Does that mean you kept the old nut?

I guess for a sixty plus year old machine, .003" difference between the ends and the middle is very acceptable. I could easily live with .003" backlash over the entire length.

I have seen this arrangement on a Chinese mill/drill, and one other machine where the nut was actually two sections that threaded into one another with a third jam nut in the design to lock it after positioning.

Another design included a spring between the two which allowed for a little more wear.

The material, well it is such a small section and I have a pretty extensive collection of smaller pieces and could be just about anything I guess. I even have some Vespel SP211 which is bearing material that has graphite.

I think it is the old "teach a man to fish and he eats forever" scenario. I have cut successfully some inside threads and have a good, bolt on chuck, pick gear lathe, that will cut about anything in either direction.

Another thought would be to go to a ball screw, while way too expensive to purchase new, I find ones that are "new old stock" or used on ebay for pretty reasonable prices.

I think the first part of any successful project is getting enthused, thanks for the stimulus.

Fact is this would be an excellent project machine, the initial cost was little, and the renovation is more about time than it is about materials and in the end one could end up with a very serviceable mill.

Yes, I kept and utilized the old nut, since with the new nut on I'm only using one flank of the old threads, and the backlash in the nut doesn't matter.

I agree 100% that you have an excellent project machine that you'll be very happy with when you're done. What I've read about these mills is that while the rigidity may not equal that of a Bridgeport in good shape, it actually comes pretty close despite it's small size. The reason for this is that the column is a single casting right up to the swivel; no turret, ram or knuckle to lose rigidity at. Robert Campbell has some pictures in an old thread of a setup he made to cut a big wooden casting pattern, where he had a big dividing head cantilevered out off the table. That setup worked fine. I don't remember his exact words, but they were something like " I'd take a heavy cut and the Index would just grin and ask for more."

One thing, I've heard some people advise against ballscrews for manual mills, because they have a tendency to backfeed without a constant source of torque, e.g. a servo. So you'd have to lock whatever axis you weren't cutting with, which could get tiresome.

The biggest part of the job with cutting the threads, for me, was grinding the tool, which I found enjoyable and educational anyway. The next biggest part was finding correct theoretical dimensions for the 7/8 - 5 thread form, because it's not in Machinery's Handbook. (Not my version, anyway).

I think you're correct that just about any material you've got on hand would work for a nut. Once you've got the operation down, you can always do it again, not that you'll probably need to.
Glad to hear you've got a project you're excited about; I share your enthusiasm with my mill as well. I look forward to seeing how it turns out.

Regards,

John

That is the point for ballscrews, very low friction and zero backlash so that they can basically operate under power in both directions. I got a chance to use a ballscrew, homebuilt modification and it was uncanny how light the touch was, you HAD to tighten up the gibs to give some resistance. While a servo is fine with no pressure feedback, the human hand has to have some to be comfortable, or at least the unaccustomed human hand.

I would guess that one could....... just a thought, machine a tap. I mean you could mess with the dimensions a bit to get it to fit the worn lead screw just right and once you had it, then it would work for any number of nuts. I would imagine that for all the Model 40 people out there this would be a very useful loaner tool.

One could tap acetal, Delrin, aluminum or other plastic to check dimensions, this might even be easier than cutting an internal thread on the lathe? Maybe not, once you have your boring bar set up, bit ground to acme dimensions and the gearing set up for the 5 tpi, the actual cutting would probably go pretty quick. Just thinking out loud.

That'd be a perfectly good way to do it, too. Making custom taps has been covered in a couple different threads here as well, so you could pick up some good information. I wonder, though, whether you'd have to make a complete set of progressives? .110" of material per side is gonna take a lot of oomph to cut by hand in one shot. Maybe not so bad if you decide to use the self- lubricating plastic you mentioned.

Interesting take on ballscrews as well. I've never seen one on a manual mill; I was just passing along some stuff I'd read here before. Apparently they were having some problems where, while cutting in one axis, the tool would push itself out of the cut in the other. IOW, if cutting in X, the cutting force of the tool would push the table forward in Y until the tool was no longer engaged in the cut. The only way around it was to lock the Y axis before the cut. If cutting in Y, the table lock would have to be engaged. That wasn't your experience?

Out of curiosity, what sort of other equipment do you have available as you tackle your mill project? Got a surface grinder or tool & cutter grinder you can use?

Can't call what I have experience, more an observation. Since the owner of the machine adjusted the gibs, and here is another observation, the way he had it adjusted the gibs was too light for me. But OK for him, and he said that over a period of a year he found that he could lighten it a bit and he found it fine. Never did the "not holding" come up.

I once own a Tikka Master-Sporter, more of a field use target rifle. The bolt action was so smooth and light it felt like it was broken. I got use to the feel after a while. But when I went back to a modified O3A3 Springfield, which locks up like you are closing a hundred pound safe door, it was real different.

I have a number of machines including a 1932 South Bend 9" lathe and B&S surface grinder, a ShopTask mill/lathe. I rail against chinese crap but I have had this machine for a decade and go back to it all the time for jobs that the others can't or won't do.

I also have a 14 x 30 Monarch lathe circa 1930, which is another project, and an Atlas 7B shaper. I would put myself on the ladder somewhere between a basement home shop hobby guy, and almost having a real machine shop. Although one local machinist made a comment about how it all looked like some kind of unkempt museum. "Don't you have any desire to own any thing that is less than half a century old?. Why don't you get rid of all this crap and buy a decent machine?" My feelings was a bit hurt for a little while, but truly I love the old iron.

Back to the tap: I was thinking one of those long progressive taps. Maybe even looking around for one factory made, you never know. It would be fun and educational to make your own but if one could be had for a hundred bucks it might be worth it.

Or a set of taps, progressive. You are right, I don't think you could go from a minor diameter to a fully tapped hole in one pass. Just too much material to remove, especially if the socket is deep, at least an inch or so.

I have heard that to do an ACME thread you start out with a standard v-thread to remove material?

I hope it OK to broaden this thread a bit, I was thinking if a new nut and screw is needed why not make the screw a few inches longer, tapered, cut troughs, and cut the threads with the end of the screw then cut it off? If the nut is soft bronze would heat treating be necessary? Or would o-1 be an appropriate choice for the screw?

Bret

A story I heard in the sixties, had a group with the task of encouraging rural families in India, to plan smaller families. The problem was that the largest families tended to be where literacy was rare.

The group came up with large signs depicting a nice family with two children in front of a nice home and obviously prosperous. The other half of the sign featured a family with many children, signs of poverty apparent.

The story goes that most of the people who saw the signs, stood and shook their heads in pity, "those poor people, only two children."

I'm with them, when I see modern, squared off machines that look like washing machines and refrigerators. I shake my head in pity, "that poor guy, not a decent hunk of iron to be seen," and go back to my shop feeling rich......

Bob

Not to say that doing it just the way you wrote couldn't work, but to produce a technically accurate inside Acme thread, you'd have to make the tapping portion of the screw bigger than the rest, with wider flats at the crests. Acme threads call for clearances on the nut at the major and minor diameters.

Plus, I don't think Machmat needs a new screw, just the nut. Unless he's going to go with ballscrews as he mentioned. My machine (actually any Index model 40 you find) was 60+ years old, roughly used, and almost all the wear was on the nut.

Machmat, I hadn't heard of starting the thread with v-threads when I cut mine, I just bored to the minor diameter and started threading. Sounds like a good idea, though.

I had to read through your message a couple of times. You are correct I would think, in this way the screw and the nut would surely have to match.

I hope the screw doesn't need replacing, that would make the job even larger.

So do you think it wouldn't work?

More I think about the idea, the more intrigued I am. This would eliminate all the error in making a screw, since the nut is custom fit to the screw made. How about using some very fine grinding compound to lap/polish the fit? Of course if you don’t have at least some clearance, on a cold day you might seize the whole thing up.

Tough call, and my honest answer is I don't know. I think the screw would run in, but as you pointed out, there's more to it than that. What you'd have, at least initially, would be a nut with very close to a zero fit on the < end > of the screw. But I don't think that that, by itself, eliminates all the potential sources of error in making the screw. For one thing, your lathe would need to have an area of travel of about 12" long, where you could do your threading, that cut absolutely no taper. With that tight a fit on the end of the screw, any taper toward the back end is going to bind it. (hmm... unless you made the big end the tapping end... but that'd kill your "perfect" fit and defeat the purpose anyway... ) ...

I agree with your initial thought about added project time making new screws. It seems like a can of worms you probably don't want to open if you don't have to, unless you just want to try your hand at it. Have you had your leadscrews out for an inspection yet? I suspect you'll find they're serviceable.

Wish I had a more knowledgeable answer for you regarding the fit on the nut. I can tell you this: a lot of forethought and experience went into the design of power transmission threads. Generally (and this is just me talkin'!) I don't go second-guessing stuff like that. If Machinery's Handbook tells me Acme nuts need clearances, I cut clearances.

Whatever you decide to do, I look forward to hearing how it goes!

Regards,
John

Me neither, sometime maybe, in the next couple of years maybe, when I have nothing left to do, I would like to give the idea a go, just to see.

I have not, and even this project is going to have to wait a while. I have been close, pulled the cover and checked out the X screw, the threads look sharp, nice edges and there seems to be no noticeable wear. The Y I checked but not as well, but it too looks pretty good. You know there has to be wear but I don't think for me it is enough to be a major concern. If I can get .003" backlash with a new nut, or spring loaded nut or double nut, I will be more than happy.

Mostly I have cutting odds and ends but now I need to fly cut a piece that is larger than what the Atlas shaper can do.

So I decided to give the mill the old fly cut check, kind of like tramming on the cut.

I am figuring that after 66 years of use the knee might be sagging a bit. So I did a fly cut in many directions X axis and Y axis locking the spindle so the Z would stay put. Three inch fly cutter.

Right to left had a ridge between passes of about .003", the head tilts so I figured I can loosen it all up, have never tried to actually tilt the head, and give it a bump and that should take care of that direction.

Front to back have a ridge of just a finger feel, measure at about .001". The Index is fixed in this direction, not sure how to go about fixing this. Do you shim the rotating head at the mating surface? Doesn't seem like this would be a good idea being this is a rotating machined and ground surface?

Could use so thoughts from the more experienced.

Now the bid deal is the thickness of my fly cut plate front to back and side to side. This was all measured on a surface plate. Side to side it was real close to zero. But front to back .012", so my first thought was that the knee was sagging and I need to shim the bottom. This is .012" in seven inches.

But I went back and checked again and I just assumed that the outboard end of the knee was low, but it wasn't. The outboard, part of the knee away from the column is actually .012" higher. This surprised me. I guess I could shim the top of the knee against the ways of the column?

I just don't know. I went back for a fresh cut and got the same results.

So the questions are. How do you tram a head that has no adjustment front to back?

And what do you do about a knee that needs to be aligned, short of having the ways on the knee reground. They could probably use this, the Y travel gets a little snug at the ends of the travel, but not by much.

Hi Machmat,

There are guys on here who can tell you a lot more about how to measure and fix the wear and error in a machine tool than I can, so I'm not even going to risk spouting off and giving you bad information. But I will say that if you're comfortable putting shims on the column ways to alter the geometry of the knee, I see no reason why you couldn't shim the surface where the head mates to the column. They're both precision machined surfaces, after all, and the swivel head is likely to see far less movement than the knee over its lifetime, if it even gets moved at all.

At any rate, it sounds like you could have one of two things going on: dovetails on the back of the knee out of parallel with the column ways, or saddle ways out of square with the column. Determining which would probably be the first step toward fixing the problem. I'll leave it to the more experienced guys to give ideas how.
Glad to hear you're getting 'er cutting.

Regards,

John

After about a half a day of measuring and trying to figured out what was what as far as the alignment of the knee with the column and how much wear was at the saddle I finally confused myself enough and gave up.

You have to pick a reference point, and no matter how much wear there is in the knee column I decided that the table was the reference plane and worked everything from it. Life started getting better and the confusion less.

The only kind of vertical mill I have every had any experience with and not all that much in total, was a Bridgeport or Bridgeport clone. Tramming the head was about a half an hour experience at most.

Before I got a granite surface plate or a cast iron surface plate I used what had to be termed a poorman's surface plate, a piece of 1/2" thick plate glass. I once saw and optical flat that used a grid and reflection to look for imperfections so I set up a large piece of graph paper and used the reflection of the glass to look for any warp or flaws and found none. I know this isn't really a surface plate but I am willing to bet that it is probably within .002" of being flat in every direction, which is certainly a lot better than nothing.

The table on this old mill is beat to crap, looks more like a fabrication table in a welding shop than a milling machine. I have put a stone to it removing not much material but at least getting rid of most of the high spot dings.

I have thought about milling the top of the table about .030", this would really clean it up, but the mill could only reach part of the table leaving about two inches on each end untouched. This might be an idea only the ignorant would come up with? What I was thinking was that this would make the spindle and the table be in reference with each other. Once this plane was established then I could take the table to a shop with a large surface grinder and have them reference to part that was milled and make the whole table flat to that plane. Any thoughts from anybody?

I supported the piece of glass at three points, using gage blocks, just like I would with a surface plate. Set up the dial indicator and swung it on a seven inch radius. Figuring larger was better.



With the indicator set to zero on the front right, everything else measures higher, the max high spot being .028" at the back left.

I then loosened the three locking bolts on the head and with the crank at the back.



I doubt that this mechanism has been touched in the twenty years. Removing a bit of paint and a couple of squirts of lube at the connection of the spindle head to column, a handle on the worm gear shaft and it moved just like it was suppose to.

I was able to zero the indicator swing in the X axis direction. Snugged up the bolts and check again. Then checked the Y axis direction and was still out by approximately .030".

Everything was going pretty well to this point so I decide to loosen up the bolts again and see how much and in what direction the head needed to move to get it to zero in the Y direction. The head needed to tilt forward at the top. Which I would guess would be what one would expect after 66 years, the table/knee/column connection has either worn or sagged a bit.

So maybe this could be fixed via shimming the bottom between the knee and the column ways?

I put a ground angle plate on top of the table to attempt to measure Z axis of both the spindle and the knee and the knee seemed to be very consistent from top to bottom. The .001" indicator just sat on zero, so if there is wear, one would expect it to be in the middle somewhere and there just didn't seem to be much. Which means that if I can get the spindle pointing in the right direction, the accuracy of the machine should increase greatly?

I found some shim stock, .018" and put it between the column and the head at the top and snugged up the bolts. Checked and adjusted the X direction again and then checked the Y direction and had gone from about .030" to .010".

Today I will go and get some proper shim stock and finish this up.

I am thinking that this has to be wrong headed. These are machined and mated surfaces and putting anything in-between them can't be good. But I am not sure what else to do? This sure is the easiest fix.

I might not ever use the ability of the head to swing right to left. Maybe I should be shimming the column to knee connection?

I am just not sure how to tram a head of a machine that has no adjustment fore and aft?

You are correct, but I will have to admit after five hours measuring and trying to figure out what was what. I ended up with three pages of notes and total confusion and a fair bit of frustration. You need something as a reference plane. This morning after thinking about it, I guess you would pick the one plane that you can't really do anything about, that is the column.

That's probably not an accurate statement, one could regrind a column I guess, if it was in really bad shape. But this one doesn't seem to be, I have not removed the knee. Any real assessment I think would require this.

The measuring that I have done so far seems to indicate that the column is not overly worn, I am assuming that this was originally a quality machine and is vertical and parallel.

As per my previous post, you have to pick something as a reference plane, so I decided to pick the table and work from there. I have a small job, and just need to flycut some cast iron plates to thickness +.010" and then will move them to a surface grinder to finish. It is not a high tolerance job and normally I would do this with the shaper but it is too small and I find I am inducing error to the cut by having to reposition the material.

I love my Atlas 7B shaper but find a lot of the times it is just too small. But that is a whole nother conversation.

So I am going like you said with shims between the head and the column, I too see this as the quickest and simplest solution. Although it does seem on the other hand to be not optimal, I hate the idea of introducing material between machined/mated surfaces.

Right now the shim stock is .018” copper, but I will go with half-hard brass shim stock. These are much softer than the steel or cast iron in then head and column so I hope I am not hosing something up.

I am sure someone with more experience is going to speak up and say this is WRONG. I am certainly open to the right way of tramming this mill?

I'm fairly certain it's going to be viewed as bad practice as well, by those in the know. I didn't mean to imply that it was the correct fix for the situation, but merely the lesser of two evils. Between sticking shims between two precision surfaces that stay fixed vs. shimming two precision surfaces that slide against each other, I'd pick the fixed surfaces.

I just thought of something also: have you had the table off yet per our previous discussion? It occurred to me that having a look at the table ways would be a good idea, to make sure you don't have any deep gouges with raised metal all around them. If you do, that could be contributing to your table being cocked out of square front to back. Same with the saddle, just to see what they look like and where the lions share of the wear is.

Regards,

John

Get rid of the glass. It introduces way too many errors. You don't need it to tram the head.

Shimming between the column and mill head is ok'fine. My old Index has a .005" stainless steel shim in there as the front of the table was high by a couple thou. The correction was to place the shim at the 12 o'clock position thereby tilting the head down and counteracting the error.

Side to side tramming on my Index is less than .001". And that was using a indicator bar that spanned about 18 inches total, 9 inches each side of the spindle. I have no complaint over less than .001" in an 18" diameter. Front to back is now under .001".

I use two solid 1-2-3 blocks on the table, each side of the spindle. The indicator tip actually slides across these two 1-2-3 blocks, not the table surface. Works much better this way.

Perfection does not reside in a 1942 milling machine in near constant use for 65+ years. I'm tickled silly at the accuracy of my Index. The knee up/down is less than .001". It has great repeatability.

I don't tilt the head on my Index. Tramming the Index is tedious due to the way the mill head is clamped to the column. It requires you keep the 2 bolts a bit snug while tramming as there's error introduced when you re-tighten them. It took me nearly an hour to get it under .001". Back and forth, front to back, up and down. I was tired by the time I got done!

Unless you can measure the flatness of that glass it really has no place in adjusting the mill accuracy. You're just introducing error under the guise that you think its ok. If you can't measure "ok" it shouldn't be there.

Live within the limitations and accuracy of this old mill or buy a newer one that makes you happy. I wouldn't spend more than a couple hundred bucks to improve it. Like a new 1 phase 220v motor and toss all that life support junk.

You said you loosened the 3 locking bolts for the head. ?? There's only two. One at 3 o'clock and one at 9 o'clock that clamps the mill head to the column where the degrees are engraved. You have something else?

The both of you keep using this term for the column/head juncture: "precision machined surfaces".

Its a well worn surface that probably stopped being "precision" about 40 years ago. Now its just a nice smooth cast iron surface that needs some help in the form of a shim to let the old boy tap dance in a straight line again.


Dutch

I agree that the glass is not optimal to a proper surface plate or some such flat surface. But in lew of not having anything else at the moment. The glass surface is flat to within any tolerance I am able to measure. To get rid of the effect of all the bumps and divots I supported the glass on three gage blocks. I am thinking that it is about as good an index plane as I am able to present.

I will say that this table shows a lot of abuse, and picking what I think is three good support spots was an exercise in doing the best one can with what one has.

Twelve o'clock is the same as my eventual shim, I am using one that is .010". Tilting the top of the head forward and moving the spindle back, wouldn't this be compensating for the table sagging, low at the front and high at the back?

I got the same results (eventually) over a radius of seven inches. Side to side less than a thousandth. Front to back a little more than a thousandth.

Tolerances are a moving target no doubt about that. I am not sure what to expect from a 1943 machine. If I can work comfortably and consistently to about .002" then for 90% of the machining I do I will probably be OK. I agree the knee operation is a pleasant surprise, it is smooth and straight. I put a shop made cylinder square 6" tall on the table and it was less than .001" vertical on the knee. Just a hair over on the spindle.

FIVE HOURS: I doubt if I will ever use the tilting function of the head. And the shim, not only the size counted. At first it was a slice of a circle about an inch tall at the center. Wanted to spread the shim surface out a bit. But the position and the size affected the amount of tilt. And every time I loosen the bolts the side to side would change and it changed the front to back also. On and on, I tried doing in a smart incremental way but that just didn't work. So I set upon my desired setup, the size and thickness of the shim and the position and then had to tighten up the head/column bolts completely, then measure. Every cycle was about half an hour to forty-five minutes. I lost count but I probably did this about ten times, five solid hours. And I was tired, and really tired of this task.

I would not argue the point, I do wonder just what kind of accuracy a piece of plate glass would have. At half an inch thick it is pretty rigid. I put a dial indicator on one edge of a piece twenty-four inches across and with finger pressure wanted to see if I could induce flex. It took a lot of pressure to see the indicator move at all. So with my three point support I am comfortable that there was not much if any sag. If there is any flaws in a piece of glass it will show up dramatically in its reflection. I tried analyzing the reflection the best I could think of and could detect no aberrations, no warping, dips or flaws. So is this the same as a ground AA or even a B surface plate? Certainly not. But I would bet it is within a couple of thousandths. I did mic along the entire edge with a .0005" mic and could detect no variation. Maybe this is an unusually good piece of glass.

I do have a friend that works at an aerospace company with a CMM that is capable of amazing accuracy, it would be interesting to have him diagnose the glass and get a real answer as to how much error there is.

I did invest in a VFD to run the machine, and that is not working as well as I hope it would. But this is another thread, the motor gets hot, there is something not right. I never let it get over about 190 degrees and shut it down and let it cool off. I can run the machine with a light load for about fifteen to twenty minutes. Like I said, there is something not right. If I thought that this was going to be a problem I would have gone the re-power path. But I needed the three phase power anyway for two other machines. So this seemed the best bet.

YEP, has three. That would be 10, 2 & 6 o'clock?



I would be willing to bet that this surface probably hasn't changed that much since 1943. As we described how much of a nuisance getting the head straight again. Although I would guess the three bolt design is far superior. I got the hang of taking off enough pressure without really loosening the bolts and could get the worm to tilt the head, of course this was after opening the space wide enough to get a shim in half a dozen times, cleaning the edge and also getting some lubrication into the space. I had it only apart maybe an eighth of an inch, which is really not enough space to inspect. But even this with a light at one end and my eyeball at the other, what I saw was a finely machined polished surface, no bumps, grit, rust, paint or grime. It was nice to see. And no shims. I would imagine that tilting the head was probably something not done often.

Since Index was counting on this surface as much as anything else for the accuracy of the machine, I will stick to the "precision machined surfaces" comment. I stuck with using copper as a shim, wanted as soft a piece of material as possible, not wanting to muck up this surface. It is good to know that someone else's machine as a shim. But I have this feeling that any moment someone is going to come forward with "Oh my GOD not a shim"

I agree with Dutchman on this point; I guess I was so hung up on the shimming issue I just missed it. On general principle:
You put your glass on top of three gage blocks, which means you must be pretty confident that the tops of those gage blocks occupy the same plane. Why not just sweep the blocks? That would remove one element of error stackup, for free. Your setup is depending on four elements for accuracy:

1. Accuracy of the gage blocks
2. Flatness of the bottom of the glass.
3. Flatness of the top of the glass.
4. Parallelism of the top to the bottom.

You can cut those four down to one, just by taking the glass off and taking your readings off the gage blocks. Not that it matters a whole lot, sounds like you've got the head trammed to a degree you're happy with. How does it cut with that shim in there, by the way? Were you able to get a better flycut test out of it?

Really interesting that your head has a three bolt connection to the column. Mine only has two like Dutchmans, with a big rib in the casting at 6 o'clock where your third bolt is. I wonder when this design change got made?

Regards,

John