Experiences with a tabletop machine shop

In 2003, I purchased a small milling machine and lathe to set up a machine shop in my bedroom. I had no metalworking or shop experience so it was a learning experience.

It's cool to machine metal if you live in a post-industrial society that often does not make the machines it uses. It is like looking behind the curtain to see where technological implements come from. When you know how to make a thing, I believe that you naturally acquire some deeper understanding of the relationships between design elements, function and cost.

When what you need is not available commercially, then it must be custom made. If you know exactly what you want, then you can hire a machinist to do it. This is often impossible as requirements are unknown until well after any parts are required. Parts must be made with incomplete understanding of how they fit into a system. There is huge value in exploring the design space through all phases of a project, especially during fabrication itself. When parts are assembled, design modification is often a necessity. It just isn't possible to anticipate all contingencies and interactions. So having your own fabrication capability offers flexibility.

Buying off-the-shelf is typically preferable to making it yourself. Commercial products are manufactured at a cost and price for level of quality that is impossible to match by amateur fabrication due to large volumes, deep manufacturing experience and specialized tooling. Even discounting the cost of your labor, you won't save money, although it may be educational and rewarding. Learning how and why is a compelling reason to do it yourself. Just keep in mind the cost tradeoffs.

The mill is necessary, the lathe is optional

You'll read that the lathe is the more basic and fundamental machine tool. This is true in the sense of bootstrapping your tools. If you were building a milling machine or more generally a machine tool industry, the lathe comes before the mill.

This is analogous to an assembler underlying a C compiler. You can't write the C compiler without an assembler. However, most programming is done with the C compiler, not the assembler. The higher level language is more practical most of the time.

So you'll need a milling machine far more than a lathe in projects. Unless you need to make rotating assemblies with cylindrical geometries, the mill is the more useful tool.

New production tabletop machine tools available in the US

The machine tool market is dominated by industrial needs. There are many options for machine tools and tooling if you are setting up a commercial machine shop. These machines are too large for the home. They may be heavy enough to require a concrete foundation to support the weight.

In contrast, there is a tiny hobbyist market in the US. As a result, there are only a few practical competing options in tabletop milling machines and lathes.

The Japanese no longer make hobby machine tools that I am aware of. Anything German made is many thousands of dollars and out of almost everyone's price range. There may be some small ROC (Taiwan) tools. But I've not seen them for sale in the US.

So your choices are between the US made Sherline and Taig versus the Chinese made Sieg models.

The Sherline is too small. The spindle is MT1 (Morse taper 1). The machine doesn't have a lot of capacity or power, enough for miniatures but nothing much larger unless you are willing to go though a lot of inconvenience. By this I mean that if you have a lot of patience, even a small and underpowered machine can remove a lot of metal. It just might take weeks of work.

The Taig is somewhat larger. It is also sold in a CNC configuration. I'm not sure of the exact capacity (numbers are deceiving unless you work with a machine). But it looks to me like it's still somewhat limited in the size of workpiece it can mill. The CNC configuration is intriguing, though.

I opted for a Sieg mill and lathe. The higher power and larger capacity was more important to me than quality. In hindsight, this was the correct choice for the kinds of projects I do. It also fit with one of my goals which was to learn. Using the Sieg tools forced me to understand their limitations.

Your choice may be different. If I needed high precision work in small parts, I'd consider the Prazi, Sherline and then Taig in that order. I'd probably choose a Sherline or Taig due to the high cost of the Prazi.

Impressions of my milling machine

My mill is from Homier. They are a good company to deal with mail-order. I'd recommend somewhat greater caution with Harbor Freight. This isn't to say I had a bad experience with them. I've been to both a retail store and ordered through a call center and just had a bad feeling. It's a good place to look around (avoid the cheap tooling) if you have a retail store nearby.

I bought the vertical mill for $400 and a 7”x12” lathe for $300 with $150 in shipping via Yellow Freight. These prices are pretty good. Typically, the mill will price at $500 and the same lathe at $400. Keep in mind that while there may be a variety of models with different brands and options, they are all essentially the same make with the same performance. The one option I'd be concerned about is the spindle size. Some mills come with a Morse taper. Others are R8. If you can get R8, you have an easier time finding compatible tooling. R8 is the industrial standard spindle size.

Note that this size of mill is about the largest that one average sized male can handle in one piece. In the crate, it's around 150 pounds. That may not seem like much if you lift weights. I don't. It's massively heavy. Worse, the center of mass is high so the balance is poor. Still, you can deal with it alone if you have to. If you go any larger, you'll need friends to help you move and install the machine. In my case, the Yellow Freight truck pulled up and the driver and I carried the crates into my apartment. To protect your hands from splinters and cuts, wear heavy leather work gloves when handling the crates. You'll need a claw hammer or pry bar to open the crates.

The main downside to the Sieg machines is the low quality. Don't count on anything being properly aligned, lubricated, or even clear of casting sand before operation. Some of this makes sense. The red "cosmoline" grease is just to prevent rust. You'll have to apply your own grease and oil. That's normal. But I found the milling column on mine to be tipped slightly forward. After a few hours experimenting with aluminum foil shims, I got it pretty close to vertical, probably as close as can be within the limits of the machine. I also found the milling table sagged downwards as it moved towards the limits of travel. So it doesn't remain level. Well, this isn't too big of a deal given how the mill is used in practice. Also, the lead screws weren't quite aligned properly with the milling table. So they'd bind towards the limits of travel. I just greased everything and worked them back and forth until things smoothed out. But really, the proper way is to disassemble the table, clean, grease, and reassemble, making any adjustments (with a file?) to get everything lined up.

Installing the machines

A waist high horizontal surface that can support several hundred pounds is required. Machine tools are typically dominated by large masses of cast iron and are very heavy. The table should be very rigid and stable. In my case, I made a worktable from softwood 2x4s and lag screws for around $75 in materials. I looked at commercial workbenches and worktables but instead made my own. I don't have a truck or any easy way to lift or transport a workbench. But eight foot 2x4s will just fit inside most compact sedans. If you don't have a shop area: no workbench, vise, etc., then just purchase a small crosscut handsaw and at least two 5 inch C-clamps. This can get you started. I'm in an apartment so clamped the 2x4s to the kitchen counter to hold them while sawing. A hand drill for pilot holes, a drill bet set, a socket wrench to drive the lag screws and wood glue completes what you need.

The lathe has a wide enough base and low enough center of mass that you don't have to bolt it down. It weighs about 70 pounds. It won't tip if you are careful.

The mill must be bolted down. It is top heavy, roughly 70 pounds in the milling column and head bolted vertically to the base. Most of that weight is at the top of the column. The base is somewhat narrow and lighter. So the mill is always in danger of tipping.

One option is to bolt it directly to the table. The other is a wide platform the mill's base is bolted to. This is what I did. The mill's base has four holes at the corners and is bolted to a ¾ inch plywood board. The board is raised off the table surface with two more boards glued to it. This is stable. Be mindful of the handle in front that moves the milling table forward and back. As the handle turns, your fingers need room to clear the platform.

What kind of results can be expected from these machines?

You get what you pay for. Buy a Prazi and have a perfect machine out of the box - no backlash (my Sieg has truly massive backlash of perhaps .02 or more - but you can compensate for this), smooth operation, everything square and aligned. At the same time, I've learned probably more than I would have by having to deal with machine limitations. You start to see that a mill is not an infinitely rigid, perfectly aligned, ideal cutting point that you move the workpiece under. A milling machine is flexible. It vibrates and changes shape under gravity. To some extent, all milling machines will have to compensate for these dynamic factors (even if it is automatically handled by the CNC firmware or software).

The reality is that tolerances down to .001 are quite difficult to achieve. There are just too many variables. Errors often add up systematically. This is why even CNC manufactured components must still be hand matched and fitted for premium products. There is too much dimensional variance. If three parts must fit together to have a total tolerance down to .002, then each part must be fabricated to well under .001 tolerances. That is difficult. In my experience, I've had to mill to within a few thousandths and ground down the rest by hand to have two parts that my dial calipers measure as identical. If you are doing this, you have to question a design than needs this kind of precision. One aspect of good design is that it doesn't push the outer boundaries of what you can fabricate.

Any machine will have some limitations. Machinists in the 19th century lacked the ability to accurately measure down to their required tolerances. They couldn't measure with enough resolution to see that identical parts really weren't. That's why there was so much hand fitting. I believe that unless you have very expensive equipment, any small shop will have to use many 19th century methods that are labor intensive to have good results. There is a saying that the only tools required for metal are: a drill; a hacksaw; a hand file. There is some truth in this.

The Sieg mini-mill was the largest that I could still handle alone in my apartment. It has sufficient power to mill T6061-T6 aircraft aluminum safely. It can also handle soft steel such as the carbon steel tubes from Home Depot. Heat treated steel is much tougher and must be approached very conservatively. But even with aluminum, you have to be cautious not exceed the machine's capability. There's not enough torque to cut very deeply. Though the Sieg mini-mill is large compared to the Sherline and Taig models, the entire column will flex noticeably unless cuts are very shallow.

What speed do I mill at? How deep do I cut?

Note to avoid confusion: colloquial terminology may use the same word “mill” with different meanings depending on the context. A “mill” may be a milling machine. A “mill” may also be the cutting tool that is held with a collet in the spindle and rotates. And “mill” may also be a verb as in the act of milling.

I mill at speeds much lower than what you'd see in Machinery's Handbook. Those speeds are for production. With my mill, I spin pretty slowly. I'm guessing the speed is 300 to 450 rpm most of the time. The depth of cut varies from 1/16” to 1/8” when cutting with the tip of an endmill to a few thousandths to .01 or .02 when cutting with the side of a mill over about 1”. This is all in 6061-T6 aluminum. When milling steel, the speeds are about the same but the depth of cut is less.

Depth and speed really depend on the cut, stiffness of the workpiece, how it's held, and the desired finish. This is an art I don't fully understand. If you are plunging an end mill into aluminum, you are asking for it to seize up. You have to be very gentle, feed as slow as you can. Sometimes, I resort to turning the mill off and turning the spindle with a spanner wrench by hand. My muscles can put out more torque than the electric motor but mostly I can feel what's happening by literally turning the spindle with my hands and backing off if I feel the mill or drill begin to seize. Vibration and sound are important. Listen and feel as you are milling. How things are going is revealed in the sound made when metal is cut. Smaller mill sizes like 1/4” are more useful than larger ones for these tabletop machines. They simply lack the power and stiffness to make effective use of larger sizes.

What about tooling and accessories?

In general, you will spend at least as much on tooling and accessories as for the machine tools alone. That's a traditional wisdom that is about right in my experience. It's tempting to economize and go cheap on tooling and accessories. My personal belief is: better to have good tooling and accessories and a mediocre machine than an excellent machine and bad tooling and accessories. It's analogous to photography. Which is more important, the body or the lens? Ultimately, the lens is the limiting factor in quality of results.

Don't skimp on measuring tools, calipers, indicators, indicator holders - if you can't measure it, then you can't make it to that tolerance. One good dial indicator, one good dial test indicator, one good 6” dial caliper and an articulated indicator holder are necessary. A machinist's rule with a scale down to .01 is very useful too. The good brands will be from the US, UK, Germany, or Japan.

Don't skimp on a hacksaw (unless you have a bandsaw). You will likely spend many hours sawing though metal stock. Get the “contractor's grade” hacksaw with the good blades.

Avoid “Kurt” style vises. Most of them are much too large for most tabletop milling machines. Get a precision toolmaker's vise, preferably one with the long slots in the base along the sides instead of the holes (which make mounting to the milling table more difficult).

ENCO is an ok company to do business with, not the cheapest but reliable and with a good selection.

Try to get a mill with a R8 spindle as collets are easier to find.

Don't skimp on end mills and drills.

Get at least one carbide spotting drill.

Get a set of parallels. Cheap ones are o.k.

A clamp set is optional. You may use this if you have something large or irregularly shaped so as to clamp the work directly to the milling table. But honestly, the vise is so convenient that you will do everything in the vise if you can get away with it.

Get some hand files, at a minimum one 6” mill file. A set of precision needle files is very useful.

I have a big bottle of blue Dykem but have never used any of it for machining. An X-acto knife can leave a mark on aluminum that I can see under good light.

Some taps are also nice to have for screws. Get plug style through taps in 6-32 and 8-32 machine screw threads at least. Those are the screws you always run into.

An articulated dial indicator holder with a magnetic mount (mine is a Noga, Japanese brand but made in Israel) is worth it. It will cost around $50 but save you hours in frustration.

I've found that electronics tweezers (the point is very narrow, almost sharp) are very useful for removing metal chips that will inevitably find their way into hands. You can try to use an X-acto knife to tease them out. But the tweezers are the best.

In short – if you have to save money, better to have fewer tools that you can rely on rather than many of questionable quality. The price of the machine tools are largely fixed by the limited selection available to the tabletop hobby market. So there's no economizing unless you elect to purchase a used tool. Be cautious as well used machine tools may have significant wear.