Building a New Lathe

[vtsteam]

Today it seemed time to try graduating the cross slide collar. I could probably do the engraving with my mill-drill in the other shed, but I decided to proceed to try it as if I didn't own one, and also don't own a working dividing head, rotary table, or even a 100 tooth gear for indexing......uh.....all of which I actually don't own.  In other words, use the lathe to build itself, and a little careful thought about what is available.

So, to get a 100 division reference, I just drew up a circle in the good old free Google SketchUp 7, and made it out of 100 line segments. I connected the center and the circumference endpoints for one quadrant, then duplicated that 3 more times and rearranged them to make a fully divided circle.

Then I printed the circle out to just fit on a standard letter size sheet of paper (minus borders), which made it 7-1/2" diameter, and therefore also fitting a half inch inside my lathe's faceplate diameter of 8-1/2".

I also included a circle a little larger than my chuck diameter to cut out, so it would fit over the chuck while bolted to the faceplate.

[vtsteam]

Then I cut the circles out with an X-acto knife

[vtsteam]

And taped the ring to my faceplate.  I found some hard piano wire, bent a loop in it, trapped that under one of the headstock bolts, and then bent it up to form a pointer over the paper divider ring.

I thought long and hard about how to fabricate a spindle lock. That seemed like another project in itself. But about that time I spied a scrap of softwood, and thought, hmmmmm, the lowly wedge...... probably mankind's first brake..... so why not try one?

A quick pass with a handsaw to shape it, inserting it between the faceplate and the front of the ways, a few light taps with a soft hammer and yes the spindle was immobilized. At least for the purposes of running a graver tool in the tool rest along a small brass collar, shaper style. There's little if any torque on the spindle for this operation so the pine brake should handle it.

I then turned an arbor to fit the brass collar in the 3-jaw. and left it in place to keep it perfectly on center. Here's the final set-up, for caveman style division engraving. Well not quite yet, I haven't ground a graving tool, or drilled and tapped the collar for a set screw, and it's now time for dinner. But tomorrow, we'll attack......

[kayzed1]

I like it a lot.

[vtsteam]

Thanks Kayzed 

After trying it out today on the new lathe, I like it a lot, too:

[tom osselton]

Nice

[Neubert1975]

nice work 

[tom osselton]

Looks good.

[vtsteam]

Thanks Tom, and Neubert.  No photos today because I spent the day removing all of the components on the headstock end of the lathe. I had noticed that the wooden wedge was able to move the faceplate a tiny bit, and I wanted to find out why. The problem turned out to be the method of headstock fixing I had used. I had made an internal clamp bolted to a plate below the ways.

If you recall, the headstock had been made to slide along the ways while boring. This required a slight amount of clearance between the bottom face of the headstock, and the retaining plate screwed to the bottom. Placing a clamp on the inside of the headstock did not eliminate this clearance, because it pinched the way strips, but had no physical connection to the headstock other than the bottom plate. It prevented fore and aft movement, but, didn't fully eliminate the clearance. It was also originally just a temporary fixing, but I had in the years since forgotten about it. 

So now I'm going to file down the lands for the bottom plate, which will remove the clearance and pinch the headstock directly to the ways. And also I'm thinking about adding redundant bolts through the ways themselves and into the headstock. We'll see what kind of wrench access I have for doing this, because there's a horizontal shears support in the area.

Anyway it's probably about time to check the lathe throughout for tolerances, and make adjustments if needed. Also the carriage halfnut has occasionally popped out of engagement, and I want to adjust the detent and spring, as well as alignment with the leadscrew. Nothing fun made today, but I'll feel good after making sure the headstock mounting is solid and the lathe is fully operational, checked and aligned.

[vtsteam]

Working on the headstock attachment, I noticed that the black paint was chipping off the corners on the carriage. I'd used sprayed Krylon, a thin lacquer type, and probably should have brushed on a true oil base enamel, like Rust-oleum. I gave a little scratch with a fingernail, and the thin paint flaked off, as if I'd used a scraper. Zinc alloys are also known to be somewhat resistant to conventional coatings. Oh, well might as well take it all off before re-painting. 

A few minutes work with a hand wire brush, and a blunt screwdriver completely cleaned the paint off the carriage. Underneath was the slightly dulled finish of the zinc as milled -- not unlike hard anodized aluminum, just a little bluer in color. I actually thought it looked better than it did painted, so I've decided to just leave it that way.

Interestingly, all of the unprotected zinc castings I've done for the lathe, all 2 to 5 years old, look completely clean and corrosion free, and look the same as the carriage does now. In the same time period I've had most ferrous lathe parts (and those on other machinery and tools) rust during off usage. There is a lot of damp in this concrete tiny shop and big swings in temperature.

So despite a few early and admittedly well-meant warnings that zinc castings in a new lathe might corrode or suffer serious electrolysis issues, experience shows that the opposite is true: they are unaffected. On the other hand, the traditional lathe building metals used have suffered the usual condensation and damp problems. Since zinc alloy has stronger physical properties than cast iron (or mild steel) it was actually advantageous in this application -- at least from my point of view.

Here is the carriage with the paint scratched off. The shiny upper T bars never had paint on them, and have been in the same shop, on the lathe for 2 years. The lower part of the carriage was painted. The apron is aluminum, and also wasn't painted. I think it all looks better without paint now than it did new. (Oh, also, the newly calibrated dial is in place on the carriage screw).

[vtsteam]

The headstock was replaced by the end of the day. I had to make a "special" tool to do it -- basically just an Allen wrench with the shorter leg ground back to only about a quarter inch long. This allowed it to fit up above one of the cross members of the lathe bed, and tighten the 4 bolts now clamping the headstock onto the ways.

I reassembled all the other parts -- spindle and roller bearings, shims, cap, dust rings, tooothed belt pulley, tightened the spindle nuts to put just a slight preload on the bearings and re-attached the leadscrew bearing. The spindle looked great, clean and shiny, and the bearings still well greased. Everything turned smoothly. I replaced the motor and belt, and ran it for a couple minutes at different speeds to check for vibration. Looks good.

I put Morse taper dead centers into the headstock and tailstock and brought them together. They matched point to point, both with the tailstock ram retracted, and extended. This means the ram bore is parallel with the ways. I'll turn a test bar at some point in the future to see how close we are overall.

Next on the to-do list though is eliminating the problem of the carriage half nut jumping out of mesh occasionally. I'll check and adjust the leadscrew bearing positions, and checki the hand-filed halfnut detent shape and the spring tension.

Even though I want to move on and make new things, this tune-up process feels very satisfying.

[vtsteam]

Today I made two way-wipers for the front of the carriage. They were bent from sheet brass. I first made paper patterns using the cut and try method. Then estimated a bend allowance and cut them out with tin snips and Dremel tool. They were bent simply in the bench vise.

I have whole sheet of felt leftover from making the headstock spindle wipers, and it was easily cut out with scissors. Then I punched the felts with a bargain bin multi-puncher -- which has turned out to be an excellent investment. I drilled and tapped the carriage 8-32, and screwed the wipers on. Seems like a fairly simple job overall, but it took a surprising amount of time -- particularly setting up the carriage to be drilled, since that had to be done off of the lathe. It's an odd shape to support.

I plan to do the rear carriage as well, but this was all I accomplished today.

Wiper pieces:



And mounted on the lathe:

[vtsteam]

Oh yes, forgot to mention, I solved the half-nut issue by relocating the halfnut dog. As it was, the dog's detent was acting a little early, before full halfnut engagement. Short of remaking a new dog, or a new halfnut,  I eventually figured out that it would be a simpler matter to just relocate the lever's  pivot so that the dog was in the detent with the nut in full engagement. So I drilled a new pivot hole in the dog, and then set everything up engaged, and transfer punched through a new pivot hole location and drilled and tapped for a new pivot.

It strikes me this is really the way I should have set up the halfnut in the first place. When building the lathe, I had logically (it seemed at the time) cast the halfnut, then bored a hole in the apron for its shaft. Then I had made the dog -- a lever with a spring at one end, drilled the lever, and apron and mounted a pivot screw. The business  end of the lever fits in a notch which, as the last operation is filed into the half nut. This filing is very tricky to get right; if you file the notch slightly wrong, the halfnut won't engage properly and you're sunk. You can't re-file the notch or move it.

If it were to do it again, I would NOT make the pivot first, but wait until after I'd filed the notch. I'd drill the pivot hole in the lever, but not the apron. Then line everything up with the dog in the notch. Then simply transfer and drill the pivot location through the hole in the lever. Easy perfect alignment! 

[awemawson]

Well Steve, when you start your production run of small lathes you can write it into the operations procedure 

[vtsteam]

Thanks, Andrew  . Maybe someone reading this thread and intending to build  a lathe will benefit from the mistakes I made.

Actually, if you're building a Gingery lathe the same thing applies, as my new lathe uses the same style halfnut and dog. And actually even more generally, if you are building anything where you need to engage a dog in a filed notch, that is in a blind location, this might be helpful as a setup procedure.

I should also add that because on my lathe (and the Gingery lathe) everything IS mounted blind behind the mounted apron, but must exactly line up with the leadscrew, and yield a detent in the exact right position, there is one additional way to ease the task of parts placement. Also it's something I learned only in correcting the carriage yesterday after two lathe builds where I had struggled with locations.

With the apron off the lathe, just engage a short (say, foot long) length of leadscrew stock in the halfnut. This is easy to set straight and parallel and to check for alignment, and also sets the proper position for the detent. Sure beats trying to do that with hidden face measurements, (and guesswork) while the apron is mounted on the lathe, and filing and testing, repeated removal and remounting, etc. Another one of those, "duh, why didn't I think of that before," moments. 

[awemawson]

It's a well known fact that in a home workshop, where virtually all parts are prototypes, the fact that the first (..second  ...maybe third !) isn't quite right is a HUGE benefit as otherwise the scrap bin (parts bin) would be empty. A bin of odds and sods from previous project is (I find) a big prompt to the imagination when working out how to make things.

[vtsteam]

Why whatever do you mean, Andrew?

[vtsteam]

Waiting impatiently for a stepper motor to arrive from a slow shipper. Meanwhile the toothed pulleys and a couple guestimated length toothed belts ordered at the same time arrived from other sources a week ago.

This stuff is for adding the electronic leadscrew drive. I guess it will be at least another couple daysfor the stepper. They just shipped yesterday. 

[vtsteam]

Well surprise, the stepper motor arrived this afternoon....that was fast! Plus I received  a hall effect sensor (I already had an uninstalled  tach readout w/out a sensor from a few years ago), so, less manual feel and guesswork, moving slightly toward the last century. 

[vtsteam]

No photos today, but I've disassembled the headstock again to add the Hall sensor. That required drilling a half inch hole through 3/4" of Zamak with a hand drill on the floor, since my drill press in the tiny shop couldn't fit the headstock in place, end-on, even without the drill table. It was slow going, but eventually accomplished. I'm glad the Hall sensor has a very long stem!

And I have removed the headstock entirely to locate the new stepper motor beneath, between the shears. This should be a very neat installation, with everything hidden except for the two small motor and lead screw pulleys at the back of the headstock. These will replace change gears. The pulleys are 16 and 24 tooth for the stepper and leadscrew respecively.

[vtsteam]

I'd installed the Hall speed sensor in the headstock and replaced it on the ways, but then I spent way too long looking for a small box of disk magnets I knew I had somewhere. Finally my daughter took pity me after I'd searched every storage location in the house at least five times repeating "I know they're here somewhere.....," and plopped a BB sized round magnet into my hand. It was from a clump of them that someone gave her as a gift a few years ago.

No complaints, but that little BB wasn't the most convenient shape, especially because a Hall effect sensor responds to only one magnetic pole, and how do you distinguish poles on a tiny sphere?

But after a little thought, I stuck it to a metal ruler, which caused it to roll with one of the poles directed toward that face. Then I waved the other side in front of a spare Hall sensor I set up with a battery. Nothing. So I rolled the magnet 180 degrees around on the ruler until it stuck fast again, and repeated my waving in front of the sensor. This time the sensor LED flickered. I put a dot of Sharpie marker on the active face of the magnet, and took it out to my tiny shed shop.

The main aluminum drive pulley was drilled to fit the sphere closely at a location that would bring it in front of the sensor in the headstock. But when I pushed the magnet into the pulley, it wanted to come back out a small way, as if it had a tiny compression spring under. I checked under it -- nothing there! Then I remembered there's some kind of induced current equilibrium effect that does this -- I kind of remember coming across something on the web to that effect. Interesting. Anyway, it didn't suit my purposes, and I squashed it down with a dot SuperGlue and a thumb! 

The magnet in the pulley:

[vtsteam]

And as a check on positioning and clearance, I mounted the pulley onto the lathe spindle and rotated it until the magnet was opposite the Hall sensor in the headstock. Looked good.

[vtsteam]

Next on the to-do list was mounting the stepper motor. I hadn't permanently affixed the headstock yet -- the stepper goes under it, and in back of an aluminum end bracket. The stepper is fairly large -- NEMA 34 case, and barely fits between the lathe shears. I needed to mark the precise location where it would need to mount, with the two pulleys in its train and the only commercial toothed drive belt size available for the small distance I had to span between the stepper and leadscrew.

To find this precise position, I bored out the leadscrew pulley to fit its shaft, then mounted it loose. I grabbed a transfer punch of a size to just fit the stepper's pulley. I stretched the belt between the two, and while holding the punch against the mounting bracket, used it like a compass point to scratch an arc through the paint.

To suit the pulleys and belt, anywhere along the arc would work. Then I scratched a vertical centerline on the bracket, and where the two lines intersected was where the both the stepper could mount between the shears and the belt drive would fit.

Scratching an arc:

[RotarySMP]

Making good progress again. I am glad you got back to this build. It is a hell of a lot of work. Respect.
Mark

[vtsteam]

Thanks Mark!  Feels good to be moving forward again. 

Where I am: I got the stepper mounted last night, and headstock back together and mounted. There's still a lot to do to get the intended electronic lead screw functional. A power supply for it arrived yesterday, I already have the stepper driver and the encoder that I bought for it awhile ago. I just also located the dusty half wired up breadboard with Arduino that I re-programmed in FORTH with a leadscrew demo application that I wrote way back at the start of this project. That program does the division and stepping. Now I have to go back and figure out everything I'd already come up with. I've found the Forth program listing, but without remarks, and no wiring diagram. It's going to be interesting reverse engineering myself. People, don't follow my example!

But once the carriage drive is operational there's only one more component needed before I can truly say I "completed" this lathe. And that's making a topslide for it.

Even now, I can use my Gingery's topslide on this lathe, since I cast an adapter plate for it earlier in this thread. And frankly, I haven't needed a topslide at all for anything I've yet turned on the lathe so far in a couple years, so the impetus isn't strong to complete a new one immediately. But maybe, just to say "We're finished!" ......we'll see.

My real drive is to start work on an engine I have in my head that I want to build. And for that I really want to concentrate on creating extended boring and milling capabilities on this lathe. But that will be a separate project in a different thread.

Oh yeah, one more thing. After I get the tach working (I hope today) I have a sneaking suspicion, I'm going to want to build a countershaft for the motor to drop it down 2 to 1. I think I have excess speed at the top end and not enough torque at the low end for bigger boring tasks. But I don't want to think about that yet. If I do anything on that score we'll make that part of the boring/milling project......