May have an issue here with the headstock

[MikeWi]

After spending a while turning hardware store steel rod down in various ways, one attempt at threading (successful) and learning how to use the DTIs correctly, I've been noticing some signs of the stock being crooked that may not have been my fault after all. I decided it was time to do some testing and truing up on the chucks and faceplate and started with the 4 way chuck's adapter plate (wow, it's off in nearly every way!lol). I noticed while facing the main "face" of the adapter that the cutter was floating as it got near the hub, and it was only slowly coming on plane as I deepened the cut. I was only taking 2-3 thousanths at a time so that I could see what was happening, so not much was removed before I saw that. I had a suspicion and decide to DTI the face of the spindle. I had tested it for runout when it first arrived, and it was as close to zero as I can determine with the cutter marks still there, but I hadn't tried testing for tilt . I'm getting consistent measurements of an almost perfect .001 difference between the back and front side of the face, in the direction (away from the operator) that I would expect from how it was cutting. The outside 3/4 of the adapter is nice and flat now and it also shows that amount of variance. Crap.

I'm assuming that this is an unacceptable amount of skew. Am I right, and how bad is this going to be to fix (referring to other pm1236 owners here.). It's difficult to determine what bolts are involved for this.
Any help?

 

[Ray C]

Many thoughts in mind...

Face cutting is a funny thing -especially if you're using carbide. Carbide relies on constant pressure -and that can't happen when SFM is gradually decreasing as with a face cut. I went nuts a couple times over figuring this out. Also, cast iron does not like to be played with. It wants a firm, even cut.

The couple thou difference your seeing in a cut part could be due to those kinds of issues.

I would be inclined to check cross-slide parallelism by attaching a good quality flat bar against the naked spindle then, mounting a DI on the crossslide and running it from out to in. First, clean and adjust your gibs. You cannot hold the flat bar with your hand you won't be steady enough. You'll need a strong magnet.

Before you go carving things up, look critically at everything involved -such as your tool post holders, QCTP, backlash in the compound... Ahhh, that's another thing to watch for, sometimes a vibration will cause the compound to back-out. Happens a lot!

Don't worry... We'll get to the bottom of it. There are many things to check first...

Also, is your lathe leveled properly? Don't even expect a lathe to run true within a couple thou over 5-6" of travel. -Ain't going to happen!


Ray

 

[MikeWi]

Many thoughts in mind...

Face cutting is a funny thing -especially if you're using carbide. Carbide relies on constant pressure -and that can't happen when SFM is gradually decreasing as with a face cut. I went nuts a couple times over figuring this out. Also, cast iron does not like to be played with. It wants a firm, even cut.

I'm using a tangential cutter.

I would be inclined to check cross-slide parallelism by attaching a good quality flat bar against the naked spindle then, mounting a DI on the crossslide and running it from out to in. First, clean and adjust your gibs. You cannot hold the flat bar with your hand you won't be steady enough. You'll need a strong magnet.

Not sure I have something I can use there. I'll look into it.

Before you go carving things up, look critically at everything involved -such as your tool post holders, QCTP, backlash in the compound... Ahhh, that's another thing to watch for, sometimes a vibration will cause the compound to back-out. Happens a lot!

Yes, the compound did have a strong tendancy to back out about 10 thousanths unless I secured it in place. Any suggestions for that? Forgot all about it until you mentioned it. If it did manage to move it was less than the width of the mark on the dial as I could not see any change. T-tool stayed in place as it should, and the other parts seem to check out.

Also, is your lathe leveled properly? Don't even expect a lathe to run true within a couple thou over 5-6" of travel. -Ain't going to happen!

Funny you should mention that. I didn't have to do anything as far as leveling goes. That's one level floor! But that's with a normal level with extra marks for when I checked it. Could a twist in the bed affect face alignment right up against the spindle? I'm getting some dogs delivered so I can try to turn between centers and check the tailstock alignment, but I'm still hazy about determining bed twist.

I thought we had consolidated the leveling related threads to a new area but I can't seem to find it. I'm not doing anything else until I get some advice. I got nothing but time to make sure it's done right, and I need all the help I can get. LOL

 

[Ray C]

You'll need true flat bar -professional quality. A good machinist square is the least starting point. Anything else... you might as well use a banana...

I'm thinking about some sort of compound dial lock -maybe just a set screw in the right place -or some other fun little gizmo. Every lathe I've ever owned had that problem.

Yes. Twist will easily throw you off that much. When I first did the Rollie method and 2 collar method, it was off 5-6 thou over 6" (approx). A few quarter turns of the built-in leveling bolts brought it right in line. -Does not take much at all!

This is a good topic for me to add to the PM1236 operator's guide. Let me noodle on the best way to get a handle on this. It's actually kinda tricky...

Find my post about YALB Yet Another Lathe Bench. I know that thread references the other links.


Ray

I'm using a tangential cutter.



Not sure I have something I can use there. I'll look into it.



Yes, the compound did have a strong tendancy to back out about 10 thousanths unless I secured it in place. Any suggestions for that? Forgot all about it until you mentioned it. If it did manage to move it was less than the width of the mark on the dial as I could not see any change. T-tool stayed in place as it should, and the other parts seem to check out.



Funny you should mention that. I didn't have to do anything as far as leveling goes. That's one level floor! But that's with a normal level with extra marks for when I checked it. Could a twist in the bed affect face alignment right up against the spindle? I'm getting some dogs delivered so I can try to turn between centers and check the tailstock alignment, but I'm still hazy about determining bed twist.

I thought we had consolidated the leveling related threads to a new area but I can't seem to find it. I'm not doing anything else until I get some advice. I got nothing but time to make sure it's done right, and I need all the help I can get. LOL

 

[Ray C]

BTW: Level to horizontal means very little. It's a good starting point but real precision happens when you make test cuts. My lathe makes very good cuts and is not perfectly level to horizontal.


Ray

You'll need true flat bar -professional quality. A good machinist square is the least starting point. Anything else... you might as well use a banana...

I'm thinking about some sort of compound dial lock -maybe just a set screw in the right place -or some other fun little gizmo. Every lathe I've ever owned had that problem.

Yes. Twist will easily throw you off that much. When I first did the Rollie method and 2 collar method, it was off 5-6 thou over 6" (approx). A few quarter turns of the built-in leveling bolts brought it right in line. -Does not take much at all!

This is a good topic for me to add to the PM1236 operator's guide. Let me noodle on the best way to get a handle on this. It's actually kinda tricky...

Find my post about YALB Yet Another Lathe Bench. I know that thread references the other links.


Ray

 

[MikeWi]

I should have remembered that the level question was a trick To me, level is horizontal in respect to gravity, and what you meant was what I think of as alignment. I remembered that horizontal or planar alignment wasn't necessary, but it was nice to see it so close as it helps with the drainage if I use the flood cooler.

I found some references to alignment from your yalb post thanks. As much as I thought I was following all those threads back before I got the lathe, I'm still unclear on how to proceed. Things like the two collar test are shown both for detecting and eliminating bed twist AND for setting the tail stock on center. I centered the tail stock using the both centers and a machinist's rule, double checking by eye with a magnifier and I can't see any issue. I still have a taper, but I was using the 3-jaw chuck, and that led me to trying to get the other chuck and plate ready so I could be more accurate, and that led me to checking the front to back difference.

Your suggestion about backlash led me to check the gibs (should have done that from the beginning, doh), and while there was no detectable play in the cross slide or compound laterally, the gib screw in the front of the cross table slide was free-turning and I adjusted it so that it provided a very small amount of drag. The new nut I got from Matt was already adjusted perfectly as it was and there is zero backlash. The compound has backlash, and the gib screw was tightened a small amount, but that of course didn't affect the backlash. I don't see that it has a way to adjust it like the cross slide has. There appears to be a set screw in the middle of both slides over the gib. I'm assuming the gibs are tapered, so what would this be for?

Anyway if I'm understanding so far, since the far end of a test bar is larger, i'd be raising the front side of the tailstock end of the lathe bed correct? And this will somehow bring it into alignment with the headstock as well?

 

[Ray C]

Fist, I think you just answered our fist questions about how to lock the compound. The little set screw in the sides of the gibs would do nicely if it was more convenient to turn. I'll take care of that really easy and put a thumb screw in there. The overall gibs are adjusted with the two big screws at the long ends. Those set overall fit while the side screw can tweak or lock it.


The Rollie method and two collar test are pretty much the same thing (in rough terms) and I think you're confusing both with a different method of aligning the tailstock. The Rollie method adjusts for alignment in the sideways and up/down directions. The two collar test is just that... a test. If the two collars don't come-out the same diameter, keep adjusting the legs until they do. The Rollie method suggests using paper shims under the legs and this is what threw me off when I tried it. Paper shims my eye! I put leveling bolts on all 4 legs and I have a leveling mechanism built into the table frame. Using those kinds of adjsutment, I was able to get the sideways and up/down (elevation, if you will) right in the ballpark. Then I did two-collar tests and nailed it using the fine-tuning bolts in the table frame.

To adjsut the tailstock, you take any old long bar with center holes on both ends. Put two aluminum rings/donuts on either end then, spin between centers and take the same cut on each ring. When the diameters are the same, the TS is aligned. Don't even try this with a jawed chuck unless you like frustration. Jawed chucks and tailstock alignment are at opposite ends of reality when it comes to turning a shaft.

And by the way, none of this work is easy -not by any means! And you would be surprised how many people just plunk a lathe down, level it with a $200 machinist level then, call the dealer to complain about the machine when it doesn't cut straight. What you're going through is what a lot of people pay others to do. When a machineshop buys a new machine, the seller usually requires authorized setup and calibration before they honor the warranty.

I can't really tell you which end to lower/raise to set it right. When I did mine, things seemed a little counterintuitive at the time but, I will admit, I was very toiled (and yes, a little frustrated) toward the end of the process.

Ray

I should have remembered that the level question was a trick To me, level is horizontal in respect to gravity, and what you meant was what I think of as alignment. I remembered that horizontal or planar alignment wasn't necessary, but it was nice to see it so close as it helps with the drainage if I use the flood cooler.

I found some references to alignment from your yalb post thanks. As much as I thought I was following all those threads back before I got the lathe, I'm still unclear on how to proceed. Things like the two collar test are shown both for detecting and eliminating bed twist AND for setting the tail stock on center. I centered the tail stock using the both centers and a machinist's rule, double checking by eye with a magnifier and I can't see any issue. I still have a taper, but I was using the 3-jaw chuck, and that led me to trying to get the other chuck and plate ready so I could be more accurate, and that led me to checking the front to back difference.

Your suggestion about backlash led me to check the gibs (should have done that from the beginning, doh), and while there was no detectable play in the cross slide or compound laterally, the gib screw in the front of the cross table slide was free-turning and I adjusted it so that it provided a very small amount of drag. The new nut I got from Matt was already adjusted perfectly as it was and there is zero backlash. The compound has backlash, and the gib screw was tightened a small amount, but that of course didn't affect the backlash. I don't see that it has a way to adjust it like the cross slide has. There appears to be a set screw in the middle of both slides over the gib. I'm assuming the gibs are tapered, so what would this be for?

Anyway if I'm understanding so far, since the far end of a test bar is larger, i'd be raising the front side of the tailstock end of the lathe bed correct? And this will somehow bring it into alignment with the headstock as well?

 

[MikeWi]

Gotcha. I'll try using those set screws as a lock and see what happens next time I turn something. I've seen the two collar method described to set the lathe way relative to the headstock as well as the tailstock and while I could see how that would work for the TS, it didn't make any sense to me the other way , and you are confirming that. I found a good description of the Rollie's Dads test and I'm finding consistent results with my earlier tests, so (heavy sigh) I have to replace the original bolts with some levelers. I knew better at the time, I just wanted to get playing with my new toy, and now I'm going to pay for it. LOL Getting those bolts lined up was such a PITA.

On the plus side, that will leave a gap between the drip pan and the lathe. Fill that with some foam backer before a light bead of silicone and you have a sealed gap that's much easier to cut loose than when they are right against each other as you found out with your new bed.

 

[Ray C]

I don't know what your bench looks like or if you even have the stock bench from PM. You could use the stock bench. You do not need the fancy thing I built into my table. I only put it in there because I had plenty of stock to work with, it was winter -and I felt like screwing around.

That said, I was getting tired of the project toward the end and the last couple days of kneeling, sitting, stand, kneeling, sitting, standing.... followed by laying in swarf... -well, it got on my nerves.

Gotcha. I'll try using those set screws as a lock and see what happens next time I turn something. I've seen the two collar method described to set the lathe way relative to the headstock as well as the tailstock and while I could see how that would work for the TS, it didn't make any sense to me the other way , and you are confirming that. I found a good description of the Rollie's Dads test and I'm finding consistent results with my earlier tests, so (heavy sigh) I have to replace the original bolts with some levelers. I knew better at the time, I just wanted to get playing with my new toy, and now I'm going to pay for it. LOL Getting those bolts lined up was such a PITA.

On the plus side, that will leave a gap between the drip pan and the lathe. Fill that with some foam backer before a light bead of silicone and you have a sealed gap that's much easier to cut loose than when they are right against each other as you found out with your new bed.

 

[MikeWi]

stock bench. Just tried a quick experiment and levered up the front edge of the right side cabinet with a bar while watching the DTI and reduced Rollie's Dad "measured difference" from 22 to 1,5. Looks like I'm on the right track.