Read on, and you could be "making chips" in no time...
Step 1: Other Options
Why not a 3D printer?
I think a lot more people have 3D printers than CNC mills and they're great tools. There's a bit of an overlap with a mill, but they both have their strengths and weaknesses. Comparing a mill to a printer we broadly have:
- You can mill PCBs. This is one of the main reasons for my choice. I was about to go down the laminator / toner transfer method of PCB manufacturing when I discover the milling option.
- You can mill different materials. Most 3D printers are limited to one or two plastic materials. I can use anything softer than steel - usually wood, acrylic and aluminium in my case.
There's also differences that can't really be described as a pro or con - just whether they suit what you're doing. With a mill you're subtracting material; with a printer you're adding material. If you want a large block with small cutaways then a mill is the best tool for the job. If you're making a single-piece hollow shape then a printer would be better.
- It can be messy. I spend a lot of time vacuuming up sawdust or ground up bits of plastic or metal.
- Rounded internal corners As you're cutting away material with a milling bit you're limited by the radius of your milling bit when doing internal corners. External corners can be perfectly square and sharp, but not inside.
Why not a laser cutter?
Laser cutters are expensive and I haven't seen any simple home builds. They're great for accurately cutting through soft sheet material like wood and plastics but can't do PCBs for instance. Basically they're a also a good tool, but not what I wanted.
Step 2: Why Not a Custom Build or an Off the Shelf Purchase?
Why not do a comlete a complete home build?
Plenty of people have done a complete home build CNC machine. In fact, you'll find some excellent ones right here on Instructables. The reasons I didn't go for one of these are:
- Rigidity and accuracy. A lot of custom build involve clamping a Dremel to some wood and making the frame yourself. They may well be good ones out there, but I didn't find anything I felt would have the accuracy to mill PCB to take small surface mount components. More to the point I didn't think I'd make a nice accurate mill!
- Time. I wanted to get making stuff rather than making the mill.
Why not buy a ready-made CNC mill?Have you seen how much these cost? They're definitely priced for company purcahse rather than the home user.
As an example, I converted a Proxxon MF70 mill for about Â£525 (â¬650 or $825). A slightly larger CNC mill from the same company (Proxxon FF500 CNC) costs about Â£4500 (â¬5600 or $7000)! There are companies that will sell a converted mill at â¬869 (e.g. Usovo in Germany) so only about 30% more than my build but where's the fun in that?
We all have different priorities. You may well decide that one of these options better suits you. If so, go ahead.
Step 3: The Manual Milling Machine Base
My mill - the Proxxon MF70I decided to base my build on the Proxxon MF70 mill. It's a fairly small mill designed for modelling work, but seemed excellent value. They seem to hold their value really well and second hand ones don't go for much less than the new price on eBay - always a good sign. I'm UK based and got mine for Â£270. I've seen one in the US for $430.
The build quality, rigidity and accuracy of this mill is incredible for the price. There are better mills and probably some cheaper mills, but this seemed to strike an excellent balance - especially for starting out.
What's not so good
The only complaint I've had so far is the small Y-axis. As standard you're limited to 46mm. The X axis is OK at 134mm and the Z axis should be more than you need. It's possible to modify the mill to increase the Y axis (see my blog but that only increases it to 82mm.
Links to builds that inspired me
I found a number of well documented conversions of this mill which inspired me to do mine. Here are some links.http://www.jarkman.co.uk/catalog/cnc/mf70.htm
Other potential machinesI've also seen coversion of some slightly larger mills such as the Seig X3 or Taig mills, but the step up in size seemed to involve a larger step up in price.
Step 4: Fixing the Stepper Motors
I've seen a lot of builds where stepper motor mounts were home-milled. Whilst these are all good, I decided against this for a couple of reasons. Firstly, they often required the use of an existing CNC mill. Secondly, I wanted to be up and running quickly and without being hampered too much by my own inexperience. Thirdly, I'd rather get a working mill and start on simple bits of wood before tackling something tricky like a stepper motor mount (manually).
I came across a mounting kit on eBay from a seller in Turkey called *mbbilici*. It looks fairly basic and I intended it as a starting point before making my own mountings and using "proper" couplers rather than something that bolts onto the manual wheel. However, I'm still using it a few months later and have no complaints. I works really well and was zero effort on my part. Thoroughly recommended.
The cost of this was $99 (= Â£63 or â¬78)
Step 5: Stepper Motors and Drivers
What I needed
There are plenty of options for stepper motors and drivers, but once again I wanted a low hassle and quick solution for my conversion. The mountings I went for required a NEMA23 sized stepper and that seemed to be about right power wise. Most people use the smaller NEMA17 motors for 3D printer builds and they might be up to the job for a mill, but I wanted a bit of oomph for cutting metal.
What I went for
Once again eBay to the rescue. I bought a kit of 3 motors, a 3-axis driver based on the Toshiba TB6560 stepper motor driver IC and a power supply. A search for "3 axis NEMA23" on eBay should bring up the same kit from a number of seller. I went with uni.supply. The cost for the PSU, motors and driver was Â£142 ($222 or â¬176).
I since found out that a lot of people HATE this driver board and have had all sorts of problems. This is what I mean!
I contacted the seller who agrees to let me return it, however the kit wasn't that much more than the motors on their own, so I decided to stick with it. So far it's behaved flawlessly for me. A real recommendation for uni.supply that they would take it back though. However, considering other people troubles I'm not sure I can actually recommend the board itself.
Step 6: Controlling the Mill
Back to the 90s
The industry standard way of controlling a CNC setup seems to be to use a PC with a parallel port and run software from ArcSoft called Mach3. There are more modern and smaller setups that are often used for 3D printers but once again I wanted to get up and running simply and quickly.
Whilst a parallel port feels like a throwback to the 90s it also means that an old PC with one should be fairly cheap - or like me you may even have one hanging around. Things I discovered whilst sorting one out were:
- USB or ExpressCard parallel port adapters for your laptop will not work.
- Dell PCs (even recent ones) have parallel ports.
Mach3 is a horrible looking piece of software. There no two ways about it. If you even used coloured button back when Visual Basic 3 was cutting edge then it will be like a blast from the past. I used it reluctantly and intended to replace it as soon as I could.
On the plus side... it works. And once I got over the style of it I found some real benefits.
- It has a useful display of the predicted toolpath so you can see what you're going to get.
- You can manually enter g code to jog the cutter into position, or even do some basic "manual" CNC milling.
- Having a PC next to the mill means it's really quick to fire up my editing software and make any quick changes if I find a problem.
So far I'm still using Mach3. The free version does restrict you to 500 lines of g-code but that's enough for you to get your feet wet. I will replace it, but I'm not in as much of a hurry as I thought I'd be.
So far the cost of my mill has been Â£270 (mill) + Â£63 (mounts) + Â£142 (motors and driver) coming to Â£475. The PC didn't cost me anything but I'm going to allow about Â£50 to get something cheap from eBay. That bring the total cost to Â£525 - $825 or â¬650.
Step 7: Design Software
The main CAD software I've used is CamBam. It's a great piece of software and has a fully featured time-limited demo version. It's simple but powerful. I can only suggest that you try it.
The only other software I've used was for PCB routing. I went with the combination of Eagle for PCB design and the pcb-gcode add in to create the g-code suitable for sending to the mill.
Step 8: Create Your Materpeice!
Once your mill is up and running it's time to create something! I was amazed how easy it was to get something from your head to a finished article. It's so satisfying to actually hold your creation in your hands too!
- I've created some tiny dolls' house furniture for my niece's birthday. I was particularly pleased with the engraving on the top of the box and that it was so accurately milled that I could press fit the pieces together with no glue. (The box is about 4cm x 4cm x 9cm.)
- I've managed to mill simple PCBs that are fine enough to take surface mount components - so far 0.05" pitch SOIC, but I'll be attempting to do TSSOP soon. The mill can easily manage 0.0025mm precision so well up to the job if you have the right milling bit! The photos shows my first ever milled item and the 1mm bit was far too large. I've done better since.
- I've made aluminium brackets for the mill itself to extend the travel of the Y axis.
- I've created some custom records for a 70s Fisher Price toy record player. See my other Instructable for details of that.
However, the big question is what could you create...