G4003G Headstock Alignment

Splat

Active User
H-M Supporter Gold Member
Joined
Mar 10, 2012
Messages
1,368
A friend has had a G4003G for months now and finally getting around to leveling it, alignment, etc. and has asked me to help. He's not PC savvy so I'm his lifeline, so to speak. He surprisingly anchored his lathe to the floor after I advised him I had done so with mine. The lathe is leveled as close as he can get it using my machinist's level. He did the 2-collars test (I gave him Richard King's writeup on 2-collars test along with a few of my tips) using a 12" rod of 2"diameter 6061 aluminum. He's using the 3-jaw chuck with about 10" stickout with 1/4"-wide collars at the ends. Yesterday I remind him to lock the cross slide and now he's getting .002" difference between collars.... .002" over roughly 10".
Running the level up and down the ways sitting upon the compound the bubble doesn't move so I know the ways are leveled. Only thing I can think of now is his headstock/spindle needs to be moved to get better alignment. The one strap we used to lift the lathe was really tight up against the front-side (closest to operator) of the spindle nose when lifting so I'm wondering if that moved the headstock over a bit. Now I didn't need to do this on my G4003G, thank God. Anyone know if there's a locating pin under the headstock or what's the best way to achieve the alignment? Thank you.
 
The only way to find out is to power off and disconnect (play it safe) and remove the front fascia panel and have a look. You'll find at least two studs or bolts inside there.
The other two are most probably accessed from outside the headstock out back.
I don't know about the G4003G but on the G0602 there are 4 studs clamping the headstock to the ways with no locating pins (tapered or not). There are 2 x grub screws on the two rear studs that give you a method of fine adjustment. If you find these are present and decide to fiddle them, then keep a small amount of tension on the studs/bolts so there is some friction to fight against while adjusting. Then nip up and test cut your collars for diameter. Repeat this procedure and continue to adjust until your happy with the results. Adjust any belting last. I scraped my headstock base flat (it was terribly warped) and adjusted it so the spindle was actually parallel to the ways using an mt4 mandrel and measuring in two planes. This got me under a thou over 10" and down to 3 tenths after test cutting collars. I was happy with that.
For a gunsmith lathe your friend should be able to achieve better than that with your help. I'd love to help but I'm in Australia, anyway someone else will chime in as well with their thoughts if their familiar with this model. cheers Alby
 
Most lathes usually need some tweaking of the headstock alignment to get the last few thousandth out. Unlikely it was changed with moving. That being said, there are other factors that can affect the two ring test such as any axial skew of the chuck. This is the information I sent to a friend recently on this. "If the lathe is leveled and it is not cutting true after leveling, then it is usually a headstock alignment issue, which is very common. The other factor is that the chuck has a slight angular skew so the piece will cut smaller at the tailstock end. You can chuck up something like a 1" round ground bar in a 4J, zero the TIR at the chuck and then check the TIR along its length looking at the +/- swing around 0. If the +/- swing in either direction of 0 is the same then you have some skewing of the piece in the chuck, but the headstock alignment is OK. If the reading is high in one direction then it is most likely a headstock an alignment. Looking beyond 12" is probably impractical, as both the ability to hold a piece true at that distance is unrealistic, you can get some slight deflection when turning, and in almost all cases you would use a tailstock. The two ring test works, but there is always some deflection by the cutter, and this will be affected by the DOC. "

Headstock alignment on these lathes is pretty straight forward, there are 4 large Allen bolts that hold down the head, two are visible under the chuck side, the other two are under the belt cover side. I only slightly loosen the bolts. On the G4003G there are two headstock jacking screws just above the motor in the back. It takes almost no movement (maybe 1 degree or a slight tightening or loosing) to make major changes in the head alignment. I use something like a 1" bar chucked up and measure the TIR along its length, looking at the +/- swing range is centered around 0 as I move down the bar, you can then gauge the movement of the headstock adjustments. Lock down the head, recheck everything and then do the two ring test. On the tailstock alignment, I use a 4J and adjustment for minimal TIR at the chuck and the tailstock end, then take a cutting pass along the bar and measure at different points. I adjust the tailstock until the cut diameter is the same as the chuck end. Sometimes for precision work I will turn my material just shy of the final diameter, check the diameters and ever so slightly tweak the tailstock so it cuts true along the length. Too long a pieces or small diameter material can deflect along the length of the material.

Just one approach.

G4003G.jpg
 
Is there a way to correct axial skew of the chuck?
 
All chucks can have some skew when holding stock, this can vary by the size of the piece held, the torque and also how the material is held. Chuck jaws are often ground with a very slight taper front to back to apply even pressure across the clamping surface once tightened, some chucks only have a single pinion to provide consistent clamping like a PBA that I use. Also materials are not always concentric and may have irregularities which affect how the piece is clamped. In some cases, the jaws can be reground to account either TIR or skew issues, but it is not a simple task. Collet chucks tend to be a bit worse in my experience, of all my chucks my Bison 8" 4J chuck has the least amount of skew which is ~0.0005 at 12". Other chucks can be +/- 0.001 at that distance or worse. The other issue can be the back plate is not running true, so I check that and the chuck body for run out for both axis.
 
Mark, thanks for the help. I haven't been able to look at the backside of either mine or my buddy's lathe (I assume they'd be the same though). I'll let you know what I find.

As for possible causes of runout, good thoughts guys. I understand about possible skewing of the work via chuck, swarf, etc... Wouldn't I see any possible skewing when slowly turning the spindle by hand with a .0001" DTI on the work?
 
Definitely if your not properly clocked in.
Definitely if your spindle is bent. (Clock your register and even the spindle bore taper seperately).
Possibly if your bearings weren't adjusted to a slight preload value. You'd know this is if chatter or uneven surface finish is present.
 
Last edited:
Definitely if your not properly clocked in.
Definitely if your spindle is bent. (Clock your register and even the spindle bore taper seperately).
Possibly if your bearings weren't adjusted to a slight preload value. You'd know this is if chatter or uneven surface finish is present.

I've clocked the spindle and he's got no real discernible run out. We're not getting a really nice surface finish but I think that's from the motor vibrations which is why I got the thick rubber mounts from McMaster for mine. That, along with the link belt, really nipped that issue in the buttinski. I advised him to do the same to his but he hasn't yet. Everything's on hold until I can get over there and assist him, so pro'lly Saturday. We'll see what happens.
 
Most lathes usually need some tweaking of the headstock alignment to get the last few thousandth out. Unlikely it was changed with moving. That being said, there are other factors that can affect the two ring test such as any axial skew of the chuck. This is the information I sent to a friend recently on this. "If the lathe is leveled and it is not cutting true after leveling, then it is usually a headstock alignment issue, which is very common. The other factor is that the chuck has a slight angular skew so the piece will cut smaller at the tailstock end. You can chuck up something like a 1" round ground bar in a 4J, zero the TIR at the chuck and then check the TIR along its length looking at the +/- swing around 0. If the +/- swing in either direction of 0 is the same then you have some skewing of the piece in the chuck, but the headstock alignment is OK. If the reading is high in one direction then it is most likely a headstock an alignment. Looking beyond 12" is probably impractical, as both the ability to hold a piece true at that distance is unrealistic, you can get some slight deflection when turning, and in almost all cases you would use a tailstock. The two ring test works, but there is always some deflection by the cutter, and this will be affected by the DOC. "

Headstock alignment on these lathes is pretty straight forward, there are 4 large Allen bolts that hold down the head, two are visible under the chuck side, the other two are under the belt cover side. I only slightly loosen the bolts. On the G4003G there are two headstock jacking screws just above the motor in the back. It takes almost no movement (maybe 1 degree or a slight tightening or loosing) to make major changes in the head alignment. I use something like a 1" bar chucked up and measure the TIR along its length, looking at the +/- swing range is centered around 0 as I move down the bar, you can then gauge the movement of the headstock adjustments. Lock down the head, recheck everything and then do the two ring test.

View attachment 235718
I just did this on my lathe. I had worked on my HS before so didn't think it would be out much. Used a linear motion rod so I know it is straight. Had about ten inches sticking out of the chuck. As I rotate the chuck I get +.004" & -.006" I continue rotating and get +.005" & -.005". You know that bugs me so much I don't know why my indicator does that. Anyway I decide to not mess with it. Obviously I have some skew going on. I ran an indicator along my 4J face plate and got ~.0004"-.0006". Sure would be nice to get rid of that skew.
 
Back
Top