Atlas 10″ runout

[martin humphreyParticipant @martinhumphrey87293]

Hi

I'm new to posting so I'll start with a simple one.

I've got a 10" atlas 1943 vintage. It's got full set gears slotted cross slide etc etc. The bed is good and everything works fine except I have about 0.04mm of runout at the spindle. I dont find this a problem as i can mostly work around this but I would like to have something a bit more accurate. First question has anybody got a 10" and what is their runout. 2nd does anybody own a warco 240 lathe and what do they think of it.. The reason for asking is to sortout the bearings for this lathe is about £250.00 and I'm not sure it's worth it.

[martin humphreyParticipant @martinhumphrey87293]

[Robert DoddsParticipant @robertdodds43397]

Martin

What method are you using to measure this 0.04mm runout Are you measuring a bar in the chuck or what?

Its hard to imagine that much error right at the spindle nose and still be able to turn anything decent.

Tell us a bit more about your checking method.

Bob D

[martin humphreyParticipant @martinhumphrey87293]

I've got a morse tapered test bar purchased from rdg tools and test at about 25mm from the end of the spindle nose on the parallel section.

[Grizzly bearParticipant @grizzlybear]

Have you got a similar runout when using a decent chuck?

If you put a sturdy bar in the chuck with six or more inches sticking out (Motor NOT running) grasp the bar and check for movement, towards and away from yourself.

With bearings costing that much money, they should be very robust.

[Keith LongParticipant @keithlong89920]

Hi Martin

Could you please tell us if you've done some of the more obvious tests:-

1 check run out on the register and direct in the morse taper end of the spindle.

2 your test bar should have centres drilled into in each end in which case, set up the bar between centres of some sort and check that the bar is true.

3 make sure there are no burrs either on the morse taper end of the bar nor in the socket

4 make sure both are scrupulously clean

Any of the above could give you an apparent run out, and are a LOT cheaper to recify than the lathe bearings.

It's vital to eliminate possible error sources such as the above before even checking the lathe bearings

Keith

[Anonymous]

Measuring using a test bar is, initially, only checking the fit of the test bar Morse taper in the spindle. It doesn't necessarily say anything about the runout of the spindle itself. To start with measure on the spindle itself. Assuming that the spindle has a screwed nose, there should still be an axial and radial ground register, right at the headstock. It would be worth checking the runout of those registers.

Regards,

Andrew

[Thor 🇳🇴Participant @thor]

Hi Martin,

I don't have an Atlas lathe, and my lathe is not nearly as old as yours (from 1976), but I get a runout about 1/4 of your result using just the old MT3 centre. Measuring about 20mm from the spindle nose. Testing for play in the bearing is good advice, you may be able to adjust how much play in the main bearings.

Regards

Thor

[martin humphreyParticipant @martinhumphrey87293]

If i put a 2" dia bar in a 3 jaw chuck I get movement on my dial gauge moving the bar around. This hasn't had new bearings yet. What I'am after really is "will new bearings bring my lathe up to spec" or do I go for broke and buy a new chinese lathe. I've been quoted spindle runouts of 0.01mm for some and don't really see this as a step up considering my lathe is 70 years old and is only sligthly worse than this.Also turning between centres on my lathe greatly decreases runout. The bearings are setup correctly and they are oiled every time I start the lathe up. I've had the spindle out and checked the bearings no pitting cages are fine and they are seated up to shoulders. I work with tapered roller bearings often and am aware what to look for. Is a lathe of this vintage ever going to be as good as a new chinese lathe or have I got this back to front.

[Robert DoddsParticipant @robertdodds43397]

Martin,

Thats a good sensible set up but on my Atlas its 3 MT but the taper is foreshortened to accomodate the through hole of the spindle so its easy to get some burr or debris in the stubby taper and throw your mandrel off.

Grizzly bear's suggestion will help you find bearing play and if you have the taper bearings these can be adjusted to some effect but not if the bearings are shot!.

Can you put a clock on the shoulder of the spindle where the chuck screws up to?

You need to separate error in the MT from bent spindle problems. Again I would suggest that a bent spindle is most unlikely. If it has been bent you should expect other collateral damage because it would take some excessive force to do that sort of damage

Does the test bar tend to describe a cone ie more error as go further away from the splindle nose?

I see a lot more similar advice has been posted whilst I was scribing

Martin , How old is your 3 jaw chuck? Are you talking runout of the bar as you rotate the chuck? may be the jaws are showing a bit of wear.

Bob D

Edited By Robert Dodds on 12/10/2012 19:21:15

Edited By Robert Dodds on 12/10/2012 19:27:27

[martin humphreyParticipant @martinhumphrey87293]

Thanks for all your suggestions. The test bar does describe a cone the further away I test. I have not tested the register yet I will do this asap. I will also test the test bar but if this is out I might not be a happy man as it's fairly new.

Edited By martin humphrey on 12/10/2012 19:39:26

[Sub MandrelParticipant @submandrel]

There is no detectable runout (i.e. I can't see any regular movement, just vibration of the needle on my 0.0005" graduated DTI) on the spindle of my mini lathe with roller bearing upgrade, I wopuld say that is better than 0.00005". That's measured at the spindle flange.

Measuring anything in the socket could just be measuring errors in the socket or what is fitted there (if not a decent test bar).

Measuring runout on anything held in a chuck is measuring the chuck, not the lathe.

<goes back out for a minute>

Just out of curiosity i thought I'd try putting the DTI on a bar in the 3-jaw. It already had a short length of 1 1/4" diameter steel bar in it (I tried out a new parting blade on it and left it there. I did not reset or retighten tthe chuck, just put the DTI on the end of the bas as I found it.

In was gobsmaked by the result – noticeably worse than the spindle flange, but only a fifth of a division by eye – thats 0.0001". Not bad for a 14 year old chuck, let alone on a cheap lathe.

Neil

[Ian S CParticipant @iansc]

You say you get movement when you chuck a bit of bar in the three jaw, you will, have you got a four jaw chuck, use that, get it running true at the chuck, then move out six inches, and measure there. An old three jaw will often be a little eccentric, and that will be different depending on the diameter of the work piece, depending on the wear on the scroll. Ian S C

[BazyleParticipant @bazyle]

Runout, as opposed to loosenes due to wear that cannot be taken up, rarely matters. If the wear contributes to chatter you have a problem but runout is only going to matter if you need to repeatedly rechuck in collets and have the chucked surface run true, as in working on clock parts or axles.

[KevinWParticipant @kevinw]

Martin,

As far as I'm aware the bearings on my Atlas 10F are in good condition. I cannot feel any play and I've never found any need to adjust them.

I've checked the runout this weekend. The spindle register (1.625"dia) has 0.001" runout.

I don't have a test bar so I fitted a 3/8" silver steel bar into an MT3/ER25 collet holder. The runout adjacent to the collet nut was 0.0015". At 3" from the collet it was 0.002" and at 6" from the collet it was 0.0045".

Hope this helps

Kevin

[Sub MandrelParticipant @submandrel]

I must don sackcloth and ashes. I found it hard to credit teh reults I reported above. I'm not sure why but something must have been sickling or otherwise.

The spindle runout is, indeed, virtually undetectable and with a 3MT socket in the bore that has runout of about a tenth of a thou, (after I cleaned it properly – twice). But the 23-jaw appears to be 2-thou eccentric, which is more in line with what would expected. That said, I have had better results in the past, maybe it's dirt or maybe it's wear.

Neil

[martin humphreyParticipant @martinhumphrey87293]

Thanks for all your comments and help.

I stripped down the spindle. Removed bearings and cones.Cleaned everthing carefully deburred all the cast iron seats re-seated the cones at 90 deg to original and rebuilt the spindle. I then added new oil to the bearings ran them for 20 mimutes and reset them. I now get no movement on my clock at the spindle shoulder and when I turned a 3" dia bar I get a max runout of less than 0.01mm 2.5" from the chuck.Sorry for the use of metric and imperial as all my clocks are metric and i'm imperial.

The test bar has been checked and is now in the bin as it's not only eccentric but bent as well.

Thanks again for all your comments.

[martin humphreyParticipant @martinhumphrey87293]

Thanks for all your comments and help.

I stripped down the spindle. Removed bearings and cones.Cleaned everthing carefully deburred all the cast iron seats re-seated the cones at 90 deg to original and rebuilt the spindle. I then added new oil to the bearings ran them for 20 mimutes and reset them. I now get no movement on my clock at the spindle shoulder and when I turned a 3" dia bar I get a max runout of less than 0.01mm 2.5" from the chuck.Sorry for the use of metric and imperial as all my clocks are metric and i'm imperial.

The test bar has been checked and is now in the bin as it's not only eccentric but bent as well.

Thanks again for all your comments.

[AndyfParticipant @andyf]

I must get one of those 23-jaw chucks, Neil. The all-round grip must be nice and even – almost as good as a collet, I hope they are self centring; independent might be a bit time-consuming to set up

Andy

[MICHAEL WILLIAMSParticipant @michaelwilliams41215]

I wonder whether your dodgy test bar came from the same place as Terry's dodgy lathe tool . There is now a vast array of cheap workshop tooling on sale – much of which is scrap metal as far as I'm concerned .

Can't help with a 23 jaw chuck but 6 jaw chucks are quite common and 12 jaw chucks have been made specially for jobs like holding thin walled glass tube .

[Ian S CParticipant @iansc]

Neil, you can sometimes get the three jaww a bit better by just slightly looseningthe chuck on its backplate, and give it a tap with a soft hammer till the test piece shows zero, then retighten the bolts, recheck the run out. There is often enough clearance in the bolt holes to allow a few thou movement. Ian S C

[KWILParticipant @kwil]

Ian, whilst that may well be true at a particular holding diameter, wear and/or precision of the scroll plate will undoubtedly mean it may well not be true at other diameters.

[Ian S CParticipant @iansc]

True Kwil, but can be handy at times. Ian S C

[Sub MandrelParticipant @submandrel]

Hi ian,

I tried that with my wonderfully crude but effectrive nylon-faced hammer, but teh regisater on mini-lathes is a very close fit to the chucks..

Neil

[david treadwellParticipant @davidtreadwell45609]

Hello this is a good thread, and I have another question along the same lines.

I have a 18"churchill grinder. If I load a 12" mt2 test bar in to the grinder head It runs out about 0.008" right on the end. With no end support from the other centre. I reamed the MT2 socket and I now have it down to 0.002-0.004"

There is conveniently a oil hole in the casting which gives access to the spindle. I inserted a dti through the hole right on to the spindle whilst running between its own bearing and it is dead straight. (not bent at all)

The reason I want it to run better is… I often make taper tooling where I grind the taper between centres, then place the taper in the grinder head directly to grind internal diameters.

Example:- the internal taper for a collet holder.

So the question is should I expect better results, or am I asking too much?

=================================================================

Just to add a bit more to the last thread.

I have a boxford with is in fair condition If I put a 0.01mm dti on the register (shoulder) it doesn't flicker.

I bought a test bar from RDG tools and it was ace dead straight and I use it to set up all my kit. Although, I think a lot of these online places change their suppliers frequently.

Lastly, the test method of putting a bar in a 3 jaw chuck seems flawed because general bar stock is not concentric. But, even a with a piece of ground bar my burnerd chucks rarely align better than 0.003" and the cheapo chineese one about 0.006"

Sorry for the change from mm to inchs. (depends which dti is to hand.)

Thanks Dave

[Author]

Posts