What's new
What's new

OT- Crazy talk Making a lathe


Hot Rolled
Dec 6, 2010
Idaho, USA
<Yoda>Electronic lead screw it is you want, hmmm?</Yoda>

Open source hardware E-LeadScrew : Lathe Electronic Lead Screw R&D

An idea I've had kicking around for a while is building a lathe from plasma or water jet cut plate steel, doing the jigging and welding robotically along an automated assembly line. That'd include annealing for stress relieving plus induction hardening and grinding of the ways, boring the headstock for bearings and boring the tailstock for its ram. Since I've been working on a Wade 8A refurb, I've come to admire how its tailstock works and wonder WTH all tailstock rams aren't copies of it. (Leblond copied Wade's asymmetric front way slope, why not nick the other good bits?)

For the ways I'd use square bar stock at a 45 degree angle to minimize the amount to cut away. The top edges of the side plates would be V grooved to hold the way bars prior to welding.

All the technology exists (and has for some years) to build a lathe production line that takes in steel plate at one end and produces ready to assemble parts at the other. Harley Davidson's production line for frames for the V-Rod motorcycle is pretty much that way.

Tube and plate in, finished frames ready for paint prep out. All the pieces are bent, cut, jigged and welded by robots.

For CNC, using a precision rack and pinion for the carriage should cost much less than a ballscrew without giving up any accuracy. Even better it would be a simple option to put a rack on the back of the bed for double sided drive like at least one early 20th or late 19th century manufacturer did with two opposite threaded screws geared together.

A screw down the center pretty much needs a closed bed top and an underslung carriage if you want to keep chips off the screw. Such designs preclude center pedestals to support long beds.

Designing with Finite Element Analysis it should be possible to make a welded steel lathe (or mill or any other machine tool) with as good or better rigidity and vibration damping as the best in cast iron. Even with cast iron if techniques from as far back as 1960's engine blocks were applied to machine tools, they could be lighter yet better instead of how most of them appear to be designed, throw more weight of iron at it, as much as possible.

Sometimes making a part lighter will fix a problem that making it heavier has failed to fix or even made worse. The US Navy found that out with torpedo firing pins in WW2. Benelli had a problem with a part in a new auto loading shotgun that kept breaking. The beefier the part was made, the faster it broke. Then some bright person thought "What if we made the part thinner so it could flex a very small amount instead of trying to completely resist all bending?". That out of box thinking made the shotgun bolt assembly stronger and the action the fastest of any auto loader.

Of course you don't want machine tool parts flexing, but applying smarter design methods instead of just "More iron! Make it heavier!" has the potential to make them better than any which have come before.

The "tops" in doing things as they've always been done, apparently just because "We've always done it this way." can be seen in the huge numbers of clones of the J head Bridgeport mill. Why so many companies have copied that thing filled with a huge amount of fiddly little parts is mind boggling. I've been doing some repairs to a real Bridgeport and a slightly larger Lagun, which is a typical BP clone. Most things I see on those give me a "WTH moment". It's like someone set out to design a milling machine head with no real clear idea of where they wanted to end up. The design of the quill power feed and auto up/down stop would have both Rube Goldberg and Heath Robinson completely giddy.

Turn a team of engine engineers loose with an old Bridgeport as an example of over-complicated, antiquated design, and a mission to design a less complex, lighter weight machine *at least* as capable and I'd bet they could do it, especially if none of them had any preconceived notions about how a mill "ought to be" built. 'Course it could backfire and they'd come up with something even more complex and fiddly, given the way car engines have gone the past 20 years.

Such a process did work with the M16 rifle, those guys weren't gunsmiths yet they set out to design a rifle that didn't repeat traditional problems. The initial problems with the M16 were caused by people who did know firearms insisting that some of Armalite's innovations were completely un-necessary. IBM shuffled off the design team for the original PC to Boca Raton, Florida then did something unheard of in the industry - left them alone to design it - and created a revolution in small computers.

But the chances of anyone bucking the heavy iron traditions in mass produced machine tools seem to be slim to none, because "That's how things are done, period. Ain't NO other way can be as good, ever.".


Nov 1, 2004
Webster Groves, MO
If there is one statement about development of a new device that can be taken as gospel, it is that it will be far harder and take much longer than the people supplying the money can believe. Management usually thinks that engineering can carefully do a design, build a prototype, do a little debugging, and be in production in a few months. Sometimes that can be achieved when doing a new version of a device that engineering already has a lot of experience with, but virtually never on a completely new design. The only worse thing than doing a smooth "dog and pony show" for the top brass is a total failure. On occasion, I have thought that failure would be better. You put together a prototype of available materials, hand fit everything, and are up till 3 AM the night before getting it to work reasonably well, and then you have to try to make them realize that the prototype is perhaps only 5% of the effort required to make it a producible and marketable item. Usually some idiot who prides himself on being a dynamic leader demands that it be on the market in weeks. If you try, it almost always means a buggy, poorly functioning device and you will be called in the office to be told that you are a bad designer.

Much of the pressure to stay with known successful designs is simple survival. Engineers stay with the incremental improvement approach because they are aware of the cost of developing a completely new design. I don't know much about the M16 development, but after Kalashnikov had a good working design, they broke it down into tiny segments and teams of engineers spent the next year researching each part. Over a hundred years later, people are still improving the 1911 .45, which, after all, is a very simple mechanism.

Re rejection of Armalite's features, one of the biggest problems in R&D is sending information on down the line, especially when there is someone who doesn't want to take the trouble. It reminds me of the game Jack Parr used to do on his show, having a series of people who spoke different languages repeat a joke, translating from one to another, then have the last person tell it again in English. Nuances in the design are lost and the cumulative result is something that often only roughly resembles the original.

I think the first place to look in designing a new lathe would be a thorough study of modern CNC machines.


Long Tom

Aug 21, 2011
Fiddlefart, Oregon
Neat thread.

I like cast iron. Why the hatin' on cast iron? A nearly perfect material for the task at hand.

Start welding stuff together..... all heck breaks loose. Steel rings like a bell, and fancy cross braces and de-resonators are all well and good, and look great on the AutoCad screen... until you get a temperature differential across the machine. Or a weld fatigues. Or........

Man, the talk above about putting prototypes in front of suits sure gave me flashbacks! Especially the part where some type-A hard charger who doesn't know a hand-wired circuit board from a hand-bent sheetmetal skin starts in on how easy it'll be to make now the it's been prototyped. They beat you into submission on accepting a deadline, then, halfway there they throw a bunch of OTHER work at the group, but expect you to somehow still make the original crazy deadline, too. I have BTDT. Good times.


Jan 27, 2016
Concrete lathe

Yeah, in crazy enough! More driven by neccesity though. Where I live in South Africa, buying an old junk lathe is incredibly expensive. People have inflated ideas of their worth. Mind you, much of South African machine shops still use 50 or 60 year old machine. The machine shop I work in has 40 year old crank shaft grinders and reborers. I lost my machine shop due to my stupidity and a poor business partnership and intend to build up again slowly. I'm bankrupt anyway! Pat Delaneys take on a concrete lathe gave me some inspiration. I'm starting off building a large swing over lathe, initially to machine brake drums, disks and clutch flywheels. I'm lucky in having access to a full machine shop and a brilliant man for a boss. Another motivation is I have designed a wood briquetting machine, but need my own lathe to fill orders for manufacturing the machine. If anyone is interested in following what will probably be a one year build, then check out my blog here. concretelathe.blogspot.co.za


May 4, 2010
Cambridge, MA
I guess that depends what kind of lathe you are talking it. If you mean a engine lathe or a screw machine, that would be difficult to do. A watchmaker's lathe would be much easier. There is a design for a homemade lathe by David Gingery available and a lot of people have made Gingery lathes. One guy said it took him 9 months of evenings and weekends. He made about 20 different wood patterns for castings. This is what it looks like when complete (notice there are no change gears):


Personally, I think a better project would be to make a small horizontal mill. It provides a better time-reward ratio. With a horizontal mill you do not need to make a bed as you do with a lathe, you can simple buy an x-y table; this greatly simplifies everything. Also, it is a lot easier to make a really powerful horizontal mill that can do big work than a lathe. Making a powerful lathe would be very difficult.

Oops, thread from the dead, sorry about that. Artibrown did it, blame him! :ack2:


May 1, 2007
Boston Hyde park Ma.
I would love to join you. I have this bug to build a vertical CNC lathe/mill and will do it some day. Right now I have trouble finding time to wipe my ass. I have had a guitar on the drawing board for about ten years now, no kidding the wood was purchased....TEN years ago and sits on a shelf in my office.

Go for it and have a good time doing it, why not......you have earned the right to knock yourself out on this after spending years learning how to machine things for others, do it for you.

Make Chips Boys !



Oct 18, 2005
Neat thread.

I like cast iron. Why the hatin' on cast iron? A nearly perfect material for the task at hand.

Start welding stuff together..... all heck breaks loose. Steel rings like a bell, and fancy cross braces and de-resonators are all well and good, and look great on the AutoCad screen... until you get a temperature differential across the machine. Or a weld fatigues. Or........
Because I can't source a 55 gallon drum of molten cast iron ?

But I can e-mail a DXF file to a local steel supplier shop, and the next day
a truck delivers my plasma cut plates ready to go (and bent in a press brake if needed).