[ryanjg117]

Interested in potentially making my own CNC router for woodworking, with a capacity around a full sheet of 4x8' plywood. I currently use a friend's dual-spindle custom-made CNC with vacuum hold down table, and would like to build something similar or perhaps a router plus plasma cutter design. I enjoy doing SolidWorks modeling and have done quite a bit of work with 80/20 extrusions (though no motion, yet) and figured it would be a fun project to take on and I would certainly learn a lot about motion. I've got a good amount of 80/20 extrusions and fasteners sitting around from a little part-out project I recently did as well.

Is there any savings to be had from building my own versus purchasing a DIY kit from, let's say, cncrouterparts.com? Assuming roughly equivalent part quality, etc?

What are some of the most "tried and tested" open designs out there? Where would be a good place to start?

Thanks and sorry for the somewhat n00berific post here.

Ryan

[ger21]

You can save a lot of money, at the expense of spending a LOT of time.
Also consider that if you buy a CNCRP kit, you have access to excellent support from them if you run into problems, vs being on your own if you build your self.

[wizard]

Well considering you already have parts, building your own makes sense. That is if the extrusions are suitably sized for this project.

One nice thing about CNC routers is that you can easily build your own and equal commercial quality machines. If you are parting out machinery with lots 80/20 it is fair to assume you have some mechsnical skills. This should put you ahead of someone just getting wet with tool building.

What you need to do now is to inventory that 80/20 to see if you have enough of the right sizes to start this project. If you end up short id suggest buying a steel beam for the gantry.

[TTalma]

Also consider used machines. I see commercial machines regularly sell for under 5k. I have even seen machines where the PC was missing/dead sell for under $1000k. Some of these larger machine have higher power requirements, and normally large space requirements.

I purchased an AXYZ millennium recently. 40x40 work area, all heavy commercial components. Designed to do wood/plastic/ and non-ferrous metals. 600ipm rapids, rack and pinion, perske spindle etc. etc. for $2k. No front end PC. Pickup, SmoothStepper, C41 VFD board, new PC and Mach 4 License was about another $500. So for the material cost of a similar sized machine with less capabilities I got great machine, and had it up and running with about 20 hr's time.

You have to decide what's important to you. I can totally understand the joy and fun of building your own, and then you'll have a real good knowledge of how it works if there is a problem in the future.

[ryanjg117]

Also consider used machines. I see commercial machines regularly sell for under 5k. I have even seen machines where the PC was missing/dead sell for under $1000k. Some of these larger machine have higher power requirements, and normally large space requirements.

I purchased an AXYZ millennium recently. 40x40 work area, all heavy commercial components. Designed to do wood/plastic/ and non-ferrous metals. 600ipm rapids, rack and pinion, perske spindle etc. etc. for $2k. No front end PC. Pickup, SmoothStepper, C41 VFD board, new PC and Mach 4 License was about another $500. So for the material cost of a similar sized machine with less capabilities I got great machine, and had it up and running with about 20 hr's time.

You have to decide what's important to you. I can totally understand the joy and fun of building your own, and then you'll have a real good knowledge of how it works if there is a problem in the future.

That's a great deal. Where/how did you find it?

[LazyMan]

I would just buy the machine, or find a used one with solid frame and retrofit if this is business venture. The most DIY I would ever consider doing is something like CNC router parts. If that machine is going to fit your needs, I would just buy it. If this is for hobby and building the machine is part of the hobby then build it.

[TTalma]

I find the machines all over, eBay, HGR, IRS, Craigslist. Patience is key though. To get the amazing deals you have to wait about once a year, great deals show up about once a week.

There was a 10'x12' Biesse on eBay a few weeks ago, real high hours, but sold for $300. I'm in NY and the machine was in VA. I considered driving down to strip some parts off of it. That was probably the best deal I've ever seen.

Small used routers 2' x 4' sized used machines seem to sell the best and highest. I don't have room for a full sheet machine, I barely have room for my machine (48" x 60" foot print) so I think the smaller ones sell better because of the smaller space demands.

Here's a great deal on ebay, no front end: https://www.ebay.com/itm/Thermwood-C...QAAOSwgCtaz4MS

Another ready to go: https://www.ebay.com/itm/BIESSE-ROVE...gAAOSwCkZZSV0i

Another no controller: https://www.ebay.com/itm/Thermwood-C...cAAOSwDsZaz4cU

The deals are not to hard to find, space (for me is)

[Goemon]

I wrestled with the build vs buy question for a while (who am I kidding, I still think about it daily!).

I think it's an individual decision based on what you need, what you expect, how soon you need it, your skills as a fabricator and how much money you have. For me, there were no machines available at the quality I wanted for the price I wanted to pay... so.... there was no decision to make.

I spent a long time researching my options and there are a lot of companies who are charging an awful lot for what is essentially a few hundred bucks worth of aluminum extrusions and corner brackets. Convenience is worth something but I will never see good value in a $3000 to $4000 benchtop machine made of 80/20 t-slots. Some charge $10,000-$20,000 for a t-slot kit.... it's madness.

The simple fact is that any design made from t-slot extrusions can be easily copied if you source your own extrusions, screws and connecting parts. In most cases that I saw, the potential savings from not buying parts as a kit were huge (well over 50%). T-slot extrusions are very easy to work with (as you already know).

With that said, there is a middle ground - i.e. You can buy your own t-slots and screws direct, then buy the gantry plates and risers as a kit from one of the many eBay sellers. Some of the savings can then be put towards better quality rails and ball screws.

I am certainly no expert. I'm still on my first and second build. My views are based on what I experienced from my own shopping experience.

I was not in a hurry with my builds. If I was, I would have bought a ready made one and sacrificed my quality expectations. A first diy project won't be quick.

[fretman_2]

I also went back and forth between build and buy. In the end it came down to the fact that I like the challenge of doing it myself. My machine is completely my own design and it actually works!!

[wizard]

Buying used can be smart and very cost effective. However it can also be a massive hassle. I spent months looking for a small lathe before eventually buying one. A lot of time and effort went down the drain looking for a deal. Just something to consider.

The one good thing about a DIY build is that you control the scedule and materials.

[fraser80]

I wouldn't recommend building your own cnc router to try and save money. As ger21 said, you are only trading your time for money. I just finished building my own machine and I am very happy I did it, but I don't think I saved much cash.

This is why I think building to save money doesn't make sense:

• If I valued my time at even a couple dollars an hour then I 'spent' much more than if I had bought a machine.
• I spent a year researching, designing, sourcing parts and fabricating my cnc router.
• In that year I easily could have completed a couple of paid projects which would offset the extra cost of the router. (once you have a cnc router people will ask you to make stuff)

I do, however, recommend building your own cnc router for the following reasons:

• To learning new skills. I can't believe how much I learned about electronics and programming while building my machine.
• Pride! The first time you get the motors or the axis to move is a pretty awesome feeling! (priceless)
• You are not planning on getting paid to use your cnc router. If you have no desire to do work for other people and get paid, then it makes sense to spend as little cash as possible.
• You have no choice because you cannot find what you are looking for.

Good luck Ryan with whatever way you choose to go!

Fraser

[Goemon]

I wouldn't recommend building your own cnc router to try and save money. As ger21 said, you are only trading your time for money. I just finished building my own machine and I am very happy I did it, but I don't think I saved much cash.

This is why I think building to save money doesn't make sense:

• If I valued my time at even a couple dollars an hour then I 'spent' much more than if I had bought a machine.
• I spent a year researching, designing, sourcing parts and fabricating my cnc router.
• In that year I easily could have completed a couple of paid projects which would offset the extra cost of the router. (once you have a cnc router people will ask you to make stuff)

I do, however, recommend building your own cnc router for the following reasons:

• To learning new skills. I can't believe how much I learned about electronics and programming while building my machine.
• Pride! The first time you get the motors or the axis to move is a pretty awesome feeling! (priceless)
• You are not planning on getting paid to use your cnc router. If you have no desire to do work for other people and get paid, then it makes sense to spend as little cash as possible.
• You have no choice because you cannot find what you are looking for.

Good luck Ryan with whatever way you choose to go!

Fraser

It's never that simple. Obviously, part of the cash saving for any diy project is in using your time instead of your money. But, whether there is a saving or not will depend on the individual, the quality of the build, how much their time is worth (which varies massively) and how well they source parts.

If you choose to buy, to compare properly, you need to factor in how long it would take you to earn the money you are putting into it. If it is for a business, you also have to factor in how much you lose from lost business in the time you spend on a diy build.

More important than any of that (IMO) for this decision is a persons current financial situation. For a business start-up, there is a lot to be said for not starting out with a mountain of machine debt. A lot of people are over-valuing their time or overestimating their financial health or the qaulity of their business opportunity because there is a huge number of relatively new CNC machines for sale. The liquidators never seem to have any shortage of stock so clearly, investing big upfront is not always the way to go.

If you want a quality small hobby machine, the key benefit of the diy route (aside from the learning experience) is that you can source superior components from the used market. Very few (if any) of the affordable benchtop machines or diy kits come with 25mm THK rails, ground ball screws or brand name motors etc. Plus, many of the cheaper off the shelf machines require assembly....

[travis_farmer]

I started with a used machine (mostly just the frame), added steppers, wired the controls, and was delighted when it made first motions. I am still working on it, but i learn a little more about it with every bit of progress, and even every set-back. i am still not a master with CNC, but i learned a lot from doing, and researching what needed doing.

I think what you start with depends entirely on your skill-level, and your time-span to get it done (if ever). starting with a kit, will give you a bunch of pre-selected parts, and all you do is put it together. no real research needed.
Building from scratch is like a trial-by-fire. you will find a lot about the machines through researching what you need. but it can get expensive if you get it wrong.
so value is in research on what you need your machine to do. what framework will be necessary to handle what you want your machine to do. and what motion components will be needed.

at least, that is what i have learned. and there is always a chance i may be wrong.

~Travis

[Goemon]

I started with a used machine (mostly just the frame), added steppers, wired the controls, and was delighted when it made first motions. I am still working on it, but i learn a little more about it with every bit of progress, and even every set-back. i am still not a master with CNC, but i learned a lot from doing, and researching what needed doing.

I think what you start with depends entirely on your skill-level, and your time-span to get it done (if ever). starting with a kit, will give you a bunch of pre-selected parts, and all you do is put it together. no real research needed.
Building from scratch is like a trial-by-fire. you will find a lot about the machines through researching what you need. but it can get expensive if you get it wrong.
so value is in research on what you need your machine to do. what framework will be necessary to handle what you want your machine to do. and what motion components will be needed.

at least, that is what i have learned. and there is always a chance i may be wrong.

~Travis

I agree and that learning experience has a huge amount of value. It means that, in future, you can also fix, maintain and upgrade your machine without huge costs.

There are degrees of diy for a build. It will take the longest if you decide to fabricate and assemble every component yourself obviously but there are plenty of options for semi-homemade though. At it's core, a basic CNC machine is just 3 linear actuators bolted to a (hopefully sturdy) frame. eBay is littered with used precision linear actuators and, if you know what to search for, there are some awesome deals available. I found a THK actuator with 20mm rails (2 rail 4 blocks), 20mm ground ball screw and 54" of travel for $450 delivered.

If I could go back and advise myself at the start of the project, I would have told myself to buy used actuators for all 3 axis and focus all my efforts on the frame. Aligning rails well is less my thing than building a quality gantry and base.

There is also a lot of options for a ready made machine frame if you have some imagination and modest diy skills. I was particularly inspired by this guy's build:

https://m.youtube.com/watch?v=herMY_2WowM

I like the smart use of off-the shelf parts, the incredibly simple construction and the level of precision he was able to achieve with a relatively modest budget.

I can't pretend that I am not tempted by this table with built in gantry either:

https://www.ebay.com/itm/Newport-30-...EAAOSwMVtZlxY8

How much time would that have saved with it's ready flat and square mounting surface with aligned predrilled screw holes. I doubt that anyone here is capable of building a more precise base than that... I want one!

[wizard]

Hi Travis;

Great to hear from you!

This is an interesting thread but I think people need to understand that even corporations struggle with the question of buying or building. Generally they never bother to build with off the shelf tooling (machines) which is what a router is in most case. But things like tooling and so forth, machine upgrades and improvements can be attacked with a combo of in house and contractor supported development. There are all sorts of factors that come into play so somebody asking this question needs to know it isn't uncommon.

For me there are numerous reasons to DIY but for a home shop or even a small commercial business, the big one is self support as you so well point out below. You can't help but to learn about your machine as you build it. If you have to pay for a field engineer to come in and fix a machine you immediately drop into the realm of high costs and downtime waiting for the tech to come in. That could mean paying not only the techs hourly rate but his plane ticket, hotel stay, meals and lets not forget the parts. Even if a DIY build ends up with a problem that perplexes the builder he can post here with a reasonable understanding of his machine, the parts it is made up of and how it works - a big communications win.

So I see the learning process leading towards ones ability to self support as being a big reason to DIY. However we also have to realize that that isn't what everybody is interested in. For those trying to decide which route to take I can only say think long and hard weighing the advantages and disadvantages, your skills and desires.

Thanks for the post.

I started with a used machine (mostly just the frame), added steppers, wired the controls, and was delighted when it made first motions. I am still working on it, but i learn a little more about it with every bit of progress, and even every set-back. i am still not a master with CNC, but i learned a lot from doing, and researching what needed doing.

I think what you start with depends entirely on your skill-level, and your time-span to get it done (if ever). starting with a kit, will give you a bunch of pre-selected parts, and all you do is put it together. no real research needed.
Building from scratch is like a trial-by-fire. you will find a lot about the machines through researching what you need. but it can get expensive if you get it wrong.
so value is in research on what you need your machine to do. what framework will be necessary to handle what you want your machine to do. and what motion components will be needed.

at least, that is what i have learned. and there is always a chance i may be wrong.

~Travis

[wizard]

I agree and that learning experience has a huge amount of value. It means that, in future, you can also fix, maintain and upgrade your machine without huge costs.

This is huge for somebody that wants a machine in their home shop. Even if you can get telephone support for a purchased machine you still need to understand the machine to communicate effectively with the on line service tech. The only problem with advocating the DIY approach is that some people just are not up to a machine build for whatever reason.

There are degrees of diy for a build. It will take the longest if you decide to fabricate and assemble every component yourself obviously but there are plenty of options for semi-homemade though. At it's core, a basic CNC machine is just 3 linear actuators bolted to a (hopefully sturdy) frame. eBay is littered with used precision linear actuators and, if you know what to search for, there are some awesome deals available. I found a THK actuator with 20mm rails (2 rail 4 blocks), 20mm ground ball screw and 54" of travel for $450 delivered.

If I could go back and advise myself at the start of the project, I would have told myself to buy used actuators for all 3 axis and focus all my efforts on the frame. Aligning rails well is less my thing than building a quality gantry and base.

I would only qualify this by saying that the world of industrial actuators covers a massive array of capabilities. So if somebody goes this route be careful about what you buy as many slides simply don't have the rigidity to be used for routers or other CNC machines.

There is also a lot of options for a ready made machine frame if you have some imagination and modest diy skills. I was particularly inspired by this guy's build:

https://m.youtube.com/watch?v=herMY_2WowM

That is one beautiful build! It is a great example of somebody going his own way and doing so economically.

I like the smart use of off-the shelf parts, the incredibly simple construction and the level of precision he was able to achieve with a relatively modest budget.

I can't pretend that I am not tempted by this table with built in gantry either:

https://www.ebay.com/itm/Newport-30-...EAAOSwMVtZlxY8

How much time would that have saved with it's ready flat and square mounting surface with aligned predrilled screw holes. I doubt that anyone here is capable of building a more precise base than that... I want one!

Those optical bench tables are an interesting breed, I'm not sure how they would work out for a machine tool build. The tables are optimized to minimize vibration. As such I'm not sure how thick the skins are on a table like this.

[Goemon]

This is huge for somebody that wants a machine in their home shop. Even if you can get telephone support for a purchased machine you still need to understand the machine to communicate effectively with the on line service tech. The only problem with advocating the DIY approach is that some people just are not up to a machine build for whatever reason.

I would only qualify this by saying that the world of industrial actuators covers a massive array of capabilities. So if somebody goes this route be careful about what you buy as many slides simply don't have the rigidity to be used for routers or other CNC machines.

That is one beautiful build! It is a great example of somebody going his own way and doing so economically.


Those optical bench tables are an interesting breed, I'm not sure how they would work out for a machine tool build. The tables are optimized to minimize vibration. As such I'm not sure how thick the skins are on a table like this.

Optical tables / optical breadboard / laser tables (and a bunch of other names for the same thing) come in a wide variety of thicknesses / strength. You can definately find ones that are (more than) thick and strong enough for a CNC router.

The thickness is usually linked to the size, ( like with surface plates) because they also need to maintain very high precision flatness for optics / laser applications. The requirement for precision and stiffness in some of those industries exceeds ours (but for different reasons). They need to maintain flatness to the same .00001" as surface plates.

They also come in a range of materials. I've seen them in stainless steel, aluminum, granite and composite honeycomb structures (which are very strong).

The limitation to be aware off for the tables with built-in active vibration isolation using pneumatics, is that there is usually a weight limitation for that function and it's quite a lot less than the max weight the table could otherwise hold. I've seen 4000lb granite tables that can only hold 600lb with active vibration isolation turned on.

I was looking at a deal on a 30" x 48" aluminum breadboard and it was 1.5" thick. That's about the same thickness as the aluminum sub plates sold for the Haas VF2. In other words, It would cover most needs here. It takes a lot of force to bend 1 1/2" of aluminum, even at 48" long....

I think the large over-looked advantage for the diy builder is that many are designed to reduce vibration using lighter materials like aluminum. Commercial machine builders use cast for it's damping properties but, even a relatively small piece of iron weighs far too much for the average diyer to move around.

The precision flatness and perfectly positioned thread holes on some of those plates is a huge plus too. Depending on what parts people are working with, I bet many people could start assembling and mounting rails as soon as they arrive. Who enjoys threading holes....

[awerby]

Goemon wrote:


"There are degrees of diy for a build. It will take the longest if you decide to fabricate and assemble every component yourself obviously but there are plenty of options for semi-homemade though. At it's core, a basic CNC machine is just 3 linear actuators bolted to a (hopefully sturdy) frame. eBay is littered with used precision linear actuators and, if you know what to search for, there are some awesome deals available. I found a THK actuator with 20mm rails (2 rail 4 blocks), 20mm ground ball screw and 54" of travel for $450 delivered.

If I could go back and advise myself at the start of the project, I would have told myself to buy used actuators for all 3 axis and focus all my efforts on the frame. Aligning rails well is less my thing than building a quality gantry and base."

I built a machine like that a few years ago; it worked out pretty well. The length of the actuators was a limiting factor, although the ones I used seem plenty rigid. It's nice not to have to deal with aligning rails and screws - the actuators have all that done already - although we did have to get the two X axis actuators lined up with each other. I'm back at the drawing board now, trying to come up with something about twice as long and a little wider.