[How do I?] Restore Morse taper in lathe spindle

NortonDommi said:
Amother 2cents worth,
As above, that should have been the first step. If the spindle has been damaged by spinning it was most likely due to have an arbor fitted without retention i.e. a drawbar through the spindle.
You have stated the lathe is old and worn, it can still be set up to work well and problems slowly worked through. A first step would be to level. I have 2 expensive levels but this method:
is cheap easy and very accurate.
After this apply 'Rollie's Dad's Method' to check headstock and bed alignment.
Align tailstock to headstock.
If you find you need to grind internal taper hold a piece of scrap in the chuck. Turn a 60 degree taper on it, fit a center to the tailstock and hold a new or good condition arbor with the correct taper between the two centers as per Bob Korves. Set you compoud to this,(taper), surface, fit your newly made/cobbled together toolpost grinder, true wheel and proceed to make very light cuts checking frequently with your good arbor and dye.
Don't rush and double check every step. Make notes as you go and you will end up with a good lathe that you can take pride in.
Click to expand...
The OP is not trying to make a MT tool, he is trying to get the MT taper in the spindle correct. Per the OP, the spindle is within .001" runout, but the spindle Morse taper is suspect.
 
gasengin said:
I have a 10 by 36 Atlas lathe with a MT#3 in the spindle. This lathe was well used. It looks like a collet may have spun in the spindle. The collet has about .005 run out. The bore is not as smooth as it should, but I don't feel any burs that would account for the run out. The outer diameter of the spindle is about 1/2 thousandth out so not likely any other issue. What is the best way to true it up? I need some guidance on this as it needs to be done correctly. Thanks

PS. I checked my 9 by 18 Logan and it was about 1/2 thousandth both on the collet and outer diameter of spindle.
Click to expand...
I looked at your original post again more carefully. You are measuring a collet that has runout. The fact that you see runout in the collet does not say anything at all about the condition of the spindle Morse taper. To qualify the spindle taper, you must test the spindle taper, and not a surrogate. Runout in the collet could be because of a bad collet, a bad collet adapter, or a bad spindle. You must do the correct tests on your spindle first, and find out for sure what you have there.
 
Here's my 2 cents: I need to sort out something before we all start making incorrect assumptions; it'll save a lot of time and off center comments.

1) have you, gasengin measured your taper from a fixed point, such as your flat-way with a test indicator, at several places in the taper????
-this is the most critical issue, and if so will tell a completely different story than I think I've heard.

2) can you purchase a new, qualtiy, MT3 stub arbor that has a cylindrical bore, such as https://www.busybeetools.com/products/stub-arbor-mt-3-x-1in.html (this is a cheap one, but mine measured a TIR of .0002, which is perfect for this test)
- this is a useful; reference, and will tell a lot if used properly. It can be inserted, measured, then turned 90 degress and repeat 3 times. Even if it is a cheap one, a lot can be inferred by taking concentricity measurements this way - but only if the above question is answered first

I can go into much more boring detail (pun intended), but let's just start with these two steps.
 
Bob Korves said:
The OP is not trying to make a MT tool, he is trying to get the MT taper in the spindle correct. Per the OP, the spindle is within .001" runout, but the spindle Morse taper is suspect.
Click to expand...
Bob,
I am probably wrong but as I read original question the O.D. of spindle has "about 1/2 thousandth" runout. The question was about the condition and truth of the internal spindle bore taper. Basic checks,(level,alignment), need to be done to ensure measurements are accurate.
I was NOT describing a way to TURN a taper I was describing how to set the compound to the requsite angle as per your post, IF, and it is an if the spindle taper needs to be ground. If the spindle bore taper is suspect it cannot be used as a center to setup an arbor to measure off, turning a center from scrap held in the chuck will be true to the axis.

Gasengin, more data please.


"
 
NortonDommi said:
I am probably wrong but as I read original question the O.D. of spindle has "about 1/2 thousandth" runout. The question was about the condition and truth of the internal spindle bore taper. Basic checks,(level,alignment), need to be done to ensure measurements are accurate.
Click to expand...
The OP stated in post #1 that the spindle bore has damage and about .005" runout, measured with a collet, which is a questionable test at best. He wants to know "What is the best way to true it up?" You are correct that the lathe needs to be geometrically correct before the existing taper can be tested (and corrected) properly and completely, using the existing components to qualify it. You are also correct that the spindle bore cannot be used to qualify anything at this point. Basic runout at several depths, with clock positions noted, can be tested accurately using an indicator with the base located on any rigid part of the lathe, and together will tell much about existing spindle issues. Spindle bearing play can be tested with the same setup. NortonDommi, your way of making an accurate arbor using an inaccurate lathe is much better than what I posted, thank you, though the OP is not trying to do that at this point.
 
BTW: I'm not a fan of "Rollies Dad Method" of truing up a lathe in my book! I have spoke of this in other threads and leave it at that.

Meanwhile, the OP has a spindle that the internal spindle taper is damaged from someone not paying attention to a boring operation an previous owner of the lathe did, while working on a work piece. And reinforcing what other have said, the OP wants to correct the runout left from the damaged caused from this incident.

Also have to remember, this is on a Atlas lathe of some vintage and the lathe has flat ways, too. I agree that some alignment checks maybe needed but not always necessary. The compound probably does not have enough travel to recut the entire taper in one pass. Also, the compound may not be in good shape for recutting the taper for straightness. The OP does not indicate if an taper attachment is available for the lathe. (Probably not if I had to guess.)

Bob has a good point in post #12. Follow his advice there and report back.

Something that will help is to take a piece of extra fine emery cloth on the end of a stick and polish out the bore of the taper. Don't use your finger to do this!
 
Bob Korves said:
I looked at your original post again more carefully. You are measuring a collet that has runout. The fact that you see runout in the collet does not say anything at all about the condition of the spindle Morse taper. To qualify the spindle taper, you must test the spindle taper, and not a surrogate. Runout in the collet could be because of a bad collet, a bad collet adapter, or a bad spindle. You must do the correct tests on your spindle first, and find out for sure what you have there.
Click to expand...

I did try several collets and centers with all about the same results with the high spot in the same place. I did not use an adapter on collets. MT#3 collet in spindle. On the center, I used a #3MT to #2MT adapter. I may still have flaws in my testing however. I will see if I have an indicator that will work to that depth on that small of internal diameter diameter and report back. I will blue it too. I do not have a taper attachment and will have to check on the travel in the compound. The main reason I need to correct this run out is when using my mill ends in the collet for the milling attachment. I almost never do center to center turning. Plus I have my smaller Logan to use also.

The lathe is a flat bed Atlas and has at least .002 wear on the front bed way under the carriage near the head. I'm sure it also has wear on the edges of the bed. Most of my work does not require a great amount of accuracy or precision, so I can live with it. I didn't pay a lot for the lathe. It does have the Atlas Pick-O-Matic, so has about every feed you could imagine. There is no way for me to justify the time and expense to fully recondition this lathe.
 
Hey gasengin,

Another way forward would be to ignore that bad internal spindle taper and use a spindle nose collet adapter.

There are many great threads here about building one (both simple and "true-adjust" types):
http://www.hobby-machinist.com/threads/retrofit-er-32-collet-chuck-to-an-atlas-craftsman-618.25476/
http://www.hobby-machinist.com/threads/building-a-set-true-er-40-collet-chuck.62089/
http://www.hobby-machinist.com/threads/true-adjustable-er-32-collets.62016/#post-511508
http://www.hobby-machinist.com/threads/er40-adjust-true-camlock-collet-chuck-build.61759/
http://www.hobby-machinist.com/threads/er-40-collet-chuck-s.7920/

I went with the Beall ER-32 collet chuck for my 1937 SouthBend 9":
http://www.bealltool.com/products/turning/colletchuck.php
they also have the bigger ER-50 system
http://www.bealltool.com/products/turning/bigchuck.php

-brino

btw, I am still looking for a couple good threads I remember, but have NOT yet found......I'll update this post if I find them.


EDIT:
Here's one: http://www.hobby-machinist.com/threads/er-40-collet-chuck-and-spanner-wrench.45930/#post-391843
and the other: http://hobby-machinist.com/threads/er25-collet-chuck-for-atlas-618.41636/
 
Last edited:
I got to thinking about Bob's comment about the need to be dead on center vertically when cutting a taper and wondered how far off dead center you could be. I set up a taper on SolidWorks and moved the cutting plane up and down from the horizontal plane through the spindle center simulating and offset cutting tool. The taper that I set up was .3"/ft over a 4" length.

At zero offset, the taper was a perfect .3"/ft. as expected. At a .005" offset, the difference in taper is not measurable (.000012"/ft.). At a .010" offset, the difference was .00009"/ft. and .020", the difference was .00021`"/ft, optimistically barely measurable. At a .030" offset, the difference was .00054"/ft.. At a .040", the difference is .00096"/ft. The resulting taper will always be smaller than the target taper.

On the other hand, the most accurate way that I know to set a compound angle is with a sine bar. At best, you will be able to set a sine bar angle to +/- .0001 over 5" which is an uncertainty of +/- .00024"/ft. Trying to get the last thousandth in cutting tool height is probably not worth the effort.

That said, I will always try to get my setup as close to specification as I can. Errors can stack in a worst case scenario and if Murphy has his way, they will.
 
Another approach to true running centers in the headstock is to use a soft center that can be trued up in place and stamping a match mark on the spindle end and the center so that it goes in to the same location when it is used; this usually ensures that the center will run very close to true. I think that careful use of a morse taper reamer is the practical way to deburr the spindle taper.
 
You must log in or register to reply here.
Back
Top