What's new
What's new

lathe leveling for dummies (long)

hackmaster J

Aluminum
Joined
Dec 24, 2005
Location
peoples republic of new jersey
while i know leveling machines can be tedious i don't think it's rocket science. that said i'm a little puzzled about this one. i have a cincinnati 17x42 hydrashift toolroom that i got a few years back and reconditioned it best i can (no refinishing of the ways or carriage). at the time i roughed the machine in. now i decided to do some maintenance on my machines and re-level them. when i checked the hydrashift i found that i had roughed it in a little better than i thought, so now i'm gonna give it the fine tune. i get out the levels and manual, clean the area and have at it. the manual says to take the twist out starting at the head then finishing at the tailstock then level it end to end. there's four points at the headstock and two at the tailstock. so i get this thing pretty much dead nutz on all the way round, make sure i don't have a soft foot and everything looks great. then (and this is where i knew things would unravel) the manual states to level the machine crossways by placing the level on the wings of the carriage and run it end to end checking for variations (which i knew would happen). huh? i figured IF the machine was new as the manual was written for wouldn't taking the twist out of the bed assure the carriage would have no variation end to end and leveling it crossways wouldn't be necessary? i forget how much it varied but it was at least .0015. of course this machine has wear and i'm not trying to work in tenths. i can get the variation out of the carriage end to end but then of course the machine is out of level when checked across the ways. i was thinking the best thing to do would try to draw a compromise between having the twist out of the ways and having the carriage travel level end to end and would "think" keeping the carriage level would tend to give better accuracy. am i just an idiot that doesn't know what they're doing or is this a matter of wear in the machine? one thing i forgot to mention is that the manual says to place the level on the vees but this requires a 1/16 spacer on the rear vee because it's shorter than the front so i just used a pair of .600 gauge blocks and leveled off the flats.
 

Attachments

  • shop 005 (Large) (Custom).jpg
    shop 005 (Large) (Custom).jpg
    45.3 KB · Views: 3,092
Every wonder how a lathe is leveled on a floating ship thats rocking back and forth in the water?

I believe the most important factor is that all leveling pads are supporting weight of the machine and that it's leveling screws are not putting the bed at a misalignment with the head stock.

If you chuck a known straight shaft that you can indicate in dead nut with NO tail stock and get it leveled in a way the your indicator reads dead zero when traveling it up and down on the carriage both top and side of the shaft is the best you can hope for.
 
Leveling off the flats is fine, even preferred in some cases as the Vee tops aren't all that precise on some lathes. With regard to level across the ways, that doesn't matter either as long as the reading is the same on both ends of the bed.
 
While "leveling" a lathe is what most of us do to dial it in what you really want is for it to be straight. In a perfect scenario the C/L of the headstock spindle should be "perfectly" parallel to the bed both horizontally and vertically. If this state could be achieved then it should be possible to machine a shaft with zero runout. This is why machining a test bar will give you the most accurate picture of the "straightness" of your lathe. If you can achieve a minimal runout variation over a foot or so of stock then that variation should interpolate to a very straight bed over its full length.

This also explains why you can have lathes on sea-going ships. If you can hold the lathe in a straight position it doesn't matter whether it's level or not...
 
Using a level gets you close but for real accuracy--to reach the potential of your machine--you need to go further. Take a look at this thread on "Shop Floor Talk"...

Machinist Level - Shop Floor Talk

Post 25 on page 3 explains how to do the final dial-in...

A 0.01mm/m level is accurate enough
Perhaps even too accurate It will show a 1 dgr twist on your leveling bolts
Problem on all used lathes is that they are worn and you will never be able to level a lathe that is used for 10 years or more
specially lenghtwise
So you have to compromise
And on a used lathe if you only do the test with a 300mm bar without a center you could have a twisted bed at the end That results in different diameters between centers for shorter or longer bars

Hackmaster
What is important is the position of your bit while turning
So levelling the carriage is important
But like I said while turning
Did you check the level of your carriage under load So pushing the carriage down here and there
Perhaps the carriage itself is twisted or worn
Perhaps the guides underneeth have to be adjusted

How accurate is your level ??

Peter from holland
 
thanks for the replies and for the most part i'm aware of the things mentioned. of course the basic goal of leveling is to get the twist out of the bed and bring it parallel with the headstock. leveling end to end is not so important. i don't have access to a test bar so right now that is not an option. i believe when i first roughed the machine in i did do a test cut. i don't remember where i read it or the exact numbers but i "think" what i read said you wanted less than .0004" taper on an unsupported length of 4 inches (maybe more)? i do remember it wouldn't hold those numbers. the machine is sitting on 6" of high strength fibermesh concrete, i had the whole shop floor poured this way because i was installing a lift and the machines. i have the original adjusting bolts but the pads were missing. i cut up six pads from 3" barstock, countersunk them for the adjusting bolts then surface ground them flat. yes the machine is originally intended to be bolted down through the adjusting bolts. the levels i'm using are .0005 per 12". i think at this point because i never really took the time to get it leveled right on to begin with i'm going to start from scratch. i'll level the machine out then make a test cut and see just how bad it is. i don't think the ways themselves have much wear i believe the wear is in the flat carriage way.
 
Don't level the ways, level the carriage.

Clean the carrige wings and stone off any burrs. Place the level across the carrige wings at 90 degrees to the ways. Then, crank the carriage back and forth along the ways and adjust the mounting feet until you kill the bubble movement. It doesn't matter what line the bubble ends up on, so long as the bubble remains on that line along the entire length of the bed. Do not touch the level until you're done and satisfied that you have removed the bed twist.

If the lathe is worn you may still need to tweak the machine further for best results with a two-collar test. Southbend's "How to Run a Lathe" discusses this method IIRC.
 
AP that's what i was driving at from the beginning. when i did what you mention which is basically what the manual says the ways were all jacked out of shape. now that said IF i can get it to cut straight by leveling off the wings then that's what i'll do but i'm kinda doubtful at this point, i'm not sure just how much i can do with this machine. it's a shame if i can't get it to do better because it's in overall good condition otherwise though of course a machine that won't hold tolerances isn't very useful. i got a really good deal on it and it came with a lot of nice accessories plus i added some. i really busted my butt cleaning, stripping, painting and massaging it the best i could. i leveled the ways figuring that would be a starting point. well the results were terrible. with a piece of aluminum bar stock i made some test cuts the last couple being nice and fine. starting at the chuck and at one inch increments out to four inches this is what i got. 1.3229, 1.3222, 1.3216 and 1.3212 for a deviation of .0017 over four inches which is terrible. i wanted to use a larger diameter stock but this is all i had handy and doubt there could have been that much deflection given how light the last cuts were. i've been thinking about putting a DRO on this machine but if i can't get it do a lot better than that there's little sense in investing in a DRO. not that i think it's an issue in this case but i've wondered why they don't put a lock on the cross slide like they do on the carriage or on all the axis of a bridgeport. i have on occasion had the cross slide move on my little sheldon while making cuts. obviously the tool should be pushing back against the cross slide but any movement could screw up a cut. i'm gonna try jacking things around going off the wings and see what that gets me. thanks again
 
Interesting thread.

Is it correct that your test cut is with stock held in a chuck on the spindle? I suppose that doing a test cut on stock held between centers might produce different, happier, results. If so, it might redirect your focus of attention.
 
If a lathe is long then you might spend some extra time on leveling but even then an hour is all you need , then check it next week.

A 17x42 is not big so get it close and tweak it later, its too short to stress over it.

Bolting it down? thats silly it wont help much on a little lathe like that.

Now if you had a cnc or screw machine with a 12' to 20 'bar feeder then you want to bolt it down and get it very level.

If its new to the shop let it settle in then tweak it right. dont expect a machine to be set for good after the first leveling.
 
"i'm gonna try jacking things around going off the wings and see what that gets me."

No.

Put the level on the shelf and go back and shim the bed till your diameter numbers start
to converge. The level at this point is leading you astray. You are pretty close being
less than two thou out over four inches, at this point you can fine tune to get that to be
one thou or less over six inches or more.
 
Let me make another silly suggestion. How do you know the spindle axis is parallel to the ways, have you done any tests ? Regards Tyrone.


You have a point there Tyrone

This is how I was tought

you level a lathe to have it straight and square
AND AFTER THAT YOU DO A 2 COLLAR TEST TO SEE IF THE SPINDLE AXIS IS PARALLEL TO THE WAYS
If you just start with a 2 collar test you only know the lathe isturning straight the lenght of your test Not how it is on longer lenghts

On a used lathe you have to tweak it here nad there but you start out with leveling as good as you can and only then the 2 collar test


Peter from holland
 
WOO-HOO this puppy is cutting now

apologize for not getting back sooner got disgusted with this machine and moved onto some of the other hundred projects i'll never finish.

anyhow i decided to take another shot at the machine tonight and the results were pleasing to say the least considering where i started (.0017 over 4 inches). i leveled the carriage across the wings to within about .0005 per ft then made a test cut on the aluminum bar stock. i had removed the bar so i could get the carriage closer to the chuck with the level on it so when i put the bar back it had a little run out. i had intended to take the run out off then make another finer cut but the first cut while not a fine as i would have wanted wasn't heavy at all. so i held my breath and miked six inches of the cut (drum roll please) and it held .0004 over the six inches with most of it being .0002. i think i started wearing the bar i miked it so much :-)

i assume like most guys the majority of the work i would be doing is closer to the chuck, i rarely have occasion to work with long lengths. i guess now i should get greedy and indicate the tailstock into the chuck then try something between centers but i'll leave that for another day. heck if i can get it to cut nice over a long length between centers i might even seriously consider treating it to a DRO.

in response to the question of the head alignment, no i'm not sure it's straight but that said if it isn't i don't think there's much i could do about it given the size and construction of the machine. while this is by no means a big machine it's not exactly a small one to have in your backyard. this bugger weighs in at about 3000#.

thanks for all the suggestions and input if/when i get brave enough to try a long piece between centers i'll post the results. i have a ROHM live center that looks to be in decent shape so maybe i'll give it a try one day.

as it stands i'm happy with the results leveling off the carriage got me and if there was as much wear in the carriage as i think i remember i doubt i could ask for better than .0004 over 6 inches.
 
more refined point

Leveling the lathe in a particular plane is not really the point. You should be using the level to determine if there is Twist,Bow, or droop in the ways. The routine of checking the ways in that order is to indicate where a condition of inaccuracy is. The point of leveliing off the bed first in one direction gives you your first datum point. Then based on that reading( on the level in this case) tells you how to adjust the rest of the bed. The leveling procedure doesnt involve the head stock/tails stock so much as the carriage. After leveling a machine , a person would then check the spindle orientation/run out to determine if the head stock is parallel to the bed ways. On a new machine that would be assumed by the recipient. Then testing of a cut or use of a test bar would follow. Then alignment of the tailstock to a "0" position would be carried out. ( basically checking to see if it is in line with the sindle, and if it's indicator reads zero.

Ether the above, or I am totally full of it!!

As for naval machines leveling the floor is done while in port, and since a ship is a large "Level" floating in water, the floor will stay level. taking into acount that the vessel is at rest. In motion, the floor stiff as it should be, will keep the machine that is bolted to it straight. All the above within in certain limits. Not in combat, or rapid motion of the vessel, of course.
In a used machine you have to take into account the wear as well.


Just my two cents.
 








 
Back
Top