Vertical Mill – Installation

Finally we can make flat things.


The lathe is called the King of Machine Tools for very good reason. There’s arguably nothing you can’t make with it. Other machine tools mainly serve as ways to make it quicker and easier to do things lathes can do. There are a few exceptions, but you have to work pretty hard to need an operation that can’t at least theoretically be done on a lathe with a sufficiently elaborate setup.

It turns out that one of those commonly-needed elaborate setups is flattening things. On a lathe, you typically do this with the somewhat-ironically-named “milling attachment”. This generally takes the form of a vise with a vertical slide attached to the cross slide on the lathe. That gives you the third dimension that the lathe is missing, thus giving you two dimensions of travel orthogonal to the spindle and one dimension of travel parallel to it. Milling attachment setup is time-consuming enough (and sufficiently lacking in rigidity) that it’s worth having a dedicated machine just for this. Enter the vertical mill.

A vertical mill is essentially a lathe stood on end with a large milling attachment permanently affixed. It’s arguably the current “standard” for multi-point machine tools (in contrast to the single-point cutting tools like lathes and shapers). The vertical mill is far from the first attempt to optimize the flattening operation on a lathe. Flat surfaces and features were previously made on things like shapers and metal planers. A shaper is particularly useful, because it’s extremely adaptable to different operations, can reach inside parts in ways other tools can’t, and can use the same tool bits as the lathe (perhaps with a slightly different grind). However, as the old machinist’s joke goes, “you can make anything with a shaper except money”. Shapers are versatile, but they are not fast. Vertical mills are fast, so as soon as metallurgy and manufacturing advanced to the point of making complex tooling like end-mills practical, milling machines came to dominate.

When most people think “mill”, they think “J-head Bridgeport”, unless they are not a machining nerd and thus think “making flour from grain”. Bridgeports are popular, well-respected, and all the cool kids seem to have one, but they are far from the only choice. The Wells Index machines, for example, are an underdog that real machinists know and love. For those of us without a lot of shop square footage, a bench-top mill is a great choice. The main difference with a typical bench top mill is that there is no “knee”. A knee-mill is the most well-regarded design, and dominates among large machines. The knee is what provides most of the vertical table movement (in addition to the quill), and it’s a very rigid way to do so. It also consumes a lot of space, so bench-top machines trade the knee for moving the head itself up and down on a column. So-called column mills are less rigid, but compact and versatile for their size.

Nowadays, for bench top mills, you’re going to be talking modern Asian-made clones like Grizzly and Harbor Freight. While there were bench-top mills back in the day, they were not common and are now hard to find. After a fair amount of research, I landed on the Precision Matthews PM-25MV. Not least because I’ve been quite impressed by the build quality on my Precision Matthews 1022V lathe. I’ll spoil it now and declare I’m extremely happy with the PM-25. The build quality, rigidity, motor, and bearings all seem excellent. Let’s dive into how mine is set up.

First things first, we need space! I had built an extension to my bench with this in mind, but there’s a catch- an overhead shelf. While the shelf will clear the mill, I need a lot more vertical space in order to get the crane in there to place the mill on the bench. Dilemma!


Sure, I have this perfect workbench area in the corner for the new mill, but there’s a shelf overhead. What to do?


After some pondering, it occurred to me that the shelf is attached to the wall on one end, which makes it a great candidate for hinging upwards. I had built it years ago as a standalone frame to fill that otherwise wasted corner, so hinging was an easy decision. By being hinged, I can move it out of the way for the rare times I need the headroom, and still take advantage of the storage 99.9% of the time.

After trimming a few areas of the shelving boards to add clearance for the soon-to-be-moving structure, it was time to take it down.


There were a lot of screws holding it up, but otherwise it was easily removed. This is a very sturdy shelf, easily holding the hundreds of pounds I had stored on it, with nary a creak.


With the shelf out, I need to modify it for more clearance. A moving thing requires a lot more personal space than a stationary thing. In particular, the end opposite the hinges will describe an arc as it moves, and any material within that arc needs to no longer exist. Otherwise, the shelf will not be able to swing up.


A circular saw was used to cut a rabbet in the end to clear a support column, and the lower trailing edge was chamfered with a jigsaw.


Okay we’re ready to attach hinges, right?


Well, not so fast. The studs are not in very helpful positions. The center one is sort of okay, but the only other choice is that rightmost one. However it’s too close to the adjacent shelf. Stringers between the joists might work, but the access to the sides of the studs is insufficient to mount them.


In my experience, virtually all problems in life can be solved with sufficient application of steel. This is no different.


The junk pile produced a chunk of 2” angle iron, which will be a nice support for my hinge. This plan was great except for the part where I drilled the mounting holes on the wrong side. The plan got a lot better after re-drilling them correctly.


See? Steel fixes everything.


For this hinging idea to be complete, the end opposite the hinges needs to be supported. Once again, steel to the rescue.


I welded up these three brackets that will support the far end. Each is one is tailored to the spot it needs to live, and the space that is available.


Here’s what the final support system looks like. It’s back to its formerly sturdy self (shelf?), but now hinges out of the way as needed.


Alright alright- enough futzing around with shelving. Get to the mill, already! Let’s do a bit of an “unboxing” blog.


All the best toys come on pallets.


I ordered my PM-25 with a high-end Taiwanese vise and the power feed option. We’ll be installing those later.


With the top of the crate lifted off, here’s what we have. Very exciting! I ordered it with the DRO, which Precision Matthews installs for you. It’s hiding in that ball of saran wrap around the column.


The next step is to get it rigged for lifting. While this is a small bench machine, it’s still about 300lbs. It needs to be lifted with proper tools and safety measures. The manual describes the proper lifting point.


Here we have a proper lifting sling placed under the head’s pivot point, as is specified in the manual.


The standard lifting safety rules apply here. Always use a proper lifting sling, not random ratchet straps or whatever ropes you have lying around. Lifting slings resist shearing, distribute load, and are reinforced in the hook areas. Make sure nothing is under the sling that can be crushed. In this case, we need to mind the electrical box and the DRO arm.

When the load is up, keep all meat-based things out from under it, control momentum, plan your moves, and don’t do anything suddenly. Never do anything in a hurry.


Ready to lift! I used a prybar to slide the mill to one edge of the pallet so the crane could reach it and ensure the initial lift is straight up. If not, the mill will want to swing and it’s easy to lose control of the situation.


Done properly, a lift like this is very low drama. Here’s our mill in its new home. You can see how the hinged shelf was critical to our success here.


You may notice that the chip tray is not shown in these photos. Truth be told, I thought it didn’t come with one. I certainly expected one, but when it wasn’t in the crate, I figured it was part of the optional stand, and so I didn’t get one. No big deal, I figured. Turns out, which I learned while dismantling the crate for disposal, that the chip tray was bolted to the underside of the lid of the crate. Whoops! Just as well, I actually prefer the mill without a chip tray. Mills throw chips like crazy (much more than lathes) so a chip tray is largely decorative in any case. A flat bench under the mill is easier to clean. I may install it later if I start running flood coolant, as the tray would be helpful for catching that.

At this point, I needed to get the mill positioned where I wanted on the bench. I wanted it as far back as possible while still having room for the front crank to be comfortable to use. Pushing it back reduces the chances of bumping into the crank handle and snagging it. The trick to maneuvering the mill into place is to leave the crane attached, with some tension on the sling. That takes most of the weight off it, so you can easily slide it around as needed.


Once positioned where I wanted, I transfer-punched the mounting holes. I’m using V-blocks to hold the transfer punches vertical, because I don’t have the perfect size for the mounting holes in the base.


After marking the holes, the mill then has to be lifted off the bench and put back on the floor again. This may seem silly, since you could just measure the mounting holes or make a template of them. However, the only way to be sure the crank handles are exactly in the right place is to put the real mill where you think you want it. It’s also important that the table have full travel in both directions, and that doing so is comfortable from the working position. I really wanted to get all of this right, so I did the extra work of physically prototyping the mill’s location.


With the holes marked and the mill moved, we can start drilling. For a job like this, I bust out the big corded drill. Cordless tools are amazing nowadays, but the big corded DeWalt can run at high torque and low speeds for long periods. It has a pretty terrifying amount of low speed grunt, which is what you need when drilling large holes in steel. I’m always mindful of body position using this drill, because I’m pretty sure it could break my wrist if it snagged and I wasn’t ready.


Now we just need to lift it back into place, and…


Shazam! With the mill back in place and bolted down, my Precision Matthews family grows ever larger. The mill is a perfect fit in its little corner.


Now it’s time for a little tune-up and customization. No machine tool is perfect, and anytime you ship something heavy across the country, damage is fairly inevitable. This machine is terrific, but it did arrive with a few issues to sort out.

The first issue was the DRO mounting.


The arm mount for the DRO is a little iffy, because it’s on an uneven surface. Precision Matthews handled this by supporting the far side with jacking screws to press against the uneven area. This works okay, but the stress of travel started to pull the cap screws out. This situation was salvageable, but I’m not crazy about this mount and will likely redo it in the future. It’ll be easy now that I have a mill!


The other problem area was the column ways cover. It’s a typical accordion-fold cover that slides on the ways.


Unfortunately the top mounting screws did not hold. On one side, the bolt had worked loose. On the other side, the bolt stripped out the threads.


One bolt just needed reseating. The other one was stripped, but the threads were deeper than the bolt was long, so a longer bolt was an easy fix.


With the shipping problems sorted, there was one customization I wanted to do right off the bat. The mill comes with a safety shield that opens and closes around the spindle. It’s interlocked, so the mill won’t run with it open. I’m not generally a fan of circumventing safety equipment. However, it was immediately obvious that this thing was going to be colossally in the way, and needed to go. Removing the shield is easy, but we need to defeat the interlock so the mill will run without the cover.


The interlock is a normally closed switch that is opened by a flat spot on the bar when the shield is opened. This is good news, because it means simply removing the bar is all that is needed. The “default” for the interlock is “unsafe”.


The next step is to clean things up. As is tradition with machine tools, everything is covered with shipping grease to protect the bare machined surfaces. A combination of WD-40 (for the easy stuff) and brake cleaner (for the hard stuff) removes it all. Parts that move against other parts get a coating of way oil, and other areas get a spray of Boeshield T-9 for rust prevention.


The machine looks ready to go now, but before we can make chips, we need to tram the head. It will be close from the factory, but “close” is only good enough for woodworkers. You can buy fancy tramming tools, or you can take ten minutes and make one from some scrap. Guess which option I took?


I grabbed some brass scrap off the junk pile to make my tramming tool. Why brass? I could make up a story about how it’s a good balance of light weight and rigidity for this application, but mostly it’s pretty and it was there.


All you need is some sort of arm that will hold a dial test indicator at some distance away from the spindle. The longer the arm, the more precise your tram will be, but the more time you’ll spend doing it. Precision always costs time (which is why it also costs money, by the transitive property), so it’s up to you to find that balance. My simple brass tramming arm is a piece of round stock turned to fit in the spindle on one end, and is press-fit into some bar stock on the other. The bar stock has a 250 thou reamed hole for a perfect snug fit on the dial-indicator’s post.

Tramming is an exercise in patience, but it isn’t difficult. You loosen the head-tilt bolts just enough that it can be moved by a tap from a soft mallet. Then you swing the arm back and forth and measure the difference in height at the ends of the table. Make small movements until you get them the same.


I’m using 1-2-3 blocks here to make it easier to find a surface to compare at each end, but make sure your blocks are really the same height. Otherwise you’re introducing error. Cheap internet 1-2-3 blocks can easily be out by a thousandth or two. If you’re a Real Machinist, you might prefer to use gage blocks for this. I only play a machinist on the internet, so 1-2-3 blocks are good enough.


With the tedium of tramming out of the way, we’re ready to make chips, right? Right? Not so fast. A vise is a supremely convenient way to hold work on the mill, and 90% of what we do will involve a vise. However, when you bolt a vise to the table, you’re introducing a whole new source of error. If it’s a pivoting or swiveling vise, you could be introducing multiple new sources of error. We minimize this by “indicating” the vise. This means adjusting it until it’s as square as we can get it, then bolting it down.

No head tram is perfect and no vise is perfectly indicated (or itself perfectly square). All these little sources of error add up and appear in your final part. The best we can do is minimize them all.


The vise is indicated by putting the dial indicator on the head and running it back and forth, making small adjustments to the position of the vise until it reads zero all the way across the fixed jaw. You should also do this for the top surface of the fixed jaw, using shims as needed to correct any error there. Cheap vises will need a lot more massaging here than expensive ones.


With our head trammed and the vise indicated, we’re finally ready to make some noise.


First chips! This is always an exciting moment for a machine tool. I took some scrap and spent time experimenting with feeds and speeds, and getting a feel for how to hit a dimension with this particular machine. Every machine tool has a personality, and will have little tricks to get the most precision out of it.


Spending some time on scrap with the machine is time well-spent. Learn the personality of the machine! For example, the manual says this mill’s worm-drive fine down-feed on the quill has zero backlash, but I found that to be incorrect. I found it has about four ten-thousandths of backlash in it, so I can hit dimensions much easier by feeding it down past where I want, then winding it back up to the right value on the quill DRO (thus removing backlash in the direction of tool pressure). Learn how to love your machine, and it will return the favor.

There you have it- a fancy new tool in the Blondihacks shop, and I really could not be happier about it. I hope you enjoyed reading about the basic process of installing and setting up a new mill. Stay tuned, because we’ll be making lots of stuff on this machine in the very near future.



13 thoughts on “Vertical Mill – Installation

  1. It’s pretty obvious that you are not going to be making any money with a mill either. You sure love to knock woodworking, but that’s because you can afford metal and and all the equipment to work it. I made every piece of wood furniture in the room where I’m typing this now, including the desk and the computer case, and yet all the wood to build it all, together with every woodworking tool I own, had cost me less than you paid just for this one mill. I don’t begrudge you your entertainment, of course; what good is having money if you can’t have fun with it? Who doesn’t like cool expensive toys? I would merely point out that woodworking is a considerably more useful skill for most people and you could reap greater benefits from doing more of it instead of complaining about it.

  2. > Pushing it back reduces the chances of bumping into the crank handle and snagging it.

    And for when you drop something, bend down to pick it up, and on standing up hit the handle with your head.

    Nice new addition to the shop!

  3. Quinn, nice job on the setup, your machine shop is growing by leaps and bounds!

    Did you get the option where you can drive the sucker with a CNC system as well? Or is the panel to the right (where you talked about the crappy arm mount I think) is just there for digital readouts?

    I’m curious what you’re criteria were for the unit and if CNC was anywhere in the list?

    1. The PM25 is a good candidate for CNC conversion, though there isn’t an official option for that from Precision Mathews. My understanding is that there are 3rd party kits for it, but I haven’t looked into their feasibility. I suspect they are like low-cost 3D printers: they work, but are a constant project to keep them that way. My machine tools are there to do work, not be projects unto themselves (mostly 🙂 )

      The panel you’re seeing is a DRO, which is exceptionally useful for mills. A DRO allows you to ignore backlash in the hand wheels, and also does math for you for laying out bolt circles, finding centers, things like that. They also have programs for milling arcs, tracking waypoints, marking dimensions, and so on. People use them on lathes as well, but for a mill, the ROI of a DRO is huge.

      I didn’t consider CNC, mainly due to cost. For an off-the-shelf robust CNC solution, you’re looking at 3x the cost just to get in the door. That gets you into the range of a low-end Tormach, which is built on steppers rather than proper servos (not that they aren’t good machines, but they are criticized widely for this). A Tormach is a huge investment in money and space, and the tooling for them is also very expensive. The other options are those CNC routers that people do some light milling with, but those are much too lightweight for much real machining work. Meanwhile, a stout robust manual mill that can do real work and uses inexpensive standard tooling is quite inexpensive now.

      In my opinion, the main advantage of CNC is for production. They are great if you need to make ten of something. If you’re making one of something, the advantage is much less compared to the cost, and you’re also taking a lot of the fun out of the process. Machining manually is fun and kinda the whole point! CNC turns everything into a software exercise, and I spend too much time on computers as it is. When I’m in the shop, I want to be making things with my hands!

      That doesn’t rule out CNC conversation later, but I’m very much enjoying manual machining right now.

      1. As for why I chose the machine I did, I basically got the biggest one that would fit in my space, and the most I could afford. 🙂 Like with lathes, bigger is generally better, and you generally get what you pay for.

      2. I run my mill manual about 2/3 of the time, but, wowie, automated drilling is sweet. Finally, my bolt-together parts are perfect. That’s especially nice for things like steam chests, where there’s a cylinder, the valve plate, the chest body, and the chest cover, all of which need four to twelve holes perfectly lined up.

      3. i totally get the manual aspect of using tools vs a cnc machine of sorts. of course in my lab the closest thing i got to a lathe is chucking up a peice of metal in a hand drill and shaping it with files and such and the closes thing i got to a cnc machine is a $400 3d printer with a tiny build area (and i think its robust enough to convert to a pcb mill). like i had to make a small pcb recently and while im pretty comfortable with doing toner transfer, i wasnt about to waste an entire $2 blue sheet for a one inch pcb. so i ended up cutting the traces with a box cutter and a dremel. i was actually rather pleased with the result.

        1. A trick I often use to avoid wasting blue sheet is to print on plain paper first, then cut out the the same size patch out of blue sheet, painter’s tape it to the plain paper and feed it back into the printer. Use your printer’s manual feed an mark a corner of the page so you know you’re feeding it the same orientation both times.

  4. A small quibble WRT your tramming arm:

    All cantilevered beams flex proportionally to the cube of the distance between the supported end and the load. Their resistance to bending is proportional to the cube of their vertical thickness.

    For a brass arm of the length and thickness shown above, surprisingly little end force will produce 0.001″ of deflection. You might want to check the repeatability of the measurement by putting a finger on the end of the beam three or four times to see how well it springs back to position.

    If there’s enough error to be a problem, turning the bar on its edge and brazing a rod along the hypotenuse of the triangle will make the arm much stiffer.

    1. Rigidity doesn’t matter a whole lot in this measurement, because you’re just comparing one side to the other. There may some lowered repeatability due to spring as you allude to, but when tramming you go back and forth about 50 times, so it averages out anyway.

Comments are closed.

Mini Cart 0

Your cart is empty.