Advice on Choosing a CNC Router

Here are some detailed thoughts on matching a new piece of CNC equipment to your needs and your budget. October 26, 2011

So, I'm starting to look at CNCs for our (currently) 3 man shop. I recently purchased and am painfully learning Cabinet Vision. We do all residential, typically face frame/inset, but have branched out to do it all in this new (read "crappy") economy. It's picking up and we're one of the lucky left standing that needs to gear up for the recovery they keep talking about. The writing is on the wall, so I'm starting to pay attention.

I am a research hound and it generally takes me about a year to make a large machine purchase. This will probably be my largest. I've read tons of past threads but it's hard to process all of them. So, what I'd like to know from current CNC owners is what features have become important to you? What components have or haven't worked? What's been a constant problem? What's been a blessing?

Forum Responses
(CNC Forum)
From contributor W:
More than anything else, it boils down to how much money you have to spend. Looks like for the most part you get what you pay for.

From contributor U:
So true on how much you want to spend, but the one thing that draws me to any manufacturer is support. Are they really there for you when you need tech, parts, etc.? Enlist a good independent tech who will have worked on nearly every type of machine. They will tell you the truth.

We don't know how we ever did without ours, and would be lost without it now. There is much to integration, software, posting code and getting everyone playing nicely together, but when you get all the right elements in place, your productivity will skyrocket.

We are a pod and rail Busellato along with a great tech to keep it all humming, as well as great support from Delmac in the way of parts. There are so many ways to do this that you have to define what you want to do, nest, beam saw and P2P, etc.

From the original questioner:
Yes, I understand that like any machine, you get what you pay for. I'm trying to figure out what I need and therefore how much I need to spend. I'm not afraid of spending money on a machine as long as I can justify what it's for. If there's a resource anyone can recommend that will educate me on the basics of CNCs, that'd help a lot. There are lots of holes in my knowledge.

From contributor D:
It will really depend on what the machine is being used for as far as what to look for. Are you carving parts or nesting images or both? That said, I am not about to argue the difference between ball screws and rack and pinion. Heck - both work fine, in my opinion. My suggestions would be to focus on more than the machine itself...

-Software can make a machine good or bad. Look for a machine with software you are comfortable with in both the interface and programming.

-Company and reps. There has been a lot of transition lately in terms of reps and manufacturers. Do your due diligence and investigate each, to assure if issues do arise there is someone to help you.

-Proprietary parts. Manufacturers are not afraid to take your money by large markups in replacement parts. Some manufacturers have a fair amount of parts which are difficult to find from outside vendors. I have even had an instance where the manufacturer of the part told me the part can only be purchased through the machine manufacturer. On the other hand, I have seen other manufacturer's supply along with their part numbers, part numbers and phone numbers of the component and manufacturer.

From the original questioner:
Right, makes sense. I guess I see it mainly processing case parts. Cutting, dadoing, drilling, etc. All so I don't have a guy standing on the slider, panel router, boring machine for a week doing those. Working solid wood would be a bonus, as we do a lot of trick stuff that most guys around here won't touch, but we're pretty proficient at doing those without a CNC and so I think that's, as I said, bonus. Though I'm sure it's like any tool in that once you have one, you start realizing all the things it can do that you never even considered.

From contributor J:
Go visit shops and machine companies. I did and one in particular had such a good relationship with the CNC provider that I jumped in with them and stayed. CNC owners are proud of these things and the good ones love to show them off. I had two by here this week.

From contributor C:
Wow, are you in for a neat experience. The heaviest machine you can buy will be a plus. The after sale service is of paramount importance. The loading and unloading of sheets and parts can negate any efficiency gained by the CNC. You do get exactly what you pay for - don't go inexpensively unless you have to. The precision, stability, and repeatability of a heavy German machine is unbelievable. The next item of equal importance is a good vacuum pump. 40 HP DEKKER is the bomb. The strength of your vacuum system will save a lot of ruined parts due to movement under full load.

From contributor M:
I don't think a 3 man shop has to spend loads on a CNC. I purchased a Techno 5x10 and it's great. The one item I'm disappointed with is the vacuum. We have a 10HP Becker and I'd say it is the bare minimum. 20HP or more would be better.

From contributor B:
Figure out what you need the machine to do, and then get the machine with the features to do it. Holding back on good features will only cost you more later. I found that out the hard way.

My machine weighs 5,000 lbs - plenty for machining my 3/4 material on nested applications. You don't need a super heavy weight machine to do it.

Not all CNCs are built the same way - as mentioned in earlier posts, you get what you pay for. Look for machines that are well built and configured with the features you need. More importantly, look for machines using parts and components that are third party supplied that are based in the U.S - not proprietary, so you can find service and support from other than just the machine manufacturer. My machine has these features, equipped with 10HP power spindle and automatic tool changing stations. I only wish I had gotten the model with the drill spindle on it now. It will have to do for now and I will get the drill spindle when I upgrade.

I got a 10HP pump and it does fine on nested cabinet parts. But too weak when I do MDF doors. If you do MDF doors, get a 20HP - it will help.

You either end up paying too much for a cheaply configured machine, or you can find a good quality machine at the same or lower price. Good luck with your research and hope you find one that fits your needs.

From contributor Y:
Here is advice I got from a friend when in the market over a year ago...

"I looked at a lot of routers in the under $100,000 price range, including MultiCam, FlexiCam, ShopBot, Techno, CNT Motion, and of course, Omnitech, which is what I bought. They all had strong points, but the Omnitech seemed like the beefiest and most well built in my opinion. It is made by Anderson, an American company well respected for their CNC machines. The Omnitech is made in Taiwan out of US parts and electronics, but this model is more of a cookie cutter approach, with standard features on all their machines to help keep the cost down. It is essentially a tank that makes cabinets.

The machine I got is a 4 x 8 table, because 99.9% of everything I do is 4 x 8 sheets. No need to spend the extra money for something I would only use on rare occasions. My slider is a 10' model, so that comes in handy for those jobs requiring 10' sheets. (Don't sell your slider even after you get the router. There will always be small things you need to cut or quick remakes of a damaged part, etc.) Also, most routers will have this, but it is essential that you get a drill block. This is for drilling 5mm shelf pin holes, mounting holes for hardware, and any other vertical drilling operations you need. I have mine set up for 1/8 pilot holes for screws. Andy has the 15mm skid bore and 10mm hinge plate holes also on his drill block. I might put those on mine as well soon. Any router without a drill block (I think ShopBot and Techno don't have them) will be much slower than a router with them. Imagine drilling one shelf pin hole at a time versus 5 at a whack with the drill block. The Omnitech has 9 drills - 5 horizontally and 5 vertically (the corner drill is on both).

Another thing I thought was important was a tool carriage that rides along on the gantry versus a tool rack at the end of the table. Every time you need to make a tool change, the machine just stops where it is, changes the tool, and continues on its merry way. With a tool rack, it has to travel all the way to the end of the table to change the tool and then move all the way back to do the next operation. This is slower and also adds more wear and tear to the servo motors and gears that move the machine. The Omnitech has 8 tools in the carousel, which is more than you will need for any single sheet.

I liked that there is a computer right at the machine on the Omnitech, which allows you to make any modifications right there, instead of running back to the office. The controller is a Fanuc controller, which is an industry standard, and could be replaced easily if need be (not that I think it will ever be necessary). Some other controllers like Siemens are more proprietary and would need to be replaced from the manufacturer. And there have been a few times I did a net meeting with tech support on the machine PC, so they were able to diagnose and fix my issues from North Carolina. And the tech support from Omnitech has been outstanding. First, they actually answer the phone! No automated voice system. And they always figure out what to do. They have also written a few new programs for me to help tweak a few things that make my life easier. It's free lifetime tech support from Omnitech - not sure you can find better than that.

My favorite feature of the router is the push off device, which pushes the completed sheet off the router and onto a panel cart, and vacuums the table in the process. This speeds up the work flow dramatically. Without a push off device, you have to manually pull each part off the machine, label it (unless you have the router do it, which still adds to the cycle time), clean off the table, and put another sheet on before you can run the next cycle. With a pushoff device, you can move the cart to the side, blow off the remaining dust crumbs, load the new sheet, and start the next cycle in a matter of seconds. You can be sorting, labeling, and edgebanding the parts from the first sheet while the second sheet is being cut. I almost never beat the router, so it is generally waiting for me to get ready to unload the second sheet while I finish up with the first. No matter what brand you end up with, I would really stress the importance of this device - the time savings really add up quickly.

We run Cabinet Vision Solid Manufacturing with "screen to machine" capabilities. Basically, from the job view it's just a few clicks to create and send the code out to the machine. Print out the sheets with the parts list and drawings of the sheet layout and you are all set to cut. If you are running KcDW, I'm sure they can integrate it with whatever router you get. I am not sure if you will need an intermediate 3rd party software, like CadCode or EnRoute to take the info from KcDw and create the G code for the router. By now almost all machines and all design software can be linked to a CNC - it would be stupid for them not to make these things compatible.

I think the thing that clinched it for me was the entire sales process and tech support. My salesman, David Paine of CNC Sales of New England, was very upfront about all the things that the machine could and couldn't do. In the 2 1/2 years that I have owned the machine, I have discovered that he never once bullshitted me about anything. I am not aware of any other salesperson I have dealt with that I can say that about. He also gave me a complete turnkey price - the total cost for my machine in May 2007 was $86,720. That included absolutely everything - the machine, tooling, tax, delivery, setup and installation. The tech guy from Anderson was at my shop for 5 full days installing it and teaching me how to run the machine, and that was part of that price. I didn't pay a dime more than the quoted price. Now of course the software people have to come in and integrate their software with your machine, but I had upgraded mine to include the CNC link and the integration was part of that price."

From contributor E:
You guys are great! I have been looking and just could not decide and now I think I have.

From contributor J:
I have a 10hp fpz and a 15 fpz. For nesting I onion skin and cut all small parts first then final pass with the cut out. This I can do with the 10 only and quite well (saves a lot of power). There are those who may not have an extra three to five minutes a sheet for this, but it serves me well. I set up the extra 15 hp but rarely use it. Good clean surface spoil board, good nest strategy and you can rock with a good mid-weight machine. I have a 5hp spindle and a multi-head setup for three bits. This works well yet I have ordered another machine with a 7.5 or 10 hp (still have the option) spindle atc this time, as MDF doors can want 4+ bits to work well. I would love a drill bank and it will come down to what you wish to afford and what you really want to do with the machine.

I am currently doing a complex trophy case for a university. Bid was tight but I got it. Once the architect sent me the dxf files for the design, I had it cut and standing in two hours. This is some of the branching out you are talking about, as this has been a good job!

Click here for higher quality, full size image

Click here for higher quality, full size image

From contributor X:
Can your machine also bore 8mm dowel holes in the sides for decks and stretchers? Also can you program it to cut all horizontal parts first, so they can be end bored while the router is cutting the sides, or are you using blind dadoes?

From contributor Y:
Blind dado construction is what we use.

From the original questioner:
All great advice. I appreciate it. Contributor Y, that was extremely helpful. I guess one of the main things I'm looking for is, "I really wish my machine had this feature/capability" or "I'm glad my machine has this much z clearance" and "I have this type of controller/drive system/HP motor/spindle/etc. and it's great (or it sucks)." I'm fairly positive I want NBM, but what features/capabilities are truly useful is difficult to know, having not used one before.

From contributor J:
I do a lot of deep 3d so a Z clearance of 10" is a must for me. The type of CNC machine I have provides for changing this, as I originally ordered an 8" and later they built me taller plates. I change them myself. My wincnc controller is a windows based industrial controller that serves my needs well. Strong vacuum is a must, full heavy steel table. Like I have said, I am ordering an ATC, as I have a multi head now and would probably have ordered an ATC initially, but at this point I am just adding another machine. I am interested in the pusher (like the Omnitech, which I do like) and can add that myself as the configuration of controller macros and machine construction allow me that flexibility with mine. I had a 10 hp spindle on the machine originally. When I burnt it I replaced it with an older Columbo 5hp and, while it performs fine, I miss the 10, so it will be on the new machine.

From contributor S:
I'm currently using a 5 x 12 pod and rail machine for panel door production and hardware prepping. This is the second router from the same maker I've put in service. I'm not well versed in the flat bed machines typically used by cabinet shops but assume they are more similar than different.

On board saw carriage. My first machine had one but the second one doesn't, so we do that using a saw aggregate tool or vertical roughing router bits. Works, but really liked having the on board saw for straight cuts in X and Y axis for fast part sizing, cutting flat panel sheets, kerfing, scoring, v grooving, etc. As a panel saw a CNC with an on board saw carriage can run rings around a traditional vertical saw and there are things a saw blade can do that a router bit can't.

The ability to run rotatable 3D simulations of your programs in your software to check order and depth of cut on your tool path. Also try to get software that is capable of letting you program using formulaic if/or statements or say repeatable toolpath loops by being able to cut and paste right in to the G code. Maybe both of these things are ho hum standard by now, but my newest machine has much more powerful software.

Unless your rendering cabinet software is writing all your programs for you, spend as much time understanding how to write parametric programs and organize or catalogue them in such a way that they are easy to access, understand, and tweak on the fly.

Most good machines will perform similarly in their respective price range so the choices become about service and software, footprint size, power and dust collection requirements and you learning how to use the machine and organizing your programs in such a way that multiple shop people can walk up and make stuff.

From contributor O:
We have had our CNC (5x12 nesting) for almost a year now. We went through this when we were deciding. We were planning to cut cabinets and some MDF doors, but the sky is the limit once you have it. Now that I have used it, I will share what I have learned.

1) Bigger is better. We have a 5x12 table and most sheets are 4x8, but we optimize 5x12 laminate sheets for commercial jobs (better yield), large countertops and architectural pieces.

2) Drill block.

3) Aerotech cutter. We use it for dust free nesting; we were shocked at how much dust was left behind with a regular bit.

3) Parametric software at the router. We have a Weeke, so it runs Woodwop. It is very easy to use and I was drawing parts after one day of training.

4) Our router has 4 smaller pumps. Keeps the noise down and can be turned on and off independently to save energy.

5) Ability to add pods. We do a lot of interesting pieces that we can mount on our pods. This has been the most fun for us to create smaller solid wood items.

Lastly, safety. We noticed a wide variation in safety standards from brand to brand. Some had no laser screens or pressure mats. Our Weeke has a fence on one end and the back. On the front and open end there is a laser field that shuts the machine down if it is broken.

Best investment ever! And so much fun.

From contributor V:
Can't say I'll think of everything here, but:

4x8 machine is adequate... 5x10 much better.
Rack and pinion on x and y are okay, but lead screw on z is important.
6" minimum z-height... 8" or 9" is better.
5 hp spindle is adequate for 90% of applications... 10 hp can be worth having but will cost a good deal more.
Tool changer is almost a must for commercial work.
I personally feel medium duty machines will do everything most people need to do. Heavy iron is nice but has cost issues for purchase, operation and maintenance.
Through vacuum with a high quality 10 hp regenerative blower for nested base. You can always screw pods down to the spoil board for the occasional solid wood components.
Single phase input inverter if you don't have 3-phase in the shop and are going with a medium duty machine.
Multi-zoned hold down table system.
Moving gantry if a medium duty machine... moving table if heavy iron.
Remote control hand unit can be nice but may never get used.
Air pressurized control cabinet and control boxes on the machine.
Well designed dust skirt.

Totally dependent upon your needs and budget, but $10k or more for this is not required.

From contributor L:
Availability of parts is an important item. Does the company keep an inventory of parts and how long would it take to get a replacement part?

One thing I consider important is documentation. How much information can you get on the machine itself? How good is the operator's manual? What kind of troubleshooting information is there? If the controller produces error codes, is there an explanation of what those codes mean? Is there a programmerís manual? There was a long discussion here about if one needs to know CNC programming or not, that you might want to check out. I donít think it hurts and if you want to do parametric CNC programs, logic statements, loops and subprograms you definitely do.

The software you use to create the CNC programs is important. You need to be comfortable with it and it needs to fit your operations. I think that a simulation feature is important and if you are going to do any manual programming I think that the simulator needs to work from the actual CNC program, not from the geometry. Will you design parts and generate CNC code in the office? If so, how does the CNC program get from the office to the router? Does the router want a G-code file, a DXF drawing file, or maybe something proprietary?

We had a lot of trouble with the power in our area and with lightning storms and learned that the router needed to be well grounded. Also, each router has a PC connected to it through a serial port connection and we had to run those serial lines through surge protectors.

It has been a long time since I have looked at new routers and I imagine that the technology has changed a lot. I didnít like the operator interfaces very much and always thought that they could be a lot more user friendly and a little less cryptic. I also felt that the machine should be able to communicate more about how it was feeling and if something went wrong, it should tell me what is wrong in plain English. At least if a fuse should blow, there could be a light that says "fuse A has blown." Any machine designers out there? I do know that there are machines that can be interrogated remotely, and that is a good thing.

From contributor K:
I have a Biesse ptp machine that we purchased 6 years ago. The only part aside from pods and tooling that we have had to replace was an air actuator on the dust collection. We have had some software issues that have always been user induced that have caused us some frustration but nothing else.

When I first purchased I tried to be cheap and use the software that came with it. Huge mistake! It's not that it didn't work; it's that we had to program each and every new or custom piece, which could take a lot of time. Two years in I realized that we were limiting ourselves on what we could do by how much programming it would take. Upgraded our Cabinet Vision and we are now full screen to machine.

Cabinet Vision has a huge learning curve and if you are like most small business owners and don't have the time to spend hundreds of hours to get proficient, find a good independent consultant to set everything up for you. My guy, Darren Johnson, had us up and running screen to machine in no time.

When we started, we put very few parts on the CNC. Now as we have evolved and our product and client have become more complex, more and more parts go on it. I am now looking at getting rid of our beam saw and ptp and going to full nested base. I have calculated that if we do that we will touch a part 3-6 less than we currently do.

As for machine, a ride along tool changer, and as big as you can get drill head.

From contributor T:
ATCs are an important part of most CNC machines. It's been said that a carousel tool changer is faster. My experience tells me the time to run a tool change is mostly a function of acceleration and deceleration time for the spindle. If it takes the spindle 3 ~ 5 seconds to wind down from 24,000 (or 18,000) to zero, can the machine travel to a stationary tool change rack in that time?

Quick calculation: 3000" per minute (rapid speed on machine) / 60 (seconds per minute) = 50" per second. Even allowing for ramp curves on the gantry, there is sufficient time to travel to a stationary rack while the spindle is stopping. So my thinking is that speed of tool change between two types is basically the same. There is less cost to a rack ATC; use the difference for other purposes.

From contributor K:
Contributor T, you are probably correct on the ATC if all you use is the router. If on the other hand you have a drill block, it's a different story. My Biesse routinely changes tools without stopping while performing other functions.

From contributor T:
No question it's faster to change drills in a drill block than auto tool change. There is an interesting thread on the CNC forum called "drill block for CNC," debating the merits of a drill block. It's interesting reading, but it is a bit of back and forth.

From contributor Q:
It took me over a year to buy a CNC as well. I went to every shop that would let me in and every manufacturer's open house that was available. In the end I bought a Weeke from Stiles. After having the machine for nearly 3 years, here is what I have to say.

1. Stiles is the best service company that I work with. I own tools from SCM, Biesse, Unique, Razor Guage, etc., and none of them have the response time or parts inventory that Stiles does. That alone would keep me from buying something different in the future. To phrase this a different way: I was not aware of the lower level of service I was receiving until I purchased this machine and had my first issue and suddenly the difference between Stiles and all other manufacturers that I work with was very obvious.

2. The Weeke is extraordinarily fast, accurate, and dependable. My machine has both a stationary and automatic tool changer and I would say that the machine operates nearly the same no matter where the next tool is located as far as time goes. That assessment probably comes from the very high acceleration/deceleration rate of the Weeke... and may not be the same as other manufacturers.

3. I believe that weight is a huge factor in this type of purchase because vibration over time is what kills the accuracy of a machine. My particular machine is roughly 14k lbs and it doesn't so much as think about shaking no matter how fast the tool is run. This is not the case with lighter machines... I have seen many look like they are ready for launch during a high feed rate.

4. Feed rate, acceleration, gantry type (closed or open), and machine weight will play a huge role in the lifespan of your tooling as well. This is something that people don't seem to focus on, but tooling costs add up really fast, especially when you are replacing them prematurely. Case in point is that I run an average of 350 sheets of 3/4 birch before I sharpen my 1/2 compression spiral. They are not dull at this point but the quality of cut begins to degrade around this point. I have heard many people say they can only get about 100 sheets out of tool. This has a lot to do with acceleration, deceleration, also effective cornering in a nest, the rigidity of the gantry, and machine weight. It should be noted that learning to figure out proper spindle speeds and feed rates of a particular tool will be a necessity for you as well.

5. Vacuum is important, as previously mentioned. I have a 10hp Becker and it does the job. I would like to have a bigger unit but that is because of climate issues more that size/performance. Also, a bigger pump doesn't necessarily equal better performance. The matrix of the flow table has a lot of input here. Sort of the old "how much water can you really push through a garden hose" scenario. There are some new machines that offer multi pump tables and I personally think that is better than one huge, expensive to run vacuum pump.

6. The only thing I wish my machine had is a higher "z" clearance, but that is because I like to tinker with the limits of my tools. If it can do it then, I like to try and make it do it so that I know the accurate limits of my offerings to a customer. I primarily build high end residential as well and "z" clearance is not an issue 99.9 percent of the time because sheet goods are mostly 3/4 and less.

7. I would not buy a machine that is sold to me under the premise that "it can be upgraded later if you outgrow it." I heard that quite often from sales reps selling less than industrial quality machines. I wanted a machine that would still be viable in 15 to 20 years as long as I maintained it well. This is a huge investment. Mine cost as much as my house... it had better perform like it.

8. I also use Cabinet Vision and so I can tell you that integration is important. Not all machines integrate with the software the same. My machine was up and running in 1 day with the Planit on-site integrator while a friend of mine who bought a different machine took nearly two weeks to have it up and running and that was not without some issues. The ironic thing is I was torn between my machine and the one he was set on buying and my gut led me to the Weeke. Good thing for me.

9. I would also find out how many service techs are located within a half day's drive from you. When you come to rely on the machine for milling (and you will) and something goes wrong you will want someone there the next day. It is also much cheaper to pay for a tech driving to you than flying to you.

10. My last piece of advice is to buy a machine that is a little more than you currently need because you will grow into it as long as you continue to grow your business. For NBM "c" axis is not important and had I bought that option I would have wasted money. Spend that money on a Schmaltz lift and make life easy on the operator.

I hope you find the right machine for your budget. There are good tools out there for under 100k but I am wary of the ones that are under 75k because at some point cost dictates that too much machine is being removed somewhere.

From contributor I:
I know that a lot of people out there will not agree with me, but for the money, ShopBot still gives you the best bang for the buck. You can get 2 or 3 ShopBot machines for what 1 big iron costs. I bought my Bot 5 + years ago and in that time I have only spent $85.00 on repairs. It's used every day and still does the job. I bought the 5 hp spindle, the drill head for drilling shelf pin holes, and the indexer for making post and legs. And I wish now that I had gotten the auto tool changer. Yes it's a little slower than big iron, but I can still cut more cabinet parts in a day then 2 men can assemble in 5 days. Tech support is fantastic; just call and you get answers. My opinion.

From contributor H:
We also use Weeke. We have got an ATC, which is really a must for commercial work. We also have the Otrix table which enables work to be elevated off table to allow edge profiling, horizontal boring, etc. We also use Magicut/Cutrite software for all nesting and Aspire for one off jobs and specials, as it is very easy to program.

From contributor Z:
I had an Onsrud Panel Pro, and it did most things that I needed to do, but I found myself constantly wishing for more horsepower and C-axis. Finally upgraded to a super duty Onsrud, aluminum table, c-axis and 18 hp. It is so much better. The only thing I hugely disliked about the Panel Pro is that it had a phenolic table. I can't say a whole lot about service, as I haven't needed it. I quite honestly never considered the smaller sign maker routers. I knew that if I went that route, I would always want something bigger, faster, and stronger.

From contributor A:
A CNC router is a printer. It will print whatever you send to it so the ease of which your design drawings go to the machine as G-code is more important and you can spend a lot of many making this thing work for your situation.

The $30 printer will print the mundane black and white document just as good as a $500 printer. The same goes to CNC routers - the $60k router will machine the nest out of a sheet of melamine just as good as a $200k machine, and it will take the same amount of time whether it has a 7hp router spindle or a 16hp spindle because your rpm and feed rates are material dependent. Ease of loading and unloading may be a factor to consider if you machine more than 20 sheets a shift.

I advise you to do your research but hold off buying one until your shop is overwhelmed with work - do not buy one in anticipation of more work down the road.