[ameinert]

For about 15 years, I have wanted to try my hand at CNC machine construction. Back then, there was limited Internet to help me and my friends working in this space told me all of the components were well out of reach price-wise of the average guy, so that was just one more crazy dream put on the backburner. (It went back there with home foundry, welder, mill, lathe, homebuilt airplane, a few software ideas.) I also did not have the space to build any of this.

Now, the prices have come down on these things and the Internet has allowed me to learn that there are gobs of crazy guys like me that want to do, and have done, all of the same things. I built myself a dedicated, heated and air conditioned workshop (aka "man cave," "hiding place," "mad scientist lab" - 36' x 20' with 9-1/2' ceilings) to feed my insanity. I am an engineer with extensive CAD knowledge and 20 years of work experience but not as much real practical experience turning ideas into reality and I use the phrase "I don't know" more than anything except possibly "Sorry, honey." I welcome all help and ideas. I have been impressed by the things I have seen people on these forums do and with everyone's helpfulness.

My objectives with this machine:
1. Primarily will do woodworking projects - plywood, hardwood, but I also want to cut large foam pieces (think single seat aircraft fuselage mold plug and wing sections). While I am at it, I want to be able to machine aluminum, but I understand there are also limitations.
2. I need it to fit in my shop and not be the only thing that fits in my shop.
3. Regarding number 2, I can make this happen more easily if I design the machine such that the table is also useful as a work table when I am not committing CNC so my concepts have started gong this direction.
4. I want 4 axis so I can cut more than just profiles and one side. See comment on aircraft fuselage.
5. Budget. I know this area is sensitive to some but I talk pretty freely about it. The Probotix guys here in town have told me that they will put together 4 axis electronics and motors (1150 oz. in) for me for about $1800. It's worth it for me to not have to try to manage all those little electrons that I can't see. I have run through about 8 concepts and all put me between $6-10k for the total cost (including the previously mentioned $1800). I would rarther stick closer to 6. The alternative is to delay it. I have a bonus check, tax return, and I am selling a motorcycle to help me finance this year's insanity but the wife is and kids are claiming some of that.

Where I am:
Configuration:
basic moving gantry style with Y straddling table and connecting outside of the X so I can hide the rails and be able to sue the table for other things.
Size: total machine envelope will be around 11' long (x), 6' wide (50' of actual Y capacity), about 8' high to accommodate 22' Z travel and room to spin something on the 4th axis

Materials:
Rails: I was planning to go with Pacific Bearing IVT V rails and carriages, but if something cheaper and capable comes up, I am open. I am excited by this prospect of 18' Thompson rails that I saw posted yesterday and have contacted that guy to see how I can snag those. It seems like the CNC Router Parts bearing carriages and cold rolled steel end up costing almost as much as the IVT stuff, at least not enough price difference to send me that way and they will not have the same strength.

Core machine Body:
Current thinking is aluminum extrusions. I had 80/20 buy me lunch and bring their demo van, but I am leaning toward the Misumi extrusions. Another guys said they make the starightest extrusions and they configure and deliver really fast for you. They also fit the IVT rails. I have also kicked around steel weldment, but that means buying a welder and learning to weld first. I want to do both but cost and time will create a delay.

Machine base:
This is where I have gotten pretty hung up. I have created CAD concepts like Joe's Hybrid using UniStrut, 80/20, LVL (laminated veneer lumber - think really thick plywood), steel I-beam / tube (like MechMate), MDF, plywood/foam sandwich, fiberglass sandwich. All of these are expensive. Welded steel seems like it may actually be the cheapest option once you make the structure stiff enough to support what ends up being a 330 lbs+ Y-Z unit. I found that all the connectors add up really fast on the options that require no welding.

Now that I have consumed the whole thread with one post, I am interested in people's input. I will post some CAD images later.

Thanks,
Andy

[prcdslnc13]

subscribed

[oneturboneeded]

Hi Andy sounds like you are in the same boat as me. Iv been looking at cnc for 10 years and am making the jump now. I have all electronics and metal to make my 5.5'x12'x2.5' beast. I'm building it to help me with my cozy IV and to do the molds on my new 7 passenger twin design. O and as soon as I get it going I'm going to 5 axis my machine. Looking forward to seeing your progress.
P.s. What plane have you built? Whatcha going to build?

[ameinert]

Thanks for the reply, oneturboneeded. I have not built any kind of plane, yet. Job, kids, money, personal physical dimensions, and health have all conspired against me over the years. I have flown the Europa and Glastar, love the Glasair, considered the Velocity before it got expensive (and probably better), kicked around Van's, Sonex and more recently Onex and Zenith CH750. Also like the wooden designs like the Falco and smaller Corby Starlet. The Vision (and Corvair Cruiser) intrigues me for its design flexibility, but I am concerned that I have not seen many constructed. I had stock options to pay for any of these back in the 90's but I held on too long and they melted away with the tech crash in '99 and now domestic things consume most of resources.

Have struggled a little with my mission - I would love to do all of it - aerobatics, cross-country, instruments, floats, etc. but really afford that many planes. More recently, I have had eyesight issues that make me think I could end up w/o a medical so the CH750 joined the list. Would love to do a single place like Onex or RV3, but my kids have insisted I be able to fly them, too. Another point for the CH750.

The "Beast" could actually help me with just about any method of construction - wood, composite, even monocoque aluminum, just not steel. Zenith uses an Onsrud CNC router to cut their aluminum sheet parts.

I tend to over-analyze things and get analysis paralysis, especially with large financial outlays and life-risking things like planes.

I look forward to hearing your progress reporst on both fronts.

Andy

[ahren]

Andy,

We'll be watching your build to see how it goes! I'm interested to know what kind of quote you got for the IVT system -- our linear system has been quite competitive with other commercial systems, but their system does look nice. Regardless of what you go with, we'd be happy to help with rack and pinion drives to move the beast -- our Nema 34 version is a good match to those Probotix motors.

Best of luck!

Ahren
CNCRouterParts

[ameinert]

Ahren,

I like what you have been doing and your rail and carriage system is very cool. I go to your site quite often. I have already downloaded several of your STEP files and included in my CAD models, including both versions of the rack drives. In some of my concepts, I use 3 of them - 2 for X and 1 for Y.

You'll probably end up with some of my money.

Andy

[ameinert]

I received a quote today from the distributor for PBC Linear's IVT rails and carriages. Linear Guides Integrated in Aluminum Extrusion Rail Systems, SIMO, Linear V Races, IVT
I inquired specifically about the IVTAAF extrusions, which the literature indicates are the strongest version. Here is what they quoted me:
2 @ 11' long rails - $150 each
2 @ 67" long rails - $77 each
1 @ 48" long rail - $55
8 carriages with wheel covers, lube holders and side cam adjustment - $88 each

I thought that seemed pretty good.

I had looked at going with CNC Router parts extended carriages as an economy option and I like Ahren's products. Those would be about $67 per rail. If I look at the cold rolled rails to mount them on, Fine Line Automation has them drilled for mounting and that comes in at $100 for 5 feet of it, which means I would spend $200+ for each X rail and another $100+ for each of two Y rails, so the IVT stuff comes in cheaper in that regard. I could just buy the cold rolled myself and come in cheaper, but I like the IVT system.

I am leaning toward a steel structure similar to MechMate rather than the aluminum extrusions, but I have been pretty schizophrenic on this whole project so far so who knows.

Andy

[ameinert]

Many months have passed since I started this thread. In that time, I have iterated my design dozens of times, built many 3D model variations in Pro/ENGINEER, and started my build. I will add many posts later to document the process and the various issues I have encountered so far. It has been a fun and interesting process so far and there is still a long way to go.

For now, here is an update on the IVT rails and carriages. Since this is a new product, the pricing was not clear to those actually providing the quotes, so my numbers changed very significantly by the time I actually got a final one and bought the parts.

Here is what I actually acquired, including actual costs:

2 @ 11' long rails (3352mm) - $240 each
2 @ 67" long rails (1700mm) - $121 each
1 @ 48" long rail (1220mm) - $87.65

I changed my design to only use 5 carriages for now. I may change my mind later but these thinsg are really massive and I may be over-built already.

5 carriages with wheel covers, lube holders and side cam adjustment - $139.97 each

I think these were a decent price and they are good quality. One consideration when working with these things is shipping. My 11' long rails are apparently too big for normal UPS or FedEx type shipping, which meant they told me I needed to use "LTL" (Less Than a full Load of a semi) shipping or make other arrangements. LTL would be anywhere from $250-600 for this stuff so I opted to pay a guy $230 to drive to the factory to pick them up in his pickup for me. Fortunately, I am only 100 miles or so from the factory.

[ameinert]

Here are a few pics I took of a carriage and (short) rail together with a tape measure included for scale. The I beam in the pic is 12 inches deep. It is one of my two primary X axis support members. The rails will be bolted to this beam in relationship to the beam as shown in 2 of the pics. However, the beam will be on edge so this whole assembly gets rotated 90 degrees to the orientation shown. The baam will sit on that white pad shown in the right side of the first pic. I'll explain the reasonf or that pad and the process for mkaing it in a later post.

[ameinert]

Way back last year at about the time I started posting, my design concept looked like the first two pictures here. My idea was to use the straight beam type gantry so that I could avoid the issues involved with making a tall gantry.

I thought if I could put together a base unit like is shown in my first 2 pics here and then just mount it on some kind of base structure, I would be good to go. My plan was to use 80/20 type aluminum extrusions for the core machine because of their versatility. Man that stuff is expensive. (I had the guys come with their demo van and they gave me a price list, CD, paper catalogs, and bought me lunch.)

It also is not the standard I had imagined. When I started looking more closely, I found that there are gobs of companies that make these extrusions (80/20, Misumi, Bosch just to name a few) and they are all different in their slot dimensions such that they are not interchangeable. The IVT rails I bought are designed to fit the Bosch extrusions and do not directly fit 80/20.

I then started looking at all the different options for bases. My approach was to make 3D models of the things and compare with my design requirements: 1) strength and stiffness 2) cost 3) "general design workability"

I actually built the price of components into each 3D model piece using actual price lists and when I duped it to a drawing, Pro/E gave me a full BOM with the item prices and total price. I liked the idea of using UniStrut/SuperStrut like Joe's Hybrid machines so I tried a couple of variations of that idea, including using the SuperStrut Perf channel, which seems a better choice than the stuff you use to hang pipes and wires. They were iteresting but the Perf stuff is pricey and the non-Perf stuff gets more expensive than I imagined once you start putting in all the fittings and bolts and make a structure adequate for this big machine.

I also tried a complete 80/20 type base, which was just hideously expensive, partly because the aluminum is not stiff enough for this big machine unless you triangulate the structure alot, use the big sections, and/or add extra intermediate pieces. That thing ended up at like $7k just for the 80/20 stuff.

My more interesting ideas were using SturdiBilt racks (those heavy steel shelf units you see in factories and Home Depot) or using engineered lumber beams. The lumber beams had to be heavily supported or freaky thick to get the stiffness I wanted and wood was always kind of a last choice for me for a serious machine anyway. Cost was decent though. The SturdiBilt is a really nice cheap option because there are loads of surplus items out there, but the dimensions were not quite right and getting into the machine to load material and stuff was going to be hard.

In the end, I found a picture of a Mech Mate style machine on the Miller Welders site that caught my eye. The idea is a simple base made of square section steel tube and I beams. The X axis is 2 big I beams.

I decided I wanted to be able to use this machine table as a usable shop work space when not routing because this thing is huge. So I decided to mount the IVT rails in the web part of the I beam so it would be out of the way. I'll include some images in a later post.

also learned from lurking on this site that using a shaft to cross the 6 foot divide to connect the two X carraiges was not workable as shown in the images, so I will be using 2 motors instead. I'll let the controllers deal with keeping things in setp.

Andy

[PaulRowntree]

also learned from lurking on this site that using a shaft to cross the 6 foot divide to connect the two X carraiges was not workable as shown in the images, so I will be using 2 motors instead. I'll let the controllers deal with keeping things in setp.

Andy

Andy, this looks great. You said the I-beams were 12" tall : what is the width and wall?
What was the problem with having the X-drive shaft running across the whole gantry?
Cheers!

[ameinert]

The I beams are "standard" 12x19 (12 inches tall by 19 lbs/foot). That means they are supposed to be 12.16" tall with a .235" thick web (the part that connects the top and bottom of the I) and .350" thick flanges (the top and bottom pieces), Flanges are 4.005" wide by standard. The beams I bought actually ended up being .125" different in height. They came from the same supplier (local steel distributor) but were made by different manufacturers who obviously did not follow the "standard" very strictly. I did not realize this until I had done a lot of work so I could not really return them and complain. They were also bowed across the width. I have had a few such issues that I have had to address, which I will post about as I get the chance.

Regarding the cross shaft, my gantry will weigh over 350 lbs. Some versions I have drawn up came in close to 600 lbs. So, if there is one big motor driving that weight thru a 6 foot long shaft, that shaft is going to twist a lot as it transmits the power, so the far side of the shaft will be behind the near side, where the motor is. I never actually ran the numbers on it, but this is what other experienced builders on the site have communicated and I believe them. Now that I have run some motor calculations, I need the power of 2 motors anyway so it's an easy choice for me.

[ameinert]

My design has settled conceptually but there are some details that are still evolving. I am attaching some images to show the final base design.

The machine base is composed of 2x2 square steel tubing welded together to form series of ladders on their sides. At the end of the ladder "longerons" are small steel pads with bolt holes drilled into them.

These ladder like pieces bolt to the legs, which are made from 5x16 (5 inch high by 16 lbs/foot) steel I beams. (see second picture for leg detail) The top and bottom of the legs are welded-on 3/8" steel plates (laser-cut pieces about $15 ea.) In the bottom, I have a single hole to attach the machine mounts.

The X rails will mount on the big I beams (see earlier post), which will sit on the pads on top of the legs. (See third image - rails not included in image.) I needed to be able to make the width of the machine adjustable to account for potential error in my gantry length, so the tops of the legs have to provide some adjustment. One of the I beams is fixed in place while the other will be able to be slide back and forth during assembly. To allow that adjustment, see the second picture, where there are slots in the leg top plate.

The legs sit on adjustable machine mounts. I used Mighty Mounts (about $35 ea). These specific ones are 5" diameter, have 3/4" of height adjustment, and support 500-2000 lbs. each. See the second picture.

[ameinert]

Now for the overall machine design. I am attaching some images. The first image is a pictorial orientation and I have included a mannequin model to provide scale. This thing is 11 feet long, about 6 feet wide, and between 8 and 9 feet tall, which is why I refer to it as "The Beast". I am still reworking the Z axis design and I have reduced the height quite a bit by relocating my motors. I had one on top of the Z axis with direct drive to an Acme screw but that motor would have run into the ceiling so I am working on moving it down low using a belt drive as shown in the image.

I was originally going to make the machine 13 feet long but that dimension really crowded my shop. I am glad I reduced it because it is tough to handle those big I beams and the long ladder pieces by myself. Two more feet would noth ave been manageable. My shop is 36 feet by 20 feet and has about a 9' 5" ceiling. I move the big stuff around using an engine hoist I bought, which was a really good choice.

My shop is a dedicated building that I had built in my backyard with in-floor radiant heat and nice lighting because I want this to be a great place to work on projects like this machine and the plane(s) I want to build. I had a window AC unit that caused too many problems so I am probably going to put in central air next year.

My gantry is moving, powered by two 1150 oz-in motors on each side. I plan to use the CNC Router parts NEMA 34 rack drives. The racks for the X axis will be mounted teeth-down under the top flange of the large I beams. They are the long thin green parts in the picture.

Running on the X axis carriages will be 4x4 steel tube uprights. I will actually make some plates that will bolt onto the carriages to extend their length and height and then bolt the uprights to those. I need the extra length so I can also mount some angled pieces to provide stability for the uprights. These will be adjustable pieces so I can adjust the upright angle. They are not designed yet but I have a basic plan.

In order to provide the 22" Z movement, I need to provide a long moment arm so I will use 2 Y axis rails, one mounted above the other, several inches apart. I was concerned about being able to keep these 2 rails parallel so I am planning to make a large plate with the holes cut in them using a friend's small CNC router. I will actually have him cut the thing from MDF in small puzzle pieces that I will attach together. I will then add full-size thin metal laminates on each face of that assembly to make a single stronger/stiffer structure. I will also stiffen that whole thing with additional bracing to resist the loads.

The second picture provides a straight on view. The large wood grain piece is used to provide additional stiffness to keep the X axis I beams from racking. I may actually make this from crossed pieces of steel angle iron or something similar instead.

Obstructed by this wooden piece is the 4th axis. The design is such that I plan to be able to remove the table top and use this rotary axis to be able to turn very large foam items or smaller items in harder materials. However, that can require the Z axis to extend so far that things will flex a lot so I have to limit that. I have run some numbers on this thing to try to keep flexing of the machine down in the thousandths but I may still have to make some changes as this evolves in iron.

This has been lots of fun so far and even if it is not entirely successful in the end, I will have built some skills, some confidence, and had a great deal of fun, not to mention the great looks from the neighbors peering into my yard when they see sparks flying out the shop door. (No worries. They all have interesting hobbies, too.)

[acondit]

I see a lot of people trying to achieve long Z travels without fully analyzing their problem. It doesn't do any good to have 22" of Z travel if the bottom of your gantry and/or the bottom of your z-axis can't clear 22".

[prcdslnc13]

I think you missed the point of the long z. The idea isnt to be able to machine 22" thick parts laying on the table. Its to remove the table and be able to access the 4th axis all the way to the center point. If you look at the model in the second picture above your post youll see a back plate behind the wood grain piece. This is what the long z will be accessing.

[PaulRowntree]

I see a lot of people trying to achieve long Z travels without fully analyzing their problem. It doesn't do any good to have 22" of Z travel if the bottom of your gantry and/or the bottom of your z-axis can't clear 22".

About 12" of his long travel is used to reach the centre of the 4th axis, which is slung below the table top.

[acondit]

That's fine. In my book that counts as part of clearing the 22". But if you build a gantry that only clears the lowest part of your machining surface by 4", then have 22" of travel, it is wasted.

[acondit]

Maybe we should have an FAQ with some starting guidelines for people. Like what is the basic calculation for necessary Z travel?

Minimum Z travel should be the distance from spindle end when the spindle is lowered to reach the lowest extreme in machining depth with the shortest reach tool to the spindle end when the spindle is raised to the point where the end of the longest tool is as high as the lowest point of the gantry and/or z-axis structure (within the cutting envelope).

This would not cover things like someone building a z-axis to cut mortise and tenon joints where the z-axis spindle can be rotated 90° to the direction of z travel, but it would give a basic guideline to someone building a simple router.

For someone building a machine with a 4th or even fifth axis, their needs and requirements may well vary. If the 4th axis centerline were coincidental with the table surface the basic rule would provide adequate guidance and if the 4th axis centerline is below the table surface, the travel would just increase to reflect the distance from the table surface to the 4th axis centerline.

What do you think?

[ameinert]

I didn't think I would set off a conversation on such a simple thing as a Z travel dimension.

I am not trying to mislead anyone with my dimensions - this project is for my own edification, not for sale or anything, so I am not really going to any special effort to make the dimensional details super clear. Essentially, I want to be able to cut something about 4" tall on my table top and I want to be able to "turn" about a 44" diameter with the 4th axis.

In the images I have included, the actual distance as modeled between the 4th axis center and the bottom of the router is just under 18" so it will need to change to get the capacity I want. I want the bottom of a common tool to be 22" from 4th axis center to that same point when the axis is raised to its zenith. The distance from the bottom of the router to the table top as modeled is 5". However, the router I built into the model at this point is just a cylindrical space claim model that I will edit some more before I finalize. I haven't actually measured (or bought) the spindle I will use so the dimensions are approximate and I actually included about 2" of tool length into that cylinder.

Also, the 4th axis was put in the most convenient location and is a rough concept model. I will add one but I am still not clear on how I will construct it. As I have actually started constructing this thing, I have more clearace above ground because of the various methods I have used to level everything, so I will probably move the 4th axis down a couple of more inches. One of the great things about a CAD system like Pro/E is I can model it and iterate the dimensions quickly to arrive at my final design.

Right now, I am still messing with Y and Z axis details so the final geometry will vary from what is here, but I made a point of buying rails large enough to allow for 22" of Z movement and my intention is to make the design accommadate that.