LinuxCNC vs Commercial

I'm new to CNC and was wondering if I could borrow a bit of your time with a few loaded questions?

I've been looking at CNC commercial routers to be used primarily for cabinet manufacturing as well as going the DIY route and building something myself.
I'm a self taught software engineer and have experience in robot building, but nothing that would require high feed rate precision. I originally looked at Mach 3/4, but the interface is unappealing and there appears to be a better community around LinuxCNC.

How does LinuxCNC compare to commercial offerings offered by companies such as Laguna tools software controller for example?
As an example, if I were to put together a machine identical to a commercial offering, would it perform the same?
Does LinuxCNC require a hardware controller to coordinate movement between servo controllers?

Ultimately, the plan is to either get in the deep end and purchase a $125,000 router with ATC and material handling, or start smaller by building a machine with high quality electronics and a sturdy 5000lb frame and add features such as ATC and material handling as it becomes necessary or I want to start taking it on.


Also, any recommendations on books on motor electronics is welcome. I've been pouring through the LinuxCNC documentation, but my inexperience is slowing the pace.

If you have time, skills, knowledge and a back workshop you can make good quality machines that do not require foundry parts. If you fail in one of those premises the best that you can make is a hobby machine that probably will not turn out as good as you want for production.
A router if one of those possible machines.

Are there any commercial distributors that use LinuxCNC?

Tormach do

What machines are you looking at in the $125,000 price range?
What sort of cabinet milling were you wanting do? (There are different machines more focused to different things, such as pod and rail for lumber and wood stock or flat vacuum tables for sheet material/nested base.)

$125k can get you a lot nicer machine than anything I saw in a quick look on the Laguna Tools website.

Personally, I am a big fan of buying a better quality higher end used machine for the same price as a lower end new one for the same price. (Usually a 5-10 year old commercial grade machine can be bought for roughly half-3/4 it's new price.)

We do a lot of nested milling of MDF sheets, and currently have 6 CNC routers with 5x10ft vacuum tables and 8-12 tool changers cutting nested parts. Most of these are running Fanuc controls, and run 24hr 6 days a week. Only 2 of those machines were bought new, and those 2 are the ones I would most like to see gotten rid of. They are Omni Tech Pal machines, that we paid $60-70k for new. We have 3 Komo Mach1 510 routers, and a Northwood IH510, these are all excellent high end machines that we purchased used for $50-$70k. The one non Fanuc large table router we have is an old SCMI Routmat we picked up dirt cheep from an auction, I was going to retrofit it to Linuxcnc if couldn't revive the old control. I was able to fix the old control, it works, that's all I'll say about it.

The Omni Tech machines are perfectly adequate machines for what they are. They are an inexpensive machine designed for a cabinet shop that will run it 4-8 hour a day, 5 days a week. Using these machines the way we are, they have been a bit of a maintenance headache. We've been told by the manufacuter that in 5 years we put more cutting hours on these machines than any other machine they've sold (and we've had them for 10yrs now).

We also have 8 other light duty CNC and gang routers that we use primarily for carving and engraving. Several of these I have converted over to Linuxcnc.

As to your basic question as to if Linuxcnc is as good as a commercial control. It all depends, on A.) what your compairing it to, and B.) The quality of the Linuxcnc install. (such as how well you integrate safety features, and other things like tool changer implementation...)

Thanks for the responses, sorry I haven't responded sooner. Life be crazy...

It would perform very similarly. It is possible that the commercial control might be slightly better optimised for wood routing.

Good enough for me. There really aren't a lot of software based options in this space. I'm just looking at this to see if this is something I can build off of. The software based option appeals to me the most because I can do things like integrate with my project system, create a labeling system, and possibly use it for other things such as material handling. My imagination is outpacing my current abilities, but I think I can get there. Just have to work on my maths...

No, LinuxCNC _is_ a motion controller.

Cool. I thought as much, but wasn't sure. It is compared to Mach 3, but I don't think they are on the same level at all. I can't figure how Mach 3 could act as a realtime controller on Windows. I don't know if they have their own kernel mode driver and that's why it required a parallel port.

You might want to consider buying such an item of orphan hardware and installing LinuxCNC on it.

This is definitely an option as well. I've been looking at used hardware based off of what Todd Zuercher wrote and I can see this as a definite possibility. Old hardware is usually made to last.

What machines are you looking at in the $125,000 price range?

Nothing from Laguna. I was looking at machines from Thermwood or C.R. Onsrud. I don't have pricing information from them, so the $125,000 was more of an upper bounds guestimate on my part. Something like this is interesting:

Personally, I am a big fan of buying a better quality higher end used machine for the same price as a lower end new one for the same price.

I agree. Our edgebander is a used holzher machine and the only real difference between the newer ones of the same caliber is you can modify settings at the panel instead of getting behind it.

As to your basic question as to if Linuxcnc is as good as a commercial control. It all depends, on A.) what your compairing it to, and B.) The quality of the Linuxcnc install. (such as how well you integrate safety features, and other things like tool changer implementation...)

I'm mainly looking to see if LinuxCNC would ever be the bottleneck in development and maintenance on the machine. This is a pretty interesting project for me, but in the end, if the machine is down a significant amount of time or required endless tweaking, it probably isn't an option.
A: I mentioned Laguna because looking at what they offer I think I can produce something at least on par with what they are.
B: This is probably something I would play around with quite a bit before jumping in and purchasing everthing I would need. I think I will be doing a lot of planning and asking a lot more questions here before I move forward and begin the build.

While I have no personal experience with either Thermwoods or CR Onsrud machines. They both seem do make decently solid hardware, in the upper mid-range price wise. The CR Onsrud Panel-Pro in your video link is very comparable to the Omni Tech machines we have.

I've not heard any bad things about Thermwoods, but in the past I steered away from them because of their proprietary control system. Knowing more about CNC now than I did then it would be less of a factor for me.

I have heard some bad things about CR Onsrud's service quality. And service should be an important factor in choosing a new machine. They fairly recently changed their controls, and older models of their machines may be very limited in the technical support available for them (or so I've heard). However due to these problems some very nice looking used CR Onsrude machines pop up on the used market regularly for very reasonable prices, some of which may be perfect candidates for a Linuscnc retrofit.

Regarding DIY: it it very well posiible to build a quality machine from the ground up. But that's not going to be cheap either if you use new components.
And a CNC router looks like a simple machine (three linear axes under 90-degree angle and a spindle motor, how hard can it be?), but looks are deceptive. It took me a year to construct my machine from the ground up. True, evening hours only, family needing time, blahblah, but I still have a LOT of time invested.

Bottom line: you should only consider a full DIY build if you enjoy the process of building a machine.

Yeah, probably not for this one. I'm probably going with Todd's recommendation of starting with an older used machine and possibly updating the controls if necessary. I'll look at doing a build once I'm more familiar with it and already have something in place.