Build or buy?

[gearsoup]

I have been a machinist for 10 years. About 7 years ago, I built my first CNC router. It was a small affair, but it was an educational experience of immense proportions. Time went on and I sold the router to open up some cash flow.
Now here I am in a new home with a job that pays well. I have a large garage and work shed from which I can do projects. I have been looking at building another one, this time much larger (4'x4' working area). For myself, there is no question about it, I will design and manufacture it. Here is the point of my post:

My father-in-law wants me to build him a small (2'x2') CNC. He is approaching retirement age, and for some reason isn't able to shingle roofs or redo kitchens very quickly anymore (he is 63, and has been swinging a hammer longer than I've been alive, lol) He wants something he can do sighns, engrave cabinets, do small gift shop/hobby work.

So, I need to examine the costs of making a foolproof CNC that anybody can run with very little instruction (he is not computer savvy by any means), or the costs of purchasing a ready made unit. If I made it for him, the costs to him would pretty much be just enough to cover the materials, plus a tad little bit for labor. Pretty certain that it will be cheaper to build it, but if one were to figure on paying wages, what would the answer be then? (assume 20 an hour for making it)

The biggest thing (in my mind) is that it needs to be 100% foolproof and reliable. Granted, you can only make a CNC as reliable as the operator, but we take a lot of things for granted with our own creations. If things have quircks, we learn to live with them.

Programming the CNC would obviously take some time to teach, so I would handle that side of things until he was comfortable enough.

[Bubba]

My .03 (inflation:})
Being of the type that believes in making anything I can rather than buying, you can guess my answer.
BUT there is another side to it also. As your experience has taught you, building means that when you have a problem, you will know HOW the machine works and is made. This allows you to make the repairs quicker. Think of it this way, if you can involve your dad in the building, he will have a learning experience and this may pay off in the future when things need fixing! Also, think of it as a "giving back" as I am sure your dad like mine tried to teach me many things (wish I had listened better)

But good luck however you decide.

[hemsworthlad]

Hi,
Bubba makes a good point involving f.i.L makes a lot of sense and could save you some technical support calls just when your in the middle of something you dont want disturbing on.?

Regards build of machine and making it fool proof, this will also depend on f.i.l mechanical or technical skills. The control or PC skills is just something he's going to have to develop, no way round that, but the mechanical workings of the machine can be made as easy or hard as your budget permits and/or his skills allow.

I,ve built or been involved with quit a few builds rangeing from MDF/skate bearing knock together types to all singing dancing 80/20/ballscrew/profiled linear bearings types with a mix of steel framed hybrid's in between.

I can say without hesitation that the best by a country mile based both on cost and ease of maintenance and reliabilty is the Steel framed machine with off the shelf linear bearings and cheap chinese import ballscrews.

If you have the skills to weld then steel is cheap compared to 80/20 and strong which helps with rigidty.
The linear bearings take away all the hassels of home made skate bearing setups which to be honest are just a pain in the arse and the single biggest flaw in most diy builds that i've seen.
Usually they get replaced a few months down the road anyway wasting time, money and hair.!! . . Often it works out more expensive than linear rails anyway because they struggle on clutching onto their design throwing more money at it only to come to the end conclusion that linear rails rule.!!

Imo the cheap chinese ballscrews with BK/BF bearing blocks also make it very easy/quick, reliable and accurate.
Yes I know multistart lead screws with a dumpster type nut work good but still are a substitute for ballscrews if people are honest.!! . . Imo with the price of chinese ballscrews i dont see the point.?

Budget will probably dictate the direction you go with regards to linear rails etc but from experience if the wallet can stretch to these components then this combination works absolutly beautifull together and very very accurate strong and has fool proof as can be regards mechanical build.!

One other point regards Labour/component cost's the use of linear rails and ballscrews with BK/BF bearing blocks considerably speed up the build process because they are virtually plugn play reducing labour.! . . . time is money to some.? . . . Really they dont work out that much more money in the grand scheme of a quality build and the differance in quality/accuracy is night and day compared to skate bearing/threaded rod setups.

IMO Better spending a few dollar's more at the start to get a machine that is up and working earning it's keep than giving me sleepless nights and draining my wallet with less than satisfactory results.? . . . . . plus it will give fault less service for years if maintained properly.

Cheers
Dean.

PS: Oh ye. . . To answer your original question YES Defiantly build just build it right the first time.!!

[gearsoup]

I hear what you are saying about the proper design. Believe me when I say this: There will be no corners cut on his, or my own machine. As I mentioned, I've got over a decade as a machininist, making parts that very few shops can competently do. I learned the errors of my last machine, and that gave me a solid foundation in designing these.

Chinese Ballscrews? Not familiar with those. I was going to use the ones available here: http://www.mcmaster.com/#ball-screws/=8eczuq

If I am not mistaken, those are made in the US for Thomson Industries. Granted the accuracy is .004 per foot, but one should be able to account for that in Mach3.
The linear motion side of things will be: http://www.mcmaster.com/#catalog/116/1070/=8ed0r9 Using open bearings on the Y and Z, and fully supported on the X axis

Frame will be entirely CNC machined aluminum, with a steel frame for it to sit on. Did I mention that I will be using the equipment where I work to make this machine? (large footprint 3 axis cnc's, cnc lathes, etc. etc.)

I'm just wondering if at the end of things, he will have more $$$ invested in this machine than he would have if he bought a ready made machine from one of the more trusted vendors on here. Not that I'm trying to talk myself out of making it, I just want to watch out for his wallet as well....


Seth

[hemsworthlad]

HI,

I use linearmotion bearings on ebay http://stores.ebay.com/linearmotionbearings?_rdc=1
The guy is called Chai and he's a great bloke very helpfull and honest, i've used him several times and recommended him to others who have had good service and prices from him.
The ball screw quality is ok for serious hobby use and you will probaly get all 3 screws with blocks and end machineing for the same or not much more than 1 of the expensive McMaster carr screws.?
Yes they are not the same quality but for the job you will be doing they are more than upto it, I have had some on my machine for over a year now cutting aluminium 99% of the time moving a largish and heavy gantry with no guards protecting them and really working in the worst enviroment and not a sign of wear.

He also sells the linear rails but i havent had any experience of those so cant advise.!!
You obviously sound like you know what your doing so dont want to come over has trying to teach dad to suck egg's but i would be reluctant to use unsupported rails if any length is involved.?

Hope this helps, goodluck.

[ger21]

The biggest thing (in my mind) is that it needs to be 100% foolproof and reliable

If you know what you're doing, it should be just as reliable as a commercially made one.

There's not a whole lot to a CNC router. Typically, you have the frame, screws, steppers, and electronics. The better made home built machines use the same (or better) motors and electronics as the more expensive ones.
Basically if you want more power than the cheap Ebay drives or HobbyCNC, your options are Gecko or Chinese.
Just about everyone ises Chinese stepers, but if you want better ones, pay 2-3 tomes more for Vexta's.

Bottom line, is can you build this for less than $5900? I would think so.

And if you're looking for something cheaper, I haven't seen anything under about $3K that I'd buy.

I just checked, and a 25"x25" K2 with servos is the same price, $5900

[HereinCS]

I agree with Ger21. You can build a machine that is just as good as a commercial machine. If you look at my build, I've spent $7k so far (without motor), and I'm going to get a machine with 4 servos, duel screw x axis, and the exact size that I need. All my parts are new or very close to new, so I think they'll last just as long as the commercial machines. Bottom line is you definitely can out do the commercial machines.

[cyclestart]

Originally Posted by gearsoup Pretty certain that it will be cheaper to build it, but if one were to figure on paying wages, what would the answer be then? (assume 20 an hour for making it)

If your employer allows personal use of large cnc at a reasonable rate you have a huge foot up on most of us. The shop will likely have material and fasteners on hand as well.

Total labor cost depends on your own efficiency, guaranteed a builder like K2 puts a LOT more value on their time. $20/hr ? Trips to metal supplier, trips to hardware store, dealing with vendors, planning, drawing, cutting, wiring, fixing mistakes (yep, made a few), configuration and calibration. Glad I haven't been keeping a time sheet

And then there is the other time factor. I've been waiting 3 months for an assembly from one vendor. These things happen.

If both commercial and home-built are running a g-code interpretor (like Mach3) there shouldn't be much difference in ease of use. Don't skip the home switches. An MPG would be a nice feature.

[gearsoup]

ah yes, the home switches....any suggestions on those?


7 years ago when I built my first machine, I had a rather lengthy list of www links taht had answers for most any quistion for the hobby cnc world. several pc's later, those links are all gone.....


Not certain if I will have complete access to machines to make this unit. I will ask my foreman this week what he thinks about the idea

[Matt Clara]

I recently did a lot of research on the subject, and if you're looking to buy one, I'd look into a zenbot. http://www.zenbotcnc.com/. Nobody who owns one has anything but good things to say about them.

[cyclestart]

Originally Posted by gearsoup ah yes, the home switches....any suggestions on those?

Was hoping someone better qualified was going to answer that one

I have used roller switches from the Source (Circuit City in America ?). Probably not the best but they do the job. The other option is to simply move the machine to home reference points without switches. For ease of use home switches are an improvement imo.

[wcarrothers1]

I've used roller micro switches from radio shack on my cnc for 3-4 years and they have been plenty accurate for homing and re-starting a piece between power cycles or the ocasional power outage. I've gotten into the habbit of always marking my machine x,y before staring a piece just incase the house looses power... They are mounted in a way where they can not be damaged by any dumb moves..