Spindle bearing replacement

All discussion about lathes including but not limited to: South Bend, Hardinge, Logan, Monarch, Clausing and other HSM lathes, including imports

Moderators: GlennW, Harold_V

Wolfgang
Posts: 184
Joined: Fri Jul 03, 2015 9:47 am
Location: S-W Ontario
Contact:

Re: Spindle bearing replacement

Post by Wolfgang »

jcfx wrote: Sun Jun 02, 2019 1:42 am Wolfgang,
I have two sets of L20 collets that came with my V10P and Maximat 7, both sets are Schaublin that American Edelstaal
offered as an accessory for the Emco lathes.
The Maximat 7's were in near mint condition in a fitted Edelstaal wood box, unfortunately the V10P set came
loose in a cardboard box. I have never checked them for runout since I don't have a set of gauge pins.

I just spun my V10P's spindle and when I'm in gear I feel the gears in the headstock meshing ever so slightly
so it feels "rough" if I diddle with the gear shift paddles till the gears don't mesh the spindle is smooth as silk.
Could it be that you're feeling the gears ? Perhaps your friends V10P has a heavier weight oil in the headstock
versus the OEM recommended 10 wt and it's "smoothing out" the gears meshing ?

It's worth a look inside, as you might know the shift forks in Emco lathes are Zamak, some batches are prone to cracking,
my Maximat 7 is out of service because of a cracked shift fork ( I caught it in time during an oil change ).
jcfx: Thanks for your reply.

I did add an oil additive to the headstock oil bath some years ago; perhaps it is time to change all the oil with the installation of new bearings.

I'll thoroughly clean out the headstock gear box before installing new bearings. I'm thinking of using automotive manual transmission oil in the headstock. I may have some synthetic oil left over from years ago.

Turning the spindle by hand: Yes indeed it is important to have the gear train in "neutral" ie. no gears engaged with any spindle gear when doing this test. Further, a very small chuck or catch plate are also necessary to check for bearing smoothness by hand. I typically have a 6" Pratt Burnert chuck on the spindle (came with the lathe) and this huge chuck (huge for a 10" lathe) has such a large moment of rotational inertia that any bearing roughness is easily obscured.

If you have a moment I'd appreciate it very much if you were to test your spindle with two standard sized collets, say 1/2", or 3/8", or 1/4" using a dowel pin or a short piece of ground rod such as drill rod, and advise your run-out test results here. When I checked my collets I used these standard dia. and was perturbed when every single stinking collet so tested ran out .0005" TIR.

armscor: Those fibre gears are incredibly tough. I have done much serious turning of square blocks on this lathe and the fibre gears were fine last time I checked. Just don't try to change gears with the spindle running as this is definitely NOT a syncromesh box.:-)

For successful parting off the lathe needs to be mounted to a rigid bench. I went one step further and built a new mounting base for the compound rest that supports the base slide of the compound rest to its very end. This new base is also 1/8" or so lower so that the quick-change tool holders I built soon after getting the lathe would have at least a 3/8" thick support under the 3/8" square tool bit. This has worked very well.

However, parting tool geometry remains important and I confess that, when I was given a commercial parting tool with tungsten carbide inserts, I use this tool preferentially when parting off steel. For aluminum and brass I use the home-ground tool. w
User avatar
Harold_V
Posts: 20231
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Re: Spindle bearing replacement

Post by Harold_V »

Wolfgang wrote: Mon Jun 03, 2019 7:32 pm If you have a moment I'd appreciate it very much if you were to test your spindle with two standard sized collets, say 1/2", or 3/8", or 1/4" using a dowel pin or a short piece of ground rod such as drill rod,
You can't determine runout in tenths when the piece you use for testing may not be round.

I can't stress enough, the importance of NOT using drill rod for such a test. It is often not round, so the readings one would achieve may not represent the true condition of the spindle. The use of a drill blank or a dowel pin is a much better choice, although neither is guaranteed to be round, but chances are so much better, due to the fact that long objects (like drill rod) are difficult to get absolutely straight, and any deviation from straightness usually results in a less than round finish grind. The shorter the piece, the better the chance it will be round, and that's assuming that the operator of the centerless grinder avoids running parts on center. Doing so often results in a three sided grind (cloverleaf), which measures precisely the same size, regardless of where measured.

H
Wise people talk because they have something to say. Fools talk because they have to say something.
armscor 1
Posts: 311
Joined: Sun Jun 12, 2016 3:12 am
Location: Philippines

Re: Spindle bearing replacement

Post by armscor 1 »

Need what Colchester calls a Talyrond, measures out of round to a millionth of an inch.
Wolfgang
Posts: 184
Joined: Fri Jul 03, 2015 9:47 am
Location: S-W Ontario
Contact:

Re: Spindle bearing replacement

Post by Wolfgang »

Harold_V wrote: Tue Jun 04, 2019 1:51 am
Wolfgang wrote: Mon Jun 03, 2019 7:32 pm If you have a moment I'd appreciate it very much if you were to test your spindle with two standard sized collets, say 1/2", or 3/8", or 1/4" using a dowel pin or a short piece of ground rod such as drill rod,
You can't determine runout in tenths when the piece you use for testing may not be round.

I can't stress enough, the importance of NOT using drill rod for such a test. It is often not round, so the readings one would achieve may not represent the true condition of the spindle. The use of a drill blank or a dowel pin is a much better choice, although neither is guaranteed to be round, but chances are so much better, due to the fact that long objects (like drill rod) are difficult to get absolutely straight, and any deviation from straightness usually results in a less than round finish grind. The shorter the piece, the better the chance it will be round, and that's assuming that the operator of the centerless grinder avoids running parts on center. Doing so often results in a three sided grind (cloverleaf), which measures precisely the same size, regardless of where measured.

H
Harold, you are, of course, correct.

But, we are hobbyists and up to a point we make do with what we have.

Here is a test I have used to confirm absence of lobality, be it three, five, or more lobes. Note that this is not a quantitative test, merely an indication of whether the pin is round or not.

On a surface plate put your test pin in a precision V-block with the test indicator directly over the high point of the pin under test. Slowly rotate the pin without moving things on the surface plate. Perhaps your wife, girlfriend, mistress etc. can lend a hand here.

If the pin is round the indicator needle will not move. Should the indiator show any run-out the the pin has lobes. By counting the number of needle movements per revolution of the pin, the number of lobes may be determined. Unfortunately this test cannot determine the amount of lobality. T measure this one would need a v-block with the included angle of 120 degrees to measure three-lobed pins. w
spro
Posts: 8016
Joined: Mon Feb 20, 2006 11:04 pm
Location: mid atlantic

Re: Spindle bearing replacement

Post by spro »

I like stuff which seems dead simple but somebody has to do it.
jcfx
Posts: 713
Joined: Sun Aug 09, 2009 1:24 pm
Location: NY

Re: Spindle bearing replacement

Post by jcfx »

Wolfgang,
I decided to check the runout on the collet sizes you mentioned. I didn't know gage pins were that inexpensive by the piece from McMaster.
so I ordered two sizes, I have some precision ground .25 dowel pins so I used those instead of ordering one. .

Runout was checked with the lathe cold and warm ( running no load for 15min ), I initially did the measurement below
with a .0001 dial indicator too, but it was acting strangely, the needle was very sluggish, so much for buying refurbished
metrology equipment, pheh ! The DTI was purchased new so I used those measurements.
I was also wrong about my L20 collets being Schaublin, my E25 collets are Schaublin, my L20's are
labeled Edelstaal - Germany, I also have some that are labeled AEI.

Mitutoyo .0005 dial test indicator ( DTI )
Edelstaal collets were checked approx 1" from face of collet.

Cold
.25" ( 2" dowel pin )
DTI = .0002 -.00025

.375 - ( Class X Vermont plug gage 2" long)
DTI = .00025

.50- ( Class X Vermont plug gage 2" long)
DTI = .00025

Warm
.25" ( 2" dowel pin )
DTI = .0005

.375 - ( Class X Vermont plug gage 2" long)
DTI = .0002

.50- ( Class X Vermont plug gage 2" long)
DTI = .0005

I'm not sure why the last .50 " warm runout measurement is so much bigger.
John Hasler
Posts: 1852
Joined: Tue Dec 06, 2016 4:05 pm
Location: Elmwood, Wisconsin

Re: Spindle bearing replacement

Post by John Hasler »

You are measuring the vector sum of spindle runout, collet runout, and gauge pin runout. Because you are measuring 1" out part of what see may also be due to misalignment of the headstock and/or crookedness of the collet and/or pin.

Measure the spindle runout by placing the ball of the DTI probe directly on the inside of the spindle taper.
User avatar
Harold_V
Posts: 20231
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Re: Spindle bearing replacement

Post by Harold_V »

John Hasler wrote: Thu Jun 06, 2019 6:06 pm Because you are measuring 1" out part of what see may also be due to misalignment of the headstock and/or crookedness of the collet and/or pin.
That won't disclose misalignment of the headstock. It may display error due to how the pin is gripped (not concentric and straight with the spindle bearings).

H
Wise people talk because they have something to say. Fools talk because they have to say something.
Wolfgang
Posts: 184
Joined: Fri Jul 03, 2015 9:47 am
Location: S-W Ontario
Contact:

Re: Spindle bearing replacement

Post by Wolfgang »

jcfx wrote: Thu Jun 06, 2019 4:59 pm Wolfgang,
I decided to check the runout on the collet sizes you mentioned. I didn't know gage pins were that inexpensive by the piece from McMaster.
so I ordered two sizes, I have some precision ground .25 dowel pins so I used those instead of ordering one. .

Runout was checked with the lathe cold and warm ( running no load for 15min ), I initially did the measurement below
with a .0001 dial indicator too, but it was acting strangely, the needle was very sluggish, so much for buying refurbished
metrology equipment, pheh ! The DTI was purchased new so I used those measurements.
I was also wrong about my L20 collets being Schaublin, my E25 collets are Schaublin, my L20's are
labeled Edelstaal - Germany, I also have some that are labeled AEI.

Mitutoyo .0005 dial test indicator ( DTI )
Edelstaal collets were checked approx 1" from face of collet.

Cold
.25" ( 2" dowel pin )
DTI = .0002 -.00025

.375 - ( Class X Vermont plug gage 2" long)
DTI = .00025

.50- ( Class X Vermont plug gage 2" long)
DTI = .00025




Warm
.25" ( 2" dowel pin )
DTI = .0005

.375 - ( Class X Vermont plug gage 2" long)
DTI = .0002

.50- ( Class X Vermont plug gage 2" long)
DTI = .0005

I'm not sure why the last .50 " warm runout measurement is so much bigger.
jcfx:
Thank you very much for taking the time and trouble to test your lathe & collet combination. As I stated much earlier above, my test results were much like yours, typically .0005" TIR (total indicator reading).

For typical lathe work this run-out is not a problem, but when semi-finished work is held in a collet for a final operation where functionality requires that the work OD be concentric to the second operation within very close limits, this run-out can be problematic.

Thanks again. w
Cary Stewart
Posts: 542
Joined: Wed Jun 08, 2011 5:54 pm

Re: Spindle bearing replacement

Post by Cary Stewart »

As for those Vermont X class pin gages you must read what the TIR, diameter tolerance and roundness tolerance are. There are very few home shop machinists that have the equipment and facility to verify that those pins are truly class X either + or - tolerance. I have a set of about 8 of those in various decimal sizes for both inch and metric diameters. If I recall the roundness of them is .000004 inches. The diameter is to .0000020 inches. Do I have the right number of 0s? To use this level of accuracy everything must be CLEAN.
Cary
John Hasler
Posts: 1852
Joined: Tue Dec 06, 2016 4:05 pm
Location: Elmwood, Wisconsin

Re: Spindle bearing replacement

Post by John Hasler »

Harold_V wrote: Thu Jun 06, 2019 7:01 pm
John Hasler wrote: Thu Jun 06, 2019 6:06 pm Because you are measuring 1" out part of what see may also be due to misalignment of the headstock and/or crookedness of the collet and/or pin.
That won't disclose misalignment of the headstock. It may display error due to how the pin is gripped (not concentric and straight with the spindle bearings).

H
You're right. It won't be affected by headstock misalignment. It would still be best to measure close to the collet, but the worst problem is that errors introduced by the collet and/or pin can cancel with spindle runout, concealing it.
User avatar
Harold_V
Posts: 20231
Joined: Fri Dec 20, 2002 11:02 pm
Location: Onalaska, WA USA

Re: Spindle bearing replacement

Post by Harold_V »

Given a properly gripped pin, one that is concentric with the bearings, it could be used to determine error, but the question then would be where the error occurs. By traversing an indicator along the side of the pin via the carriage would display any deviation. The question then would be, is the headstock disoriented, or are the ways buggered?

It's extremely tough to make determinations once one works under a thou. The slightest thing changes what you think is happening. Working to less than a thou on a lathe or mill requires a little patience, good workmanship as well as good fortune. They are, after all, not (typically) high precision machines. There are exceptions, of course.

H
Wise people talk because they have something to say. Fools talk because they have to say something.
Post Reply