[tonofsteel]

Hello All,

I have just joined the forum, and briefly read over some posts that sound similar to this one. However I have a bit of a different situtation and would like some advice if possible.

I am currently a student taking electronics engineering, and for a design project would like to build a milling machine. The electronics involved I will be able to design and build myself, so as far as stepper controllers go, I should be able to take care of it.

So the hard part is fast becomming what to use to build a CNC machine. It seems that assembling the machine is not that difficult provided you have some basic welding and fabricating skills, but finding the right parts is a challenge.

The primary use for the machine will be to machine wood and aluminum. But I would like to also machine steel. I notice that machining steel requires alot more design consideration. This is probably an unrealistic goal, but I would also like to mill pcb's. So it will have to be accurate, strong, and durable.

The desired cost for the project is, as cheap as possible. Hopefully between 1000 and 5000 dollars. So for now any ideas or requirements that the experienced people here have, just throw an estimated cost.

For the motion control I want to use a ACME leadscrew, since ballscrews are exponentially more expensive. The nuts used will be a higher quality anti-backlash type, to try and squeeze a little accuracy out of it. Would this setup be adequate for milling steel, and what kind of precision would it provide practically? Would it be adequate for milling pcb's, would it be adequate for milling pcb's that dont have ultra fine trace spacings?

How much oz-in of torque do I need to get for the stepper motors? I was going to use linear bearings running on a linear shaft that is end supported. But I am pretty sure that its going to have to be supported by a rail as well? As far as the frame of the machine goes, I have alot of scrap steel that can be thrown into it to make it massive and strong.

The style that I want to build is the overhead type.

So if anyone here has some insight as to the requirements for the different parts (ie lead screw/ ball screw, diameter of screw, diameter of linear rails, bearing requirements, stepper motor oz-in requirements) I would really appreciate the information before I buy components that are useless.

For this project plan B is to be able to just mill aluminum and pcbs, and plan C is to cast aluminum parts instead of milling them.

[DareBee]

ACME screws are adequate for straight-line milling or point-to-point CNC (if your table locks or doesn't float) but you will not be able to contour without great difficulty and inaccuracy.
You could build backlash elemination system by using springs or weights to always keep the pressure of the lead nuts on the same side. This would be satisfactory for a small machine doing very light work.
Realistically, if you are building this to actually be a usable machine (not just something to play with) you will have to use ball screws and there is not much chance you will do it for 5G (even using ACMEs). Then again when I picture a mlling machine the smallest table I see is 9 x 49. You may be able to do this in your alotted budget by making a 6 x 6 machine or something.

[tonofsteel]

Thanks for the info!

My next question is, where are some places that offer good prices on ballscrews and steppers, and the other components needed. I have googled for about 10 hours now, and found some good sources. The price varies so much from place to place, and to find some good deals is fairly hard.

So I think I am going with plan D. I have a new drill press. I am going to by a x-y axis work holder, and get some basic tooling and do some manual milling on it for awhile. Hopefully I can make some of my own parts, such as bearing holders and end plates for the various parts. The runout on this cheap press is probably horrible, so I am hoping by replacing bearings or manually compensating I should be able to get some satisfactory pieces out of it.

So i guess slowly I will source parts, and work on the pieces, to make sure i get the right stuff.

[DareBee]

That is a good idea, although the spindle of a drill press is not designed for side loading. You can get some simple "light" work done by that method if you are patient.
Older used milling machines can usually be found for about $1500 and often have readouts on them.
IMO the best way to make a professional and inexpensive CNC mill is to buy an old (80's) CNC that is mechanically sound (and are usually real cheap) and throw out all the controls and build your PC system on top of that. This machine wont compete with a new CNC in terms of speed, but will be great for 1 offs and other such things that don't need high feed rates or high spindle RPM.
Only problem with these machines is that they are usually giants.
A vertical bench mill is 2G at KBC Tools and the miniature mills start at 1G.

[Mcgyver]

tonofsteel,

pm'd ya with link to a thread here on linear motion & control.

As DaraBee said the drill press bearings aren't intended to take the sideways load, and the morse taper in the spindle won't hold very well for this. it would be a shame to spend finite shop $$$ on something who's success is tenative at best. I have never done this myself, so maybe i am being too quick to trash it, but heard enough people complain about its futility over the years that i'd encourage you to really check it out first.

DaraBee, can you quantify giants - what sort of machine are you talking about? bport or xlo size is ok, but if i came home with something larger than my house i would suffer domestic grief. This approach has a lot of merits, but with the price of manual mills in decent shape going up and up can you really get a good one for 1500?

[DieGuy]

here is one going in about 2 days for $1600 currently with ball screws installed already
http://cgi.ebay.com/ws/eBayISAPI.dll...509717524&rd=1

If you keep your eyes open lots of times you can see similar ones on E-bay. The issues are usually shipping or needs to be picked up so locale is important.

[Mcgyver]

thanks DieGuy for the link - machine looks good but two days left - isn't it the last few hours that count? Personally I'm bearish on ebay - imo ebay pricing has to be a fraction of value to compensate for risk (buying sight unseen, limited recourse and not having the vendor any kind of stakeholder in my locale) + aforementioned hassles. i still see NA made vertical mills going for 3k cdn & up if the don't need to be scraped (or maybe scrapped)

[tonofsteel]

Alright, so I have learned alot so far. I have been looking into this for awhile, and although there are many engineering challenges to creating a CNC mill that can work on steel I think I am up to it. This semester at school we had to create a microprocessor from scratch, and do a AES encryption implementation. Probably many people here do not know alot about those two subjects, but like you, I had no idea how to do it just a few months ago. So with a bit of information, and alot of hard work I think I can come up with something thats workable, mabye not to my ideal specs, but it will be fun.

So, in my current moment of over optimism I want to attempt this. I have been down the "take the cheap way out" road many times (ie drill press mill) and I know that its a recipie for disaster.

I am going to make all frame pieces, and all bearing holders, etc myself. I just need to know what are the requirements for the steppers, the linear rods, and the ballscrew. What diameters are required for this application? As well the linear shafts will not be end supported, they will be mounted on rails, what diameter do those have to be?

I like the idea of buying a milling machine, that is of an older vintage, and that would serve my purposes perfectly. However I live in Saskatchewan, Canada, which is also known as the "gap". To those unfamiliar, we do not have technology here, and we do not have access to cheap parts. For me to buy a milling machine of that vintage off e-bay, or from an auction would cost at least as much as the mill is worth in shipping. I might as well build my own and gain the experience. As well, if i mess up on this attempt, the ballscrews, steppers, etc. will be good for the next time around.

So all the smart people here, what parts do I need in what sizes? The frames, and other supporting structures will be designed to the best specs I can using information from some fellow mechanical engineers. As they do not have the experience that all of you have, I want to know your input for the critical expensive components, and the rest will be built over-spec what they would need to be, just to be sure.

So please, what parts in what size/diameter/power to handle a max load of steel at max dimensions of 2ft x 2ft @ 100 to 200 lbs. (high end extremly generous)

[DareBee]

tonofsteel,

DaraBee, can you quantify giants - what sort of machine are you talking about? bport or xlo size is ok, but if i came home with something larger than my house i would suffer domestic grief. This approach has a lot of merits, but with the price of manual mills in decent shape going up and up can you really get a good one for 1500?

One of the smaller units that comes to mind is a Maho and they kinda (very loosely) look like a bridgeport style but about 3 times the size of a 9 x 42.
Another machine I see going cheap here (when you can find them) are the Excello CNC knee mills (for you Americans these machines were a Canadian designed and built kneemill that were very common in the 80's(at least in Ontario)). I have seen these sell for under 5G, only problem is you can't get parts for the head and if it goes you have to make an adapter plate to fit a Bridgeport style head, the Excello head swivels are opposite to the Bridgeport.

Tonof...I've never built a VMC before but IMO (and automation experience) 1 of 2 things will happen:
1- You will build the machine for less than the cost of used one but it will be somewhat subpar (compared to lets say a Fadal (mine weighs 10,000#) with 40x20x28 travels a 10,000 RPM spindle, industry standard tooling, rigid tap, 4th axis, coolant and cooling systems, 21 tool ATC).

2 - You will be able to incorporate all the necessary and time proven features but it will cost you WAY more than buying a good used machine.

I would think shipping could be arrainged for less than 1G (from S Ontario)on machinery.
Here is a link of PRIME candidate for what you are looking for. http://www.machexch.com/specfs/htm/7905? A 24" Y axis is something you will not find in a purchased machine without being a very large machine.

I point to point CNC'd a 1m cubed (travels) boring mill a few years back. This machine had a central drive motor that selected individual axis via clutch packs. This machine only has ACME screws but it was originally an NC machine built in '77. The original PANELSSS had 100s of porcelin fuse holders anf these massive leaf steel relays etc LOL), anyway, I was stuck with this machine my *#%$ partner bought it and it was either scrap it (20G bye-bye) or make it usable. I didn't want to spend the 20+G it would cost to servo and ballscrew the axis so we made a point to point system using the original drivetrain and glass scales. I guess my point is that this system cost us over 50G by the time we were done.
I do not want to discourage anybody, I absolutely love the pride I feel in a job well designed and built by myself.
I just want to share my experience and say that it is really hard (if not impossible) to top what a factory makes -in cost effectiveness as well as the amount of years they have spent improving there 1 and only product.

:cheers:

[Mcgyver]

I see your point - there's a middle ground: too big for most amateurs, to slow/old/ineffiencient for biz.

I've got a manual xlo, old timer told me mine was one of the earlier ones (based on serial number) and that means much heavier castings. I believe they were also made in the USA as well as Canada - they are great machine imo. Hope the head doesn't go although not loosing sleep because i figuring belts, gears, bearings etc are standard parts and I'll be able to make parts for the vari-speed. a cnc xlo with burnt out controls would be a perfect find!

[tonofsteel]

I have been looking at some linear drive system alternatives. I am wondering about mabye using a chain drive to do the positioning. Backlash may be a problem, if you put a chain around two rollers in the front and back, and attach your platform directly to the chain, wouldnt it be tight and have little or no backlash?

Another option is a roller pinion system as shown here: http://www.nexengroup.com/rp/

I could not get a quote yet on what the costs would be, but if it is alot cheaper than ballscrews would it not be a viable alternative?

I was thinking of a belt drive, as is similar to printers, but that would probably not work at all since the loads involved would be too much for a belt to handle without flexing too much.

Mcgyver's previous posts and ideas have got me thinking out of the box. The only problem with having a feedback system for positioning is that it defeats the purpose of the stepper, and feedback systems have their own set of problems. The electronics are much more critical, and oscillations start to become a problem due to the high amplification of error signals.

Up until now I thought there were three viable and usefull options for linear drive, threaded rod, leadscrew and ballscrew. What else can be used? There has to be a linear drive system that can be made for a considerably cheaper cost, with excellent accuracy.

[musicmkr]

tonofsteel:

It takes a lot of time, patients, skill, and money, to build a milling machine that is accurate enough to be anything more than a boat anchor. Before you go spending a ton of time (and money) building your own machine there is another option you should consider.

Since it sounds like you want your project to be based on the electronics/CNC end of this machine, you should look into converting a mini or micro mill. Harborfreight sells the mini for $500, and the micro for less than $300.

That would leave you to do just the CNC conversion, and in the end you know that you'll have a useable (small) milling machine. Several people on here have converted these, so there is plenty of info on what you'll need for motors and other hardware.

[Mcgyver]

oh man, i can't sell anyone on my hydraulic fluid capstan idea....guess i'm going it have to build the sucker and see for myself. I believe (until proven wrong) that hydraulics might prove to be the holy grail of linear movement.

Don't to quick to discard a servo set up. Bigger mills would likely use servos instead of steppers, so all i'm doing is saying move from a encoder on the back of a servo (driving a ball screw) to a linear encoder telling a motor on a pump how fast to go.

Whether anyone including myself ever builds the fluid capstan aside, servo/encoder set ups are mainstream, so a linear encoder shouldn't be too much of stretch

speaking of stretch, that's the problem with chain. I suppose you could preload it, but there are so many little bearing contacts, each with its own clearance that it might be tough to keep it an accurate length as the forces on it change. dare I say, maybe with a linear encoder? ie just use the chain to move the thing as a component of it linear accuracy. or maybe in a capstan arrangement so that you could preload in both directions?

[lerman]

Alright, so I have learned alot so far. I have been looking into this for awhile, and although there are many engineering challenges to creating a CNC mill that can work on steel I think I am up to it. This semester at school we had to create a microprocessor from scratch, and do a AES encryption implementation. Probably many people here do not know alot about those two subjects, but like you, I had no idea how to do it just a few months ago. So with a bit of information, and alot of hard work I think I can come up with something thats workable, mabye not to my ideal specs, but it will be fun.

So, in my current moment of over optimism I want to attempt this. I have been down the "take the cheap way out" road many times (ie drill press mill) and I know that its a recipie for disaster.

I am going to make all frame pieces, and all bearing holders, etc myself. I just need to know what are the requirements for the steppers, the linear rods, and the ballscrew. What diameters are required for this application? As well the linear shafts will not be end supported, they will be mounted on rails, what diameter do those have to be?

I like the idea of buying a milling machine, that is of an older vintage, and that would serve my purposes perfectly. However I live in Saskatchewan, Canada, which is also known as the "gap". To those unfamiliar, we do not have technology here, and we do not have access to cheap parts. For me to buy a milling machine of that vintage off e-bay, or from an auction would cost at least as much as the mill is worth in shipping. I might as well build my own and gain the experience. As well, if i mess up on this attempt, the ballscrews, steppers, etc. will be good for the next time around.

So all the smart people here, what parts do I need in what sizes? The frames, and other supporting structures will be designed to the best specs I can using information from some fellow mechanical engineers. As they do not have the experience that all of you have, I want to know your input for the critical expensive components, and the rest will be built over-spec what they would need to be, just to be sure.

So please, what parts in what size/diameter/power to handle a max load of steel at max dimensions of 2ft x 2ft @ 100 to 200 lbs. (high end extremly generous)

Don't give up on the used machine idea. I bought a used Bridgeport clone for $250 from a used equipment dealer. Try thinking outside the box as far as transportation is concerned. I wound up taking my machine home in a dump truck and unloading it with a bucket loader. In my case, shipping was free because a friend with a dump truck owed me a favor.

Someone suggested that you could use a small car trailer -- try renting one.

Think of it as an adventure -- drive a few hundred miles, get the seller to load it onto a trailer, drive a few hundred miles back and then unload the thing.

Good luck.

Ken

[tonofsteel]

Alright, I am interested. If i can find a good used mill for 250 to 1000 dollars and then do improvements to it, I am sold. However, I am limited in which cities I can drive to. Winnepeg, Calgary, and Edmonton areas would be preferable.

So, who do I call, where do I look? I have not seen a "Mill Trader" here yet, so there isnt much listings that I can go browse. Any ideas? Also does anyone know of any mills available in the above places?

Thanks!