• Scam Alert. Members are reminded to NOT send money to buy anything. Don't buy things remote and have it shipped - go get it yourself, pay in person, and take your equipment with you. Scammers have burned people on this forum. Urgency, secrecy, excuses, selling for friend, newish members, FUD, are RED FLAGS. A video conference call is not adequate assurance. Face to face interactions are required. Please report suspicions to the forum admins. Stay Safe - anyone can get scammed.

Tailstock alignment questions

Yes, one of the classic machining tinder boxes. There are literally a lot of moving parts and tons of variables.

When you make, or at least remake which is what reconditioning is, the geometry of a machine, you need to follow a process that is exacting. There is not much equipment required, but if you don't have the right equipment, you get it, e.g. a Starrett 199 master precision level. Its different if you are just trying to set up a lathe; it's not realistic to say you have to spend $400-500 (and that's used) for a 199.

Which leads to a problem. I don't think I've used or seen a lathe upclose where headstock isn't solidly registered. They have all had inverted V ways or are pinned. You can rely on it (or least you could in the past) coming from the factory correctly aligned. This eliminated HS alignment as an issue.

One thing I learned here is that is often no longer the case. After watching how SM did it over a few days, I can see why manufactures would see it as a huge cost savings not producing lathes in this manner, but imo its a real disservice. SM had three guys full time scraping HS, TS and carriages in to the ground bed. Its imo the right way, but it is expensive.

Nevertheless, if this is your the current reality, you can't solve this problem without having a way to detect twist separate from HS misalignment. I haven't read everything that is here, but this challenge is no different than that faced by the person reconditioning the machine. The correct procedure is to remove the bed twist via leveling before working on scraping the HS in (HS alignment) because if you don't, any reading (or tests cuts) you make with the carriage moving over the bed, will be erroneous if bed has twist. If you don't have a 199, use the best you've got.

You do have an additional problem: wear. Wear can mean even with zero twist in the bed, relying on the carriage travel for detecting things relative to the headstock introduces error. There is little you can do about this, but one technique is to use a Kingway (or my version of it shown below which I think has several improvements) to slide along unworn surfaces. This is better than nothing and good for surveying as well, as advocated by scraper Richard King. (Having seen lathes been made, I do disagree with him as to how perfectly reliable these surfaces are or how exact their vectors matches those of the original ground surface as, at least at SM, they were done on different machines and setups than the way grinding).

A couple of other ideas. The best way to align the headstock, with the bed first leveled and a reliable surface to move along, is to mount a precisely ground cylinder, like a cylindrical square in the lathe. Weight reduced (holes) and a few inches in diameter is ideal. Just put it in the three jaw, alignment does not matter.

You take a reading at the chuck, then at the end - two planes. Then you rotate it 180 degrees and do so again. Looking at the differences between them tells the story. In my diagram, if A=B, its aligned. When scraping of course you have to do this in two planes, ZX and ZY but for checking alignment only ZY is looked at. Using the A=B difference method eliminates the need for test bars and eliminates any source of error in mounting - even the taper spindle mount introduces some error. You also want to sweep the indicator, not drag it along the work which can introduce error as the work may be going up or down hill in the other direction - see photo of sweep tool.

Despite all this, turning a long cylinder parallel to few tenths remains a BIG challenge. The only time I might take a test cut for alignment is with this type of job....wear, tailstock quill clearance and concrete slabs moving making things somewhat "dynamic". This is where figuring out what mounting bolt to tweak a few degrees turn pays off :)



Rx4mTqc.jpg


RuxwWpf.jpg


ClYGGr0.png





tk1OdHe.jpg

I like this very much!
 
You lost me on that one. I was describing how a lathe should be set up, has nothing to do with tool overhang afaik

Your description was good. I was mostly having fun.

The issue I was poking at is the fact that a cross slide with ways scraped to cause a tenths off course travel would cause the tool to follow that slight curve too. But depending on where the tool, tool holder or tool post are mounted on that curve, they will have a slightly different path.

Deliberately eggagerating, if the tool is at the center of the part BEFORE the cross slide even starts moving, then the tenths adjustment will have no effect. If the tool assembly is so short that the cross slide is at the far end (assuming that was even practical), the same effect cuts in. To work properly, I think that the scraping offset must coincide with the tool on-center. Since that depends to some extent on tool hangout, I wonder why it needs those few tenths at all......

Keep in mind that I'm not arguing. I'm only curious about it and wanting to learn something.

It might also help to know that when I was first contemplating the issue (before your post) I was thinking in terms of a small or weak or worn Cross-Slide that had some "wow" in its travel.
 
Curve? there's no curve....the cross slide ways scraped straight but at a slight angle, so anything above the, compound, tool, tool post etc should move along the same straight vector.

Maybe the issue is calling in convex, slightly misleading. Everyone calls it that, but technically its really not convex (which would have a curve), but a very very squat indented cone shape.
 
Curve? there's no curve....the cross slide ways scraped straight but at a slight angle, so anything above the, compound, tool, tool post etc should move along the same straight vector.

Maybe the issue is calling in convex, slightly misleading. Everyone calls it that, but technically its really not convex (which would have a curve), but a very very squat indented cone shape.

See. That splains it perfectly! I had been thinking waistline curve.

But it raises another question in my curious mind. What is the purpose of that VERY VERY VERY VERY VERY squat indented cone? Wear compensation?
 
See. That splains it perfectly! I had been thinking waistline curve.

Who told? Lets leave waste line curves out of it.....starting my diet today

But it raises another question in my curious mind. What is the purpose of that VERY VERY VERY VERY VERY squat indented cone? Wear compensation?

Now I have to quote myself :)

Why? There is no such thing as "dead on", everything is tolerances. If you tried to make it cut at exactly 90, some lathes would be 90.001 and others 89.999, i.e. some would cut convex and some concave. On those that cut convex, anything you faced wouldn't sit flat on the face but would rock. Not desirable.
 
Now I have to quote myself :)

Sorry about that. Yes you did say that.

My only excuse is that you said that back when I still thought we were talking waistlines. So it didn't compute under the fat cone scenario either. Just my stupid brain latching onto possibilities and lack thereof.

Then again, if the fat cone "biases" (good choice of words eh!) the tolerance to the concave side, why wouldnt fore/aft placement of the cutting tip relative to the minor axis of of the fat cone affect that too?

Just a thought here - if it wasn't really a fat cone but rather more of a very very very very very (LOL) slight inward angle, it would also cut a concave as long as you never cut past center - which only an idiot would do - unless their lathe was in reverse.
 
Hmmmmmm...... I just answered my own question.

If said fat cone is in the saddle ways, not the cross-slide, then the minor axis could always be on the lathe centerline. That way both forward and reverse would do that and always be where you want it no matter what the tool hang out was.

So...... Another question. Could it be that small EXPENSIVE HIGH QUALITY lathes anticipate this effect and being so light the effect might be worse so a high quality small lathe has an even fatter cone deliberately built into it to compensate for the even wider tolerances?
 
This was a really great discussion @Mcgyver. I learned something totally new. I'll have to take a very careful close look at how my own lathe is setup.
 
Back
Top