0003" -0.
The depth of the spindle taper should be .433" +/- .0157". Not that this means much.
Interesting though is the face plate has a tolerance of +.0003/-.0002 on the spindle nose and a taper depth of .3937 with the same tolerance as the spindle.

 

[Ken226]

ISO standards on the diameter of the taper should be 2.5005" +.0003" -0.
The depth of the spindle taper should be .433" +/- .0157". Not that this means much.
Interesting though is the face plate has a tolerance of +.0003/-.0002 on the spindle nose and a taper depth of .3937 with the same tolerance as the spindle.

Yea, those are the correct nominal dimensions. Where they get hard to measure is when they add features such as undercuts, relief cuts, chamfers etc, at the points where the features are dimensioned.

For example, the 2.5005" diameter is @ the point where the taper intersects the spindle face. On my spindle, and most others I've seen, there is a radiuses relief cut there, and nothing to measure.

Theoretically, if that relief cut weren't there, the taper would continue to the chuck face, intersecting it at the point where the tapers diameter is 2.5005.

A chamfer also exists on the nose of the taper, which reduces the effective contact area, but is still part of the nominal length dimension.

It's clear in tired&retireds pics above that his has a relief at the same point as mine. The theoretic point on his spindle where his taper would measure 2.5005" , has a relief cut.

 

[Ken226]

So, as an experiment, i went out to try and measure my spindle at that theoretical datum where the taper would intersect the spindle face, were that undercut not there.

Heres the undercut:


Here's my setup, using .125" thick parallels and magnets to locate the spindle face relative the the part of the taper measured:


Heres trying to measure the diameter, @ a point .125" from the spindle face. Basically, i tried to measure the taper diameter across the face of the parallels. It took a few tries to get a diameter i was confident was close to correct. After a half dozen attempts, i settled on 1.469" (probly closer to 2.4694") judging from the pic. It was challenging, trying to get the caliper jaws flat against the surface of the parallels, so it would contact the taper on the same plane as the parallel face. I ended up setting the parallels at an angle, like the letter A, so the caliper jaws could lay relatively flat against the parallels.








Then i used my CAD software to draw the taper, and continue the 7.125 degree angle along the z axis for an addition 0.125 inch. Then measured the theoretical diameter at the point the taper would intersect the chuck face, were there no undercut:



@Tired&Retired

Perhaps try a similar measurement. Arrange the parallels so you can hold the caliper jaw faces flat against the parallels. Measure the taper diameter @ that point.

 

[mikey]

@Tired&Retired, I was sad to see that you took a grinder to that spindle register. Given that you set it to 7 degrees, even if you only ground a little bit off, no standard D1-4 chuck will interface securely to that spindle because the geometry is now off. I think you should have another discussion with Matt and let him know what you did; I suspect he will agree that the spindle needs to be replaced. Since you'll be in there anyway, you might consider upgrading the spindle bearings to a precision class set (ABEC 5-7 or metric equivalent).

 

[Ken226]

Given what's stated here:
Quote:
"As I mentioned earlier, I used a 3000 grit stone rod mounted in a drill fixture which is in an MT3 adapter on my compound. I have that set to 7 degrees to match up as best I can with the angle of the taper. I only used slow speed on the lathe, and didn't apply all that much pressure, nor spend a lot of time with it. The idea was just to deburr and possibly polish the spindle nose taper a bit."

Depending on the definitions of "not much pressure" and "nor spend alot of time", he may not have removed more that a few hundred thousandths. 3000 grit doesn't take much off. Hopefully not anyway!





@Tired&Retired:

If you'll take a few measurements, I'm more than happy to help calculate the current angle of your spindle taper. To determine if at least the angle is still (or ever was) correct.

Do you have a DRO? The gear rack for Z travel is notoriously inaccurate. Especially when the pinion crosses from section to section on multi piece racks.

Anyway. I'm assuming you have either a DRO, or a dial indicator you can put on the bed to measure carriage travel.

Put an indicator on the cross slide/toolpost, and zero it against the spindle taper. Point it nice and strait to minimize sine error. Use a DRO or indicator to make accurate Z movement. Then move the carriage in the Z- direction 0.250", and record the indicated value. Should be doable with a ten thousandths indicator.

Basically, your accurately traversing an indicator a given distance along the Z axis, and recording the tapers radial change on the X. I hope that makes sense.

The angle can be calculated by using the following right triangle formula:

Arctangent multiplied by (change in X divided by travel along Z)

So, in your case, if the angle is still correct @ 7.125° (7°, 7' 30"), traversing the carriage .250" should move the indicator needle .03125".

With the calculators angular units set to °.




A .001" indicator will allow you to measure 1/4° increments. A .0005" indicator will get you 1/8° increments, and a .0001" indicator will get you about .025° increments.