What's new
What's new

Homebuilt CNC in Tokyo

Bakafish

Hot Rolled
Joined
Feb 21, 2022
Location
Tokyo Japan
I making this thread so that I can keep tangential discussion about the CNC I am building out of more specific threads where I'm asking for targeted help (and am trying to keep everyone from getting side tracked.) If this is the wrong place to put it, let me know and I will recreate this in the proper location.

Just up front, as mentioned at the top, this is for a home in the heart of Tokyo Japan and so has a lot of constraints that you are likely to be unfamiliar with, try not to make assumptions as to why I'm not doing the 'obvious' as it likely isn't something as easily done over here. To compound that, I've also chosen to try and use as much domestically sourced parts as possible out of a sense of local pride. Much of the major components have already been obtained, and while they may not be ideal, they are limitations that help focus me on making the thing rather than parts shopping.

So with that hopefully non-hostile disclaimer in place, let me lay out the goal and the basics.

I'm an Engineer living in Japan for about 15 years now, and want to build a CNC that will be able to operate in my residence for non-commercial use, but that could potentially boot strap a more serious endeavor in some form of artisan manufacturing. I do not have anything specific, but I want to be able to easily machine complex metal parts, hopefully including steel. Yes, I know that's a high hope for a homebuilt, and yes I know this design will have challenges, but this is the basic goal, let's see how close I can get.

Some straight up facts that may surprise you:

The largest residential electrical service available is a 200v 60A drop. Yes, I mean 200 (split phase 100v) not 220 or 240v, and yes only 60A for the whole house, there is a software controlled breaker at the meter that will disrupt you if you pull more.

We do not have any kind of garage or basement. This is because there is strict zoning laws in most of Japan and they impose a land utilization percentage. As a common example, a residential area may have a 60/120% limit, so if you have 100 meters of land, you can only cover 60% (60 meters) with building and that building can only have a maximum of 120% (120 meters) of internal floor space. Garages and basements directly apply to that formula, so unless you are insanely wealthy you are never going to trade off living space for an enclosure for your car. In my neighborhood a tiny single car garage would add close to a half million dollars to the price of the property.

Japan is not set up for dabblers. Things that are specialized are for specialists, Japanese stick to their lane, there is not the sort of DIY or home hobbiest communities that can support large scale secondary markets over here. Combined with an ethos of paying the price to do business, it means there are few ways for normal consumers to even buy specialized equipment or parts, and if they do they should expect to pay shockingly high prices. This applies to some of the most simple things, like basic power tools. The market is small, choice is limited, prices are very high and importation is difficult and costly. This is not the US, it doesn't work anything like the same way here.

Lastly some good news, and a most surprising fact, my wife is a patient soul who supports my eccentric ways, and mostly is tolerant of sharing our house with this monster (with the implicit assumption I keep it in my workspace and from crapping metal shavings all over the house.) As a doctor she works almost constantly, and gives me a lot of rein over my reign.

So now that we have some of the shape of the environment, lets see what we actually have on hand.

This is a fairly recent shot of the design, it's name is Kaiju.

Kaiju.png


As you can see, it is a gantry type design, using Aluminum extrusions, but before you go on a rant, these are very large and specialized beams, not typical of what is used outside of Japan. I'm not saying that they will allow me to cut steel, just that you should hold on to you comments until all the facts are known. Thanks.

The base is an old 750mm x 500mm cast iron surface plate (sorry guys, we are metric over here and I've bought into it. Units of science.)

Base_Plate.png


All rails are genuine Hiwin heavy duty 25mm units. All ball screws are 28mm NSK ground units and bearing blocks. The motors and drivers are Oriental Motors Alpha series closed-loop steppers, and the spindle is a massive 30kg 1.5kw 3-phase beast made by a Japanese specialty company. I will detail these components below over time, but I've given you guys enough to chew on. Fingers crossed this won't be a pile on... :D
 
Well, you're not going to like my thoughts one bit, but here goes:

1) For the stationary Y rail mounts, use steel if at all possible. Yes, it's not as easy as using the box extrusion, but for a mill stiffness is everything.

Everything. Really.

2) Your X screw really should be down under the two rails, not above. You're asking for racking by having the drive high and the load low. Not good.

3) I can't see the detail of the X and Y ballscrew drive bearing anchors, just the stop-the-waggle end mounts. Make the drive side mounts as stiff as possible.

4) There is no 4.

5) This is where you tell me reasons why 1-3 aren't feasible.

Carry on...
 
I would suggest cnc zone .com for this discussion.

He's using 28mm ballscrews and 25mm rails. That takes it out of hobby-farting in my book. That's why I'm encouraging taking the support structure to a matching level.

[I like big ballscrews. I've got one here that's about 8" diameter in the screw (not the nut) and over 600lbs. And about thirty feet from my desk are two ancient 3" x 10ft low-precision b-screws I want to use for a "shop made" tensile/compression load tester. I'll be shocked if I ever get it done, but a man can dream...]
 
I would suggest cnc zone .com for this discussion.

I'm just trying to keep tangental discussion about the machine from polluting threads about specific challenges. I want to be able to point people to this thread so that the signal to noise ratio can be kept high. Telling people to discuss it at another website seems a little too hostile, but I will abide by the moderators wishes if they so instruct me.
 
I like the attitude going into this. The only consideration I see off the bat is your screws and rails are in direct area to gather chips/shavings. Grit on these is bad.
 
Well, you're not going to like my thoughts one bit, but here goes:

1) For the stationary Y rail mounts, use steel if at all possible. Yes, it's not as easy as using the box extrusion, but for a mill stiffness is everything.

Everything. Really.

Yes, understood. Sourcing or fabricating these structures in steel is extremely difficult over here, and I felt the aluminum would work well enough as a proof of concept if nothing else. If it is inadequate then doing an epoxy granite fill would be next, followed by strategic reenforcement with steel plate or replacement with a custom built (expensive) weldment. The fact is I can get the 150x100mm M10 extrusions delivered to my door in a day, so I'm willing to compromise for the sake of expediency. The main spar is 100x200mm and this is a pretty tiny working envelope of 500mm square.

2) Your X screw really should be down under the two rails, not above. You're asking for racking by having the drive high and the load low. Not good.

Yes, I've contemplated putting it there as well as intuitively the lower area is applying load directly opposing the tooling forces. I just wasn't sure if the rails were strong enough to deal with the racking forces and I was making concessions to loads that didn't actually matter. I will continue to weigh options. This thing has to fit in a tiny space and I didn't want to give up the Z height, so again, lots of unfortunate compromises must be made, this is one I will keep pondering.

3) I can't see the detail of the X and Y ballscrew drive bearing anchors, just the stop-the-waggle end mounts. Make the drive side mounts as stiff as possible.

All the front bearings are proper NSK angular pairs, the end mounts are just floating deep grove. My ball screws are threaded on both sides, so I could theoretically pre-load them, but that will both increase the cost and reduce the travel. It is an option as a retrofit though. I figured these are already pretty beefy, having them constrained on both sides was overkill.

4) There is no 4.

5) This is where you tell me reasons why 1-3 aren't feasible.

Carry on...

LOL, 1 is the least feasible (although likely very easy if I was anywhere else in the world.) 2 is under serious reconsideration. 3 seems like premature optimization.
 
I like the attitude going into this. The only consideration I see off the bat is your screws and rails are in direct area to gather chips/shavings. Grit on these is bad.

Yes, I plan on making a very creative shielding for these parts as well as a full enclosure for the machine. This is the skeleton in other words, just to give some idea of how it is mechanically.
 
Japanese infrastructure is surprisingly basic, I visited a while back (Boeing work trip to Nagoya) outside of the city center the level of construction quality, wiring etc was really poor, given the size and general wealth of the economy I thought it would feel more developed. It was August and the lack of AC in the KHI office was rather noticeable :) It sounds like your property prices are a match for Oslo, my workshop is all down in France for a reason.

Anyway while the big plate is cool I don't think it will do much for the actual performance of the machine. (I suspect gantry flex etc is going to be a bigger issue than anything with the base. I would do a full frame from extrusions and use the plate as a worktable. If you do use the plate for a working surface of the mill you can just mill it aligned with the router and then drill and tap it for hold down fixtures but realistically you will want a spoil board for most things so the flatness of that plate isn't a big deal. As for the aluminum extrusion alignment off the plate, the flatness of the plate also doesn't matter that much as the extrusions are really not all that straight themselves. I would suggest ground ballscrews are seriously overkill but that isn't a problem :)

Another suggestions is that unless you access to a large milling machine the end cuts on those extrusions are possibly not going to be as straight and square as might be ideal, a configuration that allows of adjustment could make your life easier.

I have designed a couple routers but ended up buying parts for one in Germany Startseite | CNC-Alufrase the machine is a bit small but rather capable within its envelope. I run a 1.5kw Chinese water-cooled spindle and it will cut most anything. That design might give you some ideas. Do not underestimate the amount of Z travel you need, if anything I have found that to biggest limitation with my machine.

Luke
 
Japanese infrastructure is surprisingly basic, I visited a while back (Boeing work trip to Nagoya) outside of the city center the level of construction quality, wiring etc was really poor, given the size and general wealth of the economy I thought it would feel more developed. It was August and the lack of AC in the KHI office was rather noticeable :) It sounds like your property prices are a match for Oslo, my workshop is all down in France for a reason.

Anyway while the big plate is cool I don't think it will do much for the actual performance of the machine. (I suspect gantry flex etc is going to be a bigger issue than anything with the base. I would do a full frame from extrusions and use the plate as a worktable. If you do use the plate for a working surface of the mill you can just mill it aligned with the router and then drill and tap it for hold down fixtures but realistically you will want a spoil board for most things so the flatness of that plate isn't a big deal. As for the aluminum extrusion alignment off the plate, the flatness of the plate also doesn't matter that much as the extrusions are really not all that straight themselves. I would suggest ground ballscrews are seriously overkill but that isn't a problem :)

Building on top of the plate simplifies all the alignment procedures and helps lower the center of gravity. I don't expect it to contribute a lot to the structure, but it won't hurt either.

Another suggestions is that unless you access to a large milling machine the end cuts on those extrusions are possibly not going to be as straight and square as might be ideal, a configuration that allows of adjustment could make your life easier.

I have designed a couple routers but ended up buying parts for one in Germany Startseite | CNC-Alufrase the machine is a bit small but rather capable within its envelope. I run a 1.5kw Chinese water-cooled spindle and it will cut most anything. That design might give you some ideas. Do not underestimate the amount of Z travel you need, if anything I have found that to biggest limitation with my machine.

Luke

Japanese are fanatically precise with made to order industrial components like this, the parts will be square and within 0.1mm in my experience.

For Z travel, I've been super concerned about this. It may not be obvious from the picture, but the work envelope actually extends well past the front of the base plate. This will allow me to clamp things to the face or position workholding on a shelf below the surface of the baseplate.
 
I'm just trying to keep tangental discussion about the machine from polluting threads about specific challenges. I want to be able to point people to this thread so that the signal to noise ratio can be kept high. Telling people to discuss it at another website seems a little too hostile, but I will abide by the moderators wishes if they so instruct me.

Did I "Tell" you ??? No, I said "I suggest".

"Hostile" ???

I suggested it, do you need sprinkles and a cherry on top on every answer ?....:cloud9:
 
Did I "Tell" you ??? No, I said "I suggest".

"Hostile" ???

I suggested it, do you need sprinkles and a cherry on top on every answer ?....:cloud9:

I'm just explaining my reasoning, not accusing you of anything. I've got years of posts on another site, and even a few on the CNC Zone, I just thought it would be bad form (and mostly ineffective) to try and push the conversation off site.

My other post here has already spun way out of control, so my attempts at keeping things focused seems a bit fruitless. So I could use the cherry, please keep the sprinkles.
 
Building on top of the plate simplifies all the alignment procedures and helps lower the center of gravity. I don't expect it to contribute a lot to the structure, but it won't hurt either.



Japanese are fanatically precise with made to order industrial components like this, the parts will be square and within 0.1mm in my experience.

For Z travel, I've been super concerned about this. It may not be obvious from the picture, but the work envelope actually extends well past the front of the base plate. This will allow me to clamp things to the face or position workholding on a shelf below the surface of the baseplate.

I never knew that .1 mm was considered "fanatically precise". Are you sure you don't mean .01mm? Even that isn't what I would consider overly precise.
 
And here I was trying to be nice, and have the OP avoid the ensuing shitstorm that's coming.....

I'm a bit surprised it's gone as well as it has, so far.

Not to run the guy off, it's a bit beyond hobby grade... But only a bit.

We shall see what happens.

I won't mind reading along as the build continues, anyway..
 
Yes, understood. Sourcing or fabricating these structures in steel is extremely difficult over here, and I felt the aluminum would work well enough as a proof of concept if nothing else. If it is inadequate then doing an epoxy granite fill would be next, followed by strategic reenforcement with steel plate or replacement with a custom built (expensive) weldment.

Have you considered filling the extrusion with epoxy granite?
Additionally, if you add some bolts through the extrusion into the void spaces, the granite will then have something to grip to.



For keeping chips off of the ballscrews, consider a series of nested cups, like think about the cover for the cylinder on an office chair


If you havent figured out a controller yet, consider the centroid Acorn or OAK. I dont know if you can get them over there, but I used an Acorn in my build and have had a good experience.


For maximum space savings, look into low profile vises and solutions from people like Mitee-Bite. Space is a premium, so dont take up half of it with a big vise if you dont have to

sounds like a fun project :]
 
I don't. It's 2022. Skip that crap and go direct to linear motors.

You missed the part where he's severely power limited. Your hot, watt-hungry linear motors ain't so nice without the juice, and downrating them brings a loss of stiffness.

Ballscrews stay stiff as long as the drives are functioning and the motors aren't hilariously small. Ballscrews are nature's Viagra, linear motors are more saltpeter...
 
You missed the part where he's severely power limited. Your hot, watt-hungry linear motors ain't so nice without the juice, and downrating them brings a loss of stiffness.
You're living in the past. Doesn't work that way no more.

Besides, what kind of mad scientist doesn't have giant knife switches that dim the lights in his house when he puts power to the monster ? Pfffft.

So far, not going in a particularly advanced direction. I could ship him a complete machine that's about what he's describing (less the laptop for the control) for a grand. Maybe $50 shipping.

That's kinda where this is going, a tabletop toy.

Have to think that a really nice tabletop toy would be an interesting discussion tho. Could start with epoxy granite base, linear motors, beryllium structure, etc etc.
 








 
Back
Top