New G4003G Lathe - Alignment Problem

frostheave

Active User
Registered
Joined
Sep 29, 2013
Messages
130
Hi all. I am having a problem aligning my lathe. It appears to me that the lathe bed has a high spot in the middle that I cannot level out. I would appreciate any suggestions.

I didn't want to raise the height of the lathe so I riveted 2"x3/8" aluminum angle to the sides of the stands and bolted them to my concrete slab. I used 3/16" stainless rivets that have a shear strength 1170 lbs. There are 8 "Red Head" Drop-In concrete anchors, one in each stand corner and 3/8"-16 nuts for height adjustment. It remains to be seen how well they hold up over time. Hopefully they will work.


6 - Levelling Adjustment Screw.jpg


7 - Lathe Base Leveling Brackets.jpg


8 - Lathe With Stand Brackets.jpg



After a lot of reading on this forum, I decided to use a machinist level and then do the 2-collar test. It seemed the recommendation by Richard King to put the machinist level on the cross slide would be better than using 123 blocks or parallels on the ways. Below is the level and 2-collar setup I used.

9 - Lathe Alignment Level.jpg


10 - Lathe 2-Collar Test.jpg



Now on to the problem.
Using the leveling nuts on the bottom of the stands, I first leveled the head stock end, then leveled the tail stock end. When I moved the level to the center of the lathe, the right way was at least 1 division too high. I tried to get the bed level at the HS, TS, and center all at the same time but wasn't able to.

Next I got as close as I could and performed the 2-collar test to see what would happen. If you click on the image below, you can see the results from several tests.



Lathe Alignment Data.png



So my first question is, is there anything I can do to eliminate the high spot in the middle? When the lathe arrived, the bolt holding the tail stock end to the pallet had fallen out and the lathe could pivot around the head stock on the pallet. Also, the bolt fell out of the tail stock and it was free to move on the bed ways. Could the bed have been damaged due to these problems?

Since the 2-collar test shows least 0.001" out of alignment, do I need to align the head stock with the ways? Where should I level the lathe before changing the head stock alignment? Finally, this is my first lathe alignment. I may very well be doing something wrong. Any suggestions are greatly appreciated!

Bob J

6 - Levelling Adjustment Screw.jpg 7 - Lathe Base Leveling Brackets.jpg 8 - Lathe With Stand Brackets.jpg 9 - Lathe Alignment Level.jpg 10 - Lathe 2-Collar Test.jpg Lathe Alignment Data.png
 
Here's my usual "diatribe" regarding lathe alignment...

1) Level doesn't mean anything.
2) If the lathe cuts properly, don't worry about measured differences in bed dimensions etc.

The way to fine tune using the two-collar test, is to make a test cut then, make an adjustment in bench legs -followed by repeating the test. One of the write-ups about the 2-collar method suggests which legs should be adjusted given various results of a test cut. I don't happen to remember where that write-up is but, I'm sure some googling will find it.

A side note: When I built the bench for my lathe, it was positioned and intentionally "un-leveled" just enough for fluids to drain forward for easy clean-up. I'm guessing that from front to back, it's intentionally out of level by 1/16 or 1/32". Once I got the bench the way I wanted it, I did the Rollie and Two-Collar methods until it cut right. Last week, I made a 1.25" diameter, 18" long, 316 stainless shaft spun between centers. Diameter variance wavered a little (as expected) but remained +/- 0.0003" (total 6 tenths). In all honesty, I wasn't trying to get this shaft that close but, after a couple stabs at adjusting the tailstock, it just happened to come out real nice.

Ray
 
Here's my usual "diatribe" regarding lathe alignment...

1) Level doesn't mean anything.
2) If the lathe cuts properly, don't worry about measured differences in bed dimensions etc.

The way to fine tune using the two-collar test, is to make a test cut then, make an adjustment in bench legs -followed by repeating the test. One of the write-ups about the 2-collar method suggests which legs should be adjusted given various results of a test cut. I don't happen to remember where that write-up is but, I'm sure some googling will find it.

A side note: When I built the bench for my lathe, it was positioned and intentionally "un-leveled" just enough for fluids to drain forward for easy clean-up. I'm guessing that from front to back, it's intentionally out of level by 1/16 or 1/32". Once I got the bench the way I wanted it, I did the Rollie and Two-Collar methods until it cut right. Last week, I made a 1.25" diameter, 18" long, 316 stainless shaft spun between centers. Diameter variance wavered a little (as expected) but remained +/- 0.0003" (total 6 tenths). In all honesty, I wasn't trying to get this shaft that close but, after a couple stabs at adjusting the tailstock, it just happened to come out real nice.

Ray

Thanks Ray. I agree "level" is not important and it is the actual cutting that is. It is interesting that the old "South Bend" manual states that leveling the lathe and using the 2-collar methods can be used interchangeably. It describes how to do both methods. I tried to do both methods simultaneously. Given my case, do you think I need to adjust my headstock alignment with the bed or just continue tweaking the alignment by adjusting the stands? How does one know when actually changing the HS alignment is required?

Bob J
 
Adjusting the base legs can vary things by way more than a thou. You'll be surprised at how little adjustment it takes to change the results by a thou. You will find it difficult to zero in and make changes as little as 0.0002 because by the time you tighten the opposing lock nuts, it changes things in unpredictable ways. -Stick with it and you'll get there.

It's a little out of my league to make predictions about when you need to do head alignment. My theory however is: When a machine is made at a factory, they bolt it to a work bench and check all the specs. That is when coarse head alignment is probably done. If those specs fall in a normal range, fine tuning should be possible by adjusting the legs. Unless the machine suffers a terrible crash, gets dropped or disassembled, it should (probably) never need coarse head alignment assuming it was setup properly at the factory. Again, I am not a machine rebuilder and this is theory on my part.


Ray


Thanks Ray. I agree "level" is not important and it is the actual cutting that is. It is interesting that the old "South Bend" manual states that leveling the lathe and using the 2-collar methods can be used interchangeably. It describes how to do both methods. I tried to do both methods simultaneously. Given my case, do you think I need to adjust my headstock alignment with the bed or just continue tweaking the alignment by adjusting the stands? How does one know when actually changing the HS alignment is required?

Bob J
 
Adjusting the base legs can vary things by way more than a thou. You'll be surprised at how little adjustment it takes to change the results by a thou. You will find it difficult to zero in and make changes as little as 0.0002 because by the time you tighten the opposing lock nuts, it changes things in unpredictable ways. -Stick with it and you'll get there.

It's a little out of my league to make predictions about when you need to do head alignment. My theory however is: When a machine is made at a factory, they bolt it to a work bench and check all the specs. That is when coarse head alignment is probably done. If those specs fall in a normal range, fine tuning should be possible by adjusting the legs. Unless the machine suffers a terrible crash, gets dropped or disassembled, it should (probably) never need coarse head alignment assuming it was setup properly at the factory. Again, I am not a machine rebuilder and this is theory on my part.


Ray

Thanks again Ray. That makes sense to me. I am going back to tweaking and will report back with the results.

Bob J
 
Adjusting the base legs can vary things by way more than a thou. You'll be surprised at how little adjustment it takes to change the results by a thou. You will find it difficult to zero in and make changes as little as 0.0002 because by the time you tighten the opposing lock nuts, it changes things in unpredictable ways. -Stick with it and you'll get there.

+1 If you adjust the nut to make it perfect and then tighten down the lock nut, you'll lose your setting. You have to learn how much to loosen one side before tightening the other (or even vice versa) to get it where you want. Even "tight" you have wiggle room for adjustments I've found.

There are several long threads here that go into great detail on how to get everything aligned. Best thing you can do is look them up because it's a lot of information!
 
So I continued tweaking yesterday. I also helped a buddy install a new floor pan in his rusty 1976 F-250. We had his truck in the shop where it is warm and dry. I made 7 separate height adjustments to the lathe stands, cutting .002" off the 2 collars each time. After the seventh adjustment, the outer collar was 0.00015" larger than the inner collar. Ray was right, I was easily able to remove the 0.001" error by adjusting the stands.

My buddies truck was parked in the shop about 10' away from my lathe. He left last night and I went ahead and did the 2-collar test again this afternoon. Now the outer collar is 0.00025" larger than the inner. I'm not sure if the truck weight, about 7,000 lbs, had anything to do with the change but for now I'm not going to worry about it. I'll check it again in about a week after it has had time to settle in. Thank-you everyone for your help!

Bob J
 
Glad it had a happy ending.

Yes, things shift and settle a little and it's worthwhile doing the 2 collar test from time to time. I guess the floor could indeed shift but suspect the bench settling and possibly even significant temperature changes could play a part too...

EDIT: One other thing. When doing the two collar test, try to use a shaft at least 1.5" diameter and only protruding from the chuck about 6". Take a very light cut of just a couple thou. Aluminum is preferred since it requires very little cutting pressure.

Ray
 
We were having our new cnc router leveled in with the laser at work. When someone would walk past the machine (within 10') it would move 10-20 microns. This is on a 8" thick floor and the router mounted on a 36" deep pad. You need a level bed to have an accurate machine within reason. Unless it is mounted and anchored properly it will flex with temperature changes and moisture content under the floor. There is no such thing as a perfect machine alignment that stays for an extended period of time because they all flex some and wear as they are used. Get it as close and rigid as you can and work on check the ways for good scraping. I have an older lathe that is worn so I pick my projects and will eventually get a new bed and scrape the carriage to a good fit. Most people who turn something only use a few inches. If you can hold less than a thou. over several inches, it is a good lathe for the majority of hobbiest. Tim
 
Like others, I used angle iron to increase the footprint of my G4003G. I originally used rubber feet but they wouldn't stay in alignment so I made copies of the rubber feet that were adjustable and all steel. These are still right on after two plus years. I started out by leveling the bed front to back with a .0002/ft machinist level. I had to go slow because the bed wouldn't flatten out right away after an adjustment but would settle overnight. Once it was level I got a 5C test bar and a DI to get the headstock aligned. It was already close because of the leveling. Once the headstock was set I put the test bar on centers to do the tailstock. The lathe is good for .001 on 6 inches. The biggest PITA for me was adjusting all 4 legs on each end to share the weight.

DSC_0259.JPG
 
Back
Top