Sieg C3 Headstock Alignment?

[ESzczesniak]

I've fairly new to metal machining. Almost 2 months ago I got a Sieg C3 7x14. I made a few simple projects, but am trying to pause and tune/align the lathe going forward. I've done a handful of basic housekeeping adjusting the gibs, backlash, etc.

Next on my agenda was check the head stock alignment to the ways. I'm using the Rollie Dad's method of measuring and am finding 0.00275" horizontal misalignment over 5 inches using a MT3 taper bar in the spindle.

Does anyone have experience adjust the headstock? It looks like this is keyed with V-channel and held over the ways with M8 screws at each corner. These seem awfully burried. Is the best way to get to them to take the motor out?

And it looks like the only way to adjust would be with shims? I've read general advice to be really sure you need to adjust the head stock before you adjust it. I can put a MT3 to MT2 bar in to my tailstock and have virtually 0 variance from headstock end to tail stock end. But if I leave it this way, the tail stock doesn't drill centered unless it's right up near the chuck. It hits slightly away from me in the center line, consistent with the measured yaw coming towards me.

Any thoughts or suggestions?

 

[petertha]

This is potentially 'one of those subjects' that can get pretty amped up in certain forums. I will tell you what I think I know.

There is nothing wrong with the coupon cutting validation testing using RDM or the other methods that precede it. After all, the lathe is to make cylindrical parts so testing is simulating the end result. Where the fur typically flies is you have taper, so 'what then'. The usual response is jack the lathe, unfortunately often called 'lathe levelling' which has nothing to do with level. It means lathe bed twisting. But where this advise is coming from is typically older style machines where the spindle is bored within a HS integrally cast into the bed way. So adjusting the feet (twist) is pretty much the only 'lever' you get to pull on those lathes to correct HS angle. And by & large accomplishes the task because lathes in those days were typically manufactured accurately enough. So far in discussion this has nothing to do with saddle wear or other potential issues, just alignment of the HS axis to the bed axis.

Now enter typical Asian lathes, including 14" & even larger swings & I suspect also including yours. Those have a HS bolted onto the bed casting and are 'adjusted' at the factory. They vary in how they are keyed, adjusted & retained but typically there are a set of fine adjustment screws that tweak the yaw (rotation of HS looking down on the lathe). They can be easy to miss & often don't show up in manuals. Maybe out of oversight or maybe they don't want you messing with them. I have seen this case many times actually. People swear they don't have them, its not in the parts diagram, but they ultimately find them. Perhaps well hidden or quite inaccessible, but they find them. If the HS has become significantly out of alignment this can have a dramatic effect on taper cutting issue beyond what lathe twist can compensate. So much that if you do the math, you would have to twist the bed like a pretzel to compensate & now you have 2 problems not just one. Now if the HS is out 'a little bit' you are essentially in the range of foot adjusting your way back to happiness. I'm just saying, be aware that the underlying alignment could be related to HS mounting on certain qualifying lathes. And Asian lathes built this way are arguably much more predominant in hobby machine shops these days.

HS can be misaligned for many reasons. Poor job at the factory. Became bumped in transport. Was taken apart & not correctly reestablished. Loosening/settling over time. Lost shims if they went that route. Moving issues like when you see pictures of people hanging the HS in a sling. An unfortunate chuck crash that displaced the HS slightly..... etc.

I have an MT5 HS spindle so I bought one of those <cough> precision bars with a MT end. The idea was plug it into the spindle, run a DTI down the length, adjust HS into position. You can check both yaw & lift of HS axis. No coupon cut testing involved at this point. Well the one I got was within tenths diameter down the length. I thought it was pretty good from a straightness perspective and maybe even was when new. But I now know its bowed about 0.003" at the end of 12" length. So this is a problem without some specialized compensation. Its equivalent to having a worn surface plate, the concept is good but its not reliable. Real test bars cost a fortune. The majority you see in the 30-50$ range are probably like mine. Maybe you will be luckier. Personally I think mine has stress relieved over the years but didn't have a great way to validate at the time. So I'm saying use your bar with caution unless you can validate it.

Your tailstock (TS) has zero to do with HS alignment, in fact will subvert your adjustments. Keep TS completely out of the picture for now. The adjustment sequence must be 1) get HS aligned true 2) align TS to HS. Once this is done, TS is relatively painless, I can show you several methods. And its something you should check often or you can have all sorts of issues: drilling, supporting, taper cutting between centers...

 

[pete]

ESzczesniak mentioned it's one of the usual 7x14 mini lathes. Unless they've changed the design, I don't recall seeing any of those with anything like head stock alignment adjustment bolts. Afaik the head stocks are all indexed to one of the way V rails to set there alignment to the lathe bed. So that may or may not have been done well enough depending on how long its been since there machining fixtures were checked that they are still correct. For some of what I've personally seen in the past, the Chinese seem pretty lax about those checks with some lower budget machine tools.

The one fault with those MT test bars is they assume the female Morse Taper is in pristine condition. Most hobby level users don't seem to fully appreciate the utmost importance of never inserting any dinged or even partly rusty male tapers into those unhardened female tapers. And both parts need to be spotlessly cleaned before joining them together or permanent damage will be the inevitable result. Any damage at all makes checks with those test bars just about useless since they depend on a very accurate taper to provide the same for accurate dependable alignment numbers. However..................there is a 100% certain method of checking your head stocks alignment without having one of those test bars at all. And contrary to what most hobbyist's believe, even the worlds best lathes are purposely misaligned right from the factory. No it's not by very much, but that misalignment is done for extremely good and logical reasons. On very good tool room lathes for example, that misalignment amount in rough numbers equals around .001"-.0015" over 12" with the head and tail stocks pointing both up hill and towards the operator. That out of exact square condition is an effort to help compensate for both work piece weight and cutting loads. And misaligned in those directions it forces the lathe to always face just slightly concave so parts joined face to face seat properly and don't rock. The lathe also wears towards facing dead flat and not immediately towards facing convex. The misalignment in the horizontal axis would most easily be measured on a lathe capable of at least a 24" swing and the measurement done on only one half of a faced part.

So how do we use those numbers on our much smaller machines? All it requires is the face plate the lathe came with or a thick disk of metal towards the maximum size the lathe can swing, and a 10ths or metric equivalent dial test indicator. With a face plate, first clean the spindle nose and the plates seating surfaces really well. Mount the face plate to the spindle, then take a light but full clean up pass across the whole face plates surface, in most cases that would be from the 9 o'clock position in to the face plates center hole. But either direction will work. Until the checks are measured in the next step, DO NOT remove that face plate. Run your cross slide back towards the operator, now place that DTI with a magnetic base on your cross slide with its indicator tip on the spindle C/L and just barely past the face plates center hole. Zero the indicator needle. Now gently run the indicator tip using the cross slide from just past that center hole to the face plates 3 o'clock position. Any deviation you see will be DOUBLE the hopefully proper amount of concavity the lathe is facing. It has to be done in that orientation because if you were to check the face plate using the same path the tool did, then the indicator would show zero from the 9 o'clock position to that face plates center hole no matter how far out the head stock is in either direction.

With any lathe, the bed and head stock end will always be the most rigid part of the whole machine. There designed that way to be so. So what I do is get the ways at the head stock end level to each other in the X axis direction. I then set the tail stock end to be level and true to the head stock end in the Z axis. The final check is again across the tail stocks end in that X axis. With that done and with any lathe that the tail stock can be off set, I then center the tail stocks Morse Taper bore to the head stocks spindle C/L. That's my initial starting point. Any longitudinal test cuts I'll do with a heavy bar. 2" in diameter per foot of length seems to work well. And I make those test cuts between centers to remove any inaccuracy influence from either a 3 jaw or 4 jaw independent chuck. Any adjustments to remove part taper are then always done at the tail stock end and again double checking the tail stocks MT bore after any adjustment. To add a bit to what Petertha has said, a very important double check is to run an indicator along both the top and side of the tail stocks quill with it fully extended and locked to check if there's any real misalignment. I once had a lathe that came from the factory with the quill pointing up hill over .009" in 2 inches. Add a drill chuck to that length, and every center drill I tried to use instantly snapped the tips off. So the less you pay for any machine tool, the less you can take for granted it's correct even when brand new.

Without question the majority of posts about how to do machine tool alignment checks are well meaning. Unfortunately most are written without fully understanding what's being checked and exactly why it's required to fully verify what you have, or even how its done to obtain reliable results. One of the main reasons for that, is the hard to find information about it is so hard to find from an amateur's perspective. Unlikely a 7x14 mini lathe comes with anything resembling a proper test certificate. But the larger lathes and the last one I bought did. The machine tool manufacturer's universally base those test certificates from a book first written in the 1930's with multiple later editions. A PDF of that same book can be found here. https://pearl-hifi.com/06_Lit_Archi...s/Schlesinger_Georg/Testing_Machine_Tools.pdf Downloading and maybe printing the whole thing off really should be non optional reading, and doing so enough to fully understand the proper test procedures and how to interpret any numbers your getting. There's actually two methods in use today for obtaining that concave facing condition. One as I pointed out purposely misaligns the head stock towards the operators position. The second very slightly biases the cross slides dovetail to do the same by either grinding or high precision hand scraping. Either will work, but your certainly not going to see that on any 7x14 unless you do it yourself. Another advantage of using the lathes face plate to check the head stocks alignment, is it's being done under dynamic real world cutting conditions. The Morse Taper test bar method is only a static test condition and without any of the same cutting loads the lathe will see when it's being used. I happen to own some very accurate and quite expensive MT 1-MT 3 test bars. I still think there well worth having for the initial checks, but the final settings are still done with my machines by using those test cuts.

 

[petertha]

ESzczesniak mentioned it's one of the usual 7x14 mini lathes. Unless they've changed the design, I don't recall seeing any of those with anything like head stock alignment adjustment bolts. Afaik the head stocks are all indexed to one of the way V rails to set there alignment to the lathe bed. So that may or may not have been done well enough

This is a very good point. On some lathes the HS could be keyed to the bed Vee profile if it runs continually under the HS, not unlike how a saddle or tailstock is fitted to a Vee within the sliding range of bed. The parts schematic should reveal these details. Now if HS is essentially locked in a misaligned postion, there could be some remedial options but proceed with caution. Maybe enlarging the HS female Vee profile a bit & back-filling with one of those dedicated epoxy machine bedding products like Moglice? But you would have to have a very, solid pre-alignment plan so that when the HS was set down on the epoxy & brought into alignment, there was no chance for error or ambiguity because that is how it will cure & bond. This is where something like an MT ended bar would be good in principle because it would be easier to implement, but we now rely on both bar & socket being true. My Made in India bar is diametrically within tenths but has literally & physically gone banana several thou on the end, so about as useful as a French curve for straight lines

Here is the schematic for my 14x40 lathe just to add to where I was coming from. The HS resides immediately left of the removable gap section. Its been a while but I believe those rectangular pads coincide with where the set screws come against. My point is that adjustment screws are not indicated in the documentation but are there in reality. So I should qualify what I wrote if it wasn't clear enough. You may have set screws even if its not indicated, but this is only relevant to this particular style of bed.

 

[hanermo3]

To OP ...
0.05 mm over 5 inches seems like quite a lot.
Typical is 0.01 mm - 0.02 mm, from what I have seen past 20 years.

It´s quite possible to sand the steel V fitting surfaces to whatever angle you need.
You need a proper flat of some kind.
Buying a few pieces of precision flat ground steel stock might be one way.

Put 400-600 wet/dry sandpaper on the flat, use alcohol on the paper, and gently press down more on one end or the other, introducing bias.
Mark the surfaces using a permanent marker, put double/triple parallel lines where you want more metal removal.

It only takes less than half an hour.
You are aiming to remove a few grams of material only.
Sanding rounds the corners, so the fitout will be less than perfect - but it´s generally much better than what came from the factory.

If you want really perfect results, you can use specialist metal filled epoxies to finish the job, after you are really sure you have it fitted out right.
It wont really matter.
The 7x series is kind of flimsy.

Example:
I regularly make flat steel mounting surfaces to 0.02 mm - 0.01 mm using a big heavy wide festool 105 belt sander and 40-60 grit belts.
The trick is to use 100 mm wide flats, for 35 mm wide linear rails.
So the rounding occurs on the the sides of the steel flat, but the center part is extremely flat and true.
The wide belt sander wont/cannot dig in, and sharpie marks easily indicate how much metal you are removing.
Tracking with a 0.002 mm analog dti and a digital 1 micron dti confirms the results.

You can feel resistance on the linear car, if you put say a cigarette paper or alu foil (0.04 mm) under one part of a linear rail.

It´s really not hard, with a bit of practice.
(Its noisy, heavy, dirty, and a big pain in the keester.
But it´s also fast, accurate, and totally controlled.)

You can practice a bit with any piece of steel, against a flat surface of some kind.
The small 200 mm surface plates are cheap, and will give you a true reference.

With a micrometer, I prefer digital 1 micron mics, you can confirm that you made say a wedge of 0.010 mm taper over 100 mm from a piece of HSS stock.
Two wedges of 0.010 mm mm, together, should mic the same when put together, and 0.020 when swapped end to end.

 

[petertha]

Mount the face plate to the spindle, then take a light but full clean up pass across the whole face plates surface, in most cases that would be from the 9 o'clock position in to the face plates center hole. But either direction will work. Until the checks are measured in the next step....

Can you help me understand (what might be Schlesinger?) method. If the HS is rotated inward or outward (2 or 3), I can visualize the resultant convex/concave profile on the faceplate/coupon. But I am missing the bit about how to distinguish an angled HS from an angled cross slide; mode 4 looks like mode 3 from the cutting result on coupon. Can you elaborate on how these 2 effects are independently validated?

Schlesinger shows a straight edge perpendicular to axis of rotation and shows cross slide movement at some angle, but in that picture the face is already perfectly flat (no convex/concave). I've read it a few times but its not sinking in. In the picture is the bed axis aligned to axis of rotation or is it perpendicular to slide axis? In this diagram it goes unmentioned. If I mount an indicator to the faceplate coupon in modes 2,3,4 like he shows, isn't it going to read the same?

 

[pete]

I'm not sure I'd use one of the specialty way slide products under a head stock Petertha, since there a plastic product, I think you'd start seeing them slowly cold flow over some time under the pressure of firmly bolting the head stock down. There are and what's already been mentioned special and no doubt fairly expensive epoxy's today meant for use on machine tools. With a lathes head stock, everything you can do to ensure a firm evenly supported bedding of the head stock to the bed will help eliminate sources of chatter. The tough part is guaranteeing the head stock is in fact properly aligned until that epoxy sets.

And yes your correct, the method I detailed won't 100% identify a combination of inaccuracy between both the cross slide and head stock alignment or there misalignment whatever that amount might be. What it will do is allow measurements of what the lathe currently produces under those cutting loads. Its then a process of step by step elimination. And you have to start with some known baselines of what the lathe is currently doing. If I wanted to check if the cross slide was angled towards or away from the head stock, that's also easily done. Use the longest and known to be straight edge you have that's still short enough it can be swung 360 degrees in the lathes head stock. Lightly hold that and roughly centered in say the 4 jaw independent chuck between two of the jaws, I say lightly because you don't want heavy jaw pressure distoring your straight edge, and you only need to hold it tight enough to resist the tip pressure from an indicator or have it slip. Place an indicator / magnetic base on either the cross slide or lathe bed. Zero the indicator tip on the straight edges end closest to you, rotate the spindle and straight edge by 180 degrees. Adjust the straight edge ends in either direction until both ends indicate zero / zero. At that point you know the full length of the straight edge is square to the spindles rotation in the X axis. Now use the cross slide to indicate the length of that straight edge. That will show the cross slides exact orientation to the spindle. Because of the critical nature of these measurements, I think I'd want to use at least a 10ths or metric equivalent capable DTI.

In my opinion or at least it was true for me, machine tool alignments aren't the easiest to properly visualize in all three dimensions. Its non optional to be able to do so and why I recommended reading that Schlesinger PDF enough until your grasping what it's showing. Read through it a few times over a number of months while thinking about what its trying to teach, sooner or later it will start to click and make complete sense. While it's expensive at around $100 with shipping, a very good but hard to read through because its pretty dry companion book, would be the Connelly, Machine Tool Reconditioning. Machine Tool Reconditioning for Machine Tools by Machine Tool Publications I'm almost positive a great deal of what's within it is directly based on Schlesingers work. Even if you never rebuild any machine tool, what's within it will vastly change anyone's opinions about poor machine tool way cleanliness and just how important proper and adequate lubrication really is. That alone was more than worth the books cost for myself. Connelly address it in his book, but by far this diagram illustrates the possible 6 degrees of freedom within any one single machine tool slide. Renishaw: Six degrees of freedom (6DoF) explained Properly understanding that is a key concept to the rest. I'd also highly recommend the Moore Tools book Foundations of Mechanical Accuracy, a PDF of it is here, https://pearl-hifi.com/06_Lit_Archi..._Tools/Foundations_of_Mechanical_Accuracy.pdf Since this has been a real subject of interest to myself for almost 40+ years, it's also not something that I think anyone will pick up in a few days.

Yes and without question that millionths level accuracy is way outside what ESzczesniak was asking about or what any of us could even approach in our home shops. Its still in my opinion not wasted effort to educate yourself about the basics. I've used some of the principals from the information I've listed to help check and tune some of my own machines. A full check of any machine tool using Schlesinger's methods takes quite awhile, but for most of it, your only verifying exactly what you have a single time. If the machine checks out you don't have to do the more extensive checks again for multiple years. Even though my last lathe came with an industry standard test certificate I still ran the exact same checks myself to verify those numbers were real and that those checks had even been done. There's also the fact that you can't even hope to try and compensate for any machine tool inaccuracy unless you already know where they are and by how much there out. This information is also why I maintain that well over 90% of the forum posts about basic machine tool alignments are trying to be helpful, but there based on a lot of less than well understood concepts. For most of them, they recommend lathe leveling and tramming the head and vise in on a mill and that's it. Yes those are obviously correct, but are still almost the very last step after verifying the rest with any new or used machine. If far more people would go through just the information sources I've mentioned, you'd then see a whole lot less well meant but still not quite well understood machine tool alignment posts on all these forums. I'm certainly not trying to slight anyone, I'm simply stating what the facts happen to be in general. I'd also point out, that for almost all of us, using industry standards for bearing fits, thread standards and multiple other items found in Machinery's Handbook is a common and completely accepted practice, you could hardly have a home shop and not do so. Yet most at the hobby level seem to ignore what industry uses to set up and do basic alignment checks on any new machine tool they get before it's ever put into use.